DDL2016 Digital Proceedings (DDL27)

Page 1

The Aerosol Society

DDL – Shaping the next 25 years of nasal and pulmonary Drug Delivery

7th, 8th & 9th December 2016 Proceedings


Aerosol Logo

Drug Delivery to the Lungs 27

DDL27 Logo

Wednesday 7th - Friday 9th December 2016 Edinburgh International Conference Centre, Scotland, UK Drug Delivery to the Lungs 27 - Friday 9 December 2016 Wednesday Thank you to7 our Platinum Sponsors th

th

Edinburgh International Conference Centre, Scotland, UK

Thank you to our Platinum Sponsors

Add Logos 3M CHIESI Intertek Molins Oechsler Team Consulting

Visit the DDL27 Event App

Visit the DDL27 Event App

Copyright 'The Aerosol Society'


About DDL Drug Delivery to the Lungs (DDL) is Europe's premier conference and exhibition dedicated to pulmonary and nasal drug delivery. DDL is organised by a voluntary sub committee of the Aerosol Society and aims to provide a conference dedicated to those with an interest in drug delivery to the airways. The DDL scientific organising committee is made up of members representing industry and academia.

Scientific Organising Committee DDL27 The DDL committee is made up of members representing industry and academia. The current members are: Gary Pitcairn Omar Usmani Sheila Coates Nikki Evans Nick Childerhouse Michelle Dawson Ben Forbes Mark Hammond David Harris Phil Haywood Yorick Kamlag Regina ScherlieĂ&#x; Alex Slowey Geralt Williams

AstraZeneca Imperial College London Vectura Limited GlaxoSmithKline King's College London Aesica Pharma PA Consulting Group OINDP Consultant Zentiva Inhalationsprodukte GmbH Kiel University 3M Health Care Limited Aptar Pharma

Conference Chair Conference Vice-Chair Conference Organiser Conference Organiser

The Aerosol Society is a member based organisation that exists to promote the science of particle distribution and its application in aerosols. Membership is wide reaching and includes scientists and researchers in the fields of climate engineering, environmental pollution, atmospheric health, nano-technology as well as those from the pharmaceutical and healthcare industries.

Conference Abstracts Review Panel To enhance the scientific quality of the meeting and published proceedings, expert reviewers were invited to assist the scientific organising committee in reviewing abstracts submitted to the conference. Many thanks to: Francesca Buttini Shyamal Das Jolyon Mitchell Darragh Murnane Steve Newman Steven Nichols Jag Shur Jason McConville Janet Maas Sarah Zellnitz

University of Parma, Italy University of Otago, New Zealand Jolyon Mitchell, Inhaler Consulting Services Inc, Canada University of Hertfordshire, UK Consultant, UK OINDP Consultant, UK University of Bath, UK University of New Mexico Consultant Research Center Pharmaceutical Engineering, GmbH


CONTENTS Page Nos: DDL27 Programme

1-4

Conference Opening Plenary

5-6

Session 1 Presentations

7 - 19

Inhalers – Were They Really Meant For Humans? The Annual DDL Lecture Plenary Lecture Session 2 Presentations

20 - 23 24 25 - 40

Beyond Asthma And COPD Session 3 Presentations

41 - 60

The Pat Burnell New Investigator Award Session 4 Presentations

61 - 76

60 Years Of The pMDI Session 5 Presentations

77 - 92

Innovations In Aerosol Delivery Summary of Posters

93 - 101

Posters

102 - 380


WEDNESDAY 7TH DECEMBER

11.00

Registration & Refreshments

12.00

Networking Buffet Lunch in the Exhibition Hall

13.30

Welcome & Opening Remarks Gary Pitcairn, Chair, DDL Committee

13.45

Conference Opening Plenary The European Asthma Research & Innovation Partnership - tackling Europe’s high asthma prevalence and death rates Dr Samantha Walker, Director, Research & Policy and Deputy Chief Executive, Asthma UK

Page Nos 5-6

SESSION 1 INHALERS – WERE THEY REALLY MEANT FOR HUMANS? 14.15

7 - 10 Inhaler device and errors Professor Henry Chrystyn, Emeritus Professor, Inhalation Consultancy

14.45

Manned Posters (1 – 36), Exhibition and Refreshments

16.15

Human factor considerations in inhaler innovation Mr Julian Dixon, Director of Human Factors, Team Consulting Ltd

11

16.45

Clinically relevant in vitro tests for the assessment of innovator and generic nasal spray products Mrs Mandana Azimi, Virginia Commonwealth University

12 -15

17.05

Innovations in nebulized delivery Dr John N Pritchard, Respironics Respiratory Drug Delivery (UK) Ltd

16 - 19

17.30 – 20.00

17.30 - 20.00

Exhibitor Drinks Reception

Exhibitor Drinks Reception Exhibition hall hall Exhibition Drinks all attendees attendees Drinks&&Canapes Canapés for for all Sponsored by Sponsored by (Add Vectura logo)

1


THURSDAY 8TH DECEMBER :AM

08.00

Registration & Refreshments Exhibition Hall & Poster Viewing

08.55

Opening of Conference

09.00

ANNUAL DDL LECTURE The challenge of delivering drugs to the lungs Dr Stephen P. Newman, Scientific Consultant

09.45

PLENARY LECTURE Biologics in Asthma – are we turning the corner Professor Roland Buhl, Mainz University Hospital

10.15

20 - 23

24

Manned Posters (37 - 76), Exhibition and Refreshments

SESSION 2 BEYOND ASTHMA AND COPD

25 - 28

11.30

Development to approval of NARCAN® nasal spray Mr Fintan Keegan, Head of Technical Operations, Adapt Pharma Ltd

12.00

New lung-tumour-penetrating nanocarrier designed for aerosolized chemotherapy Dr Rémi Rosière, Université Libre de Bruxelles

29 - 33

12.15

Oxytocin receptor expression but a lack of response to oxytocin in airway tissues Mr Jibriil Ibrahim, Monash Institute of Pharmaceutical Sciences

34 - 36

12.30

Anti-inflammatory activity of novel trans-stilbene sulfonamide analogues as 37 - 40 potential novel therapeutic agents for lung disease Mrs Rhamiya Mahendran, University of Hertfordshire

12.45

Networking Buffet Lunch in the Exhibition Hall & Poster Viewing

2


THURSDAY 8TH DECEMBER : PM

SESSION 3 THE PAT BURNELL NEW INVESTIGATOR AWARD 14.00

Characterisation of MRP1 in human distal lung epithelial cells in vitro Mr Mohammed Ali Selo, Trinity College Dublin

41 - 44

14.15

Carrier microstructure and performance of dry powder inhalation mixtures: A step towards universal performance prediction model Mr Ahmed O. Shalash, European Egyptian Pharmaceutical Industries

45 - 48

14.30

Development of inhalable drug formulations for idiopathic pulmonary fibrosis Mr Ville Vartiainen, University of Helsinki

49 - 52

14.45

Investigating the effect of the force control agent magnesium stearate in fluticasone propionate dry powder inhaled formulations with single particle aerosol mass spectrometry (SPAMS) Mr Martin Jetzer, Novartis Pharma AG

53 - 56

15.00

Influence of blender type on the performance of ternary dry powder inhaler formulations Mr Mats Hertel, Kiel University

57 - 60

15.15

Posters, Exhibition and Refreshments

SESSION 4 60 YEARS OF pMDI 16.00

The Future of Propellants for pMDIs Dr Tim Noakes, Medical Products Technical Associate, Mexichem & Stuart Corr, Mexichem

61 - 64

16.30

Next generation formulations for pMDIs Professor Glyn Taylor, Director of Research, i2c Pharmaceutical Services

65 - 68

17.00

Novel techniques for characterising inhalers Dr Alan McKiernan, Prior PLM Medical

69 - 72

17.20

Multi-physics theoretical approach to predict pMDI spray characteristics Dr Barzin Gavtash, Loughborough University

73 - 76

17.45 20.00 17.45 –- 20.00 Entertainment Evening, Cromdale Hall, Conference Centre Poster Awards & Prize Giving Entertainment Evening by Centre Cromdale Sponsored Hall, Conference Poster Awards & Prize Giving Sponsored by (Add 3M Logo) 3


FRIDAY 9TH DECEMBER

08.00

Registration & Refreshments Exhibition Hall & Poster Viewing

SESSION 5 INNOVATIONS IN AEROSOL DELIVERY 09.00

Innovations in formulations to improve aerosol delivery Professor Peter York, Chairman and Chief Scientist, CrystecPharma

09.30

Aspergillus fumigatus in valved holding chambers: use of silver ion additive 78 - 81 technology on fungal activity and aerosol delivery characteristics Mr Mark Sanders, Clement Clarke International Ltd

09.50

In-silico lung modeling platform for inhaled drug delivery Dr Antonio Cabal, Merck & Co

10.10

Posters, Exhibition and Refreshments

11.20

m-health on asthma – friend or foe Dr Stephen Fowler, The University of Manchester

87

11.50

Polymer-protein-based dry powder for efficient pulmonary protein delivery Mr Alejandro Nieto-Orellana, University of Nottingham

88 - 91

12.10

Multiplexed optical molecular sensing and imaging in the lungs Dr Kev Dhaliwal, University of Edinburgh

92

12.40

Concluding Remarks and Conference Close Gary Pitcairn, Chair, DDL Committee

4

77

82 - 86


Drug Delivery to the Lungs 27, 2016 – Samantha Walker, Asthma UK

The European Asthma Research & Innovation Partnership - tackling Europe’s high asthma prevalence and death rates Samantha Walker, Director, Research & Policy and Deputy Chief Executive, Asthma UK Abstract Asthma is highly prevalent and associated with high morbidity and mortality. It affects 30-50 million people in Europe1, often starting in infancy and persisting throughout the lifetime. Asthma is a major global health challenge with growing impact, affecting more than 300 million people worldwide and at least 10% of all Europeans. People with both severe and mild asthma live at risk of life-threatening asthma attacks, leading to at least 500,000 hospitalisations each year2. Approximately 5-10% of asthma cases are so severe that current treatments do not work. 3 people die of asthma every day in the UK. Despite the fact that the direct and indirect costs of asthma are high and continue rising, asthma remains underprioritised in the EU research agenda. Only 0.5% of the FP7 Health research budget was devoted to asthma and COPD (€30m)3. For comparison, over €163m were spent for cardiovascular conditions (5.4 times bigger), and over €618m for brain research (20.6 times bigger). This means that asthma presents a real opportunity for the EU to address a major societal challenge whilst driving substantial economic growth. The high global prevalence of asthma, with an equivalent multibillion global market for asthma drugs, it’s historical underfunding, and the unmet need for new treatments and diagnostics represents an enormous business opportunity. We recognise that a lot of excellent work and partnerships have been developed to tackle asthma. But it’s important to recognise also that Europe’s asthma research communities are still considered fragmented, typically working at a national level on discrete and complex sub-areas of the disease, and that there is an urgent need to develop better medicines, diagnostic methods and effective self-management tools. To capitalise on this opportunity, Asthma UK was commissioned by the European Commission Framework Programme 7 to define the research priorities for asthma through coordination of the European Asthma Research and Innovation Partnership (EARIP: www.earip.eu) EARIP has brought together researchers, industry representatives, patient groups and policymakers to undertake a comprehensive analysis of asthma research, including basic biology, new treatments, health and care systems, self-management and diagnostic tools, to develop a ‘roadmap’ for investment based on unmet need in asthma. Eight expert working groups comprising of world-leading researchers, clinicians, pharmaceutical company personnel and people with asthma carried out reviews of the scientific literature and identified research priorities in key areas of basic, translational and applied asthma science through large scale surveys and pan-European eDelphi consensus-building exercises. Representatives from each group then met to agree a single, coordinated agenda for asthma research and development (R&D). Fifteen over-arching research priorities were agreed which fall into four broad priority areas; personalised medicine, triggers and risk factors for asthma and exacerbations, self-management and adherence and primary care and public health4. Underpinning this is an urgent need for greater understanding of basic asthma mechanisms to drive the development of new treatments in the ~50% people who respond poorly to steroids 5; development of novel diagnostic technologies to allow the development of personalised intervention strategies; and to maximise the potential of digital platforms to transform problems associated with adherence and self-management. This R&D ‘roadmap’ forms the basis of plans to revolutionise the way that asthma is investigated and treated in Europe by attracting significant public and private investment as well as prioritising asthma research in the European medical research agenda. With targeted and focussed investment, this research would have dramatic implications for the lives of over 30million people living with asthma in Europe, whilst reducing healthcare costs, improving competitiveness, and providing a framework through which to solve other long term conditions.

Global Initiative for Asthma (GINA), 2004 Asthma Network (2014), The Global Asthma Report. 3 Decramer M, Sibille Y, eds. European Respiratory Roadmap. Sheffield, European Respiratory Society, 2011. 4 S. C. Masefield, (Sheffield, United Kingdom) et al. , Driving Investment in Asthma Research in Europe: Priorities to Prevent, Cure and Manage Asthma More Effectively. In: American Thoracic Society conference; 2016 May 1318; San Francisco, USA. A2668 1

2 Global

McGrath KW et al. A large subgroup of mild-to-moderate asthma is persistently noneosinophilic. Am J Respir Crit Care Med. 2012 Mar 15;185(6):612-9. Epub 2012 Jan 20. 5

5


Drug Delivery to the Lungs 27, 2016 - The European Asthma Research & Innovation Partnership - tackling Europe’s high asthma prevalence and death rates

Having established the research gaps and priority areas for investment, it is vital that an integrated approach is taken to deliver these in order to maximise resource, and achieve the biggest impact for people with asthma, in the shortest space of time. Understanding how breakthroughs in one area will open the door to new developments in another is key to this; similarly, understanding who is best placed to fund and carry out R&D will enable the most coordinated approach to be adopted. Asthma UK is continuing to drive this integrated approach and through our soon-to-be launched 2016-21 research strategy will:    

Strive to fund a balanced mix of types of research. By doing this we link basic and applied science to create a pipeline of research innovation that speeds up the process through which new interventions reach people with asthma. Fund two Research Centres in the UK to develop a critical mass of researchers to collaborate, grow talent, build networks and act as hubs from which to lever greater research activity. Invest to expand the pipeline of world class asthma scientists and equip researchers across all disciplines to build more compelling asthma research applications by sharing our case for investment Actively promote collaborative working and patient and public involvement in asthma research. We will explore opportunities to organise ‘sandpit’ events with the aim of connecting researchers to write collaborative bids for funding for a specific priority research area. Patient and public involvement in research increases the quality and relevance of research and Asthma UK is committed to the promotion of this at all stages of research. Invest our funding where we think this will leverage additional financial support from other funders, and actively pursue different types of research funding partnership models to increase funding available for asthma research, subject to these meeting the same standards of governance as our existing funding.

However wisely Asthma UK directs its own limited research funding, to truly deliver better outcomes for people with asthma we must also be more active in driving European investment in asthma R&D. In the last 5 years there have been more calls on the limited research funds available globally, and we believe asthma has often been overlooked. We are proposing to make the case for asthma research as strongly as possible in the UK and internationally, to ensure asthma research receives a greater share of available research funding. We will be asking funders, institutions, researchers and academics to adopt and address the fifteen research priorities. This will improve collaboration and help us work more effectively towards common goals to improve the lives of the 300 million people living with asthma in Europe. We will be working with funders to adopt the roadmap in their funding decisions, and encouraging researchers to address the specific priorities using the funding offered in line with the priorities from the roadmap. Over the coming years we will also explore how best to encourage collaboration between researchers in the UK and across Europe to address these priorities. We will continue to work with the European Federation of Allergy and Airways Diseases Patients' Associations (EFA) to encourage organisations to adopt the Malaga-London declaration [http://www.efanet.org/malaga-londondeclaration] which is a call for the establishment of strategic framework for asthma research and for the provision of specific funding for asthma within the EU research and innovation funding programmes, as well as through specific initiatives at national level. It makes these calls for specific funding in line with the roadmap’s objectives and pulls together funder-researcher focus on to these areas. EFA will be asking its 28 members to write to members of the European Parliament and the relevant members of the council of ministers calling for EU support in the adoption of the patient declaration. Our engagement will continue with strategic partners in industry and in the Directorate-General responsible for Science, Research and Development in order to secure more investment through one of the European Commissions’ funding streams along the lines outlined in the research roadmap. Specifically, we are exploring the opportunity for a new public-private partnership between industry and the Commission that would drive the development of new therapeutic targets linked to emerging asthma phenotypes. We hope that if funding frameworks are set out alongside the roadmap we will not only see more investment, but also increased European collaboration. It is hoped that EARIP will also lay a stronger foundation for Europe's individual pharmaceutical and healthcare technology companies to grow, and create jobs, through the better identification of new therapeutic targets that will help boost competitiveness and accelerate innovations in healthcare systems and services. Lastly, for true innovation and better drug delivery to work the European Union and member nations also need to be ready for change. There is work to be done around a better regulatory framework for new or innovative treatments. National healthcare systems will have to find ways to fund drugs aimed at small patient cohorts and embrace technology to help managing ageing populations with chronic comorbidities. In summary, the identification of and cross-disciplinary agreement on the 15 most important priorities for investment in asthma R&D gives everyone the opportunity to work together across Europe to transform outcomes for people with asthma and drive growth and jobs. The development of new drugs and better diagnostic tools linked to biomarkers and a greater understanding of asthma mechanisms are fundamental to ensuring that everyone with asthma receives the medicine that works best for them and that asthma deaths are a thing of the past.

6


Drug Delivery to the Lungs 27, 2016 – Chrystyn H. Inhaler device and errors Chrystyn H Inhalation Consultancy Ltd, Tarn House, 77 High Street, Yeadon Leeds, LS19 7SP and RiRL, 5 Coles Lane, Oakington, Cambridge, CB24 3BA. Summary Inhalation errors have a significant impact on the disease control of asthma and chronic obstructive pulmonary disease (Melani AS et al. Respir Med. 2011;105(6):930-8; Al-Jahdali H et al; Allergy Asthma Clin Immunol. 2013;9(1):8. Westerik JA et al; J Asthma. 2016;53(3):321-329) with resultant economic burdens on the health care costs (Lewis A et al. BMC Health Serv Res. 2016; Jul 12, 16:251). The type and incidence of these errors has not changed over the past 40 years (Sanchis J et al. Chest. 2016;150(2):394-406). Analysis of published inhaler errors highlights that these can be categorised into dose preparation and inhalation manoeuvre errors. These categories are all generic for metered dose inhalers. Inhalation manoeuvre procedures are also generic errors when using dry powder inhalers whereas dose preparation steps are device specific. The CritiKal study, carried out throughout Western Europe and in Australia, examined inhaler errors and the level of asthma control in 3660 patients. This study has identified which inhalation technique errors are clinically significant. Also new invitro methods of identifying dose emission using patient inhalation profiles (Chrystyn H et al. Int J Pharm. 2015;491(12):268-76) are providing useful information about formulations that are not affected by how the patient inhales. Furthermore newer inhalers are now designed so that they are more intuitive to use. These latest contributions to the literature should direct healthcare professionals' counselling sessions when choosing an inhaler for a patient and training them how to use it. Introduction Problems using the correct inhaler technique have been reported to be associated with poor disease control in the management of asthma and chronic obstructive pulmonary disease [1] [2] [3]. It has been reported that errors when using either a metered dose inhaler (MDI) or dry powder inhaler (DPI) are significantly associated with an increased risk of hospitalisations, emergency room visits, increased antibiotic prescribing and increased prescribing of short course of oral prednisolone [1] as well as education level, the patient's age and follow up after inhaler training [1] [2] [3]. These studies have also shown that no previous inhaler technique training is significantly associated with errors. An economic costing model has identified the burden of poor inhaler technique in Spain, Sweden and the UK [4]. Taking into account hospitalisations, emergency room visits, increased antibiotic prescribing and oral prednisolone courses the annual cost of poor inhaler technique was €109, €55 and €21 per patient. Projecting these direct costs and adding indirect costs (such as lost productivity) revealed that the overall cost to each country was considerable. It is important that the clinical significance of inhaler errors is identified and steps are taken to focus on these because there has been no improvement with inhaler technique over the past 40 years [5] from when they were first identified [6]. Inhaler technique - scale of the problem and general classification of errors Recent systemic reviews [5] [7] and a meta-analysis [7] have shown that there are large number of patients that make errors when using their inhalers. The pooled estimate for the percentage of patients with a least one overall and one critical inhaler technique error when using MDIs was found to be 86.8% [95% confidence interval (CI) 79.4–91.9] and 45.6% [95% CI 26.0–66.6], respectively [7]. Similar error rates for DPIs were 60.9% [95% CI 39.4– 79.0] and 28.4% [95% CI 22.0–35.8], respectively [7]. Table 1 highlights that the problems when inhalers are used can be divided into dose emission, dose preparation and the inhalation manoeuvre. For MDIs all the problems shown in Table 1 are generic whereas for DPIs dose emission and dose preparation are unique to the DPI, hence they are device specific. Errors associated with the DPI inhalation manoeuvre are generic. The most common inhalation manoeuvre errors are no exhalation, not sealing the lips round the mouthpiece and not using a fast and deep inhalation as well as no breath hold. Table 1. Problems associated with inhaler use MDI devices Dose emission Consistent Dose preparation Errors are common: shake, upright Inhalation manoeuvre Errors are common: exhale, coordination, slow and deep inhalation, breath hold

DPI devices Ranges from consistent to erratic Errors are device specific Errors are common: exhale, seal lips, fast and deep inhalation, breath hold

Inhaler use problems (a) Dose emission One advantage of MDIs is that dose emission is always consistent whether this is at the beginning (BOL), middle (MOL) and end (EOL) of life for the doses available in the device [8]. In contrast dose emission from DPIs ranges from consistent [9] at BOl, MOL and EOL [10] to large inter- and intra- inhaler variability in the emitted dose [9] [11].

7


Drug Delivery to the Lungs 27, 2016 - Inhaler device and errors This DPI dose emission variability occurs despite preparing the dose according to the manufacturer's recommended dose preparation procedures in the Patient Information Leaflet. (b) Dose preparation Not shaking before actuating a dose and not holding the device upright are the common errors made by patients when using their MDIs [1]. Shaking the MDI is required because some formulations are a suspension of drug particles in the propellant while not holding the canister upright means that the next dose is not extracted from the formulation [12]. Many MDI formulations are now solutions so shaking before use may not be important but data about this is not available. Incorrect orientation of the DPI when preparing the dose is a very common error made by patients [1] [13]. Other issues during DPI dose preparation are shaking the device after the dose has been prepared [14] and exhaling into the device [1) [13]. Shaking a DPI after the dose has been metered reduces the dose available for inhalation whilst exhaling into the device introduces moisture to the formulation and thus affects dose emission [15]. These shaking and exhaling issues can only be solved by inhaler training whereas correct dose preparation can be solved by DPIs that are not dependent on the inhaler's orientation when the dose is metered [10]. Reducing the number of dose preparation steps and making DPIs more intuitive to use reduces the incidence of these dose preparation errors [14] [16]. (c) Inhalation manoeuvre Not using a slow inhalation and poor co-ordination when using a MDI are the two most common errors made by patients [17] [18]. When patients use a slow inhalation co-ordination is not important as long as the dose is actuated after the start of the inhalation [19]. Using a slow inhalation flow with a MDI improves asthma quality of life [17]. MDIs formulated so that the emitted dose emits extra-fine particles improves lung deposition and reduces the effect of the inhalation flow [20] as well as reducing the effect of poor co-ordination [21]. The generic instruction to use a 'slow and deep' inhalation is not clear and open to mis-interpretation. It has been recommended to replace this with an instruction to use an inhalation that lasts 5 seconds in an adult and 2-3 seconds in a child [22]. This objective instruction has been shown to slow down inhalation flows when patients use their MDI and during real-life use they maintain using a slow flow over the next 4 weeks [23]. Two large studies of real life inhaler technique have reported that almost a third of patients made inhalation manoeuvre errors when using their DPI [1] [13]. These common errors were not exhaling, not inhaling with a forceful and deep inhalation as well as not holding the breath at the end of each inhalation [1] [13]. A recent systematic review [5] has shown that 46% of patients[95% CI of 42%-50%] did not exhale before making an inhalation. Thirty seven percentage of patient [95% CI of 33%-40%] did not hold their breath after an inhalation. Not using a deep and forceful inhalation occurred in 16% [95% CI, of 13-20%] [5]. Another systematic review and meta analysis [7] has also revealed that the most common errors when using a DPI were no exhalation, no breath hold and not sealing the lips. What is the solution to the problems patients have using inhalers? It is essential to prescribe an inhaler that a patient can and will use [22]. A recently completed study, CritiKal [24], which is part of the iHARP asthma review service, involved 3660 adults with asthma throughout West Europe and in Australia. This study has identified which inhaler technique errors are associated with poorer asthma control. CritiKal has identified that the generic inhalation manoeuvre errors, identified above, are all related to poorer asthma control. Furthermore some clinically significant dose preparation errors have been identified [24]. The CritiKal study has, therefore, identified which clinically important (hence critical) errors to focus on during inhalation technique training. Novel in-vitro dose emission methodology which uses patient inhalation profiles to identify the type of dose the patient would have inhaled during real life use has been pioneered [25]. This method replaces the vacuum pumps traditionally used to measure dose emission. Vacuum pumps can only produce a square wave inhalation profile that no human can replicate. Using this new and novel methodology it has been shown that the dose emitted from some DPIs is not dependent on the speed of the inhalation [25]. Hence. if a patient uses a weak inhalation when using certain DPIs the emitted dose is the same as the emitted dose when using a strong inhalation [25]. Using devices that are intuitive to use [14[ [16] will reduce the incidence of dose preparation errors and help to maintain the trained inhalation technique during real life use [14] while patient satisfaction with the device improves patient compliance [26]. Wherever possible patients should be prescribed the same type of device for their medications (either all MDIs or all DPIs). It has recently been reported that patients with COPD who mix devices (MDI and DPI) have more exacerbations than those that use similar devices [27]. The cost of producing most salbutamol DPIs compared to the much lower remuneration that would be received for these salbutamol devices limits their availability. Summary The problems that patients have using their inhalers are common and result in poorer disease control and increased healthcare costs. The Critikal study has identified the clinically significant inhaler technique errors made by patients and these should be the focus during inhaler technique training. Mixing devices should be minimised but this may not be easy because of the limited availability of salbutamol DPIs. Using devices that are more intuitive, easy to use and preferred by patients as well as using devices whose emitted dose is not dependent on the inhalation flow and focusing training on the potential errors that give rise to poor disease control will help to solve the problems patients have using their inhalers.

8


Drug Delivery to the Lungs 27, 2016 – Chrystyn H.

References [1]. Melani AS, Bonavia M, Cilenti V, Cinti C, Lodi M, Martucci P, Serra M, Scichilone N, Sestini P, Aliani M, Neri M; Gruppo Educazionale Associazione Italiana Pneumologi Ospedalieri. Inhaler mishandling remains common in real life and is associated with reduced disease control. Respir Med. 2011; 105(6): pp930-938. [2]. Al-Jahdali H, Ahmed A, Al-Harbi A, Khan M, Baharoon S, Bin Salih S, Halwani R, Al-Muhsen S. Improper inhaler technique is associated with poor asthma control and frequent emergency department visits. Allergy Asthma Clin Immunol. 2013; 9(1):8. [3]. Westerik JA, Carter V, Chrystyn H, Burden A, Thompson SL, Ryan D, Gruffydd-Jones K, Haughney J, Roche N, Lavorini F, Papi A, Infantino A, Roman-Rodriguez M, Bosnic-Anticevich S, Lisspers K, Ställberg B, Henrichsen SH, van der Molen T, Hutton C, Price DB. Characteristics of patients making serious inhaler errors with a dry powder inhaler and association with asthma-related events in a primary care setting. J Asthma. 2016; 53(3): pp321-329. [4]. Lewis A, Torvinen S, Dekhuijzen PN, Chrystyn H, Watson AT, Blackney M, Plich A. The economic burden of asthma and chronic obstructive pulmonary disease and the impact of poor inhalation technique with commonly prescribed dry powder inhalers in three European countries. BMC Health Serv Res. 2016; 6: 251. [5]. Sanchis J, Gich I, Pedersen S; Aerosol Drug Management Improvement Team (ADMIT). Systematic Review of Errors in Inhaler Use: Has Patient Technique Improved Over Time? Chest. 2016; 150(2): pp394-406. [6]. Paterson IC, Crompton GK. Use of pressurised aerosols by asthmatic patients. Brit Med J. 1976; 10th January: pp76-77. [7]. Chrystyn H, van der Palen J, Sharma R, Barnes N, Delafont B, Mahajan A, Thomas M. Device errors in COPD and asthma: systematic literature review and meta-analysis. (manuscript submitted). [8]. Cripps A, Riebe M, Schulze M, Woodhouse R. Pharmaceutical transition to non-CFC pressurised metered dose inhalers. Respir Med 2000; 94 (Suppl B): ppS3-S9. [9]. Palander A, Mattila T, Karhu M, Muttonen E. In-vtro comparison of three salbutamol containing multidose dry powder inhalers. Clin Drug Invest. 2000; 20: pp25-33. [10]. Canonica GW, Arp J, Keegstra JR, Chrystyn H. Spiromax, a New Dry Powder Inhaler: Dose Consistency under Simulated Real-World Conditions. J Aerosol Med Pulm Drug Deliv. 2015; 28(5): pp309-19. [11]. Tarsin W, Assi KH, Chrystyn H. In-vitro intra- and inter-inhaler flow rate-dependent dosage emission from a combination of budesonide and eformoterol in a dry powder inhaler. J Aerosol Med. 2004;17(1): pp25-32. [12]. Byron P. Performance Characteristics of Pressurized Metered Dose Inhalers in Vitro. J Aerosol Med 1997; 10(1): pp S3-S6. [13]. Molimard M, Raherison C, Lignot M, Depont F, Abouelfath A, Moore N. Assessment of handling of inhaler devices in real life: An observational study in 3811 patients in primary care. J Aerosol Med 2003; 16: 249254. [14]. Chrystyn H, Dekhuijzen R, Rand C, Bosnic-Anticevich S, Roche N, Lavorini F, Thomas V, Steele J, Raju P, Freeman D, Small IR, Canvin J, Price DB. Evaluation of Inhaler Technique Mastery for Budesonide Formoterol Spiromax® Compared with Symbicort Turbohaler® in Adult Patients with Asthma Primary Results From the Easy Low Instruction Over Time [ELIOT] Study. Thorax. 2015 ;70 (Suppl 3): ppA154. [15]. Holmes MS, Seheult JN, O'Connell P, D'Arcy S, Ehrhardt C, Healy AM, Costello RW, Reilly RB. An Acoustic-Based Method to Detect and Quantify the Effect of Exhalation into a Dry Powder Inhaler. J Aerosol Med Pulm Drug Deliv. 2015; 28(4): pp247-53. [16]. van der Palen J, Thomas M, Chrystyn H, Sharma RK, van der Valk PDLPM, Goosens M, Wilkinson T, Stonham C, Chauhan AG, Imber V, Zhu C-Q, Svedsater H, Barnes NC. An open label crossover study of inhaler errors, preference and time to achieve correct inhaler use in patients with COPD or asthma: comparison of ELLIPTA® with other inhaler devices. NPJ Prim Care Resp Med. in press. [17]. Al-Showair RA, Pearson SB, Chrystyn H. The potential of a 2Tone Trainer to help patients use their metered-dose inhalers. Chest. 2007; 131(6): pp1776-82. [18]. Levy ML, Hardwell A, McKnight E, Holmes J. Asthma patients' inability to use a pressurised metered-dose inhaler (pMDI) correctly correlates with poor asthma control as defined by the global initiative for asthma (GINA) strategy: a retrospective analysis. Prim Care Respir J. 2013; 22(4): pp406-411. [19]. Newman SP, Pavia D, Garland N, Clarke SW. Effect of various inhalation modes on the deposition of radioactive pressurized aerosols. Eur J Respir Dis 1982; 63 (Suppl 119): pp57-65. [20]. Usmani OS, Biddiscombe MF, Barnes PJ. Regional lung deposition and bronchodilator response as a function of beta-2 agonist particle size. Am J Respir Crit Care Med 2005; 172: pp1497-1504. [21]. Leach CL, Davidson PJ, Hasselquist BE, Boudreau RJ. Influence of particle size and patient dosing technique on lung deposition of HFA-beclomethasone from a metered dose inhaler. J Aerosol Med. 2005; 18(4): pp379-85. [22]. Laube BL1, Janssens HM, de Jongh FH, Devadason SG, Dhand R, Diot P, Everard ML, Horvath I, Navalesi P, Voshaar T, Chrystyn H; European Respiratory Society; International Society for Aerosols in Medicine. What the pulmonary specialist should know about the new inhalation therapies. Eur Respir J. 2011; 37(6): pp1308-31. [23]. El Larhrib, Jones H, Rawson S, Chrystyn H. Train patients to prolong their inhalation manoeuvre when using a metered dose inhaler (MDI). J Aerosol Med Pulm Drug Del. 2015; 28(3): A11.

9


Drug Delivery to the Lungs 27, 2016 - Inhaler device and errors [24]. Price DB, Roman-Rodriguez M, McQueen RB, Bosnic-Anticevich S, Carter V, Gryffud-Jones K, Harris M, Haughney J, Henrichsen S, Infantino A, Lavorini F, Lisspers K, Papi A, Ryan D, Stallberg B, van der Molen T, Chrystyn H. Inhaler errors in the CRITIKAL study: type, frequency and association with asthma control . Manuscript in preparation. [25]. Chrystyn H, Safioti G, Keegstra JR, Gopalan G. Effect of inhalation profile and throat geometry on predicted lung deposition of budesonide and formoterol (BF) in COPD: An in-vitro comparison of Spiromax with Turbuhaler. Int J Pharm. 2015: 491(1-2): pp268-276. [26]. Chrystyn H, Small M, Milligan G, Higgins V, Gil EG, Estruch J. Impact of patients’ satisfaction with their inhalers on treatment compliance and health status in COPD. Respir Medi. 2014; 108, pp358-365. [27]. Bosnic Anticevich S, Chrystyn H, Costello R, Dolovich M, Fletcher M, Lavorini F, Rodríguez-Roisin R, Ryan D, Wan Yau Ming S, Skinner D, Earnest A, Law LM, Price DP. Impact of using multiple devices with mixed inhalation techniques on COPD outcomes: an observational study. Int J Chron Obstruct Pulmon Dis (in press)

10


Drug Delivery to the Lungs 27, 2016 - Julian Dixon Human Factor Considerations in Inhaler Innovation Julian Dixon1 1

Team Consulting Ltd, Abbey Barns, Duxford Road, Ickleton, Cambridgeshire, CB10 1SX, UK

Summary A human factors perspective on inhaled drug products illuminates a range of opportunities for meeting user needs. This paper will review and outline this range. Introduction Over recent years there has been considerable growth in the attention paid to human factors engineering/usability engineering (HFE/UE) on drug/device combination product developments, including OINDPs. This has primarily been focused on ensuring regulatory requirements regarding HFE objectives and process are met. The potential of the HFE/UE perspective for future inhaler device design, of course, goes beyond this focus on user error, patient safety and regulatory approval. This paper reviews and outlines this broader potential: the different ways in which human factors illuminates opportunities for innovation and improvement of inhaler device design. Improving the ‘human fit’ of inhalers The opportunities suggested by a focus on the human factors of inhaled products range in scale, from small readily achievable incremental improvements (‘low hanging fruit’) to those which would involve major long term technology development programmes and which might even be disruptive. Low hanging fruit will be exposed simply by engaging seriously with the HFE process throughout a development project. Better HFE will, at the very least, eliminate avoidably poor interface design decisions. Best practice HFE process is fundamental to achieving better detailed embodiment design. Sufficient and iterative formative research is key. Similarly, challenging and testing the real world effectiveness of instructional materials, including instructions for use (IFUs), will yield small improvements that benefit our users. A broader engagement with how these materials are experienced by users suggests opportunities for improvements. Industry engagement with regulatory agencies is likely to be needed to ensure that these improvements can be realised. A natural extension of this engagement with user experience is the increased and joined-up use of diverse technologies and media for training support and training aids. The established discipline of Inclusive Design focuses attention on the range of capabilities of our intended user populations and of designing interfaces that fall within the capabilities of as many individuals as possible. The drive to expand the inclusivity or universality of a general platform product design can be contrasted with the development of devices tailored or targeted to specific uses. Both represent opportunities. The HF validation study can, too readily, become the sole emphasis of HFE. Proper engagement with HFE will shift attention to its role earlier in development. Attention to preparatory HFE analysis has great potential to influence the clarity of the product concept / vision. And as device design progresses, early HF studies should be formative in terms of impact not just in name. The perspectives described above can be fruitfully applied to inhalers employing standard platform technologies. The platform technologies have well understood limitations that impact on their usability, but this does not mean that HFE cannot minimise or address many of them. High quality HF analysis at the earliest stages of development can direct effort effectively in this regard, including the choice of which platform technology to select at the outset. Giving proper weight to human factors considerations will also influence the trade-offs made between technical device performance and usability, between in-vitro performance under standardised conditions and real world performance in the hands of users. When considering potential new platform technologies, HF will again offer a different perspective on what counts as a good, or as the best, inhaler design. It will suggest different goals. HF will have important inputs to make to the direction and development of ‘connected’ inhalers, for example. In summary, human factors will be a key voice in future inhaler device development. It will carry weight alongside technical performance considerations.

11


Drug Delivery to the Lungs 27, 2016 – Mandana Azimi et al. Clinically relevant in vitro tests for the assessment of innovator and generic nasal spray products 1

Mandana Azimi , P. Worth Longest

1, 2

3

3

, Jag Shur , Robert Price & Michael Hindle

1

1

2

Department of Pharmaceutics, Virginia Commonwealth University, Richmond, VA, 23298, USA Department of Mechanical and Nuclear Engineering, Virginia Commonwealth University, Richmond, VA, 23298, USA 3 Department of Pharmacy and Pharmacology, University of Bath, Bath, BA2 7AY, UK

Summary In order to establish equivalence between innovator and generic nasal spray suspension products, similar regional drug deposition at their site of action in the nasal cavity needs to be ensured. The use of realistic models of the nasal airways together with testing using simulated patient use conditions may provide a rapid and inexpensive in vitro assessment of these nasal spray products. The regional deposition of an “in house” mometasone furoate nasal spray and a commercially available generic fluticasone propionate nasal spray test products were compared with their respective innovator products using realistic in vitro testing methods. Three sets of simulated patient use experimental conditions were chosen to compare regional deposition in two realistic nasal airways models (VCU models 1 and 2) in an attempt to simulate in vivo variability and to provide a range of expected middle passage nasal depositions. Based on the in vitro experiments, the middle passage and nasopharynx drug deposition of the “in house” mometasone furoate and fluticasone propionate generic nasal ® ® sprays were not significantly different than their respective innovator products, Nasonex and Flonase across the three different experimental conditions in the two nasal airway models. The developed in vitro test methods may offer a useful, and realistic way to differentiate nasal spray products. Future studies will evaluate the effect of nasal spray plume characteristics (droplet size, plume geometry) on the regional nasal drug deposition to assess the sensitivity of these methods to detect differences in nasal sprays. Introduction The use of realistic physical nasal airway models has become an emerging tool to investigate and compare locally acting nasal spray products. In order to establish equivalence between innovator and generic nasal spray suspension products, similar regional drug deposition at their site of action in the posterior part of the nose or [1, 2] . Realistic in vitro testing methods may offer a rapid and nasal middle passages needs to be ensured inexpensive means of screening nasal spray products compared to radiolabelled imaging and clinical studies. Furthermore, compared to the currently employed in vitro nasal spray product characterization methods which focus on the device and formulation performance, realistic in vitro test methods may also consider the interplay [3] between device, formulation, patient use variables and the resulting nasal drug deposition . Previously, an in vitro regional deposition testing method was described that revealed the importance of patient use factors and the ® geometry of the nasal cavity in determining the in vitro regional nasal deposition of the Nasonex nasal spray product. A full factorial design of experiment approach was used to assess the effects of patient related factors, including nasal spray position in the nose, head angle, the nasal spray actuation-nasal inhalation timing and applied actuation force and their interactions, together with the influence of nasal airway geometry on regional [4, 5] . Nasal middle passage deposition was observed to vary from 16.6 to 57.1% in VCU nasal drug deposition ® model 1 and 46.6 to 77.4% in VCU model 2 when tested in this realistic manner for the Nasonex nasal spray product. The utility of this realistic in vitro nasal deposition method will now be investigated using an “in house” and a commercially available generic nasal spray product, that have similar droplet size and plume characteristics compared to their respective innovator products. Therefore, the objective of this study is to compare the regional nasal drug deposition of these generic and innovator nasal spray products using previously developed realistic in vitro testing methods with a series of realistic experimental conditions in two nasal airways. Materials and Methods The regional nasal deposition of an “in house” mometasone furoate nasal spray, 50 μg (University of Bath, Bath, ® UK BN# MFM11T F9) was characterized and compared with the innovator product, Nasonex ,50 μg (Merck & Co. Inc., Whitehouse Station, NJ, USA) in the VCU nasal models 1 and 2. The “in house” formulation was designed for experimental purposes to be a generic copy of the innovator product with respect to droplet size and plume characteristics. The regional nasal deposition was tested using the experimental set-up shown in Figure 1, that includes an automated nasal spray actuator (Mighty Runt, InnovaSystems Inc., Moorestown, NJ, USA), the VCU nasal model (VCU model 1 or 2) and a programmable breath simulator (ASL 5000-XL, IngMar Medical, Pittsburgh, PA). The VCU nasal model 2 has a larger nasal vestibule surface area compared to VCU model 1 2 2 (14.94 cm vs 11.52 cm ), and the overall surface area/volume ratio for VCU model 2 was also larger than -1 -1 [5] observed in VCU model 1 (1.33 mm vs 0.74 mm ) . Based on previous studies, test conditions for VCU nasal model 1 and 2 were selected to produce low, intermediate and high combined middle passage and nasopharynx [4, 5] . For VCU model 2, only one nasal spray position was identified for testing due to its drug deposition (Table1) smaller nostril hydraulic diameter.

12


Drug Delivery to the Lungs 27, 2016 – Clinically relevant in vitro tests for the assessment of innovator and generic nasal spray products Similarly, the regional nasal deposition of fluticasone propionate nasal spray 50 μg (Roxane Laboratory, Ohio, ® USA) was characterized and compared with Flonase Nasal Spray, 50 μg, (GlaxoSmithKline, Research Triangle Park, NC, USA) using the experimental conditions and models described in Table 1, with only a single actuation [6] force of 5.8kg employed which was derived from the studies of Doughty et al . For each in vitro deposition experiment, 2 actuations of the nasal spray were delivered into a single nostril, the regional deposition of the drug was measured at four locations, (i) the nasal spray device, (ii) the anterior nose region + the amount of formulation dripped from the nose, (iii) middle passages + nasopharynx and (iv) throat + low resistant inspiratory filter at the exit of the throat. Validated HPLC assays were employed to determine the amount of drug deposited at each location and results were expressed as a percentage of recovered dose. Student t-test was used to compare the regional drug deposition of the “in house” mometasone furoate nasal spray with its innovator product and the generic fluticasone propionate with its innovator nasal spray product across the range of experimental conditions in the two nasal models (JMP Pro 12 software; p-value < 0.05).

Figure 1. Schematic diagram of experimental setup for the realistic in vitro nasal deposition studies Table 1. Experimental conditions for testing of nasal spray products using VCU nasal models 1 and 2.

Expected combined middle passage and nasopharynx regions drug deposition

Actuation force (kg)

Nasal spray a position (mm)

Head angle

Inhalation and b actuation timing

VCU Model 1 Level 1- Low (~ 20%) Level 2- Intermediate (~40%) Level 3- High (~60%)

7.5 7.5 7.5

9 5 5

50° 30° 50°

E D D

VCU Model 2 Level 1- Low (~ 50%) Level 2- Intermediate (~60%) Level 3- High (~77%)

7.5 4.5 4.5

NA NA NA

30° 30° 50°

E D D

a b

Distance between nasal spray applicator and the tip of the nose[4] Inhalation-actuation timing: E - inhalation started at the end of actuation, D - actuation during inhalation[5]

Results and Discussion ®

Middle passage and nasopharynx nasal drug deposition for the “in house” mometasone furoate and Nasonex nasal spray products in the VCU nasal models are summarized in Table 2 for low, intermediate and high, deposition conditions. The total drug recovery for these experiments ranged from 85.0 to 100.8%. The amount of drug deposited on the nasal spray device was lower than 3.0% and no drug recovered from filter positioned at the end of the nasal model. As shown in Table 2, there were no statistical differences for the amount of drug that deposited in the middle passage + nasopharynx region for paired experiments with the “in house” mometasone furoate and innovator product using the three different experimental conditions performed with VCU nasal models 1 and 2. Depending upon the experimental conditions, drug delivery to the middle passages varied from 20.2% to 59.1% (2.9-fold change) for the “in house” product and 16.6% to 57.1% (3.4-fold change) for the innovator product using VCU nasal model 1. Similar values in VCU nasal model 2 were 49.6-70.9% (1.4-fold change) and 46.677.4% (1.7-fold change), respectively. For comparison, conventional in vitro characterization of the spray droplet particle size distribution showed mean (SD) median volume droplet diameters of 47.2 (1.7) m and 44.5 (2.7) m, respectively, for the “in house” mometasone furoate and innovator nasal spray products while measured at actuation force of 7.5kg.

13


Drug Delivery to the Lungs 27, 2016 – Mandana Azimi et al. Table 2. Mean (standard deviation) in vitro middle passage and nasopharynx nasal drug deposition (expressed as a percentage of recovered dose) for the “in house” mometasone furoate and its innovator nasal spray product in VCU models 1 and 2.

Expected middle passage + nasopharynx drug deposition

Level 1-Low Level 2-Intermediate Level 3-High

Model 1 “In house” mometasone furoate 20.2 (3.2) % 37.0. (9.9) % 59.1 (6.8) %

Model 2 Nasonex

®

16.6 (2.4) % 34.1 (3.1) % 57.1 (9.4) %

“In house” mometasone furoate 49.6 (10.7) % 62.1 (13.8) % 70.9 (6.5) %

Nasonex

®

46.6(10.0) % 61.6 (9.6) % 77.4 (8.5) %

The regional drug deposition for the generic fluticasone propionate and innovator nasal spray products are summarized in Table 3. There were no statistical differences found in the amount of drug delivered to the middle passage and nasopharynx in paired experiments for these two nasal spray products using three different experimental conditions performed in VCU nasal models 1 and 2. Depending upon the experimental conditions, drug delivery to the middle passages varied from 17.5% to 39.5% (2.2-fold change) for the generic product and 21.0% to 47.2% (2.2-fold change) for the innovator product using VCU nasal model 1. Similar values in VCU nasal model 2 were 64.6-78.6% (1.2-fold change) and 55.6-79.6% (1.4-fold change), respectively. For comparison, conventional in vitro characterization of the spray droplet particle size distribution showed mean (SD) median volume droplet diameters of 69.4 (2.1) m and 70.8 (1.4) m, respectively, for the generic and innovator fluticasone propionate spray products at actuation force of 5.8kg. Comparing all the nasal spray products, there were no statistical differences in the recovered drug dose when tested using the low, intermediate and high deposition conditions in VCU models 1 and 2. Across both drug products, the variability in drug deposition in the middle passages appears more sensitive to the patient use testing parameters in VCU model 1, and overall there was a trend towards higher middle passage drug deposition observed in VCU model 2, both of which may be reflective of in vivo variability. Table 3. Mean (standard deviation) in vitro middle passage and nasopharynx nasal drug deposition (expressed as a percentage of recovered dose) for the generic fluticasone propionate and its innovator nasal spray product in VCU models 1 and 2.

Expected middle passage + nasopharynx drug deposition

Level 1-Low Level 2-Intermediate Level 3-High

Model 1 Generic fluticasone propionate 17.5 (1.0) % 42.7 (4.1) % 39.5 (5.9) %

Model 2 Flonase

®

21.0 (3.8) % 41.7 (8.4) % 47.2 (10.4) %

Generic fluticasone propionate 64.6 (4.8) % 74.5 (3.3) % 78.6 (5.6) %

Flonase

®

55.6 (7.3) % 68.1 (10.8) % 79.6 (0.3) %

Conclusions The regional drug deposition of innovator and generic nasal spray products of two drugs, mometasone furoate and fluticasone propionate, were characterized using realistic in vitro test methods. An ‘in house’ mometasone furoate nasal spray product was not significantly different from its innovator product with respect to drug deposition at the local sites of action within the middle passages of two realistic nasal airway geometries. Similarly, there was no deposition differences observed for a generic fluticasone propionate nasal spray product and its innovator using the realistic in vitro testing method. The range of local deposition observed using the patient use testing parameters and the two nasal models may be reflective of potential in vivo variability in nasal deposition. These realistic testing methods could have utility as an inexpensive tool for early evaluation of regional nasal deposition and with future in vivo validation may offer an efficient means of evaluating equivalence of nasal spray products. Acknowledgments Funding was provided by Contract # HHSF223201310220C, from the Department of Health and Human Services (DHHS), Food and Drug Administration. The content is solely the responsibility of the authors and does not necessarily reflect the official policies of the DHHS; nor does any mention of trade names, commercial practices or organizations imply endorsement by the United States Government. The technical support of IngMar Medical (ASL 5000-XL Breath Simulator) and InnovaSystems Inc (Mighty Runt Actuator Station) is gratefully acknowledged. The authors wish to thank Guenther Hochhaus, Renish Delvadia, Bhawana Saluja and Mohammad Absar for their assistance and useful discussions during the planning of this study.

14


Drug Delivery to the Lungs 27, 2016 – Clinically relevant in vitro tests for the assessment of innovator and generic nasal spray products

References 1 2 3 4 5 6

Shah S A, Berger R L, McDermott J, Gupta P, Monteith D, Connor A, Lin W: Regional deposition of mometasone furoate nasal spray suspension in humans, Allergy Asthma Proc 2015; 36: pp48-57. Djupesland P G: Nasal drug delivery devices: characteristics and performance in a clinical perspective-a review, Drug Deliv Transl Res 2013; 3: pp42-62. FDA Draft Guidance (2003) Bioavailability and bioequivalence studies for nasal aerosols and nasal sprays for local action. Azimi M, Longest P W, Hindle M: Towards clinically relevant in vitro testing of locally acting nasal spray suspension products. In: R N Dalby, P R Byron, J Peart, J Suman, P Young and D Traini (eds): Respiratory Drug Delivery Europe 2015. DHI Publisher, River Grove, IL; pp121-130, 2015. Azimi M, Longest P W, Hindle M, Walenga R L: Comparison of the in vitro deposition of nasonex nasal spray product in two realistic nasal airway models. In: R N Dalby, P R Byron, J Peart, J Suman, P Young and D Traini (eds): Respiratory Drug Delivery 2016. DHI Publishing, River Grove, IL; pp617-622, 2016. Doughty D V, Vibbert C, Kewalramani A, Bollinger M E, Dalby R N: Automated actuation of nasal spray products: determination and comparison of adult and pediatric settings, Drug Dev Ind Pharm 2011; 37: pp359- 366.

15


Drug Delivery to the Lungs 27, 2016 – John N Pritchard Innovations in nebulized delivery John N Pritchard Respironics Respiratory Drug Delivery (UK) Ltd, a business of Philips Electronics UK Limited, Chichester, PO20 2FT, UK. Summary In recent years, much of the focus of novel treatments for respiratory diseases has been in the development of inhalers containing new, longer-acting compounds in established drug classes, either alone or in combination. However, there remain patients who are unable to use such devices properly, who rely upon nebulized therapy, yet there have been few new drugs marketed in nebulized form in the last 30 years. This undoubtedly has had a negative impact on disease control. A novel approach to development is to focus first on bringing the nebulized formulation to market, and only to start development of the inhaler once the new drug candidate has successfully demonstrated safety and efficacy. Not only does this mitigate the risks associated with developing an inhaler formulation only for the drug to fail in the clinic, it also studies drug outcomes in the most severe group of patients, in whom a positive effect should be easiest to demonstrate. This can reduce the commercial barriers to the introduction of new nebulized drugs, thereby ensuring such patients have access to the latest innovations in drug development. Parallel device developments will ensure this market continues to grow at a faster rate than that of the overall inhaler market. Innovations such as small volume, breath activated mesh nebulizers are likely to see the gap narrow between the treatment times for inhaler and nebulizer use, whilst the added benefits of an electronic-based technology will lead to adoption of patient monitoring approaches in inherently more at-risk patient populations. Introduction

Estimated number of doses (Millions)

In recent years, much of the focus of novel treatments for respiratory diseases has been in the development of inhalers containing new, longer-acting compounds in the classes of ď ˘-agonists, muscarinics and combinations, either together or with a corticosteroid [1]. Accompanying this has been the introduction of new dry power inhalers and/or engineered particle formulations. These are suitable for the vast majority of patients. Nonetheless, there remain classes of patient, notably the very young, very old and very sick, who are unable to use such devices properly. These patients rely upon nebulized therapy, a delivery system that has been in existence for over 150 years. Yet there have been very few new drugs marketed in nebulized form in the last 30 years. Despite this lack of new product introductions, nebulized treatments are still estimated to be growing faster than those delivered by inhalers (Figure 1 [2]). Globally, this may in part be due to increasing longevity in the general population. However, this is also being driven by increasing access to medicines in developing economies, to the extent that nebulized treatments are now the largest value segment in China [3].

40000

MDI (+2.7% p.a.)

35000

DPI (+1.8% p.a.)

30000

Nebs (+3.3% p.a.)

25000 20000 15000 10000 5000 0 1995

2000

2005

2010

2015

Year

Figure 1 - Estimated global number of inhaled doses of medication by delivery system (adapted from [1]).

16


Drug Delivery to the Lungs 27, 2016 – Innovations in nebulized delivery Nevertheless, patients on nebulized therapy are unable to access the latest drug developments. This undoubtedly has had a negative impact on disease control. Take for example, the case of a COPD patient who exacerbates and is hospitalized in the US. Upon discharge, 50% are prescribed a nebulized treatment [4], most commonly a combination of short-acting -agonist and muscarinic drugs. Thus, their burden of disease can shift from taking treatment from an inhaler once or twice per day, to a series of 4 nebulized treatments, each lasting several minutes. It is well documented that increased treatment burden adversely affects adherence [5] and it has been estimated that a reduction in adherence of 20% leads to an increase in adverse events in COPD patients of over 10% [6]. It is not surprising, therefore, that once-daily nebulized bronchodilators have been described as a “compelling market opportunity” [4]. However, to date, this opportunity has been largely overlooked; this paper looks at key barriers to the introduction of new nebulized therapies and current initiatives to overcome them. Changing the development paradigm Pharmaceutical companies often begin inhalation drug developments with nebulizers in the early stages, to provide cheaper and faster drug development, but then switch to inhaler devices in later clinical trials to address the majority of patients. Typically it can take 10 to 12 years to launch the new drug onto the market. As a completely different delivery system and formulation, nebulized products will require fresh dose-ranging and pivotal Phase III studies, which will again take some 5 years or more to gain marketing approval. This means that there is little of the original drug patent life left by the time the nebulized formulation reaches the market. As a consequence, unless formulation patents can be obtained on the nebulized formulation, there is insufficient time left to recover the investment in developing this product before generic copies enter the market. Thus, by leaving nebulized development until late in the drug’s life-cycle makes the development of a new drug in nebulized form commercially unattractive. In order to overcome this problem, one company (Mylan) has decided to go straight to market with a new long-acting muscarinic (LAMA) drug moiety in nebulized form [7], despite having previously successfully developed a different LAMA for use in inhalers [8]. Building on this, a novel approach to development is to focus first on bringing the nebulized formulation to market, and only to start development of the inhaler once the new drug candidate has successfully demonstrated safety and efficacy. Not only does this mitigate the costs associated with developing an inhaler formulation only for the drug to fail in the clinic, it also studies drug outcomes in the most severe group of patients, in whom a positive effect should be easiest to demonstrate. This has also been modelled from a financial perspective, calculating the expected Net Present Value (eNPV, the value in today’s terms of cash flows expected from the project in the future, scaled down by the likelihood of success) under varying conditions of peak annual sales for the drug, and interest on the cash equivalent (Discount rate). This is shown in Figure 2 [9].

Figure 2 - Financial returns improve if developing a nebulized formulation before an inhaler (blue), rather than as lifecycle management (green) or an inhaler alone (red) (adapted from [9]).

17


Drug Delivery to the Lungs 27, 2016 – John N Pritchard It transpires that there is likely to be a much greater financial return than that from adopting a traditional approach to development of the inhaler first. Additionally, nebulized products can command a higher price per dose, and whilst device costs for nebulized products are higher as an initial outlay, they can be amortized over months or years of treatment, making them similar overall to those of a portable inhaler. Furthermore, the capital investment in device production is borne by the nebulizer manufacturer; it does not require an often lengthy development of a bespoke inhaler, and some nebulizers are available immediately with medical device market authorization. If this novel approach to development is adopted moving forward, then the commercial barriers to the introduction of nebulized drugs are significantly reduced, thereby ensuring patients requiring nebulized therapy have access to the latest innovations in drug development. Improving the usability of nebulizers Nebulizers are inherently relatively simple to use, in that the patient is only required to put the formulation into the device and then breathe through it using their natural tidal breathing pattern until delivery is complete. However, for jet nebulizers, there are drawbacks to the patient, including the noise of the compressor, lack of portability, treatment time and cleaning regimen. As a consequence, vibrating mesh nebulizers are gaining increasing popularity, as they address many of the above limitations [10]. As may be seen from Figure 3, they are capable of delivering a wide variety of formulations, so are suited to the delivery of most drugs. Indeed for drugs with low potency, nebulizers may be the only suitable delivery system. Mesh nebulizers are also assuming increasing use in a hospital setting, and with new developments such as mesh nebulizer systems for use in the emergency room (Aerogen UltraTM, Aerogen, Galway, Ireland), it is possible to have a single droplet generation system that can be bridged from in-line ventilator use right through to home use. If new drugs are forthcoming, then using state-of-the-art nebulizers will become increasingly important, as regulatory authorities expect a new drug only to be promoted with nebulizers for which the developer has clinical experience. Increasingly, the drug market authorization will be as a specific drug-device combination product. New, breathactivated mesh nebulizers with low residual volumes, mean that the dose from 2.5 mL of a current nebulizer formulation can be delivered in as little as 0.25 mL. Indeed, there is a novel nebulizer in development for the delivery of inhaled insulin that is capable of delivering doses as low as 50 µL with residual volumes less than 20% [11]. Such small volumes reduce treatment times to one or two breaths. The diameter of the apertures in the mesh can also be controlled to sub-micron tolerances [12], enabling the droplet size to be optimized for a particular mechanism of action. High efficiency may be particularly important in the development of treatments involving macromolecules. Not only are such drugs expensive to manufacture, they are inherently less potent, thereby requiring doses beyond those for which conventional inhalers are capable of delivering, not to mention any additional excipients required to enhance cell penetration. Furthermore, large doses will drive the development of faster output nebulizers in order to maintain an acceptable treatment time. Due to their inherently efficient use of energy, mesh nebulizers are likely to lead the way. This might be through increased hole density, larger aperture plates, or higher vibration frequencies.

Figure 3 - Variation in physicochemical properties of formulations successfully delivered using a vibrating mesh nebulizer (reproduced with permission from Respiratory Drug Delivery 2014, Virginia Commonwealth University and RDD Online [13]). Small molecules (●), proteins/peptides (●), liposomes (●), suspensions (●), and surfactants (●). Bars represent a range of values measured over a range of concentrations.

Improving adherence to therapy However good the device and drug, a therapy will be ineffective unless the patient is motivated to take their medication. Many factors influence this, so feedback from the device on correct operation can be important [14]. With

18


Drug Delivery to the Lungs 27, 2016 – Innovations in nebulized delivery a breath-activated mesh nebulizer, it is possible to log when and how the device has been used and upload the information to a server [15]. This information is helpful in understanding the challenging nature of true adherence, and may be crucial to the development and assessment of interventions to promote adherence; it allows patient, physician or support personnel such as relatives to monitor, provide feedback or coach the patient to improve adherence and technique. In the future, monitoring symptoms and capturing the data in the same system may offer even greater benefits, both by ensuring there is optimum treatment and by providing positive feedback to the patient to encourage proper adherence and manage their disease. Concluding remarks Patients requiring nebulized therapy have not benefitted from the recent spate of new products launched in inhaler format for the treatment of asthma and COPD. This situation is changing, as companies begin to recognize this unmet need. Parallel developments in the device technology will ensure this market continues to grow at a faster rate than that of the overall inhaler market. Innovations in nebulized delivery are likely to see the gap narrow between the treatment times for inhaler and nebulizer use, whilst the added benefits of an electronic-based technology will lead to rapid adoption of patient monitoring approaches in what are inherently more at-risk patient populations. References 1

Pritchard J N: Industry guidance for the selection of a delivery system for the development of novel respiratory products. Expert Opin Drug Deliv 2015; 12:11; pp1755-1765.

2

United Nations Environmental Programme: Report of the Technology and Economic Assessment Panel Volume 1. June 2015. Available at http://ozone.unep.org/en/teap-june-2015-progress-report-volume-1

3

Shen A: Market analysis of inhalation products for asthma and COPD in China. Presented at: Management Forum Inhaled Drug Delivery. London, 5-6 November 2014.

4

Winningham R E: Theravance Biopharma. Presented at: 14th annual Needham Healthcare Conference, New York, April 2015, available at https://www.sec.gov/Archives/edgar/data/1583107/000110465915027343/a15-9036_1ex99d1.htm

5

World Health Organisation: Adherence to Long-Term Therapies: Evidence for action. World Health Organisation, Geneva, Switzerland, 2003.

6

Mularksi R, McBurnie M A, Donovan J R, Walker K L, Gillespie S E, Vollmer W M: Adherence with inhaled respiratory therapeutics is associated with reduced acute exacerbations of chronic obstructive pulmonary disease. Eur Respir J 2011; 38:Suppl 55; pp4868.

7

Theravance Biopharma and Mylan Initiate Phase 3 Program for Revefenacin (TD-4208) for Treatment of Chronic Obstructive Pulmonary Disease (COPD), available at http://newsroom.mylan.com/2015-09-14-Theravance-Biopharma-andMylan-Initiate-Phase-3-Program-for-Revefenacin-TD-4208-for-Treatment-of-Chronic-Obstructive-Pulmonary-Disease-COPD

8

Mylan Announces Acquisition of Rights to Novel LAMA Respiratory Compound from Pfizer, available at http://newsroom.mylan.com/press-releases?item=123190

9

Pritchard J N: Maximizing the Effectiveness of the Inhaled Drug Development Process. In: Dalby RN, Byron PR, Peart J, Suman JD, Farr SJ, Young PM, Traini D, (eds): Respiratory Drug Delivery 2016, Vol. 3. Virginia Commonwealth University, Richmond; pp 497-502, 2016.

10

Ari A: Jet, Ultrasonic, and Mesh Nebulizers: An Evaluation of Nebulizers for Better Clinical Outcomes. Eurasian J Pulmon 2014; 16:1; pp1-7.

11

Degtyareva Y, Metcalf A, Hardaker L E A, Pritchard J N, Hatley R M H: Efficiency of the Micro mesh nebulizer tested with Three different fill volumes. Dalby RN, Byron PR, Peart J, Suman JD, Young PM, Traini D, (eds): Respiratory Drug Delivery Europe 2015, Vol 2. Virginia Commonwealth University, Richmond; pp401-404, 2015.

12

Hatley R, Hardaker L, Zarins-Tutt J, Quadrelli F, Hogan B, MacLoughlan R, Pritchard J N: Investigation of Optical Density for the Characterization of Nebulizer Meshes. In: Dalby RN, Byron PR, Peart J, Suman JD, Farr SJ, Young PM, Traini D (eds): Respiratory Drug Delivery 2016, Vol 3. Virginia Commonwealth University, Richmond; pp489-492, 2016.

13

Von Hollen D, Hatley R, Pritchard J N, Nikander K, Hardaker L, Elphick M: Factors to Consider When Selecting a Nebulizer in Drug Development. In: Dalby RN, Byron PR, Peart J, Suman JD, Farr SJ, Young PM, Traini D (eds): Respiratory Drug Delivery 2014, Vol 2: Virginia Commonwealth University, Richmond; pp 539-544, 2014.

14

Pritchard J N, Nikander K: Impact of Intervention and Feedback on Adherence to Treatment. In: Dalby RN, Byron PR, Peart J, Suman JD, Farr SJ, Young PM, (eds) Respiratory Drug Delivery 2012, Vol 1. Virginia Commonwealth University, Richmond; pp271-282, 2012.

15

Pritchard J N, Nichols C: Emerging Technologies for Electronic Monitoring of Adherence, Inhaler Competence, and True Adherence. J Aerosol Med Pulm Drug Deliv 2015; 28:2; pp69-81.

19


Drug Delivery to the Lungs 27, 2016 - Newman

The challenge of delivering drugs to the lungs Stephen P Newman Scientific Consultant, Norfolk, UK Summary The pulmonary route has an established role in the treatment of asthma and chronic obstructive pulmonary disease (COPD), and has a number of other topical and systemic applications. Successful pulmonary drug delivery requires a predictable, reproducible lung dose and clinical effect with each treatment, while minimising unwanted sideeffects, and achieving these objectives at reasonable cost. The patient presents a major barrier to achieving these goals, first because of natural lung defence mechanisms, and second because of the need to use, and to master the use of, an inhaler device. The respiratory tract has evolved in such a way as to prevent the ingress of particles and droplets, and to remove inhaled materials once deposited. Formulators strive to ensure that an adequate fraction of the delivered dose consists of particles within the fine particle range (< 5 µm) for whole lung deposition, and < 3 µm for peripheral lung deposition. Once deposited, drugs may be removed from the lungs by mucociliary clearance or by phagocytosis, or may be degraded by the action of proteases. Each patient must use an inhaler as prescribed (i.e. be adherent to the treatment regimen), and must use it correctly (i.e. prepare the device for inhalation, and then inhale from it in an appropriate way). Poor adherence and inhaler misuse are widespread problems, which can be addressed via appropriate selection of an inhaler, via use of other technology, and via education. Selection of an inhaler also needs to take into account the mass of drug to be delivered; pressurized metered dose inhalers (pMDIs) and many dry powder inhalers (DPIs) are unsuitable for delivering drug doses > 1 mg. Efficient inhalers delivering a high percentage of the drug dose to the lungs generally provide the most reproducible lung dose. The pulmonary route is a relatively complex one, but the advantages it offers justify its use for a range of treatment indications. Introduction The inhaled route is best known as a means of delivering drugs for treatment of asthma and COPD, but it can also be used to deliver a wide range of other drugs intended to achieve either a topical effect in the lungs, or a systemic effect elsewhere in the body. Topical treatment indications include use in several “orphan” diseases, such as cystic fibrosis (CF) and pulmonary arterial hypertension (PAH). The inhaled route has several advantages that derive from drug being delivered direct to its site of action (for a topically acting drug) or site of absorption (for a systemically acting drug). [1] Relatively small drug doses are often sufficient, there is generally a low incidence of systemic sideeffects, and at least for some drugs, the onset of action is rapid. Inhalation provides a way of avoiding injections for biopharmaceuticals with low oral bioavailability, and the huge alveolar surface area (> 100 m2) is an attractive target for such drugs. Several distinct types of inhaler device are used to deliver drugs to the lungs, consisting of pMDIs, DPIs and nebulizers, as well as some novel technologies that do not fit into any of the three previous categories. [2] Each of these major types of inhaler has its own advantages and limitations. Drug delivery to the lungs presents many challenges, but some of the most important of these can perhaps be summarised as trying to ensure a predictable, reproducible lung dose and clinical effect with each treatment, while minimising unwanted side-effects, and achieving these objectives at reasonable cost. Mechanical, chemical and immunological barriers Unfortunately, delivering drugs by inhalation is significantly more complicated than simply taking a tablet or capsule by mouth. The respiratory tract has evolved to prevent the entry of particles and droplets, and to remove them from the body once deposited (Figure 1). The complex anatomy of the upper airways acts as a filter for all but the smallest particles. The nose is a particularly efficient filter, and is not an ideal entry point for drugs intended to reach the lungs. The filtering effect of the oropharyngeal airways has led to definition of the “fine particle fraction” (once called the “respirable fraction”) in terms of particles having an aerodynamic diameter < about 5 µm. Any particle with an aerodynamic diameter > 5 µm is more likely to deposit in the upper airways than in the lungs. Even smaller particles (e.g. aerodynamic diameter < 3 µm) are required to target the most peripheral airways and alveoli. The inhalation mode (e.g. inhaled flow rate and volume) also has a profound effect on the amount of drug entering the lungs and its distribution within the airways. Once deposited on the airway surface, drugs can be removed from the lungs by mucociliary clearance, be acted upon by metabolic enzymes, or be engulfed by alveolar macrophages (phagocytosis). [3] Small molecules (MW < 1000 Da, e.g. analgesics) tend to have systemic bioavailability close to unity, but larger molecules (e.g. insulin) often have systemic bioavailability much less than unity, and which shows a broadly inverse correlation with molecular weight. It may be difficult to deliver an adequate amount of drug to the areas of the lung where it is most needed. The ability of aerosol particles to penetrate deep into the lung is reduced in the presence of airway narrowing in asthma and COPD. Some patients with cystic fibrosis and other conditions may have airways which are completely blocked with

20


Drug Delivery to the Lungs 27, 2016 – The challenge of delivering drugs to the lungs mucus, and beyond which it is impossible for aerosol to penetrate. Although not caused by mucus blockage, inadequate delivery to lung apices has been one factor explaining the relative failure of inhaled pentamidine in the treatment of human immunodeficiency virus (HIV) infection.

Figure 1 - Barriers to successful pulmonary drug delivery.

Inhaler misuse The patient’s behaviour presents arguably the biggest challenge to delivering inhaled drugs to the lungs. Successful pulmonary drug delivery requires a patient to use, and ideally to master the use of, an inhaler device. [4] Information about how to prepare an inhaler device for use, and then how to inhale from it, is usually presented to patients in a written package insert. Unfortunately, inhaler misuse is common, leading to a sub-optimal and potentially highly variable lung dose, or even zero lung dose. It is widely believed that pMDIs are misused to a greater degree than DPIs, but in practice this belief seems to be a myth. An increasing number of DPI devices is now available, with device-specific instructions, which could contribute to inhaler misuse by causing confusion. Inhaler misuse has clinical consequences: patients taking inhaled corticosteroids were shown to have more poorly controlled asthma if they could not use a pMDI properly, and especially if they could not coordinate actuating the pMDI with inhalation. [5] There are also financial consequences, because patients with poorly controlled asthma are likely to require more expensive treatment options. Any patient can misuse an inhaler but there are several well-recognised risk factors, including socio-economic issues, older age, severity of airway obstruction, lack of training in inhaler use, and using more than one inhaler device. Not surprisingly, young children are particularly prone to poor inhaler technique. Correct device use is now considered to be an essential factor in successful disease management. [6] Poor adherence An inhaler can only be effective if the patient takes the medication as prescribed. Poor adherence (compliance) to the treatment regimen is a separate, although related, problem to inhaler misuse. Poor adherence can involve not collecting (filling) a prescription, taking too few doses, or taking too many doses. One analysis showed that inhaled corticosteroids were underused on 24 to 69 % of days across a series of studies, compared with 2 to 23 % of days on which overuse occurred. [7] In practice, non-adherence can be either intentional or non-intentional; patients can choose not to take their medication, or can simply forget because they are too busy. The highest rates of nonadherence are often found in adolescents and young adults [8]. Poor adherence to an inhaler regimen has both clinical and financial consequences which are similar to those resulting from inhaler misuse. Several devices are available to monitor adherence objectively, and some are used to download adherence data to a computer or mobile phone. [9] Adherence monitoring systems are often incorporated into so-called “intelligent” inhaler systems. [10] The effects of poor inhaler technique and non-adherence are multiplicative, and their combination is sometimes caused “true adherence”. Addressing inhaler misuse and poor adherence The challenges of inhaler misuse and poor adherence can be tackled via both technology and education. Patients unable to use a pMDI correctly may benefit from a breath-actuated pMDI, or from the addition of a spacer device. DPIs with very simple instructions (e.g. open, inhale, close) may be used more successfully than those with more

21


Drug Delivery to the Lungs 27, 2016 - Newman complex instructions. Patients should not be switched from one DPI brand to another without adequate instruction about how to use the new inhaler. [6] Training aids are available to help patients “press and breathe” simultaneously when using a pMDI, and to adopt an inhaled flow rate appropriate to the type of inhaler. [11] Combining a bronchodilator and corticosteroid within a single inhaler may improve adherence compared with delivering the two drug classes by separate inhalers. Data logging devices that monitor adherence objectively can provide useful feedback to patients. [9] Education is a key issue in addressing problems of inhaler misuse and poor adherence. Management of chronic respiratory diseases has been described as “10 % medication, 90 % education”. [12] Education can take the form of one-on-one sessions between patient and care-giver, written treatment action plans, group training sessions, or instruction via the internet. Undertaking one-on-one sessions may be a challenge for doctors, nurses, pharmacists and other care-givers, who may have difficulty finding time in their busy schedules to meet with individual patients regularly. Unfortunately, many healthcare professionals do not understand inhaled drug delivery any better than their patients. [13] Influencing patterns of adherence has proved difficult. Overcoming the problems of non-adherence involves understanding and influencing patient behaviour. [8] This may need to take into account their concerns about the effectiveness and safety of inhaled medications, their trust in the medical profession, and factors such as family dysfunction. Conveying positive messages to patients about their disease and its treatment, including the need to take inhaled corticosteroids on a regular basis, is critical to the success of inhalation therapy. These messages need to be reinforced regularly. Choosing inhalers for different drugs Choice of inhaler is important because the patient needs to have an inhaler that he or she will use, and can use correctly, in day to day clinical practice. pMDIs and DPIs are ideal for delivering small doses of potent drugs used for asthma and COPD maintenance therapy. Ideally, the same inhaler should be used to deliver multiple drugs where this is feasible, and it should be the inhaler that a patient prefers. An inhaler must of course be affordable, but it should also be cost-effective; there is no value in a very cheap inhaler that doesn’t work. Special considerations apply to the youngest and oldest patients; in both groups delivery of drugs from nebulizers, or from pMDIs plus spacer devices, is often the best option. DPI systems should be designed to achieve a lung dose that is relatively independent of inspiratory effort. Historically, most inhalers delivered no more than 10 – 20 % of the nominal dose to the lungs; fortunately, we now have inhaler devices or device / formulation combinations capable of delivering drug to the lungs with much greater efficiency. A meta-analysis involving almost 200 data sets showed an inverse correlation between the mean value of lung deposition and its coefficient of variation. [14] Inhalers achieving a lung deposition of only c.10 % of the exvalve dose had the highest variability, and inhalers achieving a lung deposition > 30% of the ex-valve dose had the lowest variability, the latter being virtually guaranteed to have a coefficient of variation of lung deposition < 30 %, assuming the inhaler is used correctly. Inhalers delivering drug efficiently to the lungs, while also achieving a reproducible lung dose, are needed for drugs such as insulin, where pulmonary bioavailability may be limited by the action of proteases, and where the therapeutic window is narrow. The first marketed insulin product (Exubera ®, Pfizer) utilised a large “active” dry powder device, in which a novel formulation was aerosolized into a chamber above the device prior to inhalation. The device delivered insulin efficiently and reproducibly to the lungs, but its large size may have contributed to the subsequent withdrawal of the product from the market. The second marketed insulin product (Afrezza®, MannKind) utilises a simple “passive” DPI, but containing an engineered particle formulation that ensures efficient pulmonary delivery. This product is claimed to have a higher bioavailability than other inhaled insulins, and to achieve peak plasma levels more rapidly [15]. These features may result in greater acceptance by patients and physicians. Inhaled antibiotics require the delivery of very large doses that traditionally were the province of nebulizers. For instance, the dose of inhaled tobramycin (TOBI ®, Novartis) is 300 mg, delivered by specific jet nebulizers and compressors. This was the first inhaled antibiotic approved for use in patients with CF, whose lungs are colonised with Pseudomonas aeruginosa. It is also possible to deliver inhaled antibiotics by single dose DPIs where the formulation is contained within capsules or blisters. Formulation of drug powder as engineered particles can ensure that a high percentage of the nominal dose is deposited in the lungs; this in turn reduces the nominal dose required, the likely variability of the lung dose, and treatment time. A reduction in treatment time is considered likely to improve adherence. Such thinking lay behind the development of a dry powder product (TOBITM PodhalerTM, Novartis) to deliver inhaled tobramycin (Figure 2), as an alternative to nebulization. [16] The use of sophisticated nebulizer systems that control both particle size and mode of inhalation is useful for targeting some drugs reproducibly to specific sites in the lungs, e.g. the use of the AKITA® nebulizer system (Vectura) to deliver inhaled alpha-1 anti-trypsin to the alveoli of patients having a deficiency of this enzyme. [17]

22


Drug Delivery to the Lungs 27, 2016 – The challenge of delivering drugs to the lungs

Figure 2 – Mean deposition of tobramycin by DPI and by nebulizer. Delivery as PulmoSphere particles from a Podhaler DPI can achieve the same lung dose as a jet nebulizer, despite requiring a much smaller nominal dose. [16]

Concluding remarks Successful pulmonary drug delivery requires meeting a variety of challenges, which span scientific, medical, commercial and regulatory issues. Although the pulmonary route has been used for millennia, the scientific and medical issues have only been properly understood in the last few decades. Pulmonary drug delivery remains the cornerstone of asthma and COPD maintenance therapy and is becoming established and accepted for a range of other topical and systemic treatment indications, often fulfilling an unmet need. For these applications, the potential advantages offered by the pulmonary route is currently considered to justify the additional complexity. Judging by attendance at this and other conferences, interest in pulmonary drug delivery seems to be greater than ever. References 1

Rau JL: The inhalation of drugs: advantages and problems, Respir Care 2005; 50: pp367-382.

2

Dolovich MB, Dhand R: Aerosol drug delivery: developments in device design and clinical use, Lancet 2010; 377: pp10321045.

3

Clark AR: Limitations of pulmonary drug delivery. In: R Dhand, (ed): ISAM Textbook of Aerosol Medicine. On-line publication, International Society for Aerosols in Medicine, Chapter 14, 2015.

4

Newman SP: Improving inhaler technique, adherence to therapy and precision of dosing: major challenges for pulmonary drug delivery, Expert Opin Drug Deliv 2014; 11: pp365-378.

5

Giraud V, Roche N: Misuse of corticosteroid metered-dose inhaler is associated with decreased asthma stability, Eur Respir J 2002; 19: pp246-251.

6

Thomas M, Price D, Chrystyn H, Lloyd A, Williams AE, von Ziegenweidt J: Inhaled corticosteroids for asthma: impact of practice level device switching on asthma control, BMC Pulm Med 2009; 9: article 1.

7

Cochrane MG, Bala MV, Downs KE, Mauskopf J, Ben-Joseph RH: Inhaled corticosteroids for asthma therapy: patient compliance, devices and inhalation technique, Chest 2000; 117: pp542-550.

8

Morton RW, Everard ML: Adherence to aerosol medication. In: R Dhand, (ed): ISAM Textbook of Aerosol Medicine. Online publication, International Society for Aerosols in Medicine, Chapter 12, 2015.

9

Kikidis D, Konstantinos V, Tzovaras D, Usmani OS: The digital asthma patient: the history and future of inhaler based health monitoring devices, J Aerosol Med Pulm Drug Deliv 2016; 29: pp219-232.

10

Denyer J, Dyche T: The adaptive aerosol delivery (AAD) technology: past, present and future, J Aerosol Med Pulm Drug Deliv 2010; 23 (Supplement 1): ppS1-S10.

11

Sanders M, Bruin RL: Effect of a new training device on pMDI technique and aerosol performance, In: Drug Delivery to the Lungs 24. Portishead: The Aerosol Society, 2013, pp86-88.

12

Fink JB: Inhalers in asthma management: is demonstration the key to compliance? Respir Care 2005; 50: pp598-600.

13

Pritchard JN, Nicholls C: Emerging technologies for electronic monitoring of adherence, inhaler competence and true adherence, J Aerosol Med Pulm Drug Deliv 2015; 28: pp69-81.

14

Borgström L Olsson B, Thorsson L: Degree of throat deposition can explain the variability in lung deposition of inhaled drugs, J Aerosol Med 2006; 19: pp473-483.

15

Pfützner A, Forst T: Pulmonary insulin delivery by means of the Technosphere drug carrier mechanism, Expert Opin Drug Deliv 2005; 2: pp1097-1106.

16

Geller DE, Weers J, Heuerding S: Development of an inhaled dry-powder formulation of tobramycin using PulmoSphere technology, J Aerosol Med Pulm Drug Deliv 2011; 24: pp175-182.

17

Bennett WD: Controlled inhalation of aerosolised therapeutics, Expert Opin Drug Deliv 2005; 2: pp763-767.

23


Biologics in Asthma – are we turning the corner Professor Roland Buhl, Mainz University Hospital

ABSTRACT UNAVAILABLE AT TIME OF PRINTING www.ddl-conference.com

24


Drug Delivery to the Lungs 27, 2016 –Fintan Keegan et al. ®

Development to Approval of NARCAN Nasal Spray 1

2

Fintan Keegan , Michael Dey 1

2

ADAPT PHARMA Ltd., 45 Fitzwilliam Square, Dublin 2, Ireland Claidmhor Pharma Ltd, Linden Lea, Crewe Road, Sandbach, CW11 4RE, UK

Summary Opioid overdose is a major public health crisis in the United States. Treatment with naloxone has expanded, beyond use by clinicians in the hospital setting, but also beyond pre-hospital emergency medical services (EMS), into the general community. It is critical that the available naloxone dose is consistently adequate for the emergency treatment of an opioid overdose and can be readily and rapidly administered by a first responder ® without exposure to needle-stick injury. Adapt Pharma has developed NARCAN Nasal Spray supplied as a single 4 mg dose of naloxone hydrochloride in a 0.1 mL intranasal spray. Each Narcan 4 mg dose delivered by intranasal administration is approximately equivalent to a 2 mg dose of naloxone hydrochloride delivered by intramuscular injection (IM). Narcan Nasal Spray is for intranasal administration only, in the treatment of suspected or known opioid overdose and frequently must be administered by someone other than a medically trained person. This requires a simple, robust and easy to operate device. The simplicity of the nasal spray was critical in device selection. Selection of the chosen device was based on the small volume of liquid sprayed which remained in the nostril, 100 µL(microliters), regarded as the optimal for nasal administration and the higher naloxone concentration (40 mg/mL) which is optimal for nasal absorption. It is a single dose device that does not require priming before use and gives an orientation independent fine spray adhering to the nasal cavity wall, without the patient being alert. This presentation briefly describes, the rationale and stages in developing Narcan Nasal Spray to achieve these needs. Introduction Naloxone hydrochloride is an opioid antagonist developed in the 1960’s and first approved by FDA in 1971. While the mechanism of action is not fully understood it is thought that it competes for opioid receptor sites in the central nervous system (CNS), blocking the opioid effect caused by drugs such as oxycodone or heroin, reversing respiratory depression and prevention of prolonged hypoxia, organ damage and mortality. Naloxone has been used in the hospital and pre-hospital setting successfully with a wide therapeutic index of safety; initial dose of 0.4 mg to 2 mg intravenous (IV) administration, with repeated doses up to 10mg. The clinical setting, hospital, and EMS, allows for IV administration, but for many years there has been increase use of naloxone in the non clinical setting, mainly by IM injection but also by improvised intranasal administration. Opioid overdose is a critical health matter in many countries and at epidemic proportion in the US. Overdose is not confined to those people with opioid addiction and dependence, it is now the single largest cause of accidental death in the US; with over 28000 1 deaths in 2014 , and representing a 200% increase in opioid overdose related deaths since 2000. Over 60 percent of those deaths are associated with individuals taking prescription opioids. The epidemic has taken a twist in the last couple of years as illicit manufactured fentanyl (IMF) have entered the illegal opioid and counterfeit prescription opioid markets. In British Columbia, the government has declared a public health 2 emergency in response to the rise in illicit drug overdose deaths. Most deaths due to overdose are in the home, and IV and IM treatments are not always appropriate. Adapt Pharma (with our partner Opiant Pharma) and the National Institute of Drug Abuse undertook a development program for an intranasal formulation. It was approved by FDA in November 2015, for use in the non clinical community setting, and launched in February 2016. Patient Profile and Target Product Profile Narcan Nasal Spray is indicated for the emergency treatment of known or suspected opioid overdose, as manifested by respiratory and/or CNS depression. Narcan nasal spray is intended for immediate administration as emergency therapy in settings where opioids may be present. The product requires a simple, robust and easy to operate and understand administration device. One single dose of Naloxone nasal solution delivers 4 mg naloxone, which achieves exposure 5 times greater than 0.4 mg injected by IM, or 47% relative bioavailability. As the duration (naloxone plasma half-life) is longer for certain opioids compared to naloxone, and respiratory depression can reoccur, a window of 60-90 minutes is provided to obtain medical treatment. The therapeutic window of naloxone is high, IM injection repeated doses up to 10 mg can be given, and an immediate second intranasal dose of 4 mg was safe and efficacious in patients studied. Based on the use of Narcan Nasal spray for this indication, patient profile and the required product profile a high level risk assessment using a Potential Medical Harm List was carried out. This list using a severity ratings for the product risk management assessment as follows: Disastrous (D) Results in patient death, Critical (C) Results in permanent impairment or life-threatening injury, Important (I) Results in injury or impairment requiring professional medical intervention, Slight (S) Results in temporary injury or impairment not requiring professional medical intervention, Negligible (N) Inconvenience or temporary discomfort. The occurrence of the event are rated in the range, very high to very low. The severity rating along with occurrence is tabulated into an overall rating as: Acceptable, As Low as Reasonably Possible, or Unacceptable. Examples are given in Table 1

25


®

Drug Delivery to the Lungs 27, 2016 – Development to Approval of NARCAN Nasal Spray Table 1. Potential Medical Harm List and Overall Assessment of Naloxone Nasal Spray Severity Rating

Harm

Single “No dose” Critical delivered by device

Multiple “No dose” Critical delivered by device

Occurrence

Overall Rating

Comments

Low

Unacceptable requires mitigation

Potential life-critical administration; even at 1 in 50,000 defect level assumed to high; requires to be very low

Negligible

Acceptable

Second device failing at 1 in 50,000100,000 rate would be 1 in 2,500,000,000 – 10,000,000,000

A similar approach of risk assessment given in Table 2 was applied to development based on the ideal product formulation and device attributes Table 2 Target Product Profile and High Level Risks and Mitigation for Important Product Attributes for Naloxone Nasal Administration Attribute

Rationale

Risk if defective

Mitigate Risk by

easily

Must be available for immediate use

Delay in treatment may be harmful or fatal

Small discreet, package which is easily portable

Product is easy to open and make ready for use

Must be available for immediate use

Delay in treatment may be harmful or fatal

Easily opened secondary packaging (still providing adequate protection). Device ready for use.

Product must be easily non-tamper evident

Must be available for immediate use and be correct quality product

Administration of Incorrect/forged product may be harmful or fatal

Integral seals easily recognised. Quality product easily recognised.

Product portable

is

Rationale for Product Design Intranasal administration is an attractive option for local and systemic delivery of therapeutic agents. The nasal mucosa is easily accessible, administration is non-invasive, does not require needles or other penetrating devices, is essentially painless and can be performed easily by a witness/first responder, or by physicians/professionals in emergency settings, including to unconscious, or semi-conscious patients. The drug delivery offers a rapid onset of therapeutic effects, circumventing gastrointestinal degradation and hepatic firstpass metabolism of the drug. However, drug concentration and delivered volume, physical properties of drug like 3 molecular weight and lipophilicity need to be within specific ranges for the drug to be absorbed . In addition 4 Lipinski et al developed a rule of 5s for nasal delivery to be effective based on: •

Drug characteristics: molecular weight <500 Da and Log P < 5

Potency required to be high: < 5 mg per dose

Solubility need to be high: > 50 mg/mL to achieve dose in maximal nasal spray volume of 100 μL/spray

pH of solution: approximately 5.5,

Osmolality: <500 mosm/kg.

Both recognised the physical size of the nasal cavity required a small volume of maximally 100 μL with 0.2 mL divided between two nostrils to achieve effective delivery, and avoid run off beyond the nasal cavity.

Development started with laboratory bench top batches, the main objective to assess formulations with a stable high concentration of naloxone. With a stable formulation candidate a number of container closure and devices were studied. Again based on stability study data, a pilot clinical study was initiated. In parallel, spray characterisation studies were carried out. Based on optimisation of the formulation, discussion with regulatory authorities, and significantly scaled up and characterised manufacturing process, a subsequent pivotal pharmacokinetics study was initiated. The clinical report and stability from registration batches form a key part of pre-NDA meeting with the FDA. The reference product, agreed with the agency, in the PK studies, was chosen based on the lowest approved effective naloxone dose. The approved route of administration was IM injection at an approved dose of 0.4 mg naloxone.

26


Drug Delivery to the Lungs 27, 2016 –Fintan Keegan et al. PK evaluation was completed for exposure, (AUC) maximum concentration, Cmax, time to maximum concentration tmax, and initial naloxone plasma concentration. Table 3 gives examples of product attributes measured in the final product configuration. Table 3 Target Product Profile and High Level Risks and Mitigation for Important Product Attributes for Naloxone Nasal Administration – Results for Selected Formulation Attribute

Rationale

Risk if defective

Formulation Achieved

Product formulation must be delivered in volume which remains in nostril

Avoids excessive solution from being swallowed which has much lower oral bioavailability. Avoids excessive solution being inhaled and coughing/choking in nonalert or unconscious patient.

Concentration and volume of spray to be optimised including use in supine position.

Solution of 4 mg in 100 µL; retained in nostril; full dose administered easily in nasal spray.

Product must be robust and reliable when carried around in normal use

Police and first responders to emergency calls or use by members of public would require device to be carried by patient or person administering

Stable at room temperature, but also cover summer/winter use,

Stability data shows 10 to 40 mg/mL o solution stable at 25 C for 24 o months and 6 months at 40 C, o confirming excursions to 40 C in routine use would not affect the product.

Not effected by defects from walking, driving, normal work

Relative bioavailability at 47% Vs IM injection Clinical studies included supine patients with no issues of excess solution being or reported or observed.

The spray is formed by the liquid exiting the device spray orifice and has been characterised by a number of techniques such as Spray Pattern, Plume geometry and Dose Delivered. These attributes are critical to ensure the correct dose volume of solution (100 μL) is delivered from each sprayer, and that the actuator also forms the required shape and pattern for liquid spray, including limits on % < 10 μm, which is the largest droplet size which can be inhaled into the upper regions of the lung. Human Factor Usability and Label Comprehension Testing A number of studies were completed. Firstly qualitative studies were carried out to refine and improve the product label to a point where quantitative studies were initiated. An Accelerated and Compressed Evaluation (ACE), which consisted of 3 consecutive and iterative Human Factors/ Label Comprehension Pre-Tests, was conducted to assess the ability of subjects to understand the labelling, Patient Insert (PI), including a quick start guide (QSG), and to demonstrate simulated use of a naloxone nasal prototype device. The purpose of this testing schedule was to learn and adjust the labelling and materials in an iterative and accelerated manner. The objectives of the study were: (1) To evaluate the subject’s ability to correctly demonstrate the steps for evaluating a patient for the medication, administering the medication, monitoring the patient and, if appropriate, giving a second dose, as instructed in the QSG (Human Factors); (2) To evaluate the subject’s ability to comprehend key messages in the PI (Comprehension); Two Human Factor and Label Comprehension Validation Studies were conducted with the final draft Nasal Spray packaged configuration, in subjects over 12 years of age. Two subgroups were recruited within the general population sample, including Low Literacy and Adolescent (12-17 years of age) subgroups. Primary, Secondary and Exploratory Objectives were collected. In all studies subjects were not trained in the use of the device or explanation of label instructions presented. Both studies were similar but in Study 1 –one device was used, whereas in Study 2-two devices were presented in the packaging configuration. Study 2 was also differentiated by two sub sections; one cohort was given the Patient User section of product labelling and allowed to read in real time. The second cohort were not give any instructions to read, which simulates a real emergency situation. In all cases subjects were presented with a scenario of an unconscious overdose victim simulated by a life-sized mannequin similar to those used for cardiopulmonary resuscitation training. Subjects were given the product with labeling and asked to proceed as they would in a real-life emergency; no training or coaching was provided either prior to or during the simulation. Background noise, in the form of TV and radio, was introduced into the scenario to simulate voices and noise from onlookers. The primary objectives were selected because of the critical impact on delivering care to the victim of an overdose. These included the ability to demonstrate administration of the medication and to comprehend key messages in the labeling. In the Human Factors, subjects correctly completed the critical tasks in the human factors simulated use by 1) Inserting nozzle into nostril, 2) Pressing plunger to release dose into nose. The secondary objectives were important, but of lesser critical impact to the urgent administration of the medication. Subjects correctly completed the secondary tasks in the human factors simulated use: Check for response, Call 911, Move to Recovery Position after administering the dose.

27


®

Drug Delivery to the Lungs 27, 2016 – Development to Approval of NARCAN Nasal Spray

In the Comprehension section, subjects responded to the following comprehension questions from the Patient Information such as; Product Indication (product use), Product Indication (medical treatment), How Narcan Nasal Spray should be used, Necessary to get emergency medical help after using Narcan Nasal Spray, Signs of opioid overdose, Potential withdrawal symptoms after use of Narcan Nasal Spray. The primary analysis for human factors was the correct demonstration of 2 critical tasks (insert the nozzle in the nose and depress the plunger to release a dose). Subjects demonstrated the ability to correctly perform both critical tasks, meeting the success threshold in studies using either 1 or 2 devices. Using 1 device, 90.6% of subjects (n = 48 of 53) were able to correctly perform both critical tasks. Similar results were obtained in the study 2, using 2 devices: in arm 1, which included a review of the QSG, 90.6% (n =29 of 32) of subjects correctly performed both critical tasks, while in arm 2, 90.3% (n = 28 of 31) of subjects correctly performed both critical tasks without reviewing the QSG. Device History File and Reliability The device history file (DHF) is initiated at the early stages of development and maintained throughout the life of the product. Based on the FDA medical devices guidelines it uses a traditional waterfall model design process 5 Those products comprised of two or more and became applicable to Combination Products in January 2015 regulated components, as naloxone drug product and a uni-dose nasal sprayer which is physically combined and produced as a single entity, Narcan is classified as a Combination Product. As sponsor and manufacturer, one must implement a streamlined approach, demonstrating compliance with the drug CGMPs (21 CFR part 211) and selected parts of the quality system (QS) regulation (21 CFR part 820) rather than demonstrating full compliance with both. One significant QS which can catch drug product manufacturer unawares, is Design Controls (21 CFR 820.30). This system outlines the design outputs that have been established for all design inputs, design verification and validation activities must be performed to ensure that the combination product design output meets design input requirements, including user needs and intended uses. These activities must be documented in the design history file and must be subjected to design reviews. The design history file for a combination product should address all design issues relating to the combined use of the constituent parts. The file is maintained by quality department and the basis for maintenance of the file is the change control systems. As owner of the file you are responsible for all aspects. As Narcan is used in a lifesaving, emergency setting then, from both design and product control strategy perspectives, the importance and data required to prove reliability of the product, including its use in conditions which were much more demanding than normal room temperature required additional consideration of all stages of product manufacture from components, device sub-assemblies, filled and final packaged product. At each stage structured reviews including Fault Tree Analysis was used to identify and implement additional controls which increased the levels of reliability to assure patient received a full spray on device operation. Since Narcan US launch almost 300,000 doses of Nasal Spray have been dispensed. The product has been well received with no product complaints or devices failures confirming the ease and reliability in actual use. References 1

CDC Morbidity and Mortality Weekly Report (MMWR) “Increases in Drug and Opioid Overdose Deaths — United States, 2000–2014”http://www.cdc.gov/mmwr/preview/mmwrhtml/mm6450a3.htm?s_cid=mm6450a3_w

2

Province of British Columbia [CA]. BC Gov News; https://news.gov.bc.ca/releases/2016HLTH0026-000568

3

Bitter C, Zimmermann KS, Surber C (2011). Nasal Drug Delivery in Humans; Surber C, Elsner P, Farage MA (eds) (2011). Topical Applications and the Mucosa. Curr Probl Dermatol. Basel, Karger, 2011, vol 40, pp 20–35

4

Lipinski CA, Lombardo F, Dominy BW, Feeney PJ: Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Deliv Rev 1997; 23:3– 25. Lipinski et al (1997)

5

Guidance for Industry and FDA Staff: Current Good Manufacturing Practice Requirements for Combination Products. (January 2015) U.S. Department of Health and Human Services, Food and Drug Administration Office of Combination Products (OCP) in the Office of the Commissioner

28


Drug Delivery to the Lungs 27, 2016 – Rémi Rosière et al. New lung-tumour-penetrating nanocarrier designed for aerosolized chemotherapy 1

1,2

3

4

Rémi Rosière , Matthias Van Woensel , Véronique Mathieu , Thomas Mathivet , Karim Amighi 1 & Nathalie Wauthoz

1

1

Laboratory of Pharmaceutics and Biopharmaceutics, Université Libre de Bruxelles (ULB), 1050 Brussels, 2 Belgium; Research Group Experimental Neurosurgery and Neuroanatomy, Laboratory of Pediatric Immunology, 3 KULeuven, Leuven, Belgium; Laboratoire de Cancérologie et Toxicologie expérimentale, ULB, Brussels, 4 Belgium, VIB Vesalius Research Center, KULeuven, Leuven, Belgium Summary In lung cancer therapy, inhaled nanomedicine could improve systemic nanomedicine-based chemotherapy by delivering high amounts of nanocarriers directly to the lung tumour site. However, penetration of the nanocarriers into a solid tumour remains difficult, whatever the selected route of administration. The aim of this study was therefore to develop a tumour-selective nanocarrier for inhalation that is able to penetrate lung tumour cells and tissues. Solid lipid nanoparticles (SLN) coated with a folate-grafted chitosan copolymer were prepared by a nanoprecipitation method with a paclitaxel entrapment efficiency of ~100%. Efficient coating was confirmed by particle size and zeta potential (Z-average mean diameter of 160 nm to 230 nm and zeta potential of -20 mV to +30 mV before and after coating, respectively). The coated SLN were characterized by significantly higher anti® proliferative properties than Taxol , with half-maximal inhibitory concentrations of 60 and 340 µM, respectively, as observed by means of the colorimetric MTT assay. Interestingly, the nanocarrier was able to enter HeLa and M109-HiFR, two folate receptor (FR)-expressing cell lines, in vitro, and in vivo after administration by inhalation to orthotopic M109-HiFR lung-tumour-grafted mice. The SLN remained coated with the folate-grafted copolymer in lung tissues and tumours after inhalation. This study therefore demonstrated the potential for a new FR-targeted nanocarrier to reach, penetrate and distribute throughout lung tumour tissues after inhalation. Introduction Conventional chemotherapy plays an important role in lung cancer therapy, as it is used in nearly every stage of the disease. These chemotherapies are currently administered through systemic routes, which distribute the chemotherapeutic drugs to all parts of the body before reaching the tumour (systemic chemotherapy) and [1] inevitably cause severe systemic and dose-limiting toxicities to the patient . Among other therapies (e.g. targeted chemotherapy, immunotherapy), nanomedicine-based chemotherapy has been proposed to increase selectivity for lung tumours (i.e. by means of a passive targeting related to the enhanced permeability and [2] retention effect), and cells (i.e. active targeting) . Although presenting many advantages, when administered through systemic routes, nanomedicine is characterized by several limitations. These limitations include (i) low delivery efficiency (i.e. the proportion of the injected dose of nanocarriers that effectively reach the tumour site), [2] and (ii) potential poor penetration of the nanocarrier into the tumour . Inhaled nanomedicine could overcome these limitations as it presents many advantages in lung cancer therapy over both conventional and [1] nanomedicine-based systemic chemotherapy . The advantages include the delivery of high amounts of nanocarriers directly to the lung tumour site, the enhancement of drug internalization by cells, the specific [3] recognition of cancer cells, etc. However, penetration of the nanocarrier into a solid tumour remains difficult , whatever the selected route of administration. In this study, we aimed to design a nanocarrier consisting of solid lipid nanoparticles coated with a folate-grafted chitosan derivative, the folate-poly(ethylene glycol)-(N-[(2-hydroxy-3-trimethyl-ammonium) propyl] chitosan) (F[4] PEG-HTCC) copolymer and loaded with the chemotherapeutic drug paclitaxel (PTX). We intended to prove that this nanocarrier, in addition to targeting cancer cells through the folate receptors (FR), which are overexpressed in [5] many lung cancers , is able to penetrate lung tumour cells and tissues, both in vitro and in vivo. Experimental methods Materials 3-[4,5-dimethylthiazol-2-yl] diphenyltetrazolium bromide (MTT), D-α-Tocopherol poly(ethylene glycol) 1000 succinate (TPGS) and sodium taurocholate (NaTau) were purchased from Sigma-Aldrich (Diegem, Belgium). 25[N-[(7-nitro-2-1,3-benzoxadiazol-4-yl)methyl]amino]-27-norcholesterol (25-NBD cholesterol) was purchased from Avanti Polar Lipids (Alabaster, USA). Cholesterol was purchased from Fagron (Waregem, Belgium). Glyceryl® ® stearate (Geleol ) was purchased from Gattefosee (Nanterre, France). Cremophor EL was purchased from BASF (Ludwigshafen, Germany). Paclitaxel (PTX) was purchased from Carbosynth (Berkshire, UK). All the mediums and supplements for cell culture were purchased from Life Technologies (Merelbecke, Belgium). All solvents were purchased in analytical grade from Merck (Darmstadt, Germany).

29


Drug Delivery to the Lungs 27, 2016 – Rémi Rosière et al. New lung-tumour-penetrating nanocarrier designed for aerosolized chemotherapy 1

1,2

3

4

Rémi Rosière , Matthias Van Woensel , Véronique Mathieu , Thomas Mathivet , Karim Amighi 1 & Nathalie Wauthoz

1

1

Laboratory of Pharmaceutics and Biopharmaceutics, Université Libre de Bruxelles (ULB), 1050 Brussels, 2 Belgium; Research Group Experimental Neurosurgery and Neuroanatomy, Laboratory of Pediatric Immunology, 3 KULeuven, Leuven, Belgium; Laboratoire de Cancérologie et Toxicologie expérimentale, ULB, Brussels, 4 Belgium, VIB Vesalius Research Center, KULeuven, Leuven, Belgium Summary In lung cancer therapy, inhaled nanomedicine could improve systemic nanomedicine-based chemotherapy by delivering high amounts of nanocarriers directly to the lung tumour site. However, penetration of the nanocarriers into a solid tumour remains difficult, whatever the selected route of administration. The aim of this study was therefore to develop a tumour-selective nanocarrier for inhalation that is able to penetrate lung tumour cells and tissues. Solid lipid nanoparticles (SLN) coated with a folate-grafted chitosan copolymer were prepared by a nanoprecipitation method with a paclitaxel entrapment efficiency of ~100%. Efficient coating was confirmed by particle size and zeta potential (Z-average mean diameter of 160 nm to 230 nm and zeta potential of -20 mV to +30 mV before and after coating, respectively). The coated SLN were characterized by significantly higher anti® proliferative properties than Taxol , with half-maximal inhibitory concentrations of 60 and 340 µM, respectively, as observed by means of the colorimetric MTT assay. Interestingly, the nanocarrier was able to enter HeLa and M109-HiFR, two folate receptor (FR)-expressing cell lines, in vitro, and in vivo after administration by inhalation to orthotopic M109-HiFR lung-tumour-grafted mice. The SLN remained coated with the folate-grafted copolymer in lung tissues and tumours after inhalation. This study therefore demonstrated the potential for a new FR-targeted nanocarrier to reach, penetrate and distribute throughout lung tumour tissues after inhalation. Introduction Conventional chemotherapy plays an important role in lung cancer therapy, as it is used in nearly every stage of the disease. These chemotherapies are currently administered through systemic routes, which distribute the chemotherapeutic drugs to all parts of the body before reaching the tumour (systemic chemotherapy) and [1] inevitably cause severe systemic and dose-limiting toxicities to the patient . Among other therapies (e.g. targeted chemotherapy, immunotherapy), nanomedicine-based chemotherapy has been proposed to increase selectivity for lung tumours (i.e. by means of a passive targeting related to the enhanced permeability and [2] retention effect), and cells (i.e. active targeting) . Although presenting many advantages, when administered through systemic routes, nanomedicine is characterized by several limitations. These limitations include (i) low delivery efficiency (i.e. the proportion of the injected dose of nanocarriers that effectively reach the tumour site), [2] and (ii) potential poor penetration of the nanocarrier into the tumour . Inhaled nanomedicine could overcome these limitations as it presents many advantages in lung cancer therapy over both conventional and [1] nanomedicine-based systemic chemotherapy . The advantages include the delivery of high amounts of nanocarriers directly to the lung tumour site, the enhancement of drug internalization by cells, the specific [3] recognition of cancer cells, etc. However, penetration of the nanocarrier into a solid tumour remains difficult , whatever the selected route of administration. In this study, we aimed to design a nanocarrier consisting of solid lipid nanoparticles coated with a folate-grafted chitosan derivative, the folate-poly(ethylene glycol)-(N-[(2-hydroxy-3-trimethyl-ammonium) propyl] chitosan) (F[4] PEG-HTCC) copolymer and loaded with the chemotherapeutic drug paclitaxel (PTX). We intended to prove that this nanocarrier, in addition to targeting cancer cells through the folate receptors (FR), which are overexpressed in [5] many lung cancers , is able to penetrate lung tumour cells and tissues, both in vitro and in vivo. Experimental methods Materials 3-[4,5-dimethylthiazol-2-yl] diphenyltetrazolium bromide (MTT), D-α-Tocopherol poly(ethylene glycol) 1000 succinate (TPGS) and sodium taurocholate (NaTau) were purchased from Sigma-Aldrich (Diegem, Belgium). 25[N-[(7-nitro-2-1,3-benzoxadiazol-4-yl)methyl]amino]-27-norcholesterol (25-NBD cholesterol) was purchased from Avanti Polar Lipids (Alabaster, USA). Cholesterol was purchased from Fagron (Waregem, Belgium). Glyceryl® ® stearate (Geleol ) was purchased from Gattefosee (Nanterre, France). Cremophor EL was purchased from BASF (Ludwigshafen, Germany). Paclitaxel (PTX) was purchased from Carbosynth (Berkshire, UK). All the mediums and supplements for cell culture were purchased from Life Technologies (Merelbecke, Belgium). All solvents were purchased in analytical grade from Merck (Darmstadt, Germany).

30


Drug Delivery to the Lungs 27, 2016 – Rémi Rosière et al. New lung-tumour-penetrating nanocarrier designed for aerosolized chemotherapy 1

1,2

3

4

Rémi Rosière , Matthias Van Woensel , Véronique Mathieu , Thomas Mathivet , Karim Amighi 1 & Nathalie Wauthoz

1

1

Laboratory of Pharmaceutics and Biopharmaceutics, Université Libre de Bruxelles (ULB), 1050 Brussels, 2 Belgium; Research Group Experimental Neurosurgery and Neuroanatomy, Laboratory of Pediatric Immunology, 3 KULeuven, Leuven, Belgium; Laboratoire de Cancérologie et Toxicologie expérimentale, ULB, Brussels, 4 Belgium, VIB Vesalius Research Center, KULeuven, Leuven, Belgium Summary In lung cancer therapy, inhaled nanomedicine could improve systemic nanomedicine-based chemotherapy by delivering high amounts of nanocarriers directly to the lung tumour site. However, penetration of the nanocarriers into a solid tumour remains difficult, whatever the selected route of administration. The aim of this study was therefore to develop a tumour-selective nanocarrier for inhalation that is able to penetrate lung tumour cells and tissues. Solid lipid nanoparticles (SLN) coated with a folate-grafted chitosan copolymer were prepared by a nanoprecipitation method with a paclitaxel entrapment efficiency of ~100%. Efficient coating was confirmed by particle size and zeta potential (Z-average mean diameter of 160 nm to 230 nm and zeta potential of -20 mV to +30 mV before and after coating, respectively). The coated SLN were characterized by significantly higher anti® proliferative properties than Taxol , with half-maximal inhibitory concentrations of 60 and 340 µM, respectively, as observed by means of the colorimetric MTT assay. Interestingly, the nanocarrier was able to enter HeLa and M109-HiFR, two folate receptor (FR)-expressing cell lines, in vitro, and in vivo after administration by inhalation to orthotopic M109-HiFR lung-tumour-grafted mice. The SLN remained coated with the folate-grafted copolymer in lung tissues and tumours after inhalation. This study therefore demonstrated the potential for a new FR-targeted nanocarrier to reach, penetrate and distribute throughout lung tumour tissues after inhalation. Introduction Conventional chemotherapy plays an important role in lung cancer therapy, as it is used in nearly every stage of the disease. These chemotherapies are currently administered through systemic routes, which distribute the chemotherapeutic drugs to all parts of the body before reaching the tumour (systemic chemotherapy) and [1] inevitably cause severe systemic and dose-limiting toxicities to the patient . Among other therapies (e.g. targeted chemotherapy, immunotherapy), nanomedicine-based chemotherapy has been proposed to increase selectivity for lung tumours (i.e. by means of a passive targeting related to the enhanced permeability and [2] retention effect), and cells (i.e. active targeting) . Although presenting many advantages, when administered through systemic routes, nanomedicine is characterized by several limitations. These limitations include (i) low delivery efficiency (i.e. the proportion of the injected dose of nanocarriers that effectively reach the tumour site), [2] and (ii) potential poor penetration of the nanocarrier into the tumour . Inhaled nanomedicine could overcome these limitations as it presents many advantages in lung cancer therapy over both conventional and [1] nanomedicine-based systemic chemotherapy . The advantages include the delivery of high amounts of nanocarriers directly to the lung tumour site, the enhancement of drug internalization by cells, the specific [3] recognition of cancer cells, etc. However, penetration of the nanocarrier into a solid tumour remains difficult , whatever the selected route of administration. In this study, we aimed to design a nanocarrier consisting of solid lipid nanoparticles coated with a folate-grafted chitosan derivative, the folate-poly(ethylene glycol)-(N-[(2-hydroxy-3-trimethyl-ammonium) propyl] chitosan) (F[4] PEG-HTCC) copolymer and loaded with the chemotherapeutic drug paclitaxel (PTX). We intended to prove that this nanocarrier, in addition to targeting cancer cells through the folate receptors (FR), which are overexpressed in [5] many lung cancers , is able to penetrate lung tumour cells and tissues, both in vitro and in vivo. Experimental methods Materials 3-[4,5-dimethylthiazol-2-yl] diphenyltetrazolium bromide (MTT), D-α-Tocopherol poly(ethylene glycol) 1000 succinate (TPGS) and sodium taurocholate (NaTau) were purchased from Sigma-Aldrich (Diegem, Belgium). 25[N-[(7-nitro-2-1,3-benzoxadiazol-4-yl)methyl]amino]-27-norcholesterol (25-NBD cholesterol) was purchased from Avanti Polar Lipids (Alabaster, USA). Cholesterol was purchased from Fagron (Waregem, Belgium). Glyceryl® ® stearate (Geleol ) was purchased from Gattefosee (Nanterre, France). Cremophor EL was purchased from BASF (Ludwigshafen, Germany). Paclitaxel (PTX) was purchased from Carbosynth (Berkshire, UK). All the mediums and supplements for cell culture were purchased from Life Technologies (Merelbecke, Belgium). All solvents were purchased in analytical grade from Merck (Darmstadt, Germany).

31


Drug Delivery to the Lungs 27, 2016 – Rémi Rosière et al. New lung-tumour-penetrating nanocarrier designed for aerosolized chemotherapy 1

1,2

3

4

Rémi Rosière , Matthias Van Woensel , Véronique Mathieu , Thomas Mathivet , Karim Amighi 1 & Nathalie Wauthoz

1

1

Laboratory of Pharmaceutics and Biopharmaceutics, Université Libre de Bruxelles (ULB), 1050 Brussels, 2 Belgium; Research Group Experimental Neurosurgery and Neuroanatomy, Laboratory of Pediatric Immunology, 3 KULeuven, Leuven, Belgium; Laboratoire de Cancérologie et Toxicologie expérimentale, ULB, Brussels, 4 Belgium, VIB Vesalius Research Center, KULeuven, Leuven, Belgium Summary In lung cancer therapy, inhaled nanomedicine could improve systemic nanomedicine-based chemotherapy by delivering high amounts of nanocarriers directly to the lung tumour site. However, penetration of the nanocarriers into a solid tumour remains difficult, whatever the selected route of administration. The aim of this study was therefore to develop a tumour-selective nanocarrier for inhalation that is able to penetrate lung tumour cells and tissues. Solid lipid nanoparticles (SLN) coated with a folate-grafted chitosan copolymer were prepared by a nanoprecipitation method with a paclitaxel entrapment efficiency of ~100%. Efficient coating was confirmed by particle size and zeta potential (Z-average mean diameter of 160 nm to 230 nm and zeta potential of -20 mV to +30 mV before and after coating, respectively). The coated SLN were characterized by significantly higher anti® proliferative properties than Taxol , with half-maximal inhibitory concentrations of 60 and 340 µM, respectively, as observed by means of the colorimetric MTT assay. Interestingly, the nanocarrier was able to enter HeLa and M109-HiFR, two folate receptor (FR)-expressing cell lines, in vitro, and in vivo after administration by inhalation to orthotopic M109-HiFR lung-tumour-grafted mice. The SLN remained coated with the folate-grafted copolymer in lung tissues and tumours after inhalation. This study therefore demonstrated the potential for a new FR-targeted nanocarrier to reach, penetrate and distribute throughout lung tumour tissues after inhalation. Introduction Conventional chemotherapy plays an important role in lung cancer therapy, as it is used in nearly every stage of the disease. These chemotherapies are currently administered through systemic routes, which distribute the chemotherapeutic drugs to all parts of the body before reaching the tumour (systemic chemotherapy) and [1] inevitably cause severe systemic and dose-limiting toxicities to the patient . Among other therapies (e.g. targeted chemotherapy, immunotherapy), nanomedicine-based chemotherapy has been proposed to increase selectivity for lung tumours (i.e. by means of a passive targeting related to the enhanced permeability and [2] retention effect), and cells (i.e. active targeting) . Although presenting many advantages, when administered through systemic routes, nanomedicine is characterized by several limitations. These limitations include (i) low delivery efficiency (i.e. the proportion of the injected dose of nanocarriers that effectively reach the tumour site), [2] and (ii) potential poor penetration of the nanocarrier into the tumour . Inhaled nanomedicine could overcome these limitations as it presents many advantages in lung cancer therapy over both conventional and [1] nanomedicine-based systemic chemotherapy . The advantages include the delivery of high amounts of nanocarriers directly to the lung tumour site, the enhancement of drug internalization by cells, the specific [3] recognition of cancer cells, etc. However, penetration of the nanocarrier into a solid tumour remains difficult , whatever the selected route of administration. In this study, we aimed to design a nanocarrier consisting of solid lipid nanoparticles coated with a folate-grafted chitosan derivative, the folate-poly(ethylene glycol)-(N-[(2-hydroxy-3-trimethyl-ammonium) propyl] chitosan) (F[4] PEG-HTCC) copolymer and loaded with the chemotherapeutic drug paclitaxel (PTX). We intended to prove that this nanocarrier, in addition to targeting cancer cells through the folate receptors (FR), which are overexpressed in [5] many lung cancers , is able to penetrate lung tumour cells and tissues, both in vitro and in vivo. Experimental methods Materials 3-[4,5-dimethylthiazol-2-yl] diphenyltetrazolium bromide (MTT), D-α-Tocopherol poly(ethylene glycol) 1000 succinate (TPGS) and sodium taurocholate (NaTau) were purchased from Sigma-Aldrich (Diegem, Belgium). 25[N-[(7-nitro-2-1,3-benzoxadiazol-4-yl)methyl]amino]-27-norcholesterol (25-NBD cholesterol) was purchased from Avanti Polar Lipids (Alabaster, USA). Cholesterol was purchased from Fagron (Waregem, Belgium). Glyceryl® ® stearate (Geleol ) was purchased from Gattefosee (Nanterre, France). Cremophor EL was purchased from BASF (Ludwigshafen, Germany). Paclitaxel (PTX) was purchased from Carbosynth (Berkshire, UK). All the mediums and supplements for cell culture were purchased from Life Technologies (Merelbecke, Belgium). All solvents were purchased in analytical grade from Merck (Darmstadt, Germany).

32


Drug Delivery to the Lungs 27, 2016 – Rémi Rosière et al. New lung-tumour-penetrating nanocarrier designed for aerosolized chemotherapy 1

1,2

3

4

Rémi Rosière , Matthias Van Woensel , Véronique Mathieu , Thomas Mathivet , Karim Amighi 1 & Nathalie Wauthoz

1

1

Laboratory of Pharmaceutics and Biopharmaceutics, Université Libre de Bruxelles (ULB), 1050 Brussels, 2 Belgium; Research Group Experimental Neurosurgery and Neuroanatomy, Laboratory of Pediatric Immunology, 3 KULeuven, Leuven, Belgium; Laboratoire de Cancérologie et Toxicologie expérimentale, ULB, Brussels, 4 Belgium, VIB Vesalius Research Center, KULeuven, Leuven, Belgium Summary In lung cancer therapy, inhaled nanomedicine could improve systemic nanomedicine-based chemotherapy by delivering high amounts of nanocarriers directly to the lung tumour site. However, penetration of the nanocarriers into a solid tumour remains difficult, whatever the selected route of administration. The aim of this study was therefore to develop a tumour-selective nanocarrier for inhalation that is able to penetrate lung tumour cells and tissues. Solid lipid nanoparticles (SLN) coated with a folate-grafted chitosan copolymer were prepared by a nanoprecipitation method with a paclitaxel entrapment efficiency of ~100%. Efficient coating was confirmed by particle size and zeta potential (Z-average mean diameter of 160 nm to 230 nm and zeta potential of -20 mV to +30 mV before and after coating, respectively). The coated SLN were characterized by significantly higher anti® proliferative properties than Taxol , with half-maximal inhibitory concentrations of 60 and 340 µM, respectively, as observed by means of the colorimetric MTT assay. Interestingly, the nanocarrier was able to enter HeLa and M109-HiFR, two folate receptor (FR)-expressing cell lines, in vitro, and in vivo after administration by inhalation to orthotopic M109-HiFR lung-tumour-grafted mice. The SLN remained coated with the folate-grafted copolymer in lung tissues and tumours after inhalation. This study therefore demonstrated the potential for a new FR-targeted nanocarrier to reach, penetrate and distribute throughout lung tumour tissues after inhalation. Introduction Conventional chemotherapy plays an important role in lung cancer therapy, as it is used in nearly every stage of the disease. These chemotherapies are currently administered through systemic routes, which distribute the chemotherapeutic drugs to all parts of the body before reaching the tumour (systemic chemotherapy) and [1] inevitably cause severe systemic and dose-limiting toxicities to the patient . Among other therapies (e.g. targeted chemotherapy, immunotherapy), nanomedicine-based chemotherapy has been proposed to increase selectivity for lung tumours (i.e. by means of a passive targeting related to the enhanced permeability and [2] retention effect), and cells (i.e. active targeting) . Although presenting many advantages, when administered through systemic routes, nanomedicine is characterized by several limitations. These limitations include (i) low delivery efficiency (i.e. the proportion of the injected dose of nanocarriers that effectively reach the tumour site), [2] and (ii) potential poor penetration of the nanocarrier into the tumour . Inhaled nanomedicine could overcome these limitations as it presents many advantages in lung cancer therapy over both conventional and [1] nanomedicine-based systemic chemotherapy . The advantages include the delivery of high amounts of nanocarriers directly to the lung tumour site, the enhancement of drug internalization by cells, the specific [3] recognition of cancer cells, etc. However, penetration of the nanocarrier into a solid tumour remains difficult , whatever the selected route of administration. In this study, we aimed to design a nanocarrier consisting of solid lipid nanoparticles coated with a folate-grafted chitosan derivative, the folate-poly(ethylene glycol)-(N-[(2-hydroxy-3-trimethyl-ammonium) propyl] chitosan) (F[4] PEG-HTCC) copolymer and loaded with the chemotherapeutic drug paclitaxel (PTX). We intended to prove that this nanocarrier, in addition to targeting cancer cells through the folate receptors (FR), which are overexpressed in [5] many lung cancers , is able to penetrate lung tumour cells and tissues, both in vitro and in vivo. Experimental methods Materials 3-[4,5-dimethylthiazol-2-yl] diphenyltetrazolium bromide (MTT), D-α-Tocopherol poly(ethylene glycol) 1000 succinate (TPGS) and sodium taurocholate (NaTau) were purchased from Sigma-Aldrich (Diegem, Belgium). 25[N-[(7-nitro-2-1,3-benzoxadiazol-4-yl)methyl]amino]-27-norcholesterol (25-NBD cholesterol) was purchased from Avanti Polar Lipids (Alabaster, USA). Cholesterol was purchased from Fagron (Waregem, Belgium). Glyceryl® ® stearate (Geleol ) was purchased from Gattefosee (Nanterre, France). Cremophor EL was purchased from BASF (Ludwigshafen, Germany). Paclitaxel (PTX) was purchased from Carbosynth (Berkshire, UK). All the mediums and supplements for cell culture were purchased from Life Technologies (Merelbecke, Belgium). All solvents were purchased in analytical grade from Merck (Darmstadt, Germany).

33


Drug Delivery to the Lungs 27, 2016 – Jibriil Ibrahim et al. Oxytocin Receptor Expression But a Lack of Response to Oxytocin in Airway Tissues Jibriil Ibrahim, 1

1, 2, 3

1

Michelle McIntosh & Rob Bischof

2, 3

Drug delivery, Disposition and Dynamics, Monash Institute of Pharmaceutical Sciences, Parkville, VIC, Australia 2

Biotechnology Research Laboratory, Department of Physiology, Monash University, Clayton, VIC, Australia 3

The Ritchie Centre, Hudson Institute of Medical Research, Clayton, VIC, Australia

Summary Postpartum haemorrhage, the uncontrollable loss of blood after labour, is one of the greatest contributors to maternal mortality in the developing world due to poor availability of oxytocin, a nonapeptide uterotonic considered the ‘gold standard’ treatment for inducing uterine contractility and stemming blood loss. This poor availability is primarily due to restrictions in cold chain storage and access to clean needles. We recently showed that pulmonary delivery of a stable, spray-dried formulation of oxytocin is a viable alternative to traditional [1] parenteral routes of administration for the induction of uterine contractility . While the role and relationship between oxytocin and its receptor have previously been investigated in reproductive systems, little is known about the mechanisms of action in the airways. With pulmonary delivery of oxytocin set to become a future treatment option in remote areas, the ramifications of exogenous oxytocin within the respiratory system needs to be understood. The aim of this study was to examine oxytocin receptor (OXTR) expression in airway and uterine tissues in non-pregnant sheep, and to assess responses to oxytocin ex vivo and following pulmonary administration in vivo. We hypothesised that inhalation of oxytocin would have no adverse effect on respiratory function and that oxytocin would not induce changes to airway morphology nor induce an inflammatory response. Introduction [1]

We recently showed that pulmonary delivery of a stable, spray dried formulation of oxytocin is a viable alternative to traditional parenteral routes of administration for the induction of uterine contractility. The aim of this study was to examine oxytocin receptor (OXTR) expression in airway and uterine tissues in non-pregnant sheep, and to assess responses to oxytocin ex vivo and following pulmonary administration in vivo. Methods OXTR expression and localisation was examined by fluorescent immunostaining in airway and uterine tissue collected post-mortem from n=4 sheep. Airway and uterine smooth muscle contractility was tested ex vivo in -11 -6 [1] response to oxytocin (10 – 10 M) using an organ bath tension gauge apparatus . Airway resistance, airway hyper-responsiveness (AHR, in response to carbachol) and inflammation were assessed in vivo before and after [2, 3] . pulmonary delivery of oxytocin or control powder compound (10 – 100 IU) Results OXTR expression in uterine tissue was localised almost exclusively to the smooth muscle, whereas expression in the airways was seen within smooth muscle bundles, in airway epithelium, and on resident leukocytes (Figure 1). Despite this broad OXTR expression in the airway tissues, oxytocin was unable to induce contractility in airway smooth muscle. This was in direct contrast to responses in uterine smooth muscle (Figure 2). Further, pulmonary delivery of oxytocin failed to induce a significant local inflammatory response in the airways, and did not induce changes in airway resistance or AHR over an incremental dose response (10 IU, 50 IU and 100 IU; Figure 3). Control powder formulation increased the respiratory rate post-administration (breaths per minute; 100% increase from baseline) but this effect was not observed following oxytocin administration.

34


Drug Delivery to the Lungs 27, 2016 - Oxytocin Receptor Expression But a Lack of Response to Oxytocin in Airway Tissues

A

B

Figure 1 -

C

Immunostaining of airway tissue, showing (A) DAPI nuclear staining, (B) smooth muscle -actin staining and (C) oxytocin receptor (OXTR) staining. Co-localisation of OXTR with airway smooth muscle (ASM) was evident in ASM underlying the bronchiolar epithelium (white arrows); x200 original magnification.

Contraction (% of HiK maximum)

200 150 100 50

ru s te U

a Pa re nc hy m

ro B

Ex vivo effect of oxytocin on contractile activity in airway and uterine smooth muscle. Muscle contraction was normalised as a percentage of high potassium (HiK) in saline. Control powder formulation did not induce contractility in smooth muscle (data not shown)(n=4; mean±SD; *P < 0.05).

300

oxytocin control 200

100

10 IU

50 IU

100 IU in m 10

in m 5

in m 10

in m 5

in

in m

m 10

0

5

in

0 m

Percentage relative to baseline (100%)

Figure 2 -

nc

Tr ac he a

hi

0

Time post-airway delivery Figure 3 -

Oxytocin and control compound delivered into the airways (increasing doses of 10 – 100 IU at 10 min intervals over time) had no effect on airway resistance relative to baseline levels (n=4; mean±SEM; *P < 0.05).

35


Drug Delivery to the Lungs 27, 2016 - Oxytocin Receptor Expression But a Lack of Response to Oxytocin in Airway Tissues

A

B

Figure 1 -

C

Immunostaining of airway tissue, showing (A) DAPI nuclear staining, (B) smooth muscle -actin staining and (C) oxytocin receptor (OXTR) staining. Co-localisation of OXTR with airway smooth muscle (ASM) was evident in ASM underlying the bronchiolar epithelium (white arrows); x200 original magnification.

Contraction (% of HiK maximum)

200 150 100 50

ru s te U

a Pa re nc hy m

ro B

Ex vivo effect of oxytocin on contractile activity in airway and uterine smooth muscle. Muscle contraction was normalised as a percentage of high potassium (HiK) in saline. Control powder formulation did not induce contractility in smooth muscle (data not shown)(n=4; mean±SD; *P < 0.05).

300

oxytocin control 200

100

10 IU

50 IU

100 IU in m 10

in m 5

in m 10

in m 5

in

in m

m 10

0

5

in

0 m

Percentage relative to baseline (100%)

Figure 2 -

nc

Tr ac he a

hi

0

Time post-airway delivery Figure 3 -

Oxytocin and control compound delivered into the airways (increasing doses of 10 – 100 IU at 10 min intervals over time) had no effect on airway resistance relative to baseline levels (n=4; mean±SEM; *P < 0.05).

35


Drug Delivery to the Lungs, 2016–R. Mahendran et al. Anti-Inflammatory Activity of Novel Trans-Stilbene Sulfonamide Analogues as Potential Novel Therapeutic Agents for Lung Disease R. Mahendran, E. Osoba, S. Rossiter, J.P. Bassin, M.T. Cook, V. Hutter Department of Pharmacy, University of Hertfordshire, College Lane Campus, Hatfield, AL10 9AB, UK Summary Chronic obstructive pulmonary disease (COPD) is a progressive disease of the airways, leading to chronic inflammation of the lung and is the fifth major cause of mortality in the UK. Current treatment options for COPD are limited and do not prevent disease progression. Hence, there is an urgent need for alternative novel therapeutic approaches. Reactive oxygen species play a key role in oxidative stress which is associated with COPD pathogenesis. Hence, oxidative stress is a potential target for new therapies for lung inflammatory disorders. Natural peanut stilbenes such as resveratrol have been shown to exhibit anti-inflammatory properties. In this study, characterisation of a library of stilbene sulfonamide compounds for their impact on cell health and inflammation was carried out in vitro. There was no significant impact observed on cell health at concentrations below 50 µM, indicating their suitability as safe inhaled therapeutic agents. Some of the analogues have shown promising anti-oxidant potential in vitro, displaying similar activity as resveratrol at 10 µM concentrations. These preliminary investigations warrant further investigation of these trans-stilbene sulphonamide analogues as potential new therapeutic candidates for the treatment for human inflammatory lung diseases. Introduction Chronic obstructive pulmonary disease (COPD) is a progressive disease of the airways characterised by [1] destruction of lung parenchyma with loss of lung elasticity and small airway inflammation . Causative agents such as particles from tobacco smoke, biomass exposure and pollution cause airway inflammation, result in the [2] release of reactive oxygen species in the airways . Oxidative stress is an important factor in the pathogenesis of [3] COPD and as such offers an perspective target for treatment of the disease . Natural peanut stilbenes such as resveratrol (trans-3,5,4-trihydroxystilbene) and synthetic stilbene analogues have been shown to exhibit a number of clinically useful biological properties, including an anti-inflammatory [4] mode of action . The main issues preventing resveratrol and other stilbene analogues from being marketed drugs are their low oral bioavailability as a result of poor aqueous solubility and rapid liver metabolism via phase [5, 6] . However, these properties may not hinder delivery via the inhaled route for two drug metabolising enzymes the management of local airway inflammation and oxidative stress. The aim of this study was to assess the cytotoxic and anti-inflammatory properties of a library of novel primary, secondary and tertiary stilbene sulphonamide and heterocyclic analogues for their potential use as inhaled therapeutic agents. Methods and Materials Synthesis and characterisation of novel trans-stilbene benzenesulfonamide analogues: A library of novel transstilbene analogues were synthesised using an efficient three step synthesis. Briefly, chlorosulfonation of diethyl benzylphosphonate was performed to form the sulfonyl chloride intermediate which underwent aminolysis using a range of primary, secondary and cyclic alkyl amines, as well as aromatic amines including dimethylamine and piperidine. Finally, formation of the stilbene backbone with various substituted aldehydes, via the Horner[7] Wadsworth-Emmons was performed . Compounds were characterised via NMR and MS and those with purity greater than 85% were tested in vitro. Cell culture: The human type 2 alveolar epithelial cell line, A549, was purchased from LCG Standards (Teddington, Middlesex, UK) and used between passage 25 and 45. Cells were cultured in Dulbecco’s modified Eagle’s medium with 10 %v/v fetal bovine serum and supplemented with 100 IU/ml penicillin-100 µg/ml streptomycin solution and 2 mM L-glutamine. Cells were cultured on T75 flasks in a humidified atmosphere at 37°C with 5% v/v CO2 and were passaged twice weekly when they reached 80% confluence. For experiments 4 cells were seeded onto 96 well plates at a density of 1 x 10 cells/well in 100 µl of complete cell culture medium. Determination of cell viability and proliferation: 24 h after seeding, compounds (0.01 – 100 µM) were added to cells in complete cell culture medium with 1% v/v DMSO. Following a 24 h exposure period, cells were assessed for viability using the CellTiter AQueous One assay and CytoTox ONE Homogenous Membrane Integrity assays (Promega, Southampton, Hampshire, UK) according to manufacturer’s instruction. Cells were also assessed for [8] proliferation using the sulforhodamine B assay as described previously . Determination of antioxidant activity and reactive oxygen species: The 2,2-diphenyl-1-picrylhydrazyl (DPPH) assay was employed to assess novel stilbene compounds for anti-oxidant free radical scavenging activity in [9] accordance with Trotta et al. . Stilbene compound solutions were prepared in ethanol (10, 50 and 100 µM) and combined with 60 µM ethanolic DPPH solution.

37


Drug Delivery to the Lungs 27, 2016 - Anti-Inflammatory Activity of Novel Trans-Stilbene Sulfonamide Analogues as Potential Novel Therapeutic Agents for Lung Disease Resveratrol and ascorbic acid were used as positive controls. Cells were incubated at room temperature in the dark for 30 min and absorbance read at 540 nm. The (2’,7’ –dichlorofluorescein diacetate (DCFDA) cellular reactive oxygen species assay kit was used to measure reactive oxygen species within the cell following induction of oxidative stress. The protocol was conducted according to manufacturer’s guidelines (Abcam, Cambridge, UK).

Results and Discussion All piperidine substituted analogues tested displayed no significant impact (p<0.05) on cell health at concentrations below 50 µM (Table 1). A significant decrease (p<0.05) in mitochondrial activity (30% of control) and significant increase (p<0.05) in cell permeability (10%) was observed with cells exposed to compound 1A. Similarly cells incubated with compounds 1B and 1D has significantly reduced (p<0.05) cell permeability at 100 µM concentrations (14% and 24% respectively). Only compound 1D from the piperidine substituted analogues caused a significant decrease (p<0.05) in cell proliferation at 100 µM. Cell health and proliferation effects of the piperidine analogues were only noted at the highest 100 µM concentration tested. At clinically relevant concentrations for inhaled drug delivery (0.05 – 10 µM) all piperidine analogues tested had no significant impact (p>0.05) on airway epithelial cell health.

Table 1: Impact of piperidine-substituted trans-stilbene benzenesulfonamide analogues on cell health and proliferation. A549 cells were seeded 1 x 104 cells per well in a 96 well plate and incubated with piperidine-substituted trans-stilbene benzenesulfonamide analogues (1A, 1B, 1C, 1D and 1E) for 24 h (0.01 – 100 µM). Cells were assessed for viability using the CellTiter AQueous One assay and CytoTox ONE Homogenous Membrane Integrity assays (Promega, Southampton, Hampshire, UK) and proliferation using the sulforhodamine B assay. Data is displayed as arbitrary classification (% relative to untreated control): MTS and SRB assay: - denotes > 75-100%, + denotes 50-75%, ++ denotes 25-50% & +++ <25%. LDH assay: - denotes 0-10%, + denotes 10-20%, ++ 20-50% and +++ denotes >50%.

Piperidine

Compound Structure

ID

clogP

Mitochondrial Activity

Cell Permeability

Cell Proliferation

1A

5.14

+

+

-

1B

3.97

-

+

-

1C

5.14

-

-

-

1D

3.84

-

+

+

1E

5.14

-

-

-

All dimethylamine-substituted analogues tested displayed no significant impact (p>0.05) on cell health at concentrations tested (up to 100 µM). Compound 2A significantly decreased (p<0.05) cell proliferation by 25% at 50 µM and 35% respectively at the 100 µM concentration tested. At clinically relevant concentrations for inhaled drug delivery (0.05 – 10 µM) all piperidine analogues tested had no significant impact (p>0.05) on airway epithelial cell health.

38


Drug Delivery to the Lungs, 2016–R. Mahendran et al. Table 2: Impact of dimethylamine-substituted trans-stilbene benzenesulfonamide analogues on cell health and proliferation. A549 cells were seeded 1 x 104 cells per well in a 96 well plate and incubated with piperidine-substituted transstilbene benzenesulfonamide analogues (2A, 2C, 2E, 2F and 2G) for 24 h (0.01 – 100 µM). Cells were assessed for viability using the CellTiter AQueous One assay and CytoTox ONE Homogenous Membrane Integrity assays (Promega, Southampton, Hampshire, UK) and proliferation using the sulforhodamine B assay. Data is displayed as arbitrary classification (% relative to untreated control): MTS and SRB assay: - denotes > 75-100%, + denotes 50-75%, ++ denotes 25-50% & +++ <25%. LDH assay: - denotes 0-10%, + denotes 10-20%, ++ 20-50% & +++ denotes >50%. ID

Dimethylamine

Compound Structure

clogP

Mitochondrial Activity

Cell Permeability

Cell Proliferation

2A

3.68

-

-

+

2C

3.68

-

-

-

2E

3.66

-

-

-

2F

3.76

-

-

+

2G

3.63

-

-

-

150

10 µM 50 µM 100 µM

100

50

2

G

F 2

E 1

B 1

A 1

ra

tr

o

l

0

R

e

s

e

v

e

R e la t iv e D P P H f r e e r a d ic a ls c o n c e n t r a t io n ( % )

The free radical scavenging activity of resveratrol was increased in a concentration dependent manner (Figure 1). Compounds 1B, 1E and 2F caused a significant reduction (p<0.05) in free radicals in comparison with control at the highest 100 µM concentration tested. However, this is likely to be at least 10 fold greater concentration than is clinically relevant. Only compound 1E significantly reduced free radical concentrations at the lowest 10 µM concentration tested which were comparable to incubation with 10 µM resveratrol. Preliminary results from the impact of the analogues on reactive oxygen species in A549 cells indicates that compound 2F reduces the concentration reactive oxygen species by more than 56% in comparison with untreated control. These preliminary results indicate that the novel trans-stilbene sulphonamide analogues synthesised may have some potentially useful anti-oxidant activity and lowering effect on reactive oxygen species to be worthy of further investigation as potential inhaled therapeutics.

Figure 1: Assessment of anti-oxidant activity of novel trans-stilbene benzenesulfonamide analogues. DPPH anti-oxidant assay of novel trans-stilbene benzenesulfonamide analogues. A total of six wells were used per data point, data is representative of mean ± standard deviation (* denotes p<0.05). Stilbene compound solutions were prepared in ethanol (10, 50 and 100 µM) and combined with 60 µM DPPH solution (prepared in ethanol) and were left in the dark for 30 minutes, followed by the measurement of absorbance at 540 nm. Resveratrol was used as a positive control.

39


Drug Delivery to the Lungs 27, 2016 - Anti-Inflammatory Activity of Novel Trans-Stilbene Sulfonamide Analogues as Potential Novel Therapeutic Agents for Lung Disease Conclusion The assessment of the impact of the novel stilbene compounds on A549 cell health showed that none of the compounds impacted upon cell viability at concentrations below 50 µM and most importantly at clinically relevant concentrations. Some of the compounds displayed similar anti-oxidant potential as resveratrol at 10 µM. In particular compound 2F was observed to possess significant anti-oxidant potential and good safety profile throughout the study. Further structure activity relationship studies are required to help elucidate the mechanism of action of these compounds. Reactive oxygen species play an important role in the pathogenesis of airway inflammatory disorders such as COPD and targeting such a key component is a pragmatic approach to therapy. These preliminary investigations have indicated low toxicity and moderate activity against oxidative stress in vitro. This library of trans-stilbene sulphonamide analogues warrant further investigation to assess their potential use for airway diseases such as COPD and asthma.

References 1

M. Decramer, W. Janssens, and M. Miravitlles, “Chronic obstructive pulmonary disease.,” Lancet, vol. 379, no. 9823, pp. 1341–51, 2012.

2

L. G. Wood, P. a B. Wark, and M. L. Garg, “Antioxidant and anti-inflammatory effects of resveratrol in airway disease.,” Antioxid. Redox Signal., vol. 13, no. 10, pp. 1535–48, 2010.

3

I. Rahman, “Antioxidant therapies in COPD.,” International journal of chronic obstructive pulmonary disease, vol. 1, no. 1. pp. 15–29, 2006.

4

V. S. Sobolev, S. I. Khan, N. Tabanca, D. E. Wedge, S. P. Manly, S. J. Cutler, M. R. Coy, J. J. Becnel, S. A. Neff, and J. B. Gloer, “Biological activity of peanut (Arachis hypogaea) phytoalexins and selected natural and synthetic stilbenoids,” J. Agric. Food Chem., vol. 59, no. 5, pp. 1673–1682, 2011.

5

a Amri, J. C. Chaumeil, S. Sfar, and C. Charrueau, “Administration of resveratrol: What formulation solutions to bioavailability limitations?” J. Control. Release, vol. 158, no. 2, pp. 182–93, Mar. 2012.

6

C. Simon, R. G. Britton, H. Cai, A. J. Gescher, K. Brown, and P. R. Jenkins, “Novel analogues of resveratrol: Metabolism and inhibition of colon cancer cell proliferation,” Tetrahedron, vol. 69, no. 30, pp. 6203–6212, 2013.

7

E.Osoba, “Synthesis and characterisation of novel trans-stilbene benzenesulfonamide analogues,” unpublished PhD Thesis.University of Hertfordshire, 2014.

8

M. R. Boyd and K. D. Paull, “Some practical consideration and applications of the National Cancer Institute in vitro anticancer drug discovery screen,” Drug Dev Des, vol. 34, pp. 91–109, 1995.

9

V. Trotta, W.-H. Lee, C.-Y. Loo, M. Haghi, P. M. Young, S. Scalia, and D. Traini, “In vitro biological activity of resveratrol using a novel inhalable resveratrol spray-dried formulation,” Int. J. Pharm., vol. 491, no. 1–2, pp. 190–197, 2015.

40


Drug Delivery to the Lungs 27, 2016 - Mohammed Ali Selo et al. Characterisation of MRP1 in human distal lung epithelial cells in vitro 1,2

1

1

1

1

Mohammed Ali Selo , Sabrina Nickel , Maximilian Richter , Caoimhe G. Clerkin & Carsten Ehrhardt 1

School of Pharmacy and Pharmaceutical Sciences & Trinity Biomedical Sciences Institute, Trinity College Dublin, Dublin 2, Ireland 2 Faculty of Pharmacy, Kufa University, Kufa, Al-Najaf, Iraq

Abbreviations ABC transporters, ATP-binding cassette transporters; AEC, alveolar epithelial cells; AIC air-interfaced culture; ATII cells, alveolar type II cells; ATI-like cells, alveolar type I-like cells; BSA, bovine serum albumin; CF,5(6)carboxyfluorescein; CFDA, 5(6)-carboxyfluorescein diacetate; FBS, foetal bovine serum; KRB, Krebs-Ringer buffer; LCC, liquid-covered culture; MRP1, multidrug resistance-associated protein1; PBS, phosphatebuffered saline. Summary Of the more than 50 members of the ABC-transporter family, MRP1 (ABCC1) has been identified to have the highest protein expression levels in human lung tissues. The subcellular localisation and functional activity of the transporter in lung epithelia, however, remains less well investigated. The aim of this project was to study MRP1 expression and function in human respiratory epithelium using the NCI-H441 cell line as well as freshly isolated human alveolar epithelial type (AT) II and type I-like cells in primary culture in vitro. ABCC1 gene expression was comparable between ATII and ATI-like cells, while MRP1 protein expression was higher in ATI-like cells from three different patients, when compared to ATII cells. Gene and protein expression remained stable over thirty passages of NCI-H441 cells studied, independent of culture conditions. Surface biotinylation confirmed basolateral expression of the transporter in NCI-H441 cells. Efflux experiments found the transporter to be functionally active and sensitive to specific inhibitors in both, NCI-H441 and primary cells. Bi-directional transport experiments revealed a net absorptive, MK-571 and telmisartan-sensitive CF flux, underlining the basolateral localisation of the transporter. This study shows that the cell line NCI-H441 is a useful in vitro tool for the investigation of MRP1 function in distal lung epithelium. Introduction The study of drug transporters in the lung is of importance because they might influence the local [1] pharmacokinetics of inhaled drugs or can be involved in the pathophysiology of respiratory diseases . MRP1 is [2] one of the most highly expressed ABC-transporters in the human lung tissues . MRP1 has also been shown to [3] interact with inhaled drugs such as N-acetylcysteine or inhaled corticosteroids . Furthermore, there is pre-clinical and clinical evidence that changes in expression or function of the transporter are associated with the occurrence [4,5] . Unfortunately, there is a considerable lack of suitable cell culture models for the study of MRP1 of COPD [1] function or expression in the distal lung . Therefore, we aimed at establishing the human adenocarcinoma cell line, NCI-H441 as an in vitro model by comparing expression and activity of MRP1 in NCI-H441 cells with freshly isolated human alveolar epithelial cells (AEC) in primary culture. Materials and Methods All cells were grown on Transwell inserts, either under air-interfaced culture (AIC) or liquid-covered culture (LCC) conditions for expression analyses and transport studies, or on tissue culture plastic for efflux experiments with ATI-like cells. NCI-H441 cells were fed with RPMI-1640 medium supplemented with 5% foetal bovine serum (FBS), 1 mM sodium pyruvate, 100 U/ml penicillin and 100 μg/ml streptomycin, whereas ATII cells were cultured using Small Airways Growth Medium (SAGM) supplemented with 1% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin. MRP1 expression in ATII and ATI-like cells from three different donors and in at least three different passages of NCI-H441 cells was studied using real-time PCR (q-PCR) and immunoblot. Transporter localisation was confirmed by surface biotinylation. Surface biotinylation was conducted by treating cells grown in Transwell inserts with sulpho-NHS-biotin reagent either apically or basolaterally for 20 min. Bovine serum albumin (BSA, 10%) containing phosphate-buffered saline (PBS) was used in the respective opposite compartment. The reaction was quenched by incubating both compartments with 10% FBS in PBS for 20 min, followed by treating cell monolayers with lysis buffer. The biotinylated proteins were isolated by overnight incubation of cell lysate with agarose streptavidin beads with subsequent centrifugation. The biotinylated proteins were eluted from streptavidin beads by 20 min incubation with sample loading buffer and used for immunoblotting. The activity of the transporters in both cell models was investigated by carrying out bi-directional efflux and transport studies. Bidirectional efflux experiments were carried out by incubating cell monolayers grown on Transwell inserts with 100 µM CFDA with or without the inhibitor compound, MK571 (20 µM) for 60 min. Then the substrate solution was replaced with Krebs-Ringer buffer (KRB) buffer and 200 µl samples were withdrawn from both apical and basolateral compartments every 15 min up to 90 min. The fluorescence activity of samples was assessed in 96well plates using an automated plate reader at excitation and emission wavelengths of 485 and 520 nm, respectively. Bi-directional transport studies (A-B) and (B-A) were also performed by using the substrate CFDA at 100 µM as well as the inhibitor compound, MK-571 at 20 µM. The apparent permeability coefficient (Papp) was calculated using the following equation: Papp = (ΔQ/Δt)/(AC0), where ΔQ is the change in concentration of CF over the designated period of time (Δt), A is the surface area of the Transwell inserts, and C0 is the initial concentration of CFDA in the donor compartment.

41


Drug Delivery to the Lungs 27, 2016 - Mohammed Ali Selo et al. Characterisation of MRP1 in human distal lung epithelial cells in vitro 1,2

1

1

1

1

Mohammed Ali Selo , Sabrina Nickel , Maximilian Richter , Caoimhe G. Clerkin & Carsten Ehrhardt 1

School of Pharmacy and Pharmaceutical Sciences & Trinity Biomedical Sciences Institute, Trinity College Dublin, Dublin 2, Ireland 2 Faculty of Pharmacy, Kufa University, Kufa, Al-Najaf, Iraq

Abbreviations ABC transporters, ATP-binding cassette transporters; AEC, alveolar epithelial cells; AIC air-interfaced culture; ATII cells, alveolar type II cells; ATI-like cells, alveolar type I-like cells; BSA, bovine serum albumin; CF,5(6)carboxyfluorescein; CFDA, 5(6)-carboxyfluorescein diacetate; FBS, foetal bovine serum; KRB, Krebs-Ringer buffer; LCC, liquid-covered culture; MRP1, multidrug resistance-associated protein1; PBS, phosphatebuffered saline. Summary Of the more than 50 members of the ABC-transporter family, MRP1 (ABCC1) has been identified to have the highest protein expression levels in human lung tissues. The subcellular localisation and functional activity of the transporter in lung epithelia, however, remains less well investigated. The aim of this project was to study MRP1 expression and function in human respiratory epithelium using the NCI-H441 cell line as well as freshly isolated human alveolar epithelial type (AT) II and type I-like cells in primary culture in vitro. ABCC1 gene expression was comparable between ATII and ATI-like cells, while MRP1 protein expression was higher in ATI-like cells from three different patients, when compared to ATII cells. Gene and protein expression remained stable over thirty passages of NCI-H441 cells studied, independent of culture conditions. Surface biotinylation confirmed basolateral expression of the transporter in NCI-H441 cells. Efflux experiments found the transporter to be functionally active and sensitive to specific inhibitors in both, NCI-H441 and primary cells. Bi-directional transport experiments revealed a net absorptive, MK-571 and telmisartan-sensitive CF flux, underlining the basolateral localisation of the transporter. This study shows that the cell line NCI-H441 is a useful in vitro tool for the investigation of MRP1 function in distal lung epithelium. Introduction The study of drug transporters in the lung is of importance because they might influence the local [1] pharmacokinetics of inhaled drugs or can be involved in the pathophysiology of respiratory diseases . MRP1 is [2] one of the most highly expressed ABC-transporters in the human lung tissues . MRP1 has also been shown to [3] interact with inhaled drugs such as N-acetylcysteine or inhaled corticosteroids . Furthermore, there is pre-clinical and clinical evidence that changes in expression or function of the transporter are associated with the occurrence [4,5] . Unfortunately, there is a considerable lack of suitable cell culture models for the study of MRP1 of COPD [1] function or expression in the distal lung . Therefore, we aimed at establishing the human adenocarcinoma cell line, NCI-H441 as an in vitro model by comparing expression and activity of MRP1 in NCI-H441 cells with freshly isolated human alveolar epithelial cells (AEC) in primary culture. Materials and Methods All cells were grown on Transwell inserts, either under air-interfaced culture (AIC) or liquid-covered culture (LCC) conditions for expression analyses and transport studies, or on tissue culture plastic for efflux experiments with ATI-like cells. NCI-H441 cells were fed with RPMI-1640 medium supplemented with 5% foetal bovine serum (FBS), 1 mM sodium pyruvate, 100 U/ml penicillin and 100 μg/ml streptomycin, whereas ATII cells were cultured using Small Airways Growth Medium (SAGM) supplemented with 1% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin. MRP1 expression in ATII and ATI-like cells from three different donors and in at least three different passages of NCI-H441 cells was studied using real-time PCR (q-PCR) and immunoblot. Transporter localisation was confirmed by surface biotinylation. Surface biotinylation was conducted by treating cells grown in Transwell inserts with sulpho-NHS-biotin reagent either apically or basolaterally for 20 min. Bovine serum albumin (BSA, 10%) containing phosphate-buffered saline (PBS) was used in the respective opposite compartment. The reaction was quenched by incubating both compartments with 10% FBS in PBS for 20 min, followed by treating cell monolayers with lysis buffer. The biotinylated proteins were isolated by overnight incubation of cell lysate with agarose streptavidin beads with subsequent centrifugation. The biotinylated proteins were eluted from streptavidin beads by 20 min incubation with sample loading buffer and used for immunoblotting. The activity of the transporters in both cell models was investigated by carrying out bi-directional efflux and transport studies. Bidirectional efflux experiments were carried out by incubating cell monolayers grown on Transwell inserts with 100 µM CFDA with or without the inhibitor compound, MK571 (20 µM) for 60 min. Then the substrate solution was replaced with Krebs-Ringer buffer (KRB) buffer and 200 µl samples were withdrawn from both apical and basolateral compartments every 15 min up to 90 min. The fluorescence activity of samples was assessed in 96well plates using an automated plate reader at excitation and emission wavelengths of 485 and 520 nm, respectively. Bi-directional transport studies (A-B) and (B-A) were also performed by using the substrate CFDA at 100 µM as well as the inhibitor compound, MK-571 at 20 µM. The apparent permeability coefficient (Papp) was calculated using the following equation: Papp = (ΔQ/Δt)/(AC0), where ΔQ is the change in concentration of CF over the designated period of time (Δt), A is the surface area of the Transwell inserts, and C0 is the initial concentration of CFDA in the donor compartment. 42


Drug Delivery to the Lungs 27, 2016 - Mohammed Ali Selo et al. Characterisation of MRP1 in human distal lung epithelial cells in vitro 1,2

1

1

1

1

Mohammed Ali Selo , Sabrina Nickel , Maximilian Richter , Caoimhe G. Clerkin & Carsten Ehrhardt 1

School of Pharmacy and Pharmaceutical Sciences & Trinity Biomedical Sciences Institute, Trinity College Dublin, Dublin 2, Ireland 2 Faculty of Pharmacy, Kufa University, Kufa, Al-Najaf, Iraq

Abbreviations ABC transporters, ATP-binding cassette transporters; AEC, alveolar epithelial cells; AIC air-interfaced culture; ATII cells, alveolar type II cells; ATI-like cells, alveolar type I-like cells; BSA, bovine serum albumin; CF,5(6)carboxyfluorescein; CFDA, 5(6)-carboxyfluorescein diacetate; FBS, foetal bovine serum; KRB, Krebs-Ringer buffer; LCC, liquid-covered culture; MRP1, multidrug resistance-associated protein1; PBS, phosphatebuffered saline. Summary Of the more than 50 members of the ABC-transporter family, MRP1 (ABCC1) has been identified to have the highest protein expression levels in human lung tissues. The subcellular localisation and functional activity of the transporter in lung epithelia, however, remains less well investigated. The aim of this project was to study MRP1 expression and function in human respiratory epithelium using the NCI-H441 cell line as well as freshly isolated human alveolar epithelial type (AT) II and type I-like cells in primary culture in vitro. ABCC1 gene expression was comparable between ATII and ATI-like cells, while MRP1 protein expression was higher in ATI-like cells from three different patients, when compared to ATII cells. Gene and protein expression remained stable over thirty passages of NCI-H441 cells studied, independent of culture conditions. Surface biotinylation confirmed basolateral expression of the transporter in NCI-H441 cells. Efflux experiments found the transporter to be functionally active and sensitive to specific inhibitors in both, NCI-H441 and primary cells. Bi-directional transport experiments revealed a net absorptive, MK-571 and telmisartan-sensitive CF flux, underlining the basolateral localisation of the transporter. This study shows that the cell line NCI-H441 is a useful in vitro tool for the investigation of MRP1 function in distal lung epithelium. Introduction The study of drug transporters in the lung is of importance because they might influence the local [1] pharmacokinetics of inhaled drugs or can be involved in the pathophysiology of respiratory diseases . MRP1 is [2] one of the most highly expressed ABC-transporters in the human lung tissues . MRP1 has also been shown to [3] interact with inhaled drugs such as N-acetylcysteine or inhaled corticosteroids . Furthermore, there is pre-clinical and clinical evidence that changes in expression or function of the transporter are associated with the occurrence [4,5] . Unfortunately, there is a considerable lack of suitable cell culture models for the study of MRP1 of COPD [1] function or expression in the distal lung . Therefore, we aimed at establishing the human adenocarcinoma cell line, NCI-H441 as an in vitro model by comparing expression and activity of MRP1 in NCI-H441 cells with freshly isolated human alveolar epithelial cells (AEC) in primary culture. Materials and Methods All cells were grown on Transwell inserts, either under air-interfaced culture (AIC) or liquid-covered culture (LCC) conditions for expression analyses and transport studies, or on tissue culture plastic for efflux experiments with ATI-like cells. NCI-H441 cells were fed with RPMI-1640 medium supplemented with 5% foetal bovine serum (FBS), 1 mM sodium pyruvate, 100 U/ml penicillin and 100 μg/ml streptomycin, whereas ATII cells were cultured using Small Airways Growth Medium (SAGM) supplemented with 1% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin. MRP1 expression in ATII and ATI-like cells from three different donors and in at least three different passages of NCI-H441 cells was studied using real-time PCR (q-PCR) and immunoblot. Transporter localisation was confirmed by surface biotinylation. Surface biotinylation was conducted by treating cells grown in Transwell inserts with sulpho-NHS-biotin reagent either apically or basolaterally for 20 min. Bovine serum albumin (BSA, 10%) containing phosphate-buffered saline (PBS) was used in the respective opposite compartment. The reaction was quenched by incubating both compartments with 10% FBS in PBS for 20 min, followed by treating cell monolayers with lysis buffer. The biotinylated proteins were isolated by overnight incubation of cell lysate with agarose streptavidin beads with subsequent centrifugation. The biotinylated proteins were eluted from streptavidin beads by 20 min incubation with sample loading buffer and used for immunoblotting. The activity of the transporters in both cell models was investigated by carrying out bi-directional efflux and transport studies. Bidirectional efflux experiments were carried out by incubating cell monolayers grown on Transwell inserts with 100 µM CFDA with or without the inhibitor compound, MK571 (20 µM) for 60 min. Then the substrate solution was replaced with Krebs-Ringer buffer (KRB) buffer and 200 µl samples were withdrawn from both apical and basolateral compartments every 15 min up to 90 min. The fluorescence activity of samples was assessed in 96well plates using an automated plate reader at excitation and emission wavelengths of 485 and 520 nm, respectively. Bi-directional transport studies (A-B) and (B-A) were also performed by using the substrate CFDA at 100 µM as well as the inhibitor compound, MK-571 at 20 µM. The apparent permeability coefficient (Papp) was calculated using the following equation: Papp = (ΔQ/Δt)/(AC0), where ΔQ is the change in concentration of CF over the designated period of time (Δt), A is the surface area of the Transwell inserts, and C0 is the initial concentration of CFDA in the donor compartment.

43


Drug Delivery to the Lungs 27, 2016 - Mohammed Ali Selo et al. Characterisation of MRP1 in human distal lung epithelial cells in vitro 1,2

1

1

1

1

Mohammed Ali Selo , Sabrina Nickel , Maximilian Richter , Caoimhe G. Clerkin & Carsten Ehrhardt 1

School of Pharmacy and Pharmaceutical Sciences & Trinity Biomedical Sciences Institute, Trinity College Dublin, Dublin 2, Ireland 2 Faculty of Pharmacy, Kufa University, Kufa, Al-Najaf, Iraq

Abbreviations ABC transporters, ATP-binding cassette transporters; AEC, alveolar epithelial cells; AIC air-interfaced culture; ATII cells, alveolar type II cells; ATI-like cells, alveolar type I-like cells; BSA, bovine serum albumin; CF,5(6)carboxyfluorescein; CFDA, 5(6)-carboxyfluorescein diacetate; FBS, foetal bovine serum; KRB, Krebs-Ringer buffer; LCC, liquid-covered culture; MRP1, multidrug resistance-associated protein1; PBS, phosphatebuffered saline. Summary Of the more than 50 members of the ABC-transporter family, MRP1 (ABCC1) has been identified to have the highest protein expression levels in human lung tissues. The subcellular localisation and functional activity of the transporter in lung epithelia, however, remains less well investigated. The aim of this project was to study MRP1 expression and function in human respiratory epithelium using the NCI-H441 cell line as well as freshly isolated human alveolar epithelial type (AT) II and type I-like cells in primary culture in vitro. ABCC1 gene expression was comparable between ATII and ATI-like cells, while MRP1 protein expression was higher in ATI-like cells from three different patients, when compared to ATII cells. Gene and protein expression remained stable over thirty passages of NCI-H441 cells studied, independent of culture conditions. Surface biotinylation confirmed basolateral expression of the transporter in NCI-H441 cells. Efflux experiments found the transporter to be functionally active and sensitive to specific inhibitors in both, NCI-H441 and primary cells. Bi-directional transport experiments revealed a net absorptive, MK-571 and telmisartan-sensitive CF flux, underlining the basolateral localisation of the transporter. This study shows that the cell line NCI-H441 is a useful in vitro tool for the investigation of MRP1 function in distal lung epithelium. Introduction The study of drug transporters in the lung is of importance because they might influence the local [1] pharmacokinetics of inhaled drugs or can be involved in the pathophysiology of respiratory diseases . MRP1 is [2] one of the most highly expressed ABC-transporters in the human lung tissues . MRP1 has also been shown to [3] interact with inhaled drugs such as N-acetylcysteine or inhaled corticosteroids . Furthermore, there is pre-clinical and clinical evidence that changes in expression or function of the transporter are associated with the occurrence [4,5] . Unfortunately, there is a considerable lack of suitable cell culture models for the study of MRP1 of COPD [1] function or expression in the distal lung . Therefore, we aimed at establishing the human adenocarcinoma cell line, NCI-H441 as an in vitro model by comparing expression and activity of MRP1 in NCI-H441 cells with freshly isolated human alveolar epithelial cells (AEC) in primary culture. Materials and Methods All cells were grown on Transwell inserts, either under air-interfaced culture (AIC) or liquid-covered culture (LCC) conditions for expression analyses and transport studies, or on tissue culture plastic for efflux experiments with ATI-like cells. NCI-H441 cells were fed with RPMI-1640 medium supplemented with 5% foetal bovine serum (FBS), 1 mM sodium pyruvate, 100 U/ml penicillin and 100 μg/ml streptomycin, whereas ATII cells were cultured using Small Airways Growth Medium (SAGM) supplemented with 1% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin. MRP1 expression in ATII and ATI-like cells from three different donors and in at least three different passages of NCI-H441 cells was studied using real-time PCR (q-PCR) and immunoblot. Transporter localisation was confirmed by surface biotinylation. Surface biotinylation was conducted by treating cells grown in Transwell inserts with sulpho-NHS-biotin reagent either apically or basolaterally for 20 min. Bovine serum albumin (BSA, 10%) containing phosphate-buffered saline (PBS) was used in the respective opposite compartment. The reaction was quenched by incubating both compartments with 10% FBS in PBS for 20 min, followed by treating cell monolayers with lysis buffer. The biotinylated proteins were isolated by overnight incubation of cell lysate with agarose streptavidin beads with subsequent centrifugation. The biotinylated proteins were eluted from streptavidin beads by 20 min incubation with sample loading buffer and used for immunoblotting. The activity of the transporters in both cell models was investigated by carrying out bi-directional efflux and transport studies. Bidirectional efflux experiments were carried out by incubating cell monolayers grown on Transwell inserts with 100 µM CFDA with or without the inhibitor compound, MK571 (20 µM) for 60 min. Then the substrate solution was replaced with Krebs-Ringer buffer (KRB) buffer and 200 µl samples were withdrawn from both apical and basolateral compartments every 15 min up to 90 min. The fluorescence activity of samples was assessed in 96well plates using an automated plate reader at excitation and emission wavelengths of 485 and 520 nm, respectively. Bi-directional transport studies (A-B) and (B-A) were also performed by using the substrate CFDA at 100 µM as well as the inhibitor compound, MK-571 at 20 µM. The apparent permeability coefficient (Papp) was calculated using the following equation: Papp = (ΔQ/Δt)/(AC0), where ΔQ is the change in concentration of CF over the designated period of time (Δt), A is the surface area of the Transwell inserts, and C0 is the initial concentration 44 of CFDA in the donor compartment.


Drug Delivery to the Lungs 27, 2016 -Ahmed O. Shalash et al. Carrier microstructure and performance of dry powder inhalation mixtures: A step towards universal performance prediction model Ahmed O. Shalash1, Nawal M. Khalafallah2 Abdulla M. Molokhia1, Mustafa M.A. Elsayed2,3 1 European Egyptian Pharmaceutical Industries, Alexandria, Egypt, e-mail: ahmedsidala@hotmail.com 2 Department of Pharmaceutics, Faculty of Pharmacy, Alexandria University, Alexandria, Egypt 3 College of Pharmacy, University of Hail, P.O. Box 6166, Hail 81442, Saudi Arabia, e-mail: mustafa.elsayed@alexpharmres.com

Summary Performance-explaining theories are often limited. In carrier-based dry powder inhalers (DPIs), this is mainly underlain by the large number of interacting variables (device, drug and carrier). In this study, the relationships between carrier particle size and permeability, and between microstructural properties and inhalation performance are explored using eleven different carrier grades. Hydroxypropyl-βcyclodextrin, lactose anhydrous, lactose monohydrate, dextrose anhydrous, dextrose monohydrate, xylitol and sucrose were used as carriers to prepare fifteen 1% w/w fluticasone propionate inhalation mixtures. The carriers were characterized for particle size using laser diffraction and for pore-size distribution and air permeability using mercury intrusion porosimetry. The in-vitro performances after aerosolization from an Aerolizer ® at 60 L.min-1 were assessed using Next Generation Impactor ®. The carriers differed in their particle size, fine contents, particle shapes, and air permeabilities. CD had a remarkable nanoporosity. The performances of the inhalation mixtures varied widely. Optimum dispersion was achieved at carrier permeability of 3.0 Darcy. Lower FPFs were found above and below this value. However, to predict the performance of the deviating nanorough carriers in a single model, the carrier roughness scales should be taken into account. Taking porosity scales into account in a simple multivariate model, resulted in a performance predictive relationship with an adjusted R2=0.98. Introduction Dry powder inhalers (DPIs) have captured a considerable interest for their ability to locally treat pulmonary disorders and to systemically deliver active ingredients (APIs), including macromolecules and polypeptides. The performance prediction of DPIs is an obstacle that has eluded the scientific community for decades. Proposed performance-explaining theories are often limited [1]. This is mainly underlain by the large number of interacting variables (device, API and carrier), which effectively contribute to performance and create a complex multivariate environment. Carrier microstructural properties such as permeability and microporosity were found to account for the effects of carrier size distribution and microscale roughness, respectively, on the performance in a semiquantitative manner [2]. In the current study, the relationships between carrier particle size and permeability and inhalation mixture performance are explored using eleven different carrier grades belonging to seven different crystal forms or chemical entities. Materials and methods Grades of hydroxypropyl-β-cyclodextrin (Kleptose ® HPB), dextrose monohydrate (Roferose ® SF), and xylitol (Xylisorb ® 300) were from Roquette, Lestrem, France. Lactose anhydrous (Lactopress ® anhydrous 265) was from Borculo Domo Ingredients, Zwolle, The Netherlands. Lactose monohydrate grades, Lactohale ® LH200, Lactohale ® LH201, were from Friesland Foods Domo, Zwolle, The Netherlands. Lactose monohydrate 120M was from Kerry Bio-Science, Norwich, New York, USA. Lactose monohydrate 200 mesh was from The Lactose Company of New Zealand Limited, Kapuni, New Zealand. Pharmatose ® DCL14 was from DMV-Fonterra Excipients GmbH & Co.KG, Nörten Hardenberg, Germany. Dextrose anhydrous was from SunTin MediPharma Co. Ltd., Hong Kong, China. Sucrose was from Daqahlia Sugar Co., Cairo, Egypt. Fluticasone propionate was from Jayco Chemical Industries, Maharashtra, India. The eleven carrier grades were employed as carriers. They were used to prepare fifteen 1% w/w fluticasone propionate inhalation mixtures. The carriers were evaluated for particle size using laser diffraction, crystallinity using differential scanning calorimetry, particle shape using optical microscopy image analysis, and pore-size distribution and permeability of the carriers using mercury intrusion porosimetry. The in-vitro performances of the inhalation mixtures were assessed using cascade impaction analysis (Next Generation Impactor ®) after aerosolization from an Aerolizer ® at 60 L.min-1. Results Hydroxypropyl-β-cyclodextrin (CD) was amorphous while all the other carriers were crystalline. The carriers differed in their particle size with D v, mode in the range of 33.8-88.9 µm and fine contents (D<15µm) in the range of 0.46-20.1 %. The carrier permeabilities were in the range of 0.4-14.6 Darcy. Hydroxypropyl-β-cyclodextrin had a

45


Drug Delivery to the Lungs 27, 2016 - Carrier microstructure and performance of dry powder inhalation mixtures: A step towards universal performance prediction model remarkable nanoporosity. The carriers varied widely in their inhalation performance expressed in terms of fine particle fraction (FPF).

Figure 1 Relationships between carrier air permeability and size, expressed as (A) fine contents and (B) median diameter.

Discussion There were observable trends between the carrier size distribution and permeability, c.f. Fig. 1. The relationship between the permeability and either the median diameter D v, 50 or the fine contents (D<25 µm) followed a quasiexponential pattern. Carrier air permeability was derived from mercury intrusion porosimetry measurements. Several relationships between powder pore volume distribution and powder permeability were reported [2-4]. Optimum dispersion was achieved at carrier permeability of 3.0 Darcy. Lower FPFs were found above and below this value, c.f. Fig 2A. At higher air permeabilities reduced resistance to air flow leads to weaker aerodynamic (drag, lift and inertial) dispersion forces during aerosolization. At lower air permeabilities high resistance to air flow restricts effective emission and/or dispersion of an inhalation mixture. Too much fines may result in highly cohesive powders that the device separation forces may not be adequate to disperse efficiently. These effects are related to inhaler device properties (Aerolizer ®), at inhalation flow rate of 60 L.min-1. However, nanorough/nanoporous carriers, such as CD, deviate from the relationship. Inhalation performance also increased with the carrier microporosity, c.f. Fig 2B [2]. The carrier microporosity was generally defined as total volume of micropores per one gram of the powder bed with pore diameters ranging from 1 to 8 µm. Such pores are generally expected to shield drug particles from drag forces, increase drug-carrier contact. Moreover, they are not expected to influence powder fluidization behaviour. However, they were found to increase with carrier fines content and to have positive impact on performance, Fig. 2B [2]. Therefore, it is expected that microporosity contribution to performance occurred during the mixing process, mixing deagglomeration effect [2]. This effect is further supported by deviation in inhalation performance of the macroscopically rough Pharmatose ® DCL-14 carrier, despite possessing considerably high microporosity value, > 0.1 mL/g, Fig. 2B. This was also investigated and recently reported [2]. On the other hand, performance improvement associated with increasing carrier nanoporosity was attributed to the reduction in Van der Waals adhesion forces.

46


Drug Delivery to the Lungs 27, 2016 -Ahmed O. Shalash et al. Figure 2 Relationships between carrier air permeability (A) or carrier microporosity (B) and inhalation performance (FPF)

Carrier chemical entity or crystal form seems to have a minor effect on the performance, while the carrier particle shape, size and microstructural properties are key determinants of the performance, c.f. Fig. 2-3. Non-linear fitting of the permeability-performance data accounts for the effects of size distribution in a more performance descriptive manner, c.f. Fig. 2A and Fig. 3. Moreover, permeability was able to adequately predict inhalation performance of macroscopically rough Pharmatose ÂŽ DCL-14 carrier, despite its deviation from performancemicroporosity relationship, c.f. Fig. 2B. In contrast, carrier size distribution parameters, such as D v, 0.5 or percent fines (D<10 Âľm), data resulted in poor correlations with inhalation performance (FPF), c.f. Fig. 3. This is mainly attributed to the complexity of underlying interactions of air flow-induced dispersion performance with surface texture, particle shape, packing arrangement and different contribution of each size fraction, c.f. Fig. 2-3. Therefore, these effects are better represented using the carrier permeability, and porosity scales, i.e. permeability can describe carrier size and packing properties in a more performance relevant manner, c.f. Fig. 2A.

Figure 3 Relationship between carrier size and inhalation performance

Out of the carriers investigated in this study, the nanoporous (CD) is suggested as an excellent carrier; the carrier was able to achieve approximately 75% of the maximum achievable performance. Despite that CD carriers possess remarkably high air permeability values, 11.2 and 14 Darcy, and remarkably low percent fines content (D<10 Âľm), 5% and 3%. Following the permeability performance relationship trend, c.f. Fig. 2A, this should only qualify CD carriers for FPF values below < 5%. Nonetheless, relative performance for regularly shaped crystalline carriers (excluding CD carriers) can be adequately predicted, at similar testing conditions, using the carrier permeability-performance relationships, c.f. Fig. 2A. However, to predict the performance of nanorough CD carriers in their inhalation mixtures in a single model, the nanoroughness and microporosity of the carriers should be taken into account. To this end, a multiple linear regression model was developed. Three main variables were employed in the model: (1) linearized microporosity with respect to the performance [2]. (2) linearized permeability with respect to the performance, using the functions developed in figure 2, and (3) carriers nanoporosity. Fitting data of the carriers, all slightly elongated carriers, gave an adj. R 2=0.98, c.f. Fig. 4, Table 1. The future introduction of advanced quantitative shape analysis techniques and interaction effects will allow for more accurate models.

47


Drug Delivery to the Lungs 27, 2016 - Carrier microstructure and performance of dry powder inhalation mixtures: A step towards universal performance prediction model

Summary

Figure 4 Correlation between measured and predicted performance values

Carrier properties (normalized to maximum) Intercept

Value

Nanoporosity

4.474

1.565

Permeability estimate of FPF

15.211

2.078

Microporosity estimate of FPF

2.348

1.796

ANOVA

Sum of Model squares

0

Standard Error -

12301.830

Sum of error squares

95.318

F-Value

473.224

Prob>F

6.63E-12

Table 1 Multiple linear regression model parameters and statistics

Conclusions Permeability is an outstanding performance descriptor for all regular crystalline carriers. Taking porosity scales into account in a simple multivariate model, resulted in a performance predictive relationship with an adj. R2=0.98. These effects are related to inhaler device properties (Aerolizer ÂŽ), and the inhalation flow rate of 60 L.min -1. Carrier chemical entity or crystal form seems to have a minor effect on the performance. Out of the carriers investigated in this study, the nanoporous (CD) is suggested as an excellent carrier. Acknowledgements The study was funded by a Global Fellowship Award from the United States Pharmacopeial Convention (USP). References 1. 2.

3. 4.

J. Shur, H. Harris, M.D. Jones, J.S. Kaerger, R. Price, The Role of Fines in the Modification of the Fluidization and Dispersion Mechanism Within Dry Powder Inhaler Formulations, Pharmaceutical research, 25 (2008) 1631-1640. A.O. Shalash, A.M. Molokhia, M.M. Elsayed, Insights into the roles of carrier microstructure in adhesive/carrier-based dry powder inhalation mixtures: Carrier porosity and fine particle content, European journal of pharmaceutics and biopharmaceutics: official journal of Arbeitsgemeinschaft fur Pharmazeutische Verfahrenstechnik e.V, 96 (2015) 291-303. B.F. Swanson, A Simple Correlation between Permeabilities and Mercury Capillary Pressures, J Petrol Technol, 33 (1981) 2498-2504. A.E. Scheidegger, The Physics of Flow Through Porous Media, University of Toronto Press, 1974.

48


Drug Delivery to the Lungs 27, 2016 – Ville Vartiainen et al. Development of Inhalable Drug Formulations for Idiopathic Pulmonary Fibrosis Ville Vartiainen1, Janne Raula2, Katri Koli1 & Marjukka Myllärniemi1 1

University of Helsinki and Helsinki University hospital, Research Programs Unit and Haartman Institute, PO.Box 63, Haartmaninkatu 8, FI-00014 Helsinki, Finland 2

Aalto University School of Science, Puumiehenkuja 2A, FI-00076 AALTO, Finland

Introduction Idiopathic pulmonary fibrosis is a chronic progressive lung disease, in which the functional gas exchanging tissue of the lungs is replaced with connective tissue, and is characterized by histopathological pattern of usual interstitial pneumonia.[1] The mortality ranges from 78.6 to 188.6 per 1000 patient/years depending on the disease severity.[2] Currently, the IPF patients have very few treatment options. In EU and US only two drugs, pirfenidone and nintedanib, have been approved for clinical use. [3–5] The available drugs are also associated with adverse effects which significantly decrease the quality of life of the patients. In the clinical trials on pirfenidone, the CAPACITY studies, 98% of patients on pirfenidone reported at least one adverse effect, of which most common were gastrointestinal symptoms, skin related adverse effects and dizziness.[6] In the clinical trials on nintedanib, the INPULISIS studies, over 90% of the patients on nintedanib reported diarrhea as treatment-emergent adverse effect.[7] Administration of drugs directly to the lungs could result in smaller systemic and total doses, and therefore decreased adverse effects and costs. For example, the doses of inhaled corticosteroids in modern asthma treatment are only fractions of the doses used in oral administration. Local administration and reduced doses have enabled long term use of a drug class that would have severe adverse effects in systemic dosing. Aerosol flow reactor method has been successfully used, for the formulation of water soluble and insoluble drugs, as inhalable dry powders.[8,9] It is a one-step continuous process to produce highly flowable and dispersible dry powder from precursor solution. Drug particles are coated with L-leucine by physical vapor deposition, which enables formation of two qualitatively different L-leucine layers. Firstly, particles are encapsulated by continuous L-leucine layer, which protects them from moisture.[10] Secondly, L-leucine nanocrystals are formed on the particle surface. These nanocrystals allow spatial separation of the particles, which is imperative for good dispersion properties of dry powders.[11] The aim of this work is to develop inhalable drug formulations for IPF. Our research group has discovered a potential antifibrotic drug, tilorone, and demonstrated its antifibrotic properties in vivo.[12] Our hypothesis is that tilorone, pirfenidone and nintedanib can be formulated as inhalable dry powders and local administration will allow decrease of used drug doses. Experimental methods Tilorone (T, Hangzhou Pharma & Chem Co., Ltd.), pirfenidone (Pi, Wuhan Benjamin Pharmaceutical Chemical Co.), leucine (L, Alfa Aesar), mannitol (Alfa Aesar), dipalmitoylphosphatidylcholine (DPPC, Avanti Polar Lipids), Luciferin (Promega), Quick cell proliferation assay kit II (BioVision), Minimum Essential Medium Eagle (MEM, Sigma-Aldrich), RPMI-1640 media (Sigma-Aldrich), Fetal Bovine Serum (FBS, Gibco) and 96-well plate (Thermo Scientific) were used as received. Preparation of precursor solutions For formulation of tilorone, L-leucine and D-mannitol were dissolved in to deionized water at concentrations of 5 g/L, 10 g/L and 10 g/L respectively. For formulation of pirfenidone, the leucine nanosuspension was prepared by a wet-milling technique.[13] 1 g of DPPC was dissolved in 20 mL of ethanol. Then, 6 g of L-leucine was dispersed in the stabilizer solution. The drug suspension was then pipetted into the milling bowl containing 100 g of milling pearls of zirconium oxide. A planetary milling machine (Pulverisette 7 Premium, Fritsch Co.) was used for the wet-milling process. The grinding was conducted at 1100 rpm for 12 cycles of 3 minutes. After each milling cycle, there was a 15 minute pause to cool down the milling bowls and after every third cycle there was a 1 hour cooling pause. The suspension was separated from the milling pearls and collected after the grinding by sieving. After milling the resulting slurry was diluted and pirfenidone was added to final concentration of 5 g/L for L-leucine and 20 g/L of pirfenidone. Preparation of aerosol particles These precursor solutions were subjected to the aerosol flow reactor method described in detail elsewhere. [14–16] Briefly, the solute droplets were generated by an ultrasonic nebulizer (RBI Pyrosol 7901) and transferred with nitrogen gas (20 L/min) to a stainless steel aerosol reactor where temperature was set to 180°C (±1°C). After the heated zone at the reactor downstream, the aerosol was rapidly diluted and cooled with a large volume of nitrogen gas (20°C, 80 L/min; Reynolds number >3000) in a porous stainless steel tube to avoid the wall deposition of particles and to initiate the nucleation and deposition of amino acid vapor on drug particle surfaces. The solid aerosol particles were then collected by a small-scale cyclone.[17] Particle size distribution was monitored with an electrical low-pressure impactor (ELPI; Dekati Ltd.) where the particles were collected on oiled porous collection stages (Dekati Ltd.) with aerodynamic cut-off diameters from 0.03 to 7.88 µm.

49


Drug Delivery to the Lungs 27, 2016 - Development of Inhalable Drug Formulations for Idiopathic Pulmonary Fibrosis Fine powder aerosolization All the fine powders were stored over silica in a desiccator (at 0-1% of relative humidity) prior to the inhalation experiments. The aerosolization of the powders was studied using a computer-assisted inhalation simulator developed in-house. Its detailed operating principles have been discussed elsewhere [18,19] and its applicability for the aerosolization of powders has also been demonstrated previously [11,18]. Briefly, the inhalation profile is created through the interplay between vacuum and pressurized air gas, controlled by a valve system. Two commercially available inhalers; the multi-dose reservoir-type Easyhaler® inhaler (Orion Pharma) and single-dose capsule Twister™ inhaler (Aptar Pharma); were used to assess the aerosolization profiles of the combination dry powders. The Easyhaler reservoir was filled with ~ 0.5 g of fine drug powder, and for Twister HPMC capsules (Vcaps size 3, Capsugel) were filled with 5.0 ± 0.1 mg/capsule of fine drug powder. The doses from the inhalers were administered as instructed by the providers. Powder emission (10 repetitions) was determined gravimetrically by weighing the inhaler before and after each inhalation. 5 repetitions were discarded before the actual measurements. Pressure drops across the inhalers were adjusted to 2 and 4 kPa, which corresponded to inspiratory flowrates of 40 L/min and 55 L/min for Easyhaler and 43 L/min and 55 L/min for Twister, respectively. The inhalation profiles were fast, i.e. the maximum pressure drops (acceleration 16-18 kPa s−1) and flow rates (acceleration 132.3 L min-1 s−1) were achieved within two seconds, while maintaining these sets for 8 s and then stopped. The dispersed fine particles were collected isokinetically on greased aluminium collection stages of a Berner-type low pressure impactor, BLPI, (a stage aerodynamic cut-off diameters ranging from 0.03 to 15.61 µm) where mass distributions were measured gravimetrically [20]. Mass median aerodynamic diameters (MMAD) and geometric standard deviations (GSD) of the deposited powders were determined according to the formulae:  (mi ln Di )  MMAD  exp  M  

  (m D3 (ln D  ln MMAD ) 2 ½  i i i   GSD  exp   (mi Di3 )  1   

where mi is the mass fraction of particles on the collection stage and M is the sum of mass fractions and is, by definition, unity. Fine particle fractions (FPF, geometric mean diameter Dg ≤ 5.5 µm) were expressed with reference to the emitted dose (ED). Powder composition Particle composition were analyzed by 1H-NMR spectroscopy (Bruker 400 MHz UltraShield NMR) in deuterated water (tilorone sample) and deuterated ethanol (pirfenidone sample). Activity Assays Biological activity of tilorone after aerosol processing was assessed using A549 cell lines with transfected luciferase reporter for TGF-β (CAGA-luc construct) and BMP (BRE-luc construct) signaling routes described elsewhere.[12] 96-well plate was seeded with 20000 cells/well and 15000 cells/well for BRE-luc and CAGA-luc constructs respectively in MEM with 10% FBS, 1% L-glutamine, 1% streptomycin, 1% penicillin and 0.4 mg/ml G418 added and were allowed to attach for 5 hours. Tilorone stock was prepared by dissolving tilorone in sterile milli-Q water. For the final stimulation media 2% tilorone stock was added to media with 5% FBS for final tilorone concentrations of 20 µM, 10 µM, 5 µM or 1 µM. The cells were stimulated for 18 hours before the luciferase activity was measured. TGF-β and BMP-4 were used as positive controls for CAGA-luc and BRE-luc cells respectively. Three independent assays were performed. Biological activity of pirfenidone formulation was assessed by measuring dose dependent inhibition of JL-1 cell proliferation. 96-well plate was seeded with 5000 cells/well in RPMI with 10% FBS and allowed to attach for 24 hours. Stimulation media was prepared by dissolving pirfenidone in milli-Q water and mixing the solution to RPMI with 1% FBS for final pirfenidone concentrations of 750 µg/ml, 500 µg/ml and 250 µg/ml. Cells were incubated for 48 hours after which Quick cell proliferation assay kit II WST reagent was added. The cells were further incubated for 2 hours and shaked for 1 minute before measuring absorbance. Absorbance was measured at 540 nm and 690 nm was used as a reference. Two independent assays were performed. Results and Discussion Tilorone powder contained 29.4% of tilorone, 61.0% of mannitol and 9.6% of L-Leucine measured by NMR. Emissions from Easyhaler were very similar and independent of the pressure over the inhaler. Results are presented in Table 1A. ED was 2.95 mg and 2.99 mg with coefficient of variations of emissions (CVs) of 9.2% and 7.8% for 2 kPa and 4 kPa pressure drops respectively. FPF was 27.7% and 30.4% for 2 kPa and 4 kPa pressure drops respectively. For Twister, ED was 4.76 mg and 4.06 mg with CV of 6.72% and 10.8% respectively and FPF was 21.9% and 33.1% with 2 kPa and 4 kPa pressure drops respectively. The results of the activity studies are presented in Figure 1 and 2. The dose-response curves are very similar between the pure drug and the formulation, indicating that no relevant decomposition took place during the aerosol processing. According to NMR studies, the pirfenidone powder contained 91.8% of pirfenidone, 0.9% of L-leucine and 7.3% of DPPC. Table 1B shows the results of dispersion studies for pirfenidone. When using Easyhaler, ED was 2.34 mg and 2.09 mg with CV of 10.0% and 16.8% for 2 kPa and 4 kPa pressure drops respectively. FPF was 30.7% and 28.8% for 2 kPa and 4 kPa pressure drops respectively. With Twister, ED was 3.11 mg and 4.86 mg with CV of 66.0% and 49.0% for 2 kPa and 4 kPa pressure drops respectively. FPF was 31.5% and 32.4% for 2 kPa and 4 kPa pressure drops respectively.

50


Drug Delivery to the Lungs 27, 2016 – Ville Vartiainen et al. Table 1: The results of the dispersion studies for tilorone (A) and pirfenidone (B) powders. ED is the average emitted dose, FPF is the fine particle fraction (particle size <5 µm) and CV is coefficient of variance of the emitted dose

A Tilorone

ED (mg)

FPF (%)

CV (%)

EH 2 kPa

2.95

27.7

9.18

EH 4 kPa

2.99

30.4

Tw 2 kPa

4.76

Tw 4 kPa

4.06

B Pirfenidone ED (mg)

FPF (%)

CV (%)

EH 2 kPa

2.34

30.7

10

7.78

EH 4 kPa

2.09

28.8

16.8

21.9

6.72

Tw 2 kPa

3.11

31.5

66

33.1

10.8

Tw 4 kPa

4.86

32.4

49

Tilorone powder showed good aerosolization performance. Especially from Easyhaler, aerosolization behaviour is very similar despite the difference in pressure drop over the inhaler. While pirfenidone powder showed more dependence on the pressure drop and worse dose repeatability, the parameters are still well within the acceptable range. With Twister, in general more variation is seen in ED, FPF and CV. With tilorone powder the differences in ED and FPF are opposite between the pressure drops, which helps to balance out the actual delivered dose. However, with pirfenidone powder the dose repeatability was abysmal. Different performance between the inhalers is likely to result from difference in used materials and flow geometry and the results highlight the importance of selecting the optimal inhaler for each dry powder individually.

Figure 1: TGF-β (CAGA) and BMP (BRE) reporter activity for pure tilorone and the dry powder formulation

Figure 2: Cell viability in the proliferation assay for pure pirfenidone and dry powder formulation Conclusion We have successfully formulated tilorone and pirfenidone, and showed that they are active after the aerosol processing. Both of the formulations are performing well with a multidose reservoir type inhaler Easyhaler, but have problems in dose repeatability and pressure drop independency with the single dose capsule type inhaler Twister. We are also working on formulating and conducting activity assays for nintedanib. After all three drugs have been formulated and the formulations have been shown to be biologically active, we will move to in vivo – phase, where their antifibrotic activity is assessed in an experimental model of silica-induced pulmonary fibrosis in mice.

51


Drug Delivery to the Lungs 27, 2016 - Development of Inhalable Drug Formulations for Idiopathic Pulmonary Fibrosis

References 1.

Wells, A. U. The revised ATS/ERS/JRS/ALAT diagnostic criteria for idiopathic pulmonary fibrosis (IPF)-practical implications. Respir. Res. 14 Suppl 1, S2 (2013).

2.

Atkins, C. P., Loke, Y. K. & Wilson, A. M. Outcomes in idiopathic pulmonary fibrosis: A meta-analysis from placebo controlled trials. Respir. Med. 108, 376–387 (2014).

3.

FDA. Novel New Drugs 2014. Center for Drug Evaluation and Research (2015). Available at: http://www.fda.gov/Drugs/DevelopmentApprovalProcess/DrugInnovation/ucm429247.htm.

4.

EMA. Esbriet. European public assessment reports (2011). Available at: http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/medicines/002154/human_med_ 001417.jsp&mid=WC0b01ac058001d125.

5.

EMA. Vargatef. European public assessment reports (2014). Available at: http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/medicines/002569/human_med_ 001822.jsp&mid=WC0b01ac058001d124.

6.

Noble, P. W. et al. Pirfenidone in patients with idiopathic pulmonary fibrosis (CAPACITY): two randomised trials. Lancet 377, 1760–1769 (2011).

7.

Richeldi, L. et al. Efficacy and Safety of Nintedanib in Idiopathic Pulmonary Fibrosis. N. Engl. J. Med. 370, 2071–2082 (2014).

8.

Vartiainen, V., Bimbo, L. M., Hirvonen, J., Kauppinen, E. I. & Raula, J. Drug permeation and cellular interaction of amino acid-coated drug combination powders for pulmonary delivery. Int. J. Pharm. 504, 89–97 (2016).

9.

Raula, J. et al. Coated particle assemblies for the concomitant pulmonary administration of budesonide and salbutamol sulphate. Int. J. Pharm. 441, 248–254 (2013).

10.

Raula, J. et al. Investigations on the humidity-induced transformations of salbutamol sulphate particles coated with L-leucine. Pharm. Res. 25, 2250–2261 (2008).

11.

Raula, J., Lähde, A. & Kauppinen, E. I. Aerosolization behavior of carrier-free l-leucine coated salbutamol sulphate powders. Int. J. Pharm. 365, 18–25 (2009).

12.

Leppäranta, O., Tikkanen, J. M., Bespalov, M. M., Koli, K. & Myllärniemi, M. Bone Morphogenetic Protein–Inducer Tilorone Identified by High-Throughput Screening Is Antifibrotic In Vivo. Am. J. Respir. Cell Mol. Biol. 48, 448–455 (2013).

13.

Peltonen, L. & Hirvonen, J. Pharmaceutical nanocrystals by nanomilling: Critical process parameters, particle fracturing and stabilization methods. Journal of Pharmacy and Pharmacology 62, 1569–1579 (2010).

14.

Eerikäinen, H., Watanabe, W., Kauppinen, E. I. & Ahonen, P. P. Aerosol flow reactor method for synthesis of drug nanoparticles. Eur. J. Pharm. Biopharm. 55, 357–360 (2003).

15.

Lähde, A., Raula, J. & Kauppinen, E. I. Simultaneous synthesis and coating of salbutamol sulphate nanoparticles with l-leucine in the gas phase. Int. J. Pharm. 358, 256–262 (2008).

16.

Raula, J., Kuivanen, A., Lähde, A. & Kauppinen, E. I. Gas-phase synthesis of l-leucine-coated micrometer-sized salbutamol sulphate and sodium chloride particles. Powder Technol. 187, 289–297 (2008).

17.

Zhu, Y. & Lee, K. W. Experimental study on small cyclones operating at high flowrates. J. Aerosol Sci. 30, 1303–1315 (1999).

18.

Kurkela, J. A., Kauppinen, E. I., Brown, D. P., Jokiniemi, J. K. & Muttonen, E. A New Method and Apparatus for Studying Performance of Inhalers. Respir. Drug Deliv. VIII (2002).

19.

Kauppinen, E., Kurkela, J., Brown, D., Jokiniemi, J. & Mattila, T. Method and apparatus for studying aerosol sources. (2002).

20.

Hillamo, R. E. & Kauppinen, E. I. On the Performance of the Berner Low Pressure Impactor. Aerosol Science and Technology 14, 33–47 (1991).

52


Drug Delivery to the Lungs 27, 2016 – Martin Jetzer et al. Investigating the Effect of the Force Control Agent Magnesium Stearate in Fluticasone Propionate Dry Powder Inhaled Formulations with Single Particle Aerosol Mass Spectrometry (SPAMS) Martin Jetzer1,2, Bradley Morrical1, Marcel Schneider1 & Georgios Imanidis2 Novartis Pharma AG - Development, Novartis Campus, Lichtstrasse 35, 4056 Basel, Switzerland Department of Pharmaceutical Sciences, University of Basel, Klingelbergstrasse 50, 4056 Basel, Switzerland 1

2

Summary A number of researchers are investigating the role of ternary force control agents (FCA’s) in pharmaceutical inhaled products with different analytical techniques. Various application processes such as mechanofusion and particle smoothing have been published to apply FCAs on carrier particles [1–3]. Nevertheless, a thorough mechanistic understanding of the role of FCA’s in dry powder inhalation (DPI) formulations is still largely missing. In this study, the effect and impact on the formulation containing the FCA MgSt has been investigated in a FP, lactose based DPI formulation. Two different mixing methods of lactose and MgSt have been examined. The aerosol performance in terms of APSD and FPF of the DPI formulations was evaluated with cascade impaction studies with the NGI and analyzed with SPAMS. High-shear blending of the lactose carrier together with the FCA magnesium stearate (MgSt) lead to a shift of the aerodynamic particle size profile (APSD) of the active drug fluticasone propionate (FP) and the total number of particles to a higher number of smaller particles. The blending method applied to pre-blend the excipients strongly impacts the APSD and fine particle fraction (FPF) of FP. Both SPAMS and Next Generation Impactor (NGI) confirmed an increase of FPF for FP when adding MgSt to the formulation by high-shear mixing. Low-shear mixing of the excipient-blend did not increase the pharmaceutical performance of FP. The analytical techniques NGI and SPAMS successfully demonstrated that it is possible to distinguish changes in the formulation of DPI powders blended with different amounts of MgSt. Introduction Dry powder inhaled (DPI) products commonly consist of large carrier particles (usually alpha-lactose monohydrate) and a relatively low amount (0.05–10%) of active pharmaceutical ingredient (API) having a particle size typically below 6 μm. Blending lactose together with a ternary force control agent (FCA) such as MgSt prior adding the API has been shown to modify the performance of pharmaceutical inhaled products [3,4]. For this study, active blends with the cohesive drug FP [5] have been manufactured between 0 and 1% (w/w) MgSt-content in the excipient blend by high-shear and low-shear mixing the excipient-blend. A formulation of FP with lactose untreated (0% MgSt) was used as a reference. The effect of the FCA MgSt has been investigated with SPAMS and NGI to assess the aerosol performance of FP and excipient. Another aim of this study was to examine the influence of the two different mixing methods on applying the FCA on lactose and the resulting effect on aerosol performance of FP DPI formulations. For the evaluation of the aerosol performance of inhaled pharmaceutical products, the current state of the art is the use of the NGI (Copley Scientific, UK), typically with chemical analysis using high performance liquid chromatography (HPLC), to characterize the APSD and FPF of the active substance. But for the total fine particle assessment, a special NGI method has to be developed and carried out in order to determine the distribution of the excipients (e. g. lactose carrier and FCA). SPAMS is an analytical technique where the aerodynamic diameters and chemical compositions of many individual aerosol particles are determined in real-time. The SPAMS technique and its advantages have been published in several articles recently [6,7]. SPAMS is a promising method to explore effects in DPI formulations due to the fact that not only the APSD of the active drug in the formulation is determined. The total number of particles (excipient and API) is determined simultaneously and also the additional information of co-associations between two APIs. Experimental methods DPI formulation preparation: For the high-shear manufactured excipient-blends, inhalation grade lactose (Respitose® ML001, DFE Pharma) and MgSt (Peter Greven, 0-1 % w/w) were blended in a Collette MicroGral (GEA Pharma Systems) for 16 minutes at 1400 rpm. The low-shear manufactured excipient-blend was mixed for 16 min at 34 rpm in a Turbula mixer (Willy A. Bachofen AG). The excipient-blend together with the active drug FP was sieved through a sieve with a mesh size of 250 μm in a vessel. The resulting drug-excipient blend was then mixed in a Turbula mixer at 34 rpm for 16 min. The final powder was filled into size 3 HPMC capsules containing dosage strength of 200 mcg FP per capsule. The capsules were stored at 25°C and 55% relative humidity for a period of 7 days prior to the experiments. FP DPI analysis by NGI: The deposition of the dry powder formulations was investigated using an NGI equipped with a pre-separator. All measurements were conducted at 55% RH and ambient temperature (23 ± 2ºC). The cup trays of the impactor were coated with Brij ® reagent (1% v/v solution of Glycerol in Ethanol). The capsules were actuated in a Breezhaler® inhalation device at a flow rate of 90 L/min for 2.7 sec. The amount of powder deposited on the different collection cups was recovered by extracting each cup with solvent. The remaining powder in the capsule and the powder deposited in the throat and the pre-separator was also collected. After dissolution of the particles, HPLC analysis was performed. In this study the mean of three individual determinations was taken for a given NGI result.

53


Drug Delivery to the Lungs 27, 2016 - Investigating the Effect of the Force Control Agent Magnesium Stearate in Fluticasone Propionate Dry Powder Inhaled Formulations with Single Particle Aerosol Mass Spectrometry (SPAMS) FP DPI testing by SPAMS: A Livermore Instruments SPAMS 3.0 was used for this study. The experimental setup of the SPAMS instrument and sample testing has been described earlier by Morrical et al. [6]. For the calibration of the SPAMS instrument in the region of 0.1-10 μm, polystyrene (PLS) microspheres (Thermo-Fisher Scientific) were used. The dry powder formulations samples were acquired using a Breezhaler ® inhalation device. The induction port was connected to a pre-separator (Copley Scientific, UK), which was filled with 15 mL of water and then connected to a 4L relaxation chamber. This assembly of pre-separator and sampling chamber was then fitted to the SPAMS inlet. The primary purpose of the pre-separator was to filter out coarse lactose carrier particles (>10 µm) so that they cannot enter the SPAMS instrument to prevent clogging of the inlet interface. The sampling chamber allows for dilution of the particles and flow rate matching. The actuation of the DPIs was made with a simultaneous 2.7 seconds draw of air into the sampling chamber, at a flow rate 90 L/min (i.e. 4 liters of air drawn). Results and discussion Figure 1a) shows a comparison of SPAMS APSD histograms of total particles (API, fine lactose and MgSt) for three different FP blends with lactose untreated, lactose with 0.1 and 0.5% MgSt-content (w/w; excipient-blend high-shear manufactured), respectively. By adding MgSt to the formulation, the APSD profile shifts to a higher number of smaller particles with a higher number of particles in the size range from 0-1.9 μm diameter. By increasing the MgSt-content from 0.1 to 0.5% (w/w), the number of smaller particles increases even more and the size distribution shift is more pronounced.

Figure 1 - SPAMS Aerodynamic particle size distribution histograms of three FP blends with lactose untreated, 0.1 and 0.5% MgSt (w/w), respectively (high-shear manufactured). a) The plot shows the total number of particles measured (active substance FP and excipients). b) The plot shows the distribution of the active substance FP (no excipients). The size bins were scaled in order to have a comparison between the different formulations.

The APSD histogram of particles containing FP in the two high-shear manufactured formulations and the reference formulation without MgSt is shown in Figure 1b). The size bins in Figure 1b) were scaled in order to compare the three formulations with each other. It can be seen that there is a significant increase in number of FP particles with a small particle size in the range from 0-1.9 μm for the formulations containing 0.1 and 0.5% MgSt, respectively. Interestingly, the same shift observed in the total number of particles (Figure 1a) was also detected for the API FP following the same pattern. SPAMS is indicating that the high-shear manufactured blends containing the FCA MgSt have a higher dispersibility compared to the reference formulation without MgSt. In Figure 2, a series SPAMS APSD histograms of three FP blends is compared: A high-shear manufactured lactose 0.1% MgSt excipient-blend, a low-shear manufactured lactose 0.1% MgSt excipient-blend and the reference formulation without MgSt. It is clearly evident that there are significant differences between the highand low-shear manufactured blends. While the APSD of the high-shear manufactured blend shifts to a higher number of smaller particles (0-1.9 μm), the APSD of the low-shear formulation seems to have a slight shift to a larger particle size. A much lower number of particles with a size below 1.9 μm are present. The increase in number of particles in the region of 2.1 μm is probably due to MgSt agglomerates. The histogram indicates that low-shear mixing of the excipient-blend does not increase the dispersibility of the manufactured FP formulation.

Figure 2 - SPAMS Aerodynamic particle size distribution histogram of three FP blends with lactose untreated, 0.1% MgSt (w/w) high-shear and low-shear manufactured, respectively. The plot shows the total number of particles measured (active substance FP and excipients).

54


Drug Delivery to the Lungs 27, 2016 – Martin Jetzer et al. The fine particle fraction (FPF) of FP (particles with a diameter <5 μm) calculated as a percentage of total recovery, is described in Table 1. The FPF increases with the amount of added MgSt for the high-shear manufactured formulations. The tested formulations with amounts of 0.1 and 0.5% (w/w) MgSt increased about 2 and 8% in FPF, respectively. The FPF did not increase for the formulations manufactured with the low-shear excipient-blend. Formulation

FPF [%]

StDev [%]

FP Lactose untreated FP Lactose 0.1% MgSt (high-shear)

24.4 26.3

0.2 0.2

FP Lactose 0.5% MgSt (high-shear)

32.2

0.8

FP Lactose 0.1% MgSt (low-shear)

22.0

1.2

FP Lactose 0.5% MgSt (low-shear)

22.3

0.5

Table 1 - Fine particle fraction (% of declared content) obtained by NGI

v

v

v

v

Figure 3 - ToF-SIMS images of lactose 0.1% MgSt (w/w) high-shear (A) and low-shear (B) manufactured Red: Ion signal for Mg+ (from MgSt) Green: Ion signal for lactose (C3H5O2+) Orange: Overlay of MgSt (red) and lactose (green) ion signals

Characterization of changes on the lactose carrier surface or carrier coating is crucial to examine and difficult to achieve because of particle size and shape and the extremely thin coating layer. Conventional techniques such as Raman spectroscopy penetrate the coating layer and measure both, the coating layer and the host particles. Secondary ion mass spectrometry (ToF-SIMS) images showed ion signals for MgSt all over the measured area in the high-shear manufactured excipient blend (Figure 3A). Therefore, we conclude that MgSt must be evenly distributed on the lactose surface (Figure 3A: Overlay of MgSt and lactose). A thin layer of MgSt covers the lactose surface. While in the low-shear manufactured excipient blend, MgSt agglomerates (yellow circles) were detected by ToF-SIMS imaging as small intense spots (Figure 3B). Low-shear blending did not induce any surface coating of the lactose with MgSt. The overlay of MgSt and lactose shows only some intense spots with ion signals indicating MgSt agglomerates. The results collected with SPAMS and NGI indicate that there is higher dispersibility of fine particles with an aerodynamic diameter below 5 μm when adding the FCA MgSt to the excipient blend by high-shear mixing. Apparently, the MgSt is modifying the lactose carrier surface in such way that attractive forces, such as van der waals or electrostatic forces [8], between drug and carrier and also between fine excipient and coarse excipient are much reduced. This leads to easier detachment of drug and also fine excipient from the coarse carrier. The likelihood to form smaller agglomerates between drug and fine lactose increases. As a consequence, the amount of liberated drug substance increases [4]. In the investigated DPI blends, the amount of MgSt added to the blend also played an important role. Introducing more MgSt (0.5% compared to 0.1% w/w) into the system resulted in further increase of the aerosol performance in terms of FPF for the formulation when the excipient-blend was manufactured by high-shear mixing. A higher amount of MgSt seems to cover more of the lactose surface area and thus decreases interaction forces even more when applied on the lactose by high-shear mixing. It can be hypothesized that above a certain concentration of MgSt in the excipient blend, different effects may influence the aerosol performance. Various competing effects such as the saturation of activated areas on the carrier surface or the formation of drug–fines agglomerates for easier lift-off could offer an explanation for the observed behaviour [9-11]. However, no conclusive explanation can be given at this point.

55


Drug Delivery to the Lungs 27, 2016 - Investigating the Effect of the Force Control Agent Magnesium Stearate in Fluticasone Propionate Dry Powder Inhaled Formulations with Single Particle Aerosol Mass Spectrometry (SPAMS) Conclusions The blending method applied to pre-blend the excipients strongly impacts the APSD and FPF of FP. High-shear blending of the lactose carrier together with the FCA MgSt leads to a shift of the APSD profile of the active drug FP and the total number of particles to a higher number of smaller particles overall as detected with SPAMS. There is a better detachment of small particles overall and possibly co-associations of FP with fine MgSt or very fine lactose is formed during the actuation of the capsule. Both SPAMS and NGI showed an increase of FPF for FP when adding MgSt to the excipient blend by high-shear mixing. Low-shear mixing of the excipient-blend did not significantly increase the performance of FP in our formulations. Images from ToF-SIMS indicate that the distribution of MgSt in the excipient blend strongly depends on the blending method. The presence of MgSt as a thin and homogenous layer on the lactose surface in the high-shear blend appears to change the interaction forces between API and lactose carrier. The altered surface properties influence the interaction forces between API and carrier, which leads to easier detachment of FP from the lactose carrier together with higher dispersbility of total fines. For future work it would be interesting to explore, if other FCAs can have a similar effect with the used excipient high-shear mixing method. It was demonstrated with SPAMS that it is possible to distinguish changes in the formulation of DPI powders. SPAMS provided valuable additional information about the dispersion behavior of our formulations. APSD histograms of total number of particles and particles containing FP could be determined for comparison. Using an upgraded SPAMS setup with a desorption and ionization laser firing at a different wavelength, SPAMS could potentially detect co-associations not only between APIs, but also between API and excipients (lactose and FCA) in future. Acknowledgements The authors would like to thank David Fergenson (Livermore Instruments Inc., USA) for technical support with the SPAMS instrument and Rowena Crockett (EMPA, Switzerland) for ToF-SIMS imaging. References 1.

Ferrari F, Cocconi D, Bettini R, et al. The surface roughness of lactose particles can be modulated by wet-smoothing using a high-shear mixer. AAPS PharmSciTech. 2004.

2.

Kumon M, Suzuki M, Kusai A, Yonemochi E, Terada K. Novel approach to DPI carrier lactose with mechanofusion process with additives and evaluation by IGC. Chem Pharm Bull (Tokyo). 2006.

3.

Thalberg K, Ă…slund S, Skogevall M, Andersson P. Dispersibility of lactose fines as compared to API in dry powders for inhalation. Int J Pharm. 2016.

4.

Guchardi R, Frei M, John E, Kaerger JS. Influence of fine lactose and magnesium stearate on low dose dry powder inhaler formulations. Int J Pharm. 2008.

5.

Le VNP, Robins E, Flament MP. Agglomerate Behaviour of Fluticasone Propionate within Dry Powder Inhaler Formulations. Eur J Pharm Biopharm. 2012

6.

Morrical BD, Balaxi M, Fergenson D. The on-line analysis of aerosol-delivered pharmaceuticals via single particle aerosol mass spectrometry. Int J Pharm. 2015.

7.

Susz A, Morrical BD, Fergenson D. Real time determination of APSD profiles of mono and combination dry powder inhalation products using single particle aerosol mass spectrometry (SPAMS 3.0). Respiratory Drug Delivery Europe. 2015.

8.

Li Q, Rudolph V, Peukert W. London-van Der Waals Adhesiveness of Rough Particles. Powder Technology. 2006.

9.

Young P, Edge S, Traini D, Jones MD, Price R, El-Sabawi D, Urry C, Smith C. The Influence of Dose on the Performance of Dry Powder Inhalation Systems. Int Pharm. 2005.

10.

Begat P, Morton D, Staniforth JN, Price R. The Cohesive-Adhesive Balances in Dry Powder Inhaler Formulations II: Influence on Fine Particle Delivery Characteristics. Pharm Res. 2004.

11.

Louey M, Stewart P. Particle Interactions Involved in Aerosol Dispersion of Ternary Interactive Mixtures. Pharm Res. 2002.

56


Drug Delivery to the Lungs 27, 2016 - Mats Hertel et al. Influence of blender type on the performance of ternary dry powder inhaler formulations 1

2

2

2

2

Mats Hertel , Eugen Schwarz , Eva Maria Littringer , Mirjam Dogru , Sabine Hauptstein , 3 1 Hartwig Steckel & Regina Scherließ 1

Department of Pharmaceutics and Biopharmaceutics, Kiel University, Grasweg 9a, 24118 Kiel, Germany 2

Meggle Excipients and Technology, Megglestr. 6 - 12, 83512 Wasserburg am Inn, Germany 3

Deva Holding AS, Istanbul, Turkey

Summary The importance of the right blending order and the mixing parameters for ternary mixtures prepared with a high shear mixer was shown in a previous study. This project was applying a Design of Experiment (DoE) to elucidate whether the gained findings are specific for a high shear mixer or if they can also be attributed to a low shear tumble blender. ® ® For this purpose, ternary mixtures were prepared with a Turbula tumble blender and a Picomix high shear mixer using two different blending orders and the same number of total rotations of the blender. Afterwards, the blends were ® tested with an inhaler device (Novolizer ) by impaction analysis using the Next Generation Pharmaceutical Impactor (NGI). Fine particle fraction (FPF), fine particle dose (FPD) and mass median aerodynamic diameter (MMAD) were then compared with the factors blender type, blending order and rotations of the blender. Results exhibited that all factors impacted on the FPF and FPD, with the lowest effect arising from the blender type. Blending order findings and SEM pictures proved the saturation of active sites and the function of fines as a buffer during blending. Overall, improved knowledge of the exact mechanism about how fines influence the aerodynamic performance, in particular for the blending process, could be illustrated in this work. Introduction In order to facilitate powder handling, micronised active pharmaceutical ingredients (API) are often mixed with coarser ® lactose carriers. Typically, these adhesive mixtures are prepared by low shear tumble blending (e.g., Turbula blender). However, the use of high shear blenders attracts more and more attention due to more intense blending and [1] [2] shorter blending time . Several studies show the importance of process parameters in low shear and high shear [3] on the performance of binary dry powder inhalation (DPI) formulations. Some other studies focus on the mixing [4] influence of low shear mixing process parameters on the performance of ternary mixtures between API, carrier and excipient fines, which are frequently added to increase the fine particle fraction (FPF). The importance of the right blending order and the mixing parameters for ternary mixtures prepared with a high shear blender had been shown in [5] a previous study . One hypothesis is that the blending order affects the saturation of active sites. Fines occupy areas [6] of the surface which are more adhesive, leaving only weaker binding sites available for the drug particles to bind to . Grasmeijer et al. exhibited that fines coarser than the drug particles can serve as a buffer between colliding carrier [7] particles and protect drug particles from press-on forces during blending . The blending parameters are important, as there is an increase of press-on forces of the API particles on to the carrier surface with higher rotation speed and extended blending times. Hence, it appears to be obvious that the choice of blender (high-shear or low-shear blender) also influences aerodynamic performance of the resulting mixture. However, if blending is normalized to total rotation number rather than blending time or rotation speed, this might be different. To prove whether the influences of fines ® as discussed earlier are also applicable in a Turbula blender, both blenders were compared at fixed total rotations in this study. Material and methods Blender type, blending order and the total number of rotations of the blender were chosen as factors for a design of experiments (DoE). A D-optimal quadratic setup with three factors was selected for this study. Two center points ® ® (blender = Turbula , blending order = BO ([C + F] + [API]), number of rotations (Rot) = 3000 and blender = Picomix , blending order = BO ([C + API] + [F]), Rot = 3000) were repeated three times to finally gain 16 runs. The model was evaluated by neglecting insignificant terms in a backward regression. Fine particle fraction (FPF), fine particle dose (FPD), mass mean aerodynamic diameter (MMAD), homogeneity of the powder blends and recovery of the API were set as response for this study. Statistical analysis was carried out with Modde software (Version 10.0, Umetrics AB, Umea, Sweden) by using the multiple linear regression (MLR) method. ®

®

Materials: Lactose monohydrate (InhaLac 70 ) was used as a carrier, InhaLac 400 as fines (all Meggle Excipients and Technology, Wasserburg, Germany) and micronised budesonide (Farmabios, Gropello Cairoli, Italy) as a model drug. Particle sizes of the materials are given in Table 1.

57


Drug Delivery to the Lungs 27, 2016 - Influence of blender type on the performance of ternary dry powder inhaler formulations Material InhaLac 70 InhaLac 400 Budesonide

d10 [µm] ± SD 126.9 ± 0.6 1.0 ± 0.0 0.4 ± 0.0

d50 [µm] ± SD 218.7 ± 1.0 6.3 ± 0.0 1.5 ± 0.0

d90 [µm] ± SD 302.4 ± 1.3 23.9 ± 0.4 3.5 ± 0.0

Table 1 - Particle size distribution of the mixture components (average with standard deviation (SD); n = 3) ®

Preparation of model DPI formulations: A low shear tumble blender (Turbula , Typ T2C, Willy A. Bachhofen AG ® Maschinenfabrik, Basel, Switzerland) and an Alpine Picoline equipped with the Picomix high shear mixer module (Hosokawa Alpine, Augsburg, Germany) were used to prepare adhesive mixtures. Budesonide concentration was set ® to 1.5 wt% to get a delivered dose of approximately 200 µg with the Novolizer (Astellas Pharma GmbH, München, Germany). 7.5 wt% of fines were added to the blends in two different blending orders (Table 2). The rotation speed ® ® was set to 500 rpm for the Picomix and 90 rpm for the Turbula . In order to reach total rotations of 1000 to 5000 the ® blending time for every blending step was varied between 5:33 and 27:47 min for the Turbula and 1:00 to 5:00 min ® ® for the Picomix . Budesonide and InhaLac 400 were sieved through a 180 µm sieve prior blending to remove agglomerates. A sandwich-weighing-method was used at a batch size of 30 g for blending. After every blending step, the mixtures were passed through a 355 µm sieve to destroy agglomerates. Homogeneity was assessed for all formulations by measuring the budesonide content of ten randomly drawn samples by an high performance liquid chromatography (HPLC) method. The requirements for homogeneity of blends were set to a relative standard deviation (RSD) of less than 5 % and a recovery of 95 %. Each formulation was stored at ambient conditions for a minimum of one week before NGI analysis. Determination of in vitro fine particle delivery: Aerodynamic particle size distribution was determined using an NGI equipped with a critical flow controller (both Copley Scientific, Nottingham, UK) and a vacuum pump (Erweka, ® Heusenstamm, Germany) set at a flow rate of 78.3 L/min (resulting in a pressure drop of 4 kPa across the Novolizer ). To ensure to be above the limit of detection for all samples for drug quantification, 5 actuations were delivered to the ® NGI for each experiment. To avoid particle bouncing, a stage coating (Brij 35, glycerol, ethanol) was applied onto the pre-separator and each stage of the impactor before analysis. Each blend was measured in triplicate. Calculation of FPD, FPF and MMAD was performed using the CITDAS 3.0 software (Copley Scientific, Nottingham, UK). Scanning electron microscopy (SEM): The morphology of the substances and blends was examined using SEM. The powder was fixed on a carbon sticker and coated with gold using a BAL-Tec SCP 050 Sputter Coater (Leica Instuments, Wetzlar, Germany). The samples were then visualised with a Zeiss Ultra 55 plus (Carl Zeiss NTS GmbH, Oberkochen, Germany) using the SE-2 detector and a working voltage of 2 kV. Exp No

Exp Name

Run Order

Blender

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

N1 N2 N3 N4 N5 N6 N7 N8 N9 N10 N11 N12 N13 N14 N15 N16

15 2 5 4 7 12 9 11 14 10 13 6 8 3 1 16

Turbula Turbula Turbula Turbula Turbula Picomix Picomix Picomix Turbula Turbula Turbula Picomix Picomix Picomix Picomix Picomix

Factors Blending Order, BO [C + F] + [API] [C + F] + [API] [C + F] + [API] [C + F] + [API] [C + F] + [API] [C + F] + [API] [C + F] + [API] [C + F] + [API] [C + API] + [F] [C + API] + [F] [C + API] + [F] [C + API] + [F] [C + API] + [F] [C + API] + [F] [C + API] + [F] [C + API] + [F]

Rotations, Rot 1000 5000 3000 3000 3000 1000 5000 3000 1000 5000 3000 1000 5000 3000 3000 3000

FPF [%] 54.7 46.7 49.6 50.6 50.1 57.4 49.1 52.8 46.4 30.2 38.3 51.6 30.2 39.7 38.7 39.0

Responses MMAD HomoFPD [µg] [µm] geneity [%] 100.3 2.13 1.2 84.9 2.10 1.8 90.4 2.10 1.3 92.2 2.00 1.0 90.9 2.05 1.2 106.0 2.11 1.8 90.5 1.99 2.4 97.7 2.04 1.7 84.0 1.90 1.1 54.5 2.23 0.6 70.0 2.03 1.2 93.8 1.92 0.9 54.7 2.10 0.8 72.7 1.97 1.2 70.6 1.89 1.7 71.4 1.91 2.5

Recovery [%] 97.9 97.9 95.9 99.6 96.6 99.6 98.2 97.3 99.3 95.4 96.7 98.9 96.4 96.3 98.8 96.2

Table 2 – Preparation parameters and results for the various formulations tested. In the scope of DoE, C = carrier, API = budesonide and F = fines

Results and discussion ®

®

The median particle size of InhaLac 70 (Table 1) was around 219 µm. InhaLac 400 shows larger particles and a broader PSD than budesonide. This is very important for buffer effects during blending, because fine lactose particles

58


Drug Delivery to the Lungs 27, 2016 - Mats Hertel et al. being coarser than the drug particles can act as a buffer between colliding carrier particles and protect the API from press-on forces. Furthermore, fines are small enough to fill up the indentions of the large carrier (Figure 1). A

B

Figure 1 – Visualisation of representative interactive mixtures with InhaLac 70® blended with API first, then blended with fines (A) before and (B) after dispersion with the Novolizer®

Homogeneity and Recovery: No significant trends were observed. After mixing, all blends showed adequate homogeneity (all RSD < 2.47 %) and recovery (recovery > 95.36 %). However, the recovery decreased for high rotation numbers slightly. This was due to an increased deposition of drug on the walls and especially the rotor in case of the high-shear blender. In vitro fine particle delivery (Figure 2): The quality of the resulting model was excellent (percent of the variation of the response R² = 0.99, prediction quality Q² = 0.97, reproducibility RP = 0.99, model validity p = 0.22), the poor model validity is a result of an artificial lack of fit. Because of the high reproducibility the pure error was very low and not representative for the true experimental error. That resulted in significant (but not real) lack of fit. Blender, blending order and number of rotations were identified as significant factors on the FPF. The number of rotations is the most important factor influencing the FPF (-6.72 ± 0.39 %). Increased rotations led to an increase of press-on forces which caused a reduction of FPF irrespective of blender type. This reduction is more prominent, when carrier and API were blended first (- 6.33 ± 0.28 %) and the fines were added in the second blending step. When the fines were added first (6.33 ± 0.28 %), they operate as buffer between colliding carrier particles and protect the smaller drug particles from press-on forces. Furthermore, fines fill up the indentions of the carrier (Figure 2 and 3). No replacement of fines and drug can be observed for prolonged blending times. The FPF of the blends decreased considerably with higher rotation numbers. The influence of the blender is comparatively low (Turbula = -1.07 ± 0.29 %, Picomix = 1.07 ± 0.29 %). It is evident that the saturation of active sites and the reduction of press-on forces played an important role, while the used blender was less important. However, the high shear blender was advantageous as it achieved the same or even slightly better fine particle fractions with shorter mixing times. Evaluation of FPD and MMAD provided comparable results (data not shown).

Figure 2 - Coefficient plot (scaled and centred). Plotted are the regression coefficients of the FPF with confidence intervals (Confidence level 0.95)

59


Drug Delivery to the Lungs 27, 2016 - Influence of blender type on the performance of ternary dry powder inhaler formulations

Figure 3 – Scheme of blending order effects. (A): When drug is blended with carrier first, it has the opportunity to adhere to the strong binding sites (indentions) first. The high amount of fines led to complete filling of the indentions and thereby a transmission of press-on forces on to the drug. When fines were blended with the carrier first, they had the first opportunity to adhere to the strong binding sites, forcing the drug to be attached to weaker binding sites. Due to their size the fines protect the drug from press-on forces.

Scanning electron microscopy: SEM image (Figure 1A) showed high loading of the carrier surface with agglomerates of drug particles and fines. Most particles were preferably attached to the indented parts of the carrier whereas the smoother parts showed less loading. After dispersion it was evident that remaining particles were preferentially deposited in the indentions of the carrier (Figure 1B). API and fines remained in the indentions supporting the active sites theory and giving an explanation for blending order results. Conclusion The results showed that from all investigated factors the blender type has the lowest impact on the FPD and FPF. From blending order findings and SEM pictures the saturation of active sites and the function of fines as a buffer during blending can be demonstrated. This work provides a deeper insight into the exact mechanisms of how fines influence the aerodynamic performance, even by using different blending processes. References 1

Willetts J: Investigation into the Effects of High Shear Blending and Storage on Powders for Inhalation (PhD thesis) The School of Chemical Engineering, University of Birmingham 2012 Online available at:http://etheses.bham.ac.uk/3325/1/Willetts12PhD.pdf

2

Grasmeijer F, Hagedoorn P, Frijlink HW, de Boer HA: Mixing time effects on the dispersion performance of adhesive mixtures for inhalation, Plos One 2013, 8(7): 1-18.

3

Balducci AG, Steckel H, Guarneri F, Rossi A, Colombo G, Sonvico F, Cordts E, Bettini R, Colombo P, Buttini F: High shear mixing of lactose and salmeterol xinafoate dry powder blends: Biopharmaceutic and aerodynamic performances, Journal of Drug Delivery Science and Technology 2015, 30: 443-49.

4

Jones MD, Santo JGF, Yakub B, Dennison M, Master H, Buckton G: The relationship between drug concentration, mixing time, blending order and ternary dry powder inhalation performance. International Journal of Pharmaceutics 2010, 391: 137-47.

5

Hertel M, Schwarz E, Littringer E M, Dogru M, ScherlieĂ&#x; R, Steckel H: Influence of fines on commercial lactose carriers and their dry powder inhalation performance, Drug Delivery to the Lungs 2015

6

Jones M D, Price R: The influence of fine excipient particles on the performance of carrier-based dry powder inhalation formulations, Pharm. Res. 2006; 23: pp 1665-1674

7

Grasmeijer F, Lexmond A J, van den Noort M, Hagedoorn P, Hickey A J, Frijlink H W, de Boer A H: New mechanisms to explain the effects of added lactose fines on the dispersion performance of adhesive mixtures for inhalation, PLoS One 2014; 9: pp 1-11.

60


Drug Delivery to the Lungs 27, 2016 – Tim Noakes,etTim al Noakes, Stuart Corr The Future of Propellants for pMDIs Tim Noakes, Stuart Corr Mexichem, The Heath Business & Technical Park, Runcorn, Cheshire, WA7 4QX, United Kingdom Summary From its first commercial appearance in the 1950’s, the chlorofluorocarbon (CFC) propelled MDI enjoyed steady growth until the late 1980s, when the discovery of stratospheric ozone depletion forced a fundamental change to an alternative class of propellants, the hydrofluoroalkanes (HFAs). The conversion has been challenging, and it is only recently that the last CFC-propelled MDIs have been manufactured. pMDI remains a core treatment methodology for a number of inhaled medicines and continues to see growth worldwide. Having successfully met the ozone-depletion challenges and negotiated the HFA transition, additional environmental regulations relating to global warming potential, are coming to the fore. Although provision for continued pMDI usage has been made in the relevant environmental legislation, work is on-going to minimise the potential environmental impact associated with pMDI use including an appraisal of potential new propellant options. The dosage form The first pMDI appeared in the mid 1950’s, made possible by the family of the then new CFC aerosol propellants, some of the members of which are shown in Table 1. All of these molecules shared the common properties of high liquid density, non-flammability and benign toxicology, reflecting their relatively low degree of chemical reactivity. Species

Formula

B.Pt (oC)

Saturated Vapour Pressure (bara,20oC)

CFC 11 CFCl3 23.7 0.88 CFC 12 CF2Cl2 -29.8 5.66 CFC 114 CF2ClCF2Cl 3.5 1.80 HFA 134a CF3CFH2 -26.2 5.72 HFA 227ea CF3CFHCF3 -16.5 3.89 Table 1: Properties of CFCs and HFAs used in pMDIs[1]

Specific gravity (g/cc, 20oC)

Ozone Depletion Potential

1.49 1.33 1.47 1.23 1.41

1 1 1 0 0

Global Warming Potential (CO2=1) 4660 10800 8590 1300 3350

The CFC propelled MDI Medihaler-iso™ was the first of these dosage forms, produced by Riker Laboratories Inc.; an MDI based presentation of isoprenaline first introduced in 1956 [2]. Interestingly, it was a formulation containing CFC 12, ethanol and drug. Ethanol was later rarely used in CFC formulations, but the transition to HFA propellants has seen greater use in order to achieve satisfactory formulation performance Glaxo set the scene for what became the most common way of formulating and filling these aerosols when they brought their salbutamol-based Ventolin™ inhalation relief formulation to the market in 1968 [3]. It was a two-stage process where the drug was first slurried in a small quantity of the higher boiling CFC 11 and the concentrate filled to open aerosol cans, onto which the valves were then crimped. The second stage was to ‘gas’ the aerosol at the next filling point, with the lower-boiling propellant CFC 12. It is fair to say that between the late 1960’s and the early 1990’s the MDI became the dominant inhaled dosage form, replacing oral and glass bulb nebulised alternatives. Dry powder inhaler (DPI) alternatives certainly existed, but their market share was only moderate. The spectrum of actives on offer to asthmatics widened to consist of βagonist relief medications such as salbutamol, ‘preventer’ medications based on corticosteroids such as beclamethasone diproprionate and formulations mainly targeted at COPD sufferers based on antichologenics such as Ipratropium bromide. Generally most formulations only contained a single drug, but there were a few combination therapies. However, as the mid-1980’s arrived, so did a worrying cloud on the horizon of the MDI’s future, stratospheric ozone depletion. The identification of the ozone depletion problem, both in the lab and stratosphere, was a great scientific success of its time and is well described elsewhere[4]. This concern lead to the Montreal Protocol (MP), an unprecedented regulatory framework which mandated the phase out of the manufacture and use of certain classes of chemicals, including the CFCs. Recognising the time and difficulty associated with any reformulation or new technology development for pMDI applications, the Protocol contained a mechanism (Essential Use Allowances (EUAs)) to permit continuing use of CFCs for such essential uses beyond the phase out date for the regulated substance. It was only recently that the very final EUAs for the use of CFCs in pMDIs were approved for 2015 consumption in China.

61


Drug Delivery to the Lungs 27, 2016 - The Future of Propellants for pMDIs The Future of Propellants for pMDIs In looking to develop alternatives to CFC-propelled medical inhalers, a number of approaches were explored including multidose DPIs and ambulatory pocket nebulisers. However, there was also a major push to develop one or more non-ozone depleting propellants for pMDI use. HFA MDIs From this effort, HFAs 134a and 227ea, as shown in table 1, appeared to be viable candidates but their adoption necessitated full safety studies that were carried out by two pharmaceutical company consortia, IPACT-1 and IPACT-2[5], which studied 134a and 227ea respectively. This work still provides the bedrock for current medical 134a specifications. From a manufacturing perspective, the advancement in standards of GMP compliance expectation for these excipients also caused a move to the establishment of separate purification assets operating to cGMP, which acted to purify industrial grade HFAs to the required levels and with appropriate control measures. Typically these ‘polishing plants’ have been based upon high performance distillation columns, and represented a major departure from the long- established practice with their CFC forerunners, which was to just take product from the industrial plant and ‘test the quality in.’ The absence of a suitable alternative for CFC 11 at the time gave rise to a number of alternative formulation and aerosol filling approaches, both single- and two-stage. Nowadays, these two principal approaches have many variants, including 3M’s cold-fill process such that there is almost a continuum of options between the ‘pure’ versions. The transition to HFAs 134a and 227ea presented numerous technical and formulation challenges including the risk of particles agglomerating and adhering to the can walls resulting in sub-optimal drug delivery through the life of the can. Whilst addition of ethanol can mitigate this effect, it is not without problems of its own and other techniques such as the use of inert can coatings have also been developed. 3M introduced the first HFA 134a pMDI in 1995[1] MDIs propelled by HFAs 134a and 227ea are arguably the workhorses of respiratory medication delivery. They provide a strong platform from which virtually all of the small molecule respiratory drugs can be delivered. However, the bulk of MDI formulations these days are relatively mature with much of the latest developments in inhalation medicine and delivery focussed on DPI products and associated delivery devices, often still covered by IP. In the market, this results in pMDIs achieving around 20% of the market sales value from around two-thirds of the total standard unit sales with DPIs having around 50% of the market value from only around 25% or so of the total market volume.

The Future There are now over 100 companies worldwide making pMDIs, with a combined 2016 output estimated at 750M units[6]. Worldwide production in unit terms is increasing at between 5-8% p.a. in the short-term, with that growth concentrated in the developing world. However, environmental regulation is now once more on the horizon, this time as part of international efforts to minimise the rise in global temperatures as a consequence of anthropogenic enhancement of natural greenhouse warming. A standardised measure of the potential for a molecule to enhance global warming is the Global Warming Potential (GWP) with CO2 having a GWP of 1. It is clear from Table 1 that HFA 134a already has a significantly lower GWP than CFC 12 but despite this, there are some concerns that the GWPs of HFAs 134a and 227ea are still unacceptably high taking into consideration their projected increase in pMDI consumption going forward. HFAs are already the subject of regulation in the EU via the ‘F’-gas Regulation[7], and in the USA, via ‘SNAP’[8]. Further, the Parties to the Montreal Protocol (MP) are discussing extending that regulatory framework to cover HFA emissions with a programme aimed at achieving a phase-down to around 15% of today’s rate of CO2- equivalent emission by 2035 (Figure 1). Unlike the CFCs, this will not be a phase-out.

62


Drug Delivery to the Lungs 27, 2016 – Tim Noakes,et Tim al Noakes, Stuart Corr

80 60

non-Art.5

40

Art.5

20

EU

0

2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046

% of CO2e baseline

100

Figure 1 Montreal Protocol amendment North American Proposal and EU F-Gas Regulation schedule The precise impact of these MP proposals on pMDI propellants is not yet clear. However, the regulations currently in place give a degree of protection for pMDI usage and the environmental regulators are of the view that there will be adequate supplies of HFAs available to support this use under the phase-down limits. Although the UNEP Technical and Economic Assessment Panel (TEAP) recognizes that any environmental benefits associated with transitioning from the existing propellants to say DPI technology would be small [9] we can still look for ways of minimising the impact of pMDI technology and perhaps to further improve performance. Minimising the environmental impact of pMDI There are a number of ways that the environmental impact of pMDIs can be minimised. These range from recycle schemes such as GSK’s “Complete the Cycle” scheme where residual and unused pMDIs are collected for propellant recovery, through to the development of new reduced-GWP propellants. As evidenced by the CFC to HFA transition, identification and implementation of a new medical propellant is no easy task and any new reduced-GWP propellant needs to: 1. 2. 3. 4.

Be at least as toxicologically safe as the current propellants Function effectively for a range of solution and suspension pMDI drug formulations: ideally offer performance benefits Be of acceptable cost and available at appropriate scale and purity Be sustainable

Two candidates that have been proposed as reduced-GWP pMDI propellants are the unsaturated HFCs (HFOs) particularly HFO 1234ze(E) (trans-CF3CH=CFH) which is already being used being used commercially as an aerosol propellant in some applications. Although HFO 1234ze(E) has GWP<4 and is regarded as safe for use across a range of non-medical applications, it has not yet been found to have acceptable toxicological safety for the more demanding inhalation medicine application sector. Other potential reduced-GWP propellants that have been proposed are the hydrocarbons, particularly propane, butanes and pentanes and whilst these have been used successfully in formulations for some topical uses, they again have not yet been found to have acceptable toxicological safety for inhalation medicine use. After screening over 400 candidate molecules, Mexichem has embarked on a detailed program to investigate the potential for a new medical propellant: 1,1-difluoroethane (HFA 152a). This program encompasses a number of aspects including toxicological safety, end-user safety, device component compatibility and formulation performance in vitro. Some properties of HFA 152a are illustrated in Table 2. Although at a relatively early stage, to date, the pre-clinical safety studies are promising and have raised no grounds for concern and work continues. Like the HFOs 1234yf and 1234ze(E), HFA 152a is classed as flammable, although less so than hydrocarbons. Detailed end-user safety studies based on computational fluid dynamic simulations have shown no significant increase in end-user risk.

63


Drug Delivery to the Lungs 27, 2016 - The Future of Propellants for pMDIs The Future of Propellants for pMDIs Property

HFA 134a

HFA 227ea

HFA 152a

BP (oC)

-26.2

-17.1

-24.7

Vap. Press (barg, 25oC)

5.65

2.89

4.99

Liq. Density (g/cc, 25oC)

1.22

1.41

0.91

Flammability

No

No

LFL 3.8%

GWP

1430

3220

124

VOC

No

No

No

Table 2: Properties of HFA 152a compared to HFA 134a and HFA 227ea On the formulation performance front, HFA 152a has shown surprisingly good performance and a number of benefits over existing formulations. For example, despite having a significantly lower density than HFA 134a, HFA 152a shows significant improvements in suspension settlement and re-suspension behaviour with salbutamol sulphate even in the absence of additional excipients: Propellant/excipients HFA 134a HFA 134a/EtOH

Time to sediment(minutes) < 0.5 1.5

HFA 134a/EtOH/Oleic acid

3

HFA 152a

2

Table 3 Turbiscan™ Salbutamol sulphate formulation sedimentation times HFA 152a versus HFA 134a From an environmental impact perspective, initial estimates suggest that an MDI formulated with HFA 152a will have a comparable carbon footprint to many DPI devices and a more detailed environmental impact study is in progress. It is perhaps an understatement to say that these are early days and much work remains to be done, not least on toxicological safety. However, based on the data generated so far, HFA 152a shows sufficient promise for Mexichem to continue its research and development program. If ultimately successful, HFA 152a could eventually expand the portfolio of propellants available to pMDI formulators and help to further reduce the environmental impact associated with pMDI use. If HFA 152a is found to be unsatisfactory for whatever reason, at least it is an avenue that has been explored and strengthens the case for robust maintenance of the current pMDI propellants going forward.

1

Mexichem Chemical Safety Data Sheets. GWPs currently calculated values from IPCC AP5

2

50th Anniversary of the first pMDIs, The Pharmaceutical Journal, 277, p795

3

Ventolin remains a breath of fresh air for asthma sufferers, after 40 years, The Pharmaceutical Journal, 279, p404

4

Freuh S, Depletion of the Earth’s Ozone Layer Discovery and Response, National Academy of Science Online, 2015

5

International Pharmaceutical Aerosol Consortium for Toxicity (IPACT) (I) and (II)

6

Mexichem estimate derived from 2016 propellant and can usage data

7

Regulation (EU) No 517/2014 of the European Parliament and of the Council

8 9

Significant New Alternatives Program. Federal Register Vol. 80 No.138, July 20, 2015, 42870-4295 IPCC/TEAP Special Report: Safeguarding the Ozone Layer and the Global Climate System, Chapter 8, 2005

64


Drug Delivery to the Lungs 27, 2016 - Glyn Taylor et al. Next Generation Formulations for pMDIs Glyn Taylor, Simon Warren, Cuong H Tran i2c Pharma Services, Cardiff Medicentre, Heath Park, Cardiff, CF10 4UJ, UK Summary Given its 60-year history, it might be tempting to see similarities in physical appearance of the originally marketed pressurised metered dose inhaler (pMDI) with its present form and conclude that little has changed during its lifetime. This however would be specious, since significant advancements have been made in all aspects of the pMDI, including hardware, formulation and performance. The pMDI remains the most widely used inhaler device worldwide for a variety of reasons. Clearly the introduction of the Montreal Protocol with the switch from CFC to ozone-friendly HFA propellants stimulated new research into the fundamental aspects of particle science, device technology and aerosol generation. This has resulted in new strategies to modulate the performance and broaden the applicability of pMDIs. The use of: engineered particles; solubilisation aids including cyclodextrins, and nanoparticles; suspension stabilisers including porous phospholipid particles; lactose and amino acid carrier systems; have all been applied to optimise lung delivery and extend the scope of pMDIs to a wider range of drugs, drug combinations and doses. In addition to small molecule applications, strategies are evolving to formulate macromolecules and biologics, including proteins, DNA, vaccines and bacteriophages. The potential evolution of new propellants with even lower environmental impact will offer future challenges but the wealth of research conducted over the past two decades will help overcome the task ahead. One key driver for the continued development of the pMDI is its global appeal to a broad spectrum of patients. Introduction Some key advantages of the pMDIs reside in qualities of being highly popular with patients, robust, compact and portable, highly reliable and quick to use. The pMDI platform is also versatile in many aspects, including the ability to deliver drugs with disparate physicochemical properties in doses ranging from a few micrograms to several milligrams. Clearly there are some disadvantages, and the problem of poor inhaler technique by patients, in particular ineffective co-ordination of inspiration and actuation, is an often cited disadvantage for pMDI use. Other inhaler types, however, also suffer from problems of sub-optimal handling and dosing by patients. The introduction of “smart” pMDIs has been tried in the past but it is hoped that current regulatory ethos and healthcare payment systems will facilitate the widespread use of, devices such as the 3M Intelligent Control Inhaler™ and also that of training aids including Trainhaler™ and Flo-tone™, which are designed to improve pMDI inhaler techniques in patients. A review of these devices is beyond the scope of this article, which will focus on formulation issues. Despite the environmental success of the Montreal Protocol, the formulation challenges of changing from CFC to HFA propellants with their lower boiling points, different densities, other physicochemical properties and especially the insolubility problems of legacy surfactants, presented significant challenges in re-formulation. The two decades taken post-Montreal Protocol, even in the US, to phase out certain CFC pMDIs, reflects the complexity of replacing certain formulations. From a positive aspect, the improvements in pMDI performance compared to the CFC formulations have been significant and strategies for enhancements are still progressing. It is perhaps easy to overlook that standard pharmaceutical and medical textbooks, written little more than a decade ago, referred to pMDIs as devices delivering only 5-10% of the emitted dose to the lungs. The performance of modern pMDIs has dramatically st increased and they have evolved to be clearly fit for purpose in the 21 Century. The objective of this article is to review and highlight some of the recent developments in pMDI formulation approaches for both small molecule drugs and also for macromolecules. It attempts to illustrate that the pMDI will remain at the forefront in device choice, offering a viable and highly significant contribution to the management of respiratory diseases and certain systemic therapies.

Discussion Drugs are formulated as either solutions or suspensions in pMDI products. The challenges of solution formulations are generally those associated with solubility and/or chemical stability of drug in the propellant. In contrast the challenges with suspension formulations are in achieving and maintaining physical and/or physicochemical stability during pMDI manufacture, storage and patient dosing.

65


Drug Delivery to the Lungs 27, 2016 - Next Generation Formulations for pMDIs Solution pMDI Formulations Most drugs and biologics are not soluble in HFAs but where a drug has partial solubility, such as beclomethasone, then a co-solvent such as ethanol is needed to ensure that the drug is fully dissolved at the therapeutic doses needed. Other steroids such as ciclesonide and flunisolide are also presented in solution formulations containing ethanol. The amounts of ethanol in the formulation will influence aerosol quality, with high [1] concentrations decreasing the fine particle fraction (FPF) . Other excipients with low volatility (compared to HFAs) such as glycerol can also be included to modify the aerosol particle size distribution. In addition, pH modifiers are used to maintain chemical stability of some solution pMDIs, such as certain formoterol and ipratropium products. Other excipients include novel functionalized methylated polyethyleneglycols and oligolactic [1] acid, both acting as solubilisation aids when used in combination with ethanol . The principle of using cyclodextrins as solubilizers in pMDI has also reported, for example with sildenafil (for the treatment of pulmonary [2] arterial hypertension) with FPFs in the range of 45-81% . Suspension pMDI Formulations The majority of marketed pMDI products contain drugs formulated as suspensions. Whilst this approach reduces the problem of chemical instability, most micron-sized drug suspensions are inherently physically unstable, to greater or lesser extents, in HFA propellants. Optimizing suspension pMDI formulations to ensure reproducible and effective therapeutic dosing can be achieved using different strategies. One consequence of the Montreal Protocol was the finding that most regulatory approved surfactants at that time were insoluble in HFAs. A number of those surfactants are however soluble to some degree in ethanol, and hence ethanol is a useful excipient in suspension pMDI formulations not only to impart sufficient surfactant solubility, but also functioning alone, as a wetting agent, to aid suspension formation during manufacture and patient use. Most pMDI suspension formulations contain jet-milled micronised drug particles, however particle engineering technologies including sonication methods, have been developed to generate homogenous particles of welldefined shape and size. In addition, high purity crystals containing different combinations of drugs can be [3] manufactured into a single particle using the “sonocrystallization” technology . Additionally, this approach does not require additional functional excipients, adjuvants, or co-suspension agents for optimal pMDI performance. A fluticasone pMDI product (Fliveo®, Circassia) employing the sonocrystallization technology has received recent UK and Swedish generic product approvals based upon equivalence data from in vitro performance alone. Suspension pMDIs with Particulate Excipients: TM

Co-Suspension

Delivery Technology

A different approach to stabilize pMDI suspensions and improve performance is to employ spray-dried porous phospholipid particles, as seen in the Co-Suspension Delivery Technology developed by Pearl Therapeutics/AstraZeneca. The technology uses phospholipid microparticles
with aerodynamic diameters of [4] 1−2 μm to irreversibly
associate with drug microcrystalline particles during manufacture . This results to minimize drug-drug interactions and improve suspension stability. Using this strategy, formulations containing either a single drug or combinations of drugs, for example, a corticosteroid, a long-acting beta-agonist and a longacting muscarinic antagonist, have been developed and can give high aerosol performance with FPFs in excess [4] of 60% . A glycopyrronium/formoterol product (Bevespi Aerosphere™) using the Co-Suspension technology has recently received FDA approval for use in COPD patients, whilst a glycopyrronium pMDI and a triple combination budesonide/glycopyrronium/formoterol pMDI product are in Phase III of clinical trials. TM

Opt2Fill

Technology

The Opt2Fill technology employs a “second particulate” such as micronised lactose, other sugars or amino acids with particle sizes in the range of tens of microns and are readily dispersible in HFAs. The second particulate is first admixed with micronised drug, promoting adsorption of the drug and forming an ordered blend. The second particulate acts as a stabiliser within the HFA environment associating with the drug and minimising aggregation between the high-energy micronised drug particles (Figure1). Upon pMDI actuation and aerosolisation, the micronised drug detaches from the larger second particulate “carrier”. The Opt2Fill technology has applications for both low and high dose drugs such as tiotropium and budesonide, polar and lipophilic drugs such as salbutamol and beclomethasone, and combination formulations including formoterol/ budesonide and salmeterol/ fluticasone. In vitro FPFs in excess of 60% have been reported for several drugs using [5] this type of formulation . Figure 1 - Opt2Fill Carrier with Adsorbed Drug Particles; (left) after Blending and (right) Suspended in a Model HFA.

66


Drug Delivery to the Lungs 27, 2016 - Glyn Taylor et al. In a recent advancement of Opt2Fill pMDI manufacturing methods, the ordered blends of micronised drug and second particulate have been formulated as tablets with the inclusion of a regulatory approved HFA soluble [6] dispersing agent . The Opt2Fill tablets are then dispensed into cans and crimped. Propellant is then added at a later stage. In comparison with established pressure fill single- and two-stage manufacturing of suspension pMDIs, the Opt2Fill approach has advantages including the avoidance of homogenization/mixing vessels and pressure vessels, with associated challenges of propellant top-up throughout batch filling. Batch sizes can be readily varied from clinical trial scale to large commercial batches, ensuring process continuity. Additionally, cleaning programmes between different products are greatly simplified. In one example, the in vitro performance of a combination Opt2Fill tablet pMDI formulation of salmeterol xinafoate and fluticasone propionate is shown to be [7] similar to the marketed product and the effect of formulating the Opt2Fill powder into a tablet is negligible (Figure 2).

Figure 2 – Aerosol Performance of Salmeterol/Fluticasone pMDI Formulations in the NGI

Macromolecule and Biologic Formulations The environment within a pMDI formulation is sometimes considered very challenging for macromolecules, especially for proteins to maintain their conformational stability. Despite this, some successes in the formulation of macromolecules have been reported. A very early example is that of crystalline insulin zinc (CFC) pMDI formulations that demonstrated good physicochemical stability over several months, and a predicted shelf-life, [8] with respect to chemical stability, of 19 years . Clearly many macromolecules and biologics are not readily crystallized and will require sophisticated techniques to maintain their essential structure and potency and produce particles of a suitable size for lung delivery. Studies with nanoparticles of insulin prepared from water-inoil emulsions formulated into a pMDI with HFA134a and cineole have been reported to maintain primary, [9] secondary and tertiary structures of insulin and gave an extra-fine particle fraction of around 45% . Successes have also been reported with a number of other macromolecules including DNase, lysozyme, alkaline phosphatase, and bovine serum albumin. Excipients, including vinyl polymers, trehalose and carboxymethylcellulose to either protect the macromolecules during spray-drying, maintain the three-dimensional [10] structure, or to stabilise the suspensions in HFA formulations have been investigated . Porous (poly (dl-lactideco-glycolide) microparticles with a density similar to that of HFA227 have also been reported as potential carriers [11] for the delivery of proteins in pMDI formulations . Research on formulations which may prove suitable for other biologics has also been undertaken. Bains et al developed methods using a low-energy microemulsion process to prepare surfactant-coated plasmid DNA (pDNA) nanoparticles. Transfection studies demonstrated that pDNA biological function was preserved following aerosol generation from a HFA134a pMDI formulation. The flocculated system also remained readily dispersible after 5 months and follow-on accelerated stability studies at 40ď‚°C/75%RH demonstrated minimal loss of activity [12] over 1 month . Chitosan nanoparticles have also been used to prepare pDNA in HFA227 formulations and the [13] incorporation into HFA-philic engineered oligolactide core shell particles resulted in FPFs in excess of 50% . These studies highlight the potential for gene delivery using pMDI formulations. Bacteriophage delivery from [14] pMDIs has also been studied and the potential for delivery of vaccines and phages should not be discounted . Overall, these studies indicate that the shear forces associated with aerosolisation of macromolecules and biologics from pMDI propellants are not greater, and probably cause less degradation than has been reported previously using air-jet nebulization.

67


Drug Delivery to the Lungs 27, 2016 - Next Generation Formulations for pMDIs

Conclusions “The pMDI is not dead!” it lives on as an efficient, adaptable and cost-effective method for pulmonary drug delivery. The pMDI is the inhaler device most often used by patients worldwide, and in the UK 70% of inhaler devices sold between 2002 and 2008 were pMDIs. The percentage is however much lower in some other [15] European countries . Data for recent years (from IMS MIDAS) shows similar trends. In the 60 years since the introduction of the pMDI, the “mechanical marvel” has faced many challenges, of these the Montreal Protocol might be viewed as its “mid-life crisis”. To accept this however, would be to infer that the pMDI is now in decline and entering the twilight years of its life. Clearly this is not the case and activities in both commercial and academic research are as vibrant now as they have ever been over the past decades. New propellants may evolve and will present further challenges for the formulators. Given the wealth of knowledge gleaned during the past 25 years we have a much better understanding of the key parameters which [16] influence pMDI performance . Hence it is inevitable that new strategies will be developed to embrace any future propellant changes and formulators, as previously, will rise to the challenge and seize opportunities for further improvements in pMDI efficiencies. Clearly the formulation of many macromolecules and biologics will require bespoke solutions but evidence from the radical improvements in pMDI formulations since the CFC-HFA conversion demonstrates that research and innovation in this field will adapt to overcome the inevitable hurdles. This will in part be driven by the inherent advantages of the pMDI for the pulmonary delivery of locally- and systemically-acting drugs. References 1

Myrdal P B, Sheth, P, Stein S W: Advances in metered dose inhaler technology: formulation development, AAPS PharmSciTech 2014; 15: pp434-455.

2

Sawatdee S, Phetmung H, Srichana T: Sildenafil citrate monohydrate–cyclodextrin nanosuspension complexes for use in metered-dose inhalers. Int J Pharm. 2013; 455: pp248-258.

3

Parikh D, Karki S, Hipkiss D: Engineered combination respiratory medicines for localized lung delivery, Respir Drug Del 2012; 3: pp699-704.

4

Vehring R, Lechuga-Ballesteros D, Joshi V, Noga B, Dwivedi S K: Cosuspensions of microcrystals and engineered microparticles for uniform and efficient delivery of respiratory therapeutics from pressurized metered dose inhalers, Langmuir 2012; 28: pp15015-15023.

5

Tran C H, Davies K, Medina A M, Warren S, Taylor G: An evaluation of formulation variables on the performance of an innovative budesonide pMDI suspension. Drug Del Lung 2012; 23: pp154-157.

6

Taylor G, Warren S J, Tran C H: Pressurised metered dose inhalers and method of manufacture, Patent WO 2015/121653.

7

Taylor G, Tran C H, Zheng C, Warren S: Application of an Opt2Fill Dispersible Tablet to the Production of a Novel Salmeterol/Fluticasone pMDI, Respir Drug Del 2016; 2: pp381-384.

8

Lee, S-W, Sciarra J J: Development of an aerosol dosage form containing insulin, J Pharm Sci 1976; 65: pp567-572.

9

Nyambura B K, Kellaway I W, Taylor K M G: Insulin nanoparticles: Stability and aerosolization from pressurized metered dose inhalers, Int J Pharm 2009; 375: pp114-122.

10

Fathe K, Ferrati S, Moraga-Espinoza D, Yazdi A, Smyth H D: Inhaled biologics: From preclinical to product approval. Curr Pharm Des 2016; 22: pp2501-2521.

11

Cocks E, Somavarapu S, Alpar A, Greenleaf D: Influence of suspension stabilisers on the delivery of protein-loaded porous poly (dl-lactide-co-glycolide) (PLGA) microparticles via pressurized metered dose inhaler (pMDI), Pharm Res 2014; 31: pp2000-2009.

12

Bains B K, Birchall J C, Toon R, Taylor G: In vitro reporter gene transfection via plasmid DNA delivered by metered dose inhaler, J Pharm Sci 2010; 99: pp3089-3099.

13

Conti DS, Bharatwaj B, Brewer D, da Rocha SRP: Propellant-based inhalers for the non-invasive delivery of genes via oral inhalation, J Control Rel 2014; 14: 495-530.

14

Hoe S, Boraey M A, Ivey J W, Finlay W H, Vehring R: Manufacturing and device options for the delivery of biotherapeutics, J Aerosol Med Pulm Drug Del 2014; 27: pp315-328.

15

Lavorini F, Corrigan C J, Barnes P J, Dekhuijzen P R N, Levy M L, Pedersen S, Roche N, Vincken W, Crompton G K: Retail sales of inhalation devices in European countries: So much for a global policy, Resp Med 2011; 105: pp1099-1103.

16

Ivey J W, Vehring R, Finlay W H: Understanding pressurized metered dose inhaler performance, Expert Opin Drug Deliv 2014; 12; pp901-916

68


Drug Delivery to the Lungs 27, 2016 –A.P. McKiernan Novel techniques for characterising inhalers A.P. McKiernan 1

Prior PLM Medical, IDA Business & Technology Park, Carrick-on-Shannon, N41 WK46, Ireland

Summary Background: While the initial plume formation and expansion during pressurised metered dose inhaler (pMDI) dose release is considered important to drug transport to the lungs, it is not well understood, in part due to the transient nature of the event and the difficulties in accessing the metering chamber. Plume velocity is an important measure of inhaler device performance as large or very fast drug particles tend to deposit in the oropharyngeal region due to the sharp directional flow change. Methods: Schlieren imaging is an optical technique that is sensitive to refractive index gradients which are often present in pMDI plumes due to gas density variations. We have developed a Schlieren setup and have observed plume expansion up to 400 mm from commercial pMDIs for various actuator/canister/formulation combinations. We have also used phase contrast X-ray imaging at a synchrotron to investigate the stem/sump/orifice of the same pMDIs and a dry powder inhaler (DPI) during dose release. Results: The plume leading edge dynamics are well described by a high-speed drag model. It was observed that HFA 134a drives faster plumes than HFA 227ea, particularly as the orifice diameter increases. Propellant cavitation behaviour appears dependent on the canister/formulation while the dose duration appears additionally dependent on the actuator design. Individual carrier particles were observed to travel counter to the general airflow inside a DPI. Conclusions: High speed Schlieren imaging and phase contrast X-ray imaging are suitable techniques to evaluate formulation or actuator modifications, metering chamber behaviour or tooling modifications during inhaler device development. Introduction The World Health Organisation estimates that over 300 million people worldwide suffer from respiratory diseases such as asthma and chronic obstructive pulmonary disorder. Typically, inhaled medicine, in the form of pMDIs and DPIs, is used to treat these diseases. Nevertheless, these technologies have their limitations and in particular [1] device/treatment efficacy is often quite poor . Efficacy is a complex problem but the plume speed and particle [2, 3] . While APSD is routinely measured in laboratories size distribution are thought to be important contributors during device development and manufacture, the expelled plume physical characteristics are often not as easy to [4] is an optical method that is sensitive to the refractive index gradient (first measure. Schlieren imaging derivative) of visible light (Figure 1). For gases, this is directly related to the density by , where is gas density and is the Gladstone-Dale coefficient. Essentially, a region with a refractive index gradient causes the light beam passing through it to deviate slightly and this is captured by a camera. We have observed the external plume from a number of commercial pMDI actuator / canister combinations and present plume velocity behaviour.

Figure 1: Schlieren imaging technique

Figure 2: phase contrast X-ray imaging

The dynamics of the plume up to the point at which it exits the device are thought to influence velocity and APSD. Various groups have studied the effects on both parameters of different nozzle orifice diameters and metering [5, 6] [7] and of inlet geometry for DPIs . Some have used transparent models chamber designs in the case of pMDIs [8, 9] . X-rays are capable of and high speed cameras to study the flow inside model actuators during dosing penetrating inhaler devices to visualise internal features but high intensities are required to investigate fast events such as dose release. When the atomic numbers in the constituents of polymers, drugs and propellants are [10] at a synchrotron facility is more sensitive than conventional X-ray similar, phase contrast X-ray imaging imaging making it well suited to characterising dosing event dynamics (Figure 2). We present first results of a study of the dynamics of the dosing event internally in some commercial devices.

69


Drug Delivery to the Lungs 27, 2016 - Novel techniques for characterising inhalers Experimental Methods Prior PLM Medical’s large field of view z-type mirror-based Schlieren system was used to image the inhaler plume onto a high speed CMOS camera running at 4100 fps (240 µs exposure time) using a 4000 lumen white LED and [11] 413 mm diameter parabolic mirrors. The spatial resolution was measured using a knife-edge technique resulting in an absolute error of  0.45 mm on measured plume positions. Video post-processing was performed [12] using FIJI . The leading edge of the plume was tracked using the Multi Kymograph plugin, followed by edge detection. This is illustrated in Figure 3 (a, b), where the same (horizontal) line of pixels from each successive frame is stacked side-by-side (vertically) leading to a position vs. time plot such as Figure 3 (c). A monochromatic X-ray beam of energy 26 keV and bandwidth 0.4% was used at beamline ID19 at the European Synchrotron Radiation Facility, Grenoble, France. The pMDI devices were first shaken at 3Hz for 5 seconds using a pneumatic shaking feature on the fixture before being actuated by a custom-built fixture with a linear actuator. They were fired under 28 l/min flow for 3 s using a vacuum pump and critical flow controller while the multi-dose DPI was triggered by a 60 l/min flow for 3 s. The camera (4000-12000 fps, 71-227 µs exposure time), linear actuator, shaking function and flow controller were driven by an Arduino microcontroller. An absolute positional error of  24.29 µm was measured for this setup. The pMDI devices were investigated with canisters containing actives and either HFA 134a or HFA 227ea propellants and capillary retention (CR) or rapid fill / rapid drain (RF/RD) valves inserted into the actuator. In CR valves, after each dose event, the liquid suspension/solution is forced through capillary channels from the reservoir to metering chamber where it is retained through capillary action. In RF/RD the channel diameter is greater so that the metering chamber is refilled by shaking the device prior to use. Actuators 1 to 4 have orifice diameters of 0.30, 0.33, 0.42 and 0.53 mm respectively. Due to X-ray beam time constraints, four pMDI actuator types/designs were studied with one actuator per design coupled in turn with each of the three available canister options. Each device was primed/ dosed according to manufacturer’s instructions for both techniques. [13, 14, 15]

– a blast/shockwave model and two A number of common flow models were fit to the Schlieren data classical drag force theoretical aerodynamic models. The drag force model is proportional to velocity and applies to laminar flow while the high-speed drag force model is proportional to the square of the velocity and applies for , the drag model:

turbulent flow. The shockwave model is of the form: , where

and the high-speed drag model:

is elapsed time,

energy is released

is ambient air density, is a constant that depends on the specific heat capacity of the plume in the explosion, is the stopping distance and is the slowing coefficient. gas, Results and discussion Typical plots of leading edge position vs. time are displayed in Figure 4 and it can be observed that the blast model does not describe the plume behaviour very well but the high-speed drag model provides the closest fit to the data, particularly over the first 300 mm of travel. Figure 5 presents plume velocity at 100 mm from the mouthpiece, often considered to correspond to the back of the throat. The HFA 227ea plumes are generally of lower velocity than the HFA 134a plumes (regardless of canister valve type). This may be partly due to the lower boiling temperature of HFA 134a but is more marked with larger orifice actuators suggesting that dose duration (not shown) and velocity are dependent on both formulation and actuator design.

Figure 4: Plume leading edge vs. time for Actuator 1 with 3 canisters. Some common flow models are fit to the data. Figure 3: Schlieren stills (a,b) and resulting kymograph (c)

70


Drug Delivery to the Lungs 27, 2016 –A.P. McKiernan

Figure 5: Plume velocity (n=3) for HFA 227ea CR, HFA 134a CR and HFA 134a RF/RD canisters in each actuator type (actuators in order of increasing orifice diameter).

Figure 6: DPI nozzle region early and late during dose release event. Note the suspended shrinking blister reservoir (blurred region, circled).

Figure 7: Particle (circled) travelling in the opposite direction to general flow (arrow).

Figure 6 displays two X-ray stills taken during dose release from a dry powder inhaler. Individual particles (probably carrier particles due to their size) leave the powder blister/reservoir on the right and travel towards the nozzle exit on the left (out of view). In some cases, flow reversal appears to take place for regions of the flow where particles were seen travelling from right to left (Figure 7). Pulsations in the numbers of particles ejected were also seen (not shown). Video footage shows the powder reservoir mass lifts approximately 100 Âľm away from its holding crucible and powder is then stripped from the reservoir mass at similar rates from all sides. The pulsations appear to be related to the powder lifting off as surface layers at times. Typical X-ray stills as the propellant/drug mixture flows through the canister stem and actuator orifice are shown in Figure 8 for each pMDI canister/actuator combination. While the propellant and valve type varies as indicated in the figure, so too does the formulation. Cavitation behaviour appears to be quite different across the canisters in terms of the boiling/bubbling. A larger quantity of smaller bubbles are seen with HFA 134a from the RF/RD canister, suggesting that this could be driven by the formulation rather than the valve, for example ethanol or other excipient content may well influence it. The contrast achieved with the current setup was not sufficient to make detailed observations of cavitation/ligament behaviour in the orifice channel. Stem

Flash boiling of propellant in sump Nozzle orifice channel

Figure 8: Dosing event 85 ms after dose commencement from 4 different actuators each with 3 canisters containing HFA 227ea and HFA 134a propellants and capillary retention (CR) or rapid fill rapid drain valves (RF/RD).

Conclusion We have developed a large field of view high speed Schlieren imaging system and have presented initial results obtained with this setup. Plume positions, velocities and durations have been measured for various commercial pressurised metered dose inhalers (pMDI) actuator / canister combinations. The plumes measured are reasonably well described by a high-speed drag model and it was observed that HFA 134a drives faster plumes than HFA 227ea, particularly as the orifice diameter increases.

71


Drug Delivery to the Lungs 27, 2016 - Novel techniques for characterising inhalers Phase contrast X-ray imaging has been used to observe the propellant mixture behaviour inside the canister stem and actuator for commercial pMDI devices. Cavitation appears to depend on formulation more than actuator. Powder dosing from a DPI was also investigated and both flow reversal and pulsations in powder density during the dosing event have been observed. High speed Schlieren imaging is a suitable tool to aid actuator design modifications or tooling modifications during device development. In addition to positional and velocity information, other measurements such as plume fire angle can also be carried out using this apparatus. The technique provides images of the whole gaseous plume along the direction of motion at each time-point. In contrast, techniques such as PIV provide access to particulate motion and velocity fields but with increased system complexity and cost. Thermal imaging is another technique suitable for plume imaging but at lower frame rates. Phase contrast X-ray imaging is also a useful method to characterise inhaler actuators and in addition it can provide access to the metering chamber, stem and gasket behaviour. It also enables the systematic study of nozzle geometry and flow rates on powder and flow behaviour inside DPI devices. Acknowledgments We acknowledge the European Synchrotron Radiation Facility for provision of synchrotron radiation facilities and we thank Alexander Rack for assistance in using beam line ID19. We also acknowledge the engineering and toolmaking staff of Prior PLM Medical for the design and manufacture of the Schlieren setup and experimental fixture. References 1

O.S. Usmani, ‘Small-Airway Disease in Asthma: Pharmacological Considerations’, Current Opinion in Pulmonary Medicine 21, no. 1 (January 2015): 55–67, doi:10.1097/MCP.0000000000000115.

2

T.G.D. Capstick and I.J. Clifton, ‘Inhaler Technique and Training in People with Chronic Obstructive Pulmonary Disease and Asthma’, Expert Review of Respiratory Medicine 6, no. 1 (February 2012): 91–103, doi:10.1586/ers.11.89.

3

R. Nave and H. Mueller, ‘From Inhaler to Lung: Clinical Implications of the Formulations of Ciclesonide and Other Inhaled Corticosteroids’, International Journal of General Medicine, March 2013, 99, doi:10.2147/IJGM.S39134.

4

G.S. Settles, “Schlieren and shadowgraph techniques: visualizing phenomena in transparent media”. Berlin ; New York: Springer, 2001.

5

J. Berry, S. Heimbecher, J.L. Hart and J. Sequeira, ‘Influence of the Metering Chamber Volume and Actuator Design on the Aerodynamic Particle Size of a Metered Dose Inhaler’, Drug Development and Industrial Pharmacy 29, no. 8 (January 2003): 865–76, doi:10.1081/DDC-120024182.

6

H. Smyth, G. Brace, T. Barbour, J. Gallion, J. Grove and A. J. Hickey , ‘Spray Pattern Analysis for Metered Dose Inhalers: Effect of Actuator Design’, Pharmaceutical Research 23, no. 7 (July 2006): 1591–96, doi:10.1007/s11095-006-0280-z.

7

J. Shur, S. Lee, W. Adams, R. Lionberger, J. Tibbatts and R. Price, ‘Effect of Device Design on the In Vitro Performance and Comparability for Capsule-Based Dry Powder Inhalers’, The AAPS Journal 14, no. 4 (December 2012): 667–76, doi:10.1208/s12248-012-9379-9.

8

H.K. Versteeg and G.K. Hargrave, ‘Near-Orifice Spray and Valve Flow Regime of a Pharmaceutical Pressurised Metered Dose Inhaler’, Zaragoza 9 (2002): 11.

9

H.K Versteeg, G.K Hargrave, and M Kirby, ‘Internal Flow and Near-Orifice Spray Visualisations of a Model Pharmaceutical Pressurised Metered Dose Inhaler’, Journal of Physics: Conference Series 45 (1 July 2006): 207–13, doi:10.1088/17426596/45/1/028.

10

S.C. Mayo, A.W. Stevenson and S. W. Wilkins, ‘In-Line Phase-Contrast X-ray Imaging and Tomography for Materials Science’, Materials 2012, 5(5), 937-965 ; doi:10.3390/ma5050937

11

C.S. Williams and O.A. Becklund, “Introduction to the Optical Transfer Function”. SPIE - The International Society for Optical Engineering, 1989.

12

J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, J.-Y. Tinevez, D. J. White, V. Hartenstein, K. Eliceiri, P. Tomancak, and A. Cardona, ‘Fiji: an open-source platform for biological-image analysis’, Nat. Methods, vol. 9, no. 7, pp. 676–682, Jun. 2012.

13

D.B. Geohegan, “Physics and diagnostics of laser ablation plume propagation for high-Tc superconductor film growth”, Thin Solid Films, 220, 1-2, 1992, 138-145, doi:10.1016/0040-6090(92)90562-P

14

G. Falkovich, “Fluid Mechanics (A short course for physicists)”. Cambridge University Press, 2011.

15

P Hough, TJ Kelly, C Fallon, C McLoughlin, P Hayden, ET Kennedy, JP Mosnier, SS Harilal and JT Costello, “Enhanced shock wave detection sensitivity for laser-produced plasmas in low pressure ambient gases using interferometry”, Meas. Sci. Technol., 23 (2012), 125204, doi:10.1088/0957-0233/23/12/125204

72


Drug Delivery to the Lungs 27, 2016 - B. Gavtash1, et el. Multi-physics theoretical approach to predict pMDI spray characteristics B. Gavtash1, H K Versteeg1, G. Hargrave1, B. Myatt1, D. Lewis2, T. Church2, G. Brambilla3 1

Wolfson School of Mechanical, Electrical and Manufacturing Engineering, Loughborough University, Loughborough, LE11 3TU, United Kingdom 2

Chiesi Limited, Bath Road Industrial Estate, Chippenham, Wilts, SN14 0AB, United Kingdom 3

Chiesi Farmaceutici SpA, Via Palermo, 43122 Parma, Italy

Summary Continued success in treatment of asthma and COPD requires development of new formulations, which may alter spray characteristics and atomisation quality of atomiser devices such as pMDI. Fundamental understanding of the underlying physical phenomena and the dynamic nature of pMDI aerosol plumes is essential to maintain pMDI device atomisation quality. In this paper, we describe a simulation of pMDI aerosol generation and plume development using a model of two-phase flow and atomisation of HFA134/ethanol formulation. The model is implemented within a CFD simulation to study the dynamic aerosol development and predict spray velocity and temperature. The CFD result shows plume velocity slows down over a relatively short distance and droplet temperature settles at a steady value of 291 K approximately at a distance of 30 mm from the spray orifice. This means that droplets reach their final size before reaching the exit of the mouthpiece. Introduction 1

pMDI is the most widely used aerosol device to treat asthma, but it has one significant drawback, which is the high amount of oropharyngeal drug deposition. This deficiency is a function of droplet size and velocity. Evolution of these parameters can be addressed by understanding the thermo-fluid dynamic processes governing (i) the flow inside the pMDI actuator, (ii) the droplet generation mechanism, and, (iii) the interactions between aerosol droplets and inhaled air outside the actuator. These processes determine the characteristics of the dynamic pMDI aerosol plume. To the knowledge of these authors there is currently no approach from first principles that is able to tackle this problem. This work presents the main features of a pMDI CFD simulation coupled with a model of the flow inside the actuator and an atomisation model to produce a well-defined spray source. The development of the aerosol plume is tracked along with inhaled air flow through a model of the USP Induction Port (USP-IP). Internal flow/atomisation model Flow visualisations using transparent models of pMDI actuator suggest that the dominant two-phase flow regime in the actuator sump is a well-mixed, bubbly flow. Our model of the internal flow inside metering and expansion 2 chambers assumes such a flow regime. The quasi-steady, homogeneous frozen model (HFM) of Fletcher and 3 Clark is used to predict the mass flow rate and velocity through pMDI orifices. The formulation considered here is a 90:10 (w/w %) binary mixture of HFA134/ethanol whose saturated vapour pressure is defined in the model 4 using an empirical correlation published in reference . For the purposes of predicting droplet size, it is assumed that the two-phase flow regime inside the spray orifice is predominantly annular, which is supported by flow 5 visualisations (see also figure 2 (a)). Using a flat sheet approximation and velocity estimates from the internal 6,7 framework, the growth rate and wavelength of flow model, within the linear instability sheet atomisation (LISA) the most unstable disturbances on the annular liquid film is estimated using equation 1. 1 where is the density ratio between the gas and liquid phase (i.e. ), and are the kinematic viscosity and surface tension of the liquid and (= ) is the wave number. Numerical maximisation of equation 1 with respect to wave number yields the maximum growth rate, as a function of wave number, . The wave number corresponding to the maximum growth rate is denoted by . Amplification of unstable waves is responsible for sheet disintegration and formation of unstable cylindrical ligaments with diameter of: 2

Figure 1 (a). High resolution image of flow inside expansion chamber

Figure 1 (b). Schematic of internal flow model

73

Figure 1 (c). Schematic of frozen flow model inside orifices


Drug Delivery to the Lungs 27, 2016 - Multi-physics theoretical approach to predict pMDI spray characteristics

Figure 2 (a). High resolution image of flow inside spray orifice

Figure 2 (b). Schematic of atomisation model

where is ligament constant, which takes a value of 0.5. The diameter of the therapeutic droplets, 6 obtained based on capillary instability analysis of the initial ligaments :

can be

3 Previous implementation of such internal flow and atomisation models to pMDIs showed good agreement with 7,10 . PDA measurement of spray velocity and droplet size, conducted in near-orifice regions External flow model The physical problem involves the injection of a pMDI into a standard USP-IP accompanied by ambient co-flow of 30 l/min, induced by a vacuum pump. Due to evaporation of formulation inside the pMDI actuator, the emerging spray consists of liquid (droplets) and propellant vapour phase. Droplets and spray vapour phase (HFA vapour) are introduced by means of a transient aerosol generator - the so-called “pseudo spray source”-, using information from the internal flow and atomisation models described earlier. Modelling parameters describing the actuator are summarised in table 1.

Figure 3. Schematic of CFD simulation

Device parameters

Unit

Value

Metering chamber volume

µl

25

Expansion chamber volume

µl

25

Valve orifice diameter

mm

0.7

Spray orifice diameter

mm

0.3

Spray orifice length

mm

0.75

Formulation composition

(w/w %)

HFA134 90 ethanol 10

Ambient temperature

°C

≈ 22

Table 1. Actuator modelling parametrs and geometric characteristics

Results The contours of velocity magnitude along with streamlines of the steady air flow prior to injection of the aerosol spray source are shown in figure 4 (a-b), rendered on the symmetry plane through the actuator and USP-IP. The air flow enters with a mean velocity of 2.5 m/s and travels through the gaps around the canister down towards the stem block where the flow direction sharply changes. Here the flow becomes skewed towards the bottom surface of the mouthpiece. Next, the air moves around the stem block and upwards, leading to the formation of a complex three dimensional flow motion as shown by the visualisation of flow streamlines in figure 4 (b). This rapid change in flow direction accelerates the flow to around 3.5 m/s at the mouthpiece exit plane. Subsequently, the flow enters into the converging section of the USP-IP, where recirculation regions appear at the top and bottom corners of the USP-IP converged section. After the flow reaches the end of USP-IP horizontal section, another rapid change of flow direction occurs, which skews the flow towards the rear wall of the vertical section of the USP-IP. The skewed flow profiles near the bottom of the mouthpiece as well as around the outer wall of the USP-IP vertical section region cause the creation of pairs of vortices. As shown in figure 4 (c-e), these secondary motions can be best visualised in a series of cross-sectional planes perpendicular to the main flow path named alphabetically. The velocity magnitude is indicated by colour coding and velocity vectors illustrate secondary flow 8,9 motion. The main features of our flow calculations closely resemble those of reported in .

74


Drug Delivery to the Lungs 27, 2016 - B. Gavtash1, et el.

(a). mid-section and cross-sectional planes

(b). Near-orifice region on mid-section plane

Figure 4. (a-b) . Contour of air flow distribution coloured by air flow velocity

(c). Cross-sectional plane (a)

(d). Cross-sectional plane (c)

(e). Cross-sectional plane (d)

Figure 4 (c-e) . Contour and vector field of air flow distribution coloured by air flow velocity, on cross-sectional planes

The next phase of simulation includes spray injection into the stabilised air flow field. Temporal evolution of spray location is shown in the time-ordered images of figure 5 (a-d), which are all rendered on the symmetry plane of the pMDI actuator and the USP-IP. The droplets are coloured based on their magnitudes of velocity and are superimposed on the contour of gas phase velocity. Early stages of the plume spatial development include the formation of a shield-like particle embedded configuration at 0.01 s after the actuation. This flow structure appears at the plume tip (indicated by means of a red arrow), is due to the plume front edge experiencing the largest drag force; the droplets at the front are pushed away by the more energetic droplets emerging from the upstream source. During the first 0.01 s, the plume direction appears to be diverted upwards where deflection is indicated by red arrow with dashed tail. This is due to the influence of air flow mean direction (see figure 4 (b)) on droplet trajectory, at the initial stages of actuation when the mass loading of the droplets is still small. From 0.03 s onwards, however, the mass loading of the droplets has become more significant and the spray source overwhelms the inhaled air flow and follows the injection axis. At 0.03 s, and 0.05 s, strong recirculation regions are generated as a consequence of air entrainment at the plume edges. At 0.03 s the droplets at the plume tip are carried by this recirculation zone and travel backwards through the inlet region of the USP-IP. The spray pattern is relatively similar at 0.05 s and 0.07 s. At these times the plume front has progressed into the vertical section of the USP-IP where the majority of droplets are pushed towards the outer wall (indicated by arrows in figure 5 (c-d)). This is due to the high inertia of the droplets, as well as the air/propellant vapour mixture forcing the flow outwards o through the 90 â—Ś bend. At any instant during the injection, it can be seen that the plume slows down to around 1 m/s over almost 45-50 mm of travelling distance, indicating the large momentum transfer between the spray and the air.

Figure 5 (a-d) Contour of gas flow and droplet velocity inside USP-IP

75


Drug Delivery to the Lungs 27, 2016 - Multi-physics theoretical approach to predict pMDI spray characteristics

Figure 6. (a-d) Contour of gas flow and droplet temperature inside USP-IP

Figure 7. Evolution of sampled droplets temperature with respect to droplet travelling distance inside USP-IP

Figure 6 (a-d) show droplet temperatures in combination with gas phase temperature rendered on the symmetry plane. The plume temperature starts with a prescribed temperature around 250 K, corresponding to normal boiling point of multicomponent droplets at atmospheric pressure. As the injection starts, it can be seen that a pulse of colder vapour travels through the USP-IP and gradually mixes with the air. Such interaction results in HFA vapour/air mixture temperature to settle at an approximate value of 291 K. Figure 7, shows the temperature of an arbitrary chosen group of droplets as a function of its axial distance travelled, taking the mouthpiece exit location as reference. It can be seen that droplet temperature experiences an initial decrease, which corresponds to evaporation of remaining HFA134 content of the droplets. The minimum point on the curve corresponds to 13 mm from the spray orifice (i.e. 7 mm behind the mouthpiece exit) when all HFA134 in droplets is evaporated. From this point onwards the ethanol mass fraction of the droplets is larger than 0.95 and, subsequently the droplet temperature increases monotonically as until it reaches thermal equilibrium with the surrounding air/HFA vapour mixture. The location of this equilibrium state is around 30 mm from spray orifice (i.e. 10 mm outside the mouthpiece exit). Such findings suggest that heat and mass transfer to and from the droplets are significant only in the near-orifice region. Concluding remarks This work has shown an overview of a CFD simulation coupled with a model of the flow inside the actuator and an atomisation model to evaluate the development of the aerosol plume produced by a pMDI. Flow distributions show that the high-momentum of the spray source dominates the flow after 30 ms when sufficient mass of aerosol has entered the USP-IP confirming the need for an accurate description of the spray source. Our internal flow model and spray source calculations indicates that droplets containing a HFA/ethanol mixture emerge from the spray orifice with a typical size range 3-5 µm and that their size rapidly reduces within the mouthpiece region due to evaporation of the remaining volatile HFA, as suggested by temperature trace of figure 7. After an axial distance of 30 mm from spray orifice the droplets have reached their final size and temperature, and the aerosol and air/HFA vapour gas phase carry the droplets towards the USP-IP vertical section and exit face. Such findings may indicate that key plume dynamics in terms of interphase phenomena takes place only in near-orifice region. Current CFD model of pMDI may be linked with alternative models such as (i) turbulent diffusion of droplets, and (ii) droplet-wall impaction, to estimate oropharyngeal deposition on the walls of USP-IP as the first approximation, and realistic mouth-throat geometry as the ideal version. The outcome of this work may also serve as a design tool to predict evolution of spray characteristics as a function of device geometry and formulation composition. Acknowledgment The authors would like to thank Dr. Konstantinos Karantonis and Dr. Kristian Debus from CD-Adapco, for providing required licensing and relevant Star-CCM+ CFD modules. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

Dolovich, M.B. &Fink, J.B. : Respiratory Care Clinics of North America,2001; 7(2):pp131-174. Fletcher, G.E. PhD thesis, Loughborough University of Technology, Loughborough, UK, 1975. Clark, A.R. PhD thesis, Loughborough University of Technology, Loughborough, UK, 1991. Gavtash, B., Myatt, B., O’shea, H., Mason, F., Lewis, D., Church, T., Versteeg, H.K., Hargrave, G., And Brambilla, G.: Saturated vapour pressure (SVP) measurement of ethanol/HFA binary mixtures. Journal of Aerosol Medicine and Pulmonary Drug Delivery, 2016, 29(3), A1-A25. Versteeg HK, Hargrave GK, Kirby M. Internal Flow and Near-Orifice Spray Visualisations of a Model Pharmaceutical Pressurised Metered Dose Inhaler. J. Phys. Conf. Ser. 2006;45:207-213. doi:10.1088/1742-6596/45/1/028. Senecal, P., Schmidt, D.P., Nouar, I., Rutland, C.J., Reitz, R.D. and Corradini, M. International Journal of Multiphase Flow, 1999; 25(6):1073-1097. Gavtash, B., Versteeg HK, Hargrave, G.K., Lewis, D, Church, T., Brambilla, G.: Linear Instability Sheet Atomisation (LISA) model to predict drplet size issued from a pMDI. In: ILASS – Europe 2014, 26th Annual Conference on Liquid Atomization and Spray Systems, Sep. 2014, Bremen, Germany. Gjellerup, C., Frederiksen, S.O.: CFD and PIV investigation of the flow inside the USP throat and in a replica of the human upper airways. Joint Dissertation (MSc), Techical University of Denmark, 2007 Worth Longest, P., Hindle, M., Das Choudhuri, S., Xi, J.: Comparison of ambient and spray aerosol deposition in a standard induction port and more realistic mouth–throat geometry. Journal of Aerosol Science, 2008, 39(7), 572-591. Gavtash, B., Versteeg, H.K., Hargrave, G., Lewis, D., Church, T., Brambilla, G., Myatt, B., O’Shea, H. and Mason, F., 2016, June. CFD Simulation of pMDI Aerosols in Confined Geometry of USP-IP using Predictive Spray Source. Journal of Aerosol Medicine and Pulmonary Drug Delivery, 2016, 29(3), A1-A25.

76


Drug Delivery to the Lungs 27, 2016 – Peter York Innovations in formulations to improve aerosol delivery Peter York Chief Scientist, CrystecPharma Whilst innovation in inhaler devices and device components has been a regular occurrence over recent years, the same degree of novelty has not taken place in formulation science for drug delivery systems for inhalation. Perhaps a major constraint in innovation for formulation design is the limited number of excipients currently approved by the regulatory authorities for inhalation medicines, and the somewhat demanding requirements associated with the approval of additional excipients by regulators to this list even for those materials already approved for alternative routes of delivery. There are however a small number of exceptions to this scenario. The regulatory enforced changes associated with the new approved propellants for MDIs have been managed with the introduction of reformulated products, and through the approval of MDIs containing new NCEs onto the market. Also, with the growing desire to formulate inhaled medicines containing biotech based therapeutic agents, the drive to identify other suitable excipients to enable the preparation of stable, quality and efficacious inhalable medicines has increased. And there has been success in this particular field, although great care must be taken when generating micron sized particles of such sensitive materials. For DPIs, the ubiquitous sugar lactose remains the primary filling/bulking agent, which in most formulations is blended with the milled 1 – 5 micron sized drug particles, with milling achieved almost universally for such systems by micronisation. Whilst micronisation is an effective particle size reduction process, its disadvantages such as generating highly charged cohesive products, introducing amorphous hygroscopic domains, and interbatch variation, have been well recognised for some time, and efforts to mitigate and manage these effects have met with only limited success. Fruitful scientific studies have researched a range of options for manipulating lactose particles to achieve improved aerosol generation characteristics, such as modification of particle surface rugosity and energetics, and sequential blending of drug powder with different particle sized fractions of lactose, as well as incorporating low levels of nanosized powder flow aids. Further helpful research has introduced the concept of ‘cohesive-adhesive balance’ (CAB), which quantifies requirements for the ‘smart’ drug – lactose blend in terms of a creating a mix with sufficient interparticle strength to prevent demixing on handling and packaging, whilst able to demix and free adhering drug particles once the aerosol is generated by the patient operating the inhaler device. It is perhaps in the area of drug particle design that the most innovative features for improving the formulation of aerosols for inhaled drug delivery has taken place. New ‘bottom up’ approaches which transform a drug solution to dry, respirable sized drug particles have been researched, developed, and are operating at scaled-GMP level. These processes are in contrast to the ‘top-down’ milling of conventionally crystallised large drug particles, as exemplified by micronisation. The ‘bottom up’ particle formation procedures, including controlled crystallisation, spray drying, spray drying followed by sonification and supercritical fluid technologies, overcome many of the challenges resulting from high energy milling operations. Indeed in some cases the need for a carrier can be eliminated, inter-batch variability is dramatically reduced, free-flowing micron sized particles are obtained, and formation of composite drug-excipient and drug-drug particles is possible. Such methodologies are proving particularly beneficial when generating inhalable particles of biotech based therapeutic agents. The presentation will highlight the key points referred to above and provide examples. Additionally, the innovative features of the ‘bottom up’ processes will be illustrated and assessed, together with comments on current limitations and perceived requirements when designing formulations for respiratory drug delivery.

77


Drug Delivery to the Lungs 27, 2016 - Sanders MJ et al. Aspergillus fumigatus in valved holding chambers: use of silver ion additive technology on fungal activity and aerosol delivery characteristics 1

1

2

Sanders MJ , Bruin R & Tran CH 1

Clement Clarke International Limited, Edinburgh Way, Harlow, Essex, CM20 2TT, UK i2c Pharmaceutical Services, Cardiff Medicentre, Heath Park, Cardiff, CF14 4UJ, UK

2

Summary Valved holding chambers (VHCs) are an essential auxiliary device for pressurised metered dose inhalers (pMDIs). However, VHCs are capable of hosting fungi, thereby possibly promoting pulmonary aspergillosis via contaminated devices. VHC-polymers incorporating silver ions have been shown to resist the growth of certain microbes. The present research examined the in vitro effects on (i) the spored activity of Aspergillus fumigatus and (ii) the aerosol particle size distribution (APSD) and dose characteristics of salbutamol (Ventolin® Evohaler® pMDI) from the VHC Able Spacer® 2 with either ~1wt% (the standard commercially-available device) or 4wt% silver ions incorporated. Fungal activity was determined using a modified ISO 22196:2011 methodology for contaminant splashes onto non-porous surfaces, using flat discs of ~1% and 4% silver ion polymer material. APSD was determined from standard Next Generation Impactor (NGI) testing. Fungal activity of A. fumigatus was reduced at 24h when silver ions were present, compared to Control, and there was a five-fold Log10 difference between the ~1% and 4% silver ion material. There were no differences in salbutamol aerosol characteristics and APSD delivered via the standard Able Spacer 2 (~1% silver ions) and the test Able Spacer 2 (4% silver ions). The increase in silver ions did not alter the performance of the Ventolin pMDI. In vivo, the connection between VHC contamination and infection, and the aerosolisation of chamber contaminants remain unanswered. However, the availability of a new spacer with enhanced silver ion concentration for at-risk patients could extend the potential for prevention of concomitant fungal lung disease. Introduction Valved holding chambers (VHCs) are an essential auxiliary device for improving drug delivery from, and reducing local corticosteroid side effects attributable to, pressurised metered dose inhalers (pMDIs). The presence of microbes in VHC bodies is well known, with evidence of contamination with Pseudomonas, Klebsiella and [1-3] Staphylococcus spp., and also fungal Candida spp. We have previously conducted research demonstrating 72h survival of the bacterium K. pneumoniae — a common opportunistic pathogen for patients with chronic pulmonary disease — on two types of VHC polymer material, determining a potentially positive effect from an [4] antimicrobial additive; plus drug- and bacteria-specific growth promotion in the presence of commonly inhaled [5] drugs. Of potentially greater importance, however, is the ubiquitous opportunistic fungal pathogen Aspergillus [6] fumigatus. Inhalation is the main route of infection, with fungal spores capable of reaching the alveoli. The hundreds of spores inhaled on a daily basis do not challenge a robust immune system but immunologically compromised patients (neutropenics, transplantees, AIDS and chronic obstructive pulmonary disease [COPD] patients, particularly those receiving chronic high-dose steroid therapy) are at risk of Aspergillus lung diseases. [7] Invasive aspergillosis carries a high mortality rate, with COPD being a potential risk factor in up to 20% of cases. Allergic bronchopulmonary aspergillosis (ABPA), a non-invasive form, is most commonly seen in asthma and [8] cystic fibrosis patients. A case of non-invasive laryngeal aspergilloma, thought to be secondary to long term [9] inhaled steroid therapy, has also been reported. With spacer-type devices capable of hosting fungi, the possibility exists that, with a single breath from a contaminated VHC, a patient would receive a high-level fungal challenge in a matter of a few seconds. It is troubling to think that the high-dose inhaled steroid therapy used in the treatment of ABPA could conceivably promote pulmonary colonisation with Aspergillus via contaminated spacers. Prevention should be the preferable strategy. The presence of silver ions is known to confer anti[10,11] microbial properties, and a VHC-polymer incorporating silver ions has been shown in vitro to resist the growth [4] of certain respiratory microbes. Our present research is in two parts: a determination of the effect of two concentrations of silver ion additive on the spored activity of A. fumigatus on VHC body polymer in vitro and on the aerosol performance of two VHCs devices. Experimental Methods The Clement Clarke Able Spacer® 2 VHC was used as the exemplar device. The VHC has the dimensions 155 x 55 x 66mm, a weight of 62g, and a volume of 210ml. The transparent ABS polymer plastic body incorporates an anti-microbial silver ion additive (Figure 1) at a concentration of approximately 1 wt% (~1%). A new version of the Spacer, retaining the same physical properties, with 4 wt% (4%) silver ion additive technology has been developed (Figure 2), the purple colouration serving only to make the distinction between identical objects. [12]

The test The test methodology to determine A. fumigatus spore activity was a modification of ISO 22196:2011 materials (n=3 per sample group) were flat discs of Able Spacer® 2 body polymer with ~1% and 4% concentration of silver ion additive and the same polymer without silver ion additive (sterile Control, n=6). 100L distilled water 5 aliquots of A. fumigatus (5.0 x 10 spores/mL), were placed on the surface of each material sample. Both sets of silver ion-containing samples and three Control samples were incubated in 35C/≥95% relative humidity chambers for 24 hours.

78


Drug Delivery to the Lungs 27, 2016 - Aspergillus fumigatus in valved holding chambers: use of silver ion additive technology on fungal activity and aerosol delivery characteristics All samples were neutralised with soybean-casein-digest-lecithin-polysorbate 80 (SCDLP) medium to inactivate silver ions, with the three unincubated Control samples providing 0 hour data. Numbers of A. fumigatus colony 2 forming units (CFU) were established from spiral dilution. Data were converted to Log10 CFU/cm .

Figure 2 – Standard Able Spacer® 2 (left) and Able Spacer® 2 with 4% silver ions (right)

Figure 1 - Able Spacer® 2 connected to a pMDI

The dose characteristics and aerosol particle size distribution (APSD) of Ventolin® Evohaler® pMDI (n=5, GSK, 100g of salbutamol as salbutamol sulphate) were determined using the Next Generation Impactor (NGI, Copley Scientific Limited, Nottingham, UK) operated at 30 L/min, comparing the standard, commercially available Able Spacer 2 (~1% silver ion additive) and the test Able Spacer 2 with 4% silver ion additive (Figure 2). Prior to assessment, the VHCs were washed in 25°C warm soapy water: disassembling into mouthpiece, main section and actuator adaptor components; gently agitating for 30 seconds before air drying and reassembly. pMDIs were primed, shaken for 5 seconds, and used according to the Patient Instruction Leaflet. Once primed, five actuations per pMDI were delivered through the Able Spacer 2 to the NGI, with measurements taken before and after sampling to determine total actuation weights. Able Spacer components, Ventolin® pMDI actuator, induction port and NGI stages were washed with recovery solution (10-50mL) and drug quantitation determined from validated HPLC analysis, using standard solutions and methodologies. Data were presented and reported using a validated spreadsheet tool, permitting the calculation of individual actuation aerosol properties including g dose delivered, fine particle dose (FPD, g <5.0 m particles), and fine particle fraction (FPF, % <5.0 m). Particle size (MMAD, m) and GSD (σg) were also determined. Results Spored activity of A. fumigatus on VHC body polymer samples was reduced at the 24 hour time point compared with sterile Control when silver ion additive was present (Table 1). A silver ion concentration effect was apparent, with a five-fold Log10 difference between ~1% and 4% of the additive. 2

Table 1. Viability of A. fumigatus expressed as CFU/cm at 24 hours Sample material (n=3)

2

24h A. fumigatus CFU/cm (geometric mean)

Log10 reduction compared with Control (%)

4.20 x 10

3

Able Spacer 2 (~1%)

2.80 x 10

3

0.2 (33.5%)

Able Spacer 2 (4%)

4.50 x 10

2

1.0 (89.3%)

Sterile Control

There were no differences in the aerosol properties (Table 2) of salbutamol delivered via the standard Able Spacer 2 (~1% silver ion additive) and the test Able Spacer 2 (4%). Similar APSD patterns (Figure 3) were determined for the standard Able Spacer 2 (~1% silver ion additive) and the test Able Spacer 2 (4%). The increase in silver ions did not alter the performance of the Ventolin pMDI.

79


Drug Delivery to the Lungs 27, 2016 - Sanders MJ et al. Discussion The inclusion of ~1% and 4% concentration silver ion polymer additive to otherwise identical VHC devices did not affect the in vitro aerosol performance characteristics of salbutamol pMDI. The 4% silver ion additive demonstrated a superior performance in reducing Aspergillus fumigatus spored activity in vitro when compared with sterile control and a ~1% additive, evidenced by a modest but relevant Log10 reduction. This result suggests that a higher concentration of silver ions is required to combat the fungus Aspergillus than that which is helpful to combat bacteria. Table 2. Summary of NGI aerosol performance per actuation, n=5 pMDIs (mean ± SD) Aerosol property Metered weight (mg)

Able Spacer 2 ~1% silver ion additive

4% silver ion additive

74.0 ± 0.3

73.8 ± 0.6

Metered dose (g)

108.6 ± 9.4

107.4 ± 9.2

Emitted dose (g)

95.9 ± 11.0

94.9 ± 9.1

54.0 ± 4.3

53.6 ± 1.9

MMAD (m)

2.22 ± 0.05

2.18 ± 0.04

GSD (σg)

1.66 ± 0.02

1.67 ± 0.01

FPD (g <5.0 m)

52.0 ± 8.7

50.7 ± 3.8

On Actuator (g)

12.8 ± 3.6

12.6 ± 1.6

FPF (% <5.0 m)

Missing from our current research are data reporting culture experiments of A. fumigatus and local surface moisture from VHCs/spacers. A review of the literature suggests that culture data have not yet been reported. The [13,14] [14] literature does however report contamination of nebulisers and humidifiers. These are moist environments but the use in VHCs of anti-static plastics that work through hydrophilic processes does create local surface moisture that could favour conditions for fungal growth in what may appear to be a dry chamber. There is a tension between VHC hygiene recommendations (manufacturers’ weekly versus UK/US Guidance for monthly washing) and device effectiveness owing to static build-up whereby decreasing the number of VHC washes will allow drug sediment to accumulate in the chamber, decreasing the static load and potentially promoting microbial growth. We also do not yet know if microbes aerosolise within the chamber and through the one-way valve, or the in vivo effectiveness of antimicrobial plastics in this clinical scenario. Answers to these questions may impact recommended wash/dry cycles, and the relative importance of the use of anti-static polymers with or without anti-microbial additives.

Figure 3 – Aerosol particle size distribution recovery of salbutamol via Able Spacer® 2 with ~1% and 4% silver ion additive (mean values ± SD, n=5)

80


Drug Delivery to the Lungs 27, 2016 - Aspergillus fumigatus in valved holding chambers: use of silver ion additive technology on fungal activity and aerosol delivery characteristics Conclusion Previous research indicates that VHCs, spacers, and face masks suffer microbial contamination; K. pneumoniae can survive on VHC-body samples and drug residue can promote microbial growth in vitro. The current work extends these findings, demonstrating Aspergillus fumigatus spore activity at 24h and indicating a reduction in activity proportional to VHC silver ion additive concentration. Questions remain unanswered: the connection between contamination and infection; the aerosolisation of chamber contaminants; washing and/or use of antistatic plastics. Clement Clarke are, however, now exploring the possibility of introducing a new Able Spacer with enhanced silver ion concentration for minimal additional cost (no more than 5%) on the basis that at-risk patients should have access to the potential for prevention. References 1

Cohen H A, Cohen Z: Bacterial contamination of spacer devices used by children with asthma, J Amer Medical Assoc 2003; 290 (2): pp195-196.

2

Cohen H A, Cohen Z, Pomeranz A S, Czitron B, Kahan E: Bacterial contamination of spacer devices used by asthmatic children, J Asthma 2005; 42 (3): pp169-172.

3

De Vries T W, Rienstra S R, van der Vorm E R: Bacterial contamination of inhalation chambers, results of a pilot study, J Aerosol Med 2004; 17 (4): pp354-356.

4

Sanders M, Bruin R: Klebsiella pneumoniae survival on plastic valved holding chamber (VHC) bodies. Presented at: The International Society for Aerosols in Medicine / Woolcock Institute of Medical Research Conference, Sydney, Australia, September 25-26, 2014; J Aerosol Med Pulmon Drug Del 2014; 27 (6): A-8; doi:10.1089/jamp.2014.ab02.abstracts

5

Sanders M, Bruin R: Valved holding chambers (VHCs): Can drug residue promote microbial growth? Presented at: 8th International Primary Care Respiratory Group Conference, Amsterdam, Netherlands, May 25-28, 2016; Primary Care Respir Med 2016; 26 (16022): 16; doi:10.1038/npjpcrm.2016.22

6

O’Gorman C M: Airborne Aspergillus fumigatus conidia: a risk factor for aspergillosis, Fungal Biology Reviews 2011; 25 (3): pp151-157. doi:10.1016/j.fbr.2011.07.002

7

Ruiz-Camps I, Aguado J M, Almirante B, Bouza E, Ferrer-Barbera C F, Len O, Lopez-Cerero L, Rodriguez-Tudela J L, Ruiz M, Sole A, Valejo C, Vazquez L, Zaragoza E, Cuenca-Estrela M: Guidelines for the prevention of invasive mould diseases caused by filamentous fungi by the Spanish Society of Infectious Diseases and Clinical Microbiology (SEIMC), Clin Microbiol Infect 2011; 17 (suppl 2): pp1-24. doi:10.1111/j.1469-0691.2011.03477.x

8

Kousha M, Tadi R, Soubani A O: Pulmonary aspergillosis: a clinical review, Eur Respir Rev 2011; 20 (121): pp156-174. doi:10.1183/09059180.00001011

9

Darley D, Lowinger D, Plit M: Laryngeal aspergilloma: a complication of inhaled fluticasone therapy for asthma, Respir Case Reports 2014; 2 (4): pp123-125. doi:10.1002/rcr2.70

10

Prabhu S, Poulose E K: Silver nanoparticles: mechanism of antimicrobial action, synthesis, medical applications, and toxicity effects, Int Nano Lett 2012; 2: pp32-41. doi:10.1186/2228-5326-2-32

11

Panáček A, Kolář M, Večeřová R, Prucek R, Soukupová J, Kryštof V, Hamal P, Zbořil R, Kvítek L: Antifungal activity of silver nanoparticles against Candida spp., Biomaterials 2009; 30 (31): pp6333-6340. doi:10.1016/j.biomaterials.2009.07.065

12

International Organization for Standardization: Measurement of antibacterial activity on plastics and other non-porous surfaces, ISO 22196:2011. Edition 2, July 21 2011.

13

Peckham D, Wynne S, Denton M, Pollard K, Barton R: Fungal contamination of nebuliser devices used by people with cystic fibrosis, J Cystic Fibrosis 2013; 12 (Suppl1): ppS88, doi:10.1016/S1569-1993(13)60297-0

14

Jadhav S, Sahasrabudhe T, Kalley V, Gandham N: The microbial colonization profile of respiratory devices and the significance of the role of disinfection: a blinded study, J Clin Diagn Res 2013; 7 (6): pp1021-1026, doi:10.7860/JCDR/2013/5681.3086

81


Drug Delivery to the Lungs 27, 2016 - Antonio Cabal et al. In-Silico Lung Modeling Platform for Inhaled Drug Delivery 1

1

1

2

Antonio Cabal , Guido Jajamovich , Khamir Mehta , Peng Guo , & Andrzej Przekwas

2

1

2

Merck & Co., Inc., Kenilworth, NJ USA CFD Research Corp., 701 McMillian Way NW, Huntsville, AL, 35806, USA

Summary The inability to measure local lung concentrations responsible for lung efficacy is the main challenge common to any inhalation drug delivery program targeting the lungs. The model described in this work is a multiscale mechanism-based integrated computational platform (lung platform) to provide mechanistic insights into key physiological elements associated with pulmonary drug delivery: deposition, mucociliary clearance, dissolution, absorption, transport, distribution, partition, and action. Two versions of the lung platform (LP) were developed for translational purposes, one for rats (RLP) and one for humans (HLP) to account for the species specific physiological based differences in airway morphology and drug distribution throughout the body. All these components facilitate a prediction of regional distribution of drug within the lungs. Published data for two inhaled corticosteroids (mometasone furoate, budesonide), a short-acting beta-agonist (salbutamol), and a long-acting beta-agonist (formoterol) in both rats and humans were used to qualify the model and illustrate its prediction ability. The translational pharmacokinetic benefits of the lung platform are illustrated using the selected compounds, where the physicochemical properties and the drug delivery details constitute the main input to the model. Clearance was the only parameter adjusted using weight based allometric scaling to translate from rats to humans. The results show the feasibility of the proposed modeling approach, once it has been properly qualified using existing data, toward bridging the gap between inhaled data in preclinical species and the prediction of human lung and systemic exposure. Introduction The major challenge in the development of inhalable compounds is limited understanding of the relationship [1] between pharmacokinetic (PK) and pharmacodynamic (PD) effects in the lung . Currently, determining effective drug concentration in the lung after an inhaled drug delivery is semi-quantitative at best, making accurate [1, 2, 3] . Modeling PK/PD relationships accurately requires an prediction of the clinical dose very challenging understanding of free drug concentrations in the lung at any given time. Several attempts to develop in silico [4] have provided a useful tool for simulating systemic PK models, such as the work by Weber and Hochhaus following inhaled treatment with corticosteroids in humans. The goal of the work presented here is to develop and utilize a mathematical framework to understand the deposition, dissolution, distribution, and kinetics of a compound following inhaled dosing. This framework is used to develop two species specific, rat and human, versions of the model with the purpose of leveraging the in-vitro and preclinical data generated in the early phases of research to address the main translational challenge of inhaled drug development: inability to measure lung PK in humans in order to make reasonable human PK/PD projections. Lung Platform Figure 1 shows a schematic of the different modules that constitute the LP. The drug can be delivered into the lungs in two ways, either intratracheally (gavage needle, attached to a syringe is inserted into the rat’s trachea) or through inhalation. Assuming a dry powder formulation, the inhaled modeled particles will be deposited in the LP into three main regions of the respiratory tract: the upper tract (nasal cavity, oral cavity, pharynx, and larynx), tracheobronchial region (trachea, bronchi, bronchioles, and terminal bronchioles), and pulmonary region (respiratory bronchioles, alveolar ducts, alveolar sacs, and alveoli). Once the particles are deposited in the different regions of the respiratory tract according to their size and momentum, they start dissolving in the surface lining fluid while simultaneously being cleared from the airway region (generation 0 to 15) due to the action of the mucociliary escalator. The dissolved drug partitions into seven different lung tissue compartments (surface lining liquid, macrophages, epithelial cells, interstitium, smooth muscle cells, immune cells, and endothelial cells) before reaching the systemic circulation. The physiologically based PK (PBPK) module accounts for the drug distribution and partition throughout the body, as well as elimination. The compartmental absorption and transit (CAT) module is responsible for oral drug dissolution, absorption and transit through the gastrointestinal tract (GIT). It also manages the portions of the inhaled doses that end up in the GIT during inhalation or are swallowed after inhaled delivery due to mucociliary clearance. Mentioned below are the key assumptions for the model. 1.

A typical path lung model is used assuming airways branch 24 times where the bronchial tree is [5] considered to be a series of parallel semi-rigid cylindrical tubes connected in series in a bifurcation tree

2.

All deposited particles are monodisperse spheres which dissolve at a rate proportional to the difference [6, 7] . between the amount of solid active pharmaceutical ingredients (API) and its concentration in solution

82


Drug Delivery to the Lungs 27, 2016 - In-Silico Lung Modeling Platform for Inhaled Drug Delivery 3.

The transport of chemical species through the lung airway walls is determined by passive diffusion of neutral molecules (driven by concentration gradient), and the diffusion generated by the chemical [8] potential of ionized molecules

4.

A PBPK model reasonably describes the drug disposition throughout the body

[9, 10]

The PD module is used as an independent add-on module which will be different for compounds with different [11] mechanisms of action. Details for the mometasone specific PD-module were published earlier this year .

Figure 1: Schematic of the multi-scale generic lung platform

Results The key inputs driving the LP are the compound-specific physicochemical properties (permeability, solubility, lipophilicity, acid-ionization, molecular weight, unbound fraction, blood to plasma ratio, etc.) and the details about the specific type of lung delivery: inhaled or intratracheal (IT). Rat and human data for four different compounds; mometasone furoate (MF), budesonide (BD), salbutamol (SB), and formoterol (FT), was identified and digitized from published literature in order to qualify the RLP and HLP. i.

Model Qualification [11, 12, 13]

and We qualified the model using PK data from diverse compounds (MF, BD, SB, and FT) in both rats [14, 15, 16, 17] . For each case, the input data for the model constituted the drug physicochemical properties humans while the PK clearance parameter was either adjusted based on the data or used from literature if available. Single dose data then was used to check the agreement of the model with the observed PK profile in the systemic circulation for both rat and humans. Figure 2 shows the systemic total drug concentration as predicted by the model (solid lines) following IT or inhaled delivery. The measured concentration observations are shown as individual data points (color matched symbols). No rat PK data was available for salbutamol.

Figure 2: Comparison between preclinical (left) and clinical (right) data with the corresponding RLP and HLP predictions

Since the only adjustable parameter in the model was the clearance, we checked the model estimated human clearance parameter and compared that with the values as estimated from a typical weight based allometric scaling using the values for rat.

83


Drug Delivery to the Lungs 27, 2016 - Antonio Cabal et al. We found our results to be similar to the ones computed from allometric scaling, which fortifies further, our assessment that the model has the requisite elements for translating drug candidates from preclinical species to clinical scenarios while enabling the prediction of local lung concentrations for both the species. ii.

Lung vs. Systemic PK Profiles The selective abundance of the drug in the lung compared to the systemic circulation, after pulmonary delivery is one of the key advantages of inhaled drug delivery. For inhaled therapeutics where the drug is delivered directly to the target, the lung, it is essential to know the difference between the local lung PK and the systemic PK. This difference allows for smaller inhaled doses than corresponding oral ones, accelerates the onset of local lung effect, and improves safety profile of drugs (systemic adverse effect are less severe and less frequent) by significantly reducing systemic exposure. The ratio of lung concentration to its corresponding systemic concentration after inhaled drug delivery can be strongly influenced by the physicochemical properties of the compounds evaluated. For instance in the case of dry powder formulations, low solubility can effectively mete out the drug at a lower rate in the lungs, yielding a [11] . Increases in lipophilicity (logP) influence the transport rate of the drug lower systemic peak concentration through various physiological compartments after it has been dissolved, resulting in the drug traversing the various lung compartments more rapidly and entering into the systemic circulation. Low LogP yields slower [8, 11] . The LP facilitates the evaluation of these transport rates, suggesting an improvement in lung residency time types of “what if� scenarios to generate testable hypothesis that could be used to optimize the desired drug characteristics. Similar analyses done through experimental trial and error tests would be very costly and ineffective due to the difficulties associated with varying a single property of chemical entities while keeping all the other properties unchanged.

Figure 3: Comparison between preclinical (left) and clinical (right) predicted systemic (thin lines) and lung (thick lines) PK

Once the dimensions of the respiratory tree are defined, drug deposition in the rat and human LP can be [5] calculated from the typical path lung model . Three mechanisms are considered to predict the deposition fraction [5] in each generation: deposition by diffusion, sedimentation, and inertial impaction . The overall deposition probability at each airway branch generation in one breathing cycle is the sum of the deposition probability during inhalation, breath holding, and exhalation. The deposition module was qualified comparing experimental [5,18-20] with regional depositions of monodisperse particles with measurements collected from literature aerodynamic diameters ranging from 0.1 to 10 Âľm. After drug microparticles are deposited in the respiratory tract, which is divided into 25 generations from trachea to the alveoli, they start dissolving in the liquid lining the surface of the lungs. Drug microparticles are at the same time cleared from the airway region (generations 0 through 15) due to mucus motion because of cilia beating. The mucociliary escalator does not exist in the alveolar region (generations 15 through 24). It carries the undissolved [21] . Drug cleared from the airways could be drug toward the oral region with a species-specific mucus velocity swallowed, in which case it will enter the GIT through the CAT module, shown on the right side of the LP schematic (see figure 1). Figure 3 shows the difference between the simulated lung and systemic concentrations for all the compounds evaluated in this analysis. Our results show at least two orders of magnitude difference in pulmonary and systemic drug concentrations during the terminal decay phase of the PK for both rat and humans, except for salbutamol which is about 50-fold. The Cmax and exposure values for each of the compounds, species, and doses plotted in figure 3 are listed in Table 1. For compounds where the magnitude of the PD response is proportional to the PK exposure, large differences between lung and systemic exposure after inhalation widens the therapeutic window of the compound when compared with oral or intravenous delivery of the same drug. .

84


Drug Delivery to the Lungs 27, 2016 - In-Silico Lung Modeling Platform for Inhaled Drug Delivery LP helps mitigate some of the many difficulties associated with the development of inhaled drugs by incorporating the interspecies differences in airway morphology which have a significant influence in interspecies variability observed after drug inhalation The addition to the LP of normal defence mechanisms like mucociliary clearance supports the need to account for an important physiological function that affects the regional distribution of drug within regions of the respiratory tract. Integrating every one of these modules (figure 1) and all the knowledge associated with them provides a valuable tool to help support decision making in the challenging field of research and development for inhalation delivery of drugs. Table 1: Cmax and exposure (24-hour AUC) for each of the compounds plotted in figure 3. A nominal 1 mg dose was delivered in all the human simulations.

Compound

Plasma

Plasma

Lung

1.85E+03

6.25E+05

9.59E+02

2.39E+05

2.55E+03

7.30E+05

4.38E+03

9.21E+05

4.69E+03

5.17E+05

2.96E+03

1.93E+05

AUC0-24hr [pg*h/mL]

3.54E+03

2.73E+05

1.02E+04

5.83E+05

Cmax [pg/mL]

1.34E+05

3.17E+07

5.10E+01

1.71E+04

2.02E+05

3.18E+07

2.36E+02

4.84E+04

Cmax [pg/mL]

4.02E+03

1.65E+06

8.91E+02

2.46E+05

AUC0-24hr [pg*h/mL]

6.85E+03

2.22E+06

3.10E+03

7.44E+05

max

AUC

Mometasone

Budesonide

max

AUC

[pg/mL]

0-24hr

C

Salbutamol

HUMAN

Plasma

C

Formeterol

RAT

[pg*h/mL]

[pg/mL]

0-24hr

[pg*h/mL]

Conclusions    

The lung platform provides an in-silico option to compensate for many of the difficulties associated with the measurement of the local drug concentration in target tissues of the lung. Rat and human versions of the lung platform were qualified using existing inhaled literature data on multiple compounds: mometasone furoate, budesonide, salbutamol, and formoterol. The lung platform predicted local lung concentrations vs. time profiles for rats and humans. In both species the lung concentrations are at least two orders of magnitude higher than the corresponding systemic concentration during the drug terminal decay phase. PK translation from rats to humans was done using the exact same compound specific physicochemical properties, allometrically scaled clearance, and the species specific dose delivery information. We hence demonstrate the ability of the LP to be used as a translational platform to make more accurate human projections from preclinical observations.

85


Drug Delivery to the Lungs 27, 2016 - Antonio Cabal et al.

References 1

Forbes B, Asgharian B, Dailey L A, Ferguson D, Gerde P, Gumbleton M, Gustavsson L, Hardy C, Hassall D, Jones R, Lock R, Maas J, McGovern T, Pitcairn G R, Somers G, Wolff R K: Challenges in inhaled product development and opportunities for open innovation. Adv Drug Deliv Rev 2011; 63: pp69-87.

2

Cooper A E, Ferguson D, Grime K: Optimisation of DMPK by the inhaled route: challenges and approaches. Curr Drug Metab. 2012; 13: pp457-73.

3

Cheng Y S: Mechanisms of pharmaceutical aerosol deposition in the respiratory tract. AAPS Pharm Sci Tech 2014; 15: pp630-40.

4

Weber B, Hochhaus G: A pharmacokinetic simulation tool for inhaled corticosteroids. AAPS J. 2013; 15: pp159–171.

5

Schum M, Yeh H: Theoretical evaluation of aerosol deposition in anatomical models of mammalian lung airways. Bull Math Biol. 1980; 42: pp1-15.

6

Wang J, Flanagan D R. General solution for diffusion-controlled dissolution of spherical particles. 1. Theory. J Pharm Sci. 1999; 88(7): pp731-8.

7

Wang J, Flanagan DR. General solution for diffusion-controlled dissolution of spherical particles. 2. Evaluation of experimental data. J Pharm Sci. 2002; 91(2): pp534-42.

8

Yu J, Rosania G. Cell-Based Multiscale Computational Modeling of Small Molecule Absorption and Retention in the Lungs. Pharm Res. 2010; 27: pp457-67.

9

Davies B, Morris T. Physiological parameters in laboratory animals and humans. Pharm Res.1993; 10: pp1093-5.

10

Rodgers T, Rowland M. Mechanistic approaches to volume distribution predictions: understanding the processes. Pharm Res. 2007; 24: pp918-33.

11

Caniga M, Cabal A, Mehta K, Ross D S, Gil M A, Woodhouse J D, Eckman J, Naber J R, Callahan M K, Goncalves L, Hill S E, Mcleod R L, McIntosh F, Freke M C, Visser S A G, Johnson N, Salmon M, Cicmil M: Preclinical experimental and mathematical approaches for assessing effective doses of inhaled drugs, using mometasone to support human dose predictions. J Aerosol Med Pulm Drug Deliv 2016; 29: pp1-16.

12

Jendbro M, Johansson C J, Strandberg P, Falk-Nilsson H, Edsbacker S. Pharmacokinetics of budesonide and its major ester metabolite after inhalation and intravenous administration of budesonide in the rat. Drug Metab Disp. 2001; 29: pp76976.

13

Tronde A, Norden B, Marchner H, Wendel A K, Lennernas, H, Bengtsson U H. Pulmonary absorption rate and bioavailability of drugs in vivo in rats: structure–absorption relationships and physicochemical profiling of inhaled drugs. J Pharm Sci. 2003; 92: pp1216-33.

14

Lecaillon J B, Kaiser G, Palmisano M, Morgan J, Della Cioppa G. Pharmacokinetics and tolerability of formoterol in healthy volunteers after a single high dose of Foradil dry powder inhalation via aerolizer. Eur J Clin Pharmcol. 1999; 55: pp131-8.

15

Affrime MB, Cuss F, Padhi D, Wirth M, Pai S, Clement R P, Lim J, Kantesaria B, Alton K, Cayen M N. Bioavailability and metabolism of mometasone furoate following administration by metered-dose and dry-powder inhalers in healthy human volunteers. J Clin Pharmacol. 2000; 40: pp1227-36.

16

Ward J K, Dow J, Dallow N, Eynott P, Milleri S, Ventresca G P. Enantiomeric disposition of inhaled, intravenous and oral racemic-salbutamol in man — no evidence of enantioselective lung metabolism. Br J Clin Pharmacol. 2000; 49: pp15-22.

17

Lähelmä S, Kirjavainen M, Kela M, Herttuainen J, Vahteristo M, Silvasti M, Ranki-Pesonen M. Equivalent lung deposition of budesonide in vivo: a comparison of dry powder inhalers using a pharmacokinetic method. Br J Clin Pharmacol. 2004; 59: pp167-73.

18

Kelly J T, Bobbitt C M, Asgharian B. In vivo measurements of fine and coarse aerosol deposition in the nasal airways of female long-evans rats. Toxicological Sci, 2001; 64: pp253-58.

19

Zhang L and Yu C P. Empirical equations for nasal deposition of inhaled particles in small laboratory animals and humans. Aerosol Sci Technol. 1993; 19: pp51-6.

20

Raabe O, Yeh H, Newton G, Phalen R, Velasquez D. Deposition of Inhaled Monodisperse Aerosols in Small Rodents. 1977 In: Inhaled Particles. IV (Walton, W.H. Ed.), New York: Pergamon Press: pp3-21.

21

Hoffman W, Asgharian B. The Effect of Lung Structure on Mucociliary Clearance and Particle Retention in Human and Rat Lungs. Toxicological Science. 2003; 73: pp448-56.

86


Drug Delivery to the Lungs 27, 2016 - Fowler

mHealth in asthma – friend or foe? Stephen J Fowler Division of Infection, Immunity and Respiratory Medicine, School of Biological Sciences, Faculty of Biology, Medicine and Health, The University of Manchester and University Hospital of South Manchester NHS Foundation Trust, Manchester UK. Summary: mHealth systems are promising tools for supporting asthma self-management and a user-centred design may improve compliance. We have determined the opinions of people with asthma and healthcare professionals (HCPs) on the use and functionality of mHealth self-management systems, initially by recruiting to focus groups. After framework analysis of the transcripts the resultant questionnaire was sent out to 186 people with asthma (91 uncontrolled) and 63 HCPs. Patients most frequently responded that they would like a mHealth system that allows them to monitor their asthma over time (72%) and that can collect data to present to their healthcare teams (70%). A system to alert patients to deteriorating asthma control (86%) and advising them when to seek medical attention (87%) were most frequently selected by HCPs. Patients were less likely than HCPs to believe that measuring medication adherence, inhaler technique and respiratory symptoms could help them achieve better asthma control. Furthermore, patients with uncontrolled asthma were more likely to believe that alerts to adverse environmental conditions, and low/high temperature and humidity, would be helpful in managing their asthma. In summary, this study provides insights from the perspective of patients and HCPs with regards to the use of mHealth systems for asthma self-management, which should be considered during the development of new mHealth systems. Main body: mHealth has the potential to revolutionize the self-management of long-term medical conditions such as asthma. However, the utility and efficacy of mHealth for asthma self-management has not been shown. A user-centred design is integral for mHealth to meet end-user expectations and may improve health outcomes. The myAirCoach project (http://www.myaircoach.eu/myaircoach/) aims to design and develop mHealth solutions for people living with asthma as well as healthcare professionals, in order to help them to gain better understanding and control of the disease. Such solutions would need to be tailored to individual needs, as it is well recognised that triggers and other risk factors for poor asthma control differ between individuals. The first phase of myAirCoach comprised a study in which we sought to obtain the perspectives and expectations of prospective end-users of myAirCoach – people with asthma and healthcare professionals who treat the disease on the use of mHealth for asthma self-management. We used a sequential exploratory mixed methods design. Initially focus groups (two involving individuals with asthma, and one involving healthcare professionals) were held. The analysis of this information then informed the development of questionnaires, which were widely disseminated to individuals with asthma and healthcare professionals. We identified differences in response frequency between patients with asthma and healthcare professionals and between patients with controlled and uncontrolled asthma. Participants included individuals with a doctor’s diagnosis of asthma from primary and secondary care in Manchester (UK), London (UK) and Leiden (the Netherlands), and healthcare professionals involved in the treatment of patients with asthma in the North West of England. The main outcome categories related to i) perceived uses; ii) useful measurements; and iii) barriers, of mHealth for asthma self-management Focus group participants (18 asthma patients and five healthcare professionals) identified 12 potential uses of mHealth. Questionnaire results showed that asthmatics (n = 186) most frequently requested an mHealth system to monitor asthma over time (72%) and to collect data to present to healthcare teams (70%). In contrast, a system alerting patients to deteriorating asthma control (86%) and advising them when to seek medical attention (87%) was most frequently selected by healthcare professionals (n = 63). Individuals with asthma were less likely than healthcare professionals (P<0.001) to believe that assessing medication adherence and inhaler technique could help them achieve better asthma control. In conclusion, our data provide strong support for mHealth for asthma self-management, but highlight fundamental differences between the perspectives of patients and healthcare professionals. Both end-users perspectives should be carefully considered for the successful integration of mHealth for asthma self-management.

87


Drug Delivery to the Lungs 27, 2016 - Alejandro Nieto-Orellana et al. Polymer-protein-based dry powder for efficient pulmonary protein delivery 1

2

2

1

2

Alejandro Nieto-Orellana , David Coghlan , Malcolm Rothery , Cynthia Bosquillon , Nick Childerhouse , 1 1 Giuseppe Mantovani & Snow Stolnik 1

2

School of Pharmacy, University of Nottingham, Nottingham NG7 2RD, United Kingdom Vectura Group plc, One Prospect West, Chippenham, Wiltshire, SN14 6FH, United Kingdom

Summary Pulmonary delivery of protein therapeutics has considerable clinical potential for treating many local and systemic diseases. However, poor protein conformational stability, immunogenicity and protein degradation by proteolytic enzymes are major challenges to overcome. To address these, a family of structurally related copolymers with linear A-B and miktoarm A-B3 macromolecular architectures were non-covalently complexed with the model protein lysozyme to generate well-defined nano-sized assemblies, which were then formulated into dry powders by spray-drying. Trehalose and leucine were used as bulking agent and disperser agent, respectively. The spraydried powders displayed high protein encapsulation efficiencies (80-100 %) and excellent aerodynamic properties with a fine particle fraction up to 68 %. In aqueous media, the dry powders rapidly released the polymer-protein nanocomplexes. Importantly, these retained their original size upon release from the dry powders. Finally, polymer-protein nanocomplexes provided protection of their protein payloads towards proteolytic degradation. The present study, therefore, shows that our approach could potentially lead to more efficient delivery systems for pulmonary administration of proteins. Introduction The design of polymer-protein nanosystems represents a potentially effective strategy to overcome barriers [1] and clearance by proteolytic associated with the pulmonary delivery of proteins such as protein stability enzymes in the lungs. Typically, polymers are covalently conjugated to proteins - e.g. in PEGylated protein therapeutics. However, this approach is not suitable for all proteins, as polymer conjugation might induce loss of their intended bioactivity. Within this context, nanocomplexes where synthetic polymers bind to proteins via non[2] covalent interactions are emerging as very valuable alternatives . However, the size of polymer-protein nanocomplexes is not suitable for direct delivery into the lungs. Therefore, their formulation into carrier systems [3] designed for pulmonary delivery is a requisite . Nowadays, dry powder inhalers are progressively outperforming [4] metered doses inhalers and nebulisation systems on the market due to their higher stability and efficiency . In this work, we aimed to develop inhalable spray-dried powders of a model protein complexed by biocompatible synthetic polymers into nanosized structures. The polymers are a novel family of A-B copolymers with either a linear A-B, or miktoarm A-B3 structure, where hydrophilic A polymer block provides stability to the nanocomplexes and B is a polyanion responsible for protein complexation through ionic interactions. The feasibility of preparing polymer-protein nanocomplexes into dry powders exhibiting excellent aerosolisation properties was investigated with the aim of identifying structure-function relationship for this novel family of complexing copolymers. Furthermore, the release of the complexes and following protein protection against proteolytic enzymes was evaluated. Experimental Methods Materials Sodium phosphate dibasic (99.9 %), sodium phosphate monobasic dehydrate (≥99.0 %), phosphoric acid, micrococcus lisodeikticus lyophilised cells, trypsin from bovine pancreas, poly(allyl amine) solution (Mw: 17,000 g -1 mol ), L-leucine and D-(+)-trehalose dihydrate were purchased from Sigma Aldrich. Lysozyme molecular biology grade and BCA protein assay kit were supplied by AppliChem and ThermoFisher Technologies, respectively. Preparation of polymer-protein nanocomplexes The copolymers used in this study possessed linear A-B or miktoarm A-B3 structures where the length of the hydrophilic A block was kept constant and that of protein-binding B arms was systemically varied. The number of monomer units “b” in B block arms were either 10 or 30, to give a library of Linear A-b10 (L10) and A-b30 (L30); and Miktoarm A-m-(b10)3 (M30) and A-m-(b30)3 (M90) copolymers. The copolymers and the lysozyme model protein were mixed in phosphate buffer (PB) (10 mM, pH 7.4) at relative molar charge ratios of 2.5 (ratio between the overall charge of the protein and charged “b” units of the copolymers).

88


Drug Delivery to the Lungs 27, 2016 - Polymer-protein-based dry powder for efficient pulmonary protein delivery Spray-drying conditions A laboratory scale spray-drying equipment developed by Vectura Group PLC (Chippenham, United Kingdom) was used to prepare the dry powders. In vitro pulmonary deposition The aerodynamic properties of the dry powders were assessed using a fast screening impactor (FSI; Copley -1 Scientific, UK). A standard dispersion procedure was conducted for 4 s at an air flow rate of 60 L min . The cut-off aerodynamic diameter was 5 µm. An amount of 12 ± 1 mg of powder was loaded into blisters and aerosolised using a proprietary unit dose DPI device (Vectura, UK). Particles with an aerodynamic diameter lower than 5 µm deposited on a filter in the fine fraction collector. This filter was weighed before and after the air actuation, in order to determine the fine particle fraction (FPF). The emitted dose (ED) was calculated by accurately weighing the blister before and after aerosolisation. Each powder was tested in triplicate. Protein incorporation efficiency The amount of protein encapsulated in the dry powders was evaluated by BCA assay. Dry powders were incubated in PB for 1 h under gentle stirring at room temperature - theoretical protein concentration 15-25 µg mL 1 . Then 150 µL of particle samples and 150 µL of BCA solution prepared as described by the manufacturer were mixed and heated at 56 ºC for 60 min. Once cooled at 25 °C, the absorbance was measured at λ=562 nm using a micro-plate reader SpectraMax M2. All the absorbance values were corrected by subtracting the values measured for relevant blank samples. Each measurement was performed in quadruplicate. The final protein encapsulation efficiency (%) was calculated from the ratio between the amount of protein detected and that of protein utilised to prepare the dry powders. Recovery of nanocomplexes from the dry powders in aqueous medium To investigate whether the nanocomplexes could be recovered from the dry powders, the latter were incubated in -1 PB for 1 h under gentle stirring at room temperature at a protein concentration of 70 µg mL (as quantified by BCA assay). The hydrodynamic diameter was measured by nanoparticle tracking analysis (NTA) using a Nanosight LM14. All measurements were performed at 25 °C. The NTA 2.0 Build 127 software was used for data capturing and analysis. The samples were measured for 80 s. Enzymatic activity In order to assess the enzymatic activity of lysozyme, a turbidimetric enzymatic assay was performed by measuring the decrease in optical density of a suspension of Micrococcus lysodeikticus lyophilized cells, a natural substrate for lysozyme, in PB. Nanocomplexes and also free lysozyme were extracted from the dry powders as -1 previously described, at a protein concentration of 37 µg mL (as quantified by BCA assay). 10 mg of lyophilised cells were added to 20 mL of PB. 300 µL of this substrate suspension were added to 150 µL of particle solution and the decrease in optical density was measured at λ=460 nm as a function of time for 60 sec. The absorbance decay plots from 0 to 10 sec were fitted to a linear equation and the enzymatic activities were then determined from the slope of the fitted line. Not spray-dried free lysozyme was used in this work as reference (100 % enzymatic activity). Protein protection against proteolytic enzymes Nanocomplexes were recovered from the dry powders as previously described, at a protein concentration of 37 . -1 µg mL (as quantified by BCA assay). 657 µg of trypsin were added to 4 mL of particle solution and the mixture was kept at room temperature for 3 h under gentle stirring. 480 µg of poly(allyl amine) were then added to the solution, and the mixture was kept under gentle stirring for 1 h. Lysozyme activity was then quantified as previously described. All kinetic experiments were carried out in triplicate. Results Spray-dried powders were characterised by suitable d50 diameter for inhalation ( 2.5 μm) and spherical morphology with smooth surface (data not shown). Polymer-protein nanocomplexes were found to be efficiently encapsulated into the dry powders achieving up to 100 % when the shorter polymers were used. However, spraydried powders with the longer polymers decreased the protein encapsulation efficiency to around 80 % (Figure 1A). In the powder dispersion experiment, the emitted dose (ED) and the fine particle fraction (FPF) were on average 95 and 65 %, respectively which showed that around 65 % of the dry powders were assumed to be fine particles optimal for inhalation (Figure 1-B). Interestingly, linear L10 polymers showed a significant improvement of the aerodynamic properties reaching FPFs values up to 68 %. Nevertheless, M90 copolymers decreased these values to 57 %. The hydrodynamic diameter of the reconstituted polymer-protein nanocomplexes was not significantly different from that of nanocomplexes before spray-drying (Figure 1-C), indicating that the spray-

89


Drug Delivery the Lungs 2016 - AlejandroNieto-Orellana Nieto-Orellana et Drug Delivery to thetoLungs 27, 27, 2016 - Alejandro et al. al. Polymer-protein-based dry powder for efficient pulmonary protein delivery

drying process did not affect the stability of these nanoassemblies. Importantly, the polydispersity of these redispersed nanocomplexes decreased. 1 2 2 1 2 Alejandro Nieto-Orellana , David Coghlan , Malcolm Rothery , Cynthia Bosquillon , Nick Childerhouse , 1

Giuseppe Mantovani & Snow Stolnik

1

1

A

2 120

School of Pharmacy, University of Nottingham, Nottingham NG7 2RD, United Kingdom B 120 **** * * Vectura *** Group plc, One Prospect West, Chippenham, Wiltshire, SN14 6FH, United Kingdom 100 E D a n d F P F (% )

P r o te in e n c a p s u la tio n

100 Summary

80

M

M

9

3

0

0

0

3

L

1

L

m

zy

o

s

y

L

M

M

9

3

0

0

0

3

L

L

1

0

e

m

L

y

s

o

zy

0

e

e ffic ie n c y %

80 Pulmonary delivery of protein therapeutics has considerable clinical potential for treating many local and systemic 60 diseases. However, poor protein conformational stability, immunogenicity and protein degradation by proteolytic 60 enzymes are major challenges to overcome. To address these, a family of structurally related copolymers with 40 linear4 0A-B and miktoarm A-B3 macromolecular architectures were non-covalently complexed with the model 20 protein lysozyme to generate well-defined nano-sized assemblies, which were then formulated into dry powders 20 by spray-drying. Trehalose and leucine were used as bulking0agent and disperser agent, respectively. The spraydried powders displayed high protein encapsulation efficiencies (80-100 %) and excellent aerodynamic properties 0 with a fine particle fraction up to 68 %. In aqueous media, the dry powders rapidly released the polymer-protein nanocomplexes. Importantly, these retained their original size upon release from the dry powders. Finally, polymer-protein nanocomplexes provided protection of their protein payloads towards proteolytic degradation. The present study, therefore, shows that our approach could potentially lead to more efficient delivery systems for D pulmonary administration of proteins.

C

b e fo r e s p r a y -d r y in g

N a n o c o m p le x e s a ft e r s p r a y -d r y in g N o r m a liz e d P a r tic le C o n c e n tr a tio n

N o r m a liz e d P a r tic le C o n c e n tr a tio n

N a n o c o m p le x e s Introduction 100

100

The design of polymer-protein nanosystems represents a potentially effective strategy to overcome barriers L10 [1] and clearance by proteolytic associated with the pulmonary delivery of proteins such 8 as protein stability L30 80 0 enzymes in the lungs. Typically, polymers are covalently conjugated to proteins - e.g. in PEGylated protein M 30 therapeutics. However, this approach is not suitable for all proteins, as polymer conjugation might induce loss of 60 60 M 90 their intended bioactivity. Within this context, nanocomplexes where synthetic polymers bind to proteins via non[2] . However, the size of polymer-protein covalent interactions are emerging as very valuable alternatives 40 40 nanocomplexes is not suitable for direct delivery into the lungs. Therefore, their formulation into carrier systems [3] designed for pulmonary delivery is a requisite . Nowadays,2 0dry powder inhalers are progressively outperforming 20 [4] metered doses inhalers and nebulisation systems on the market due to their higher stability and efficiency . 0

0

In this0 work, we aimed to develop inhalable6 0spray-dried powders of a model protein complexed by biocompatible 200 400 0 0 200 400 600 synthetic polymers into nanosized structures. The polymers are a novel family of A-B copolymers with either a H y d ro d y n a m ic d ia m e te r (n m ) H y d ro d y n a m ic d ia m e te r (n m ) linear A-B, or miktoarm A-B3 structure, where hydrophilic A polymer block provides stability to the nanocomplexes and B is a polyanion responsible for protein complexation through ionic interactions. The feasibility of preparing Figurepolymer-protein 1. (A) Proteinnanocomplexes encapsulationinto efficiency, (B)exhibiting (solid bars) final aerosolisation particle fraction (FPF) was andinvestigated (hatched bars) dry powders excellent properties emittedwith dose powders.structure-function (C, D) particle size distribution of nanocomplexes and after spray-drying the(ED) aimofofdry identifying relationship for this novel family of before complexing copolymers. as measured by NTA. ‘Lysozyme’ powders contained no protein polymer,. All formulations contained 6.6 was % (w/w) Furthermore, the release of the complexes and following protection against proteolytic enzymes evaluated. phosphate buffer salts. One-way ANOVA was used for statistical analysis. * p < 0.1, *** p < 0.001, **** p < 0.0001.

All nanocomplexes recovered from the dry powders exhibited a decrease in protein activity as compared to Experimental Methods lysozyme spray-dried with trehalose and leucine (Figure 2-A). Following incubation of the particles in the presence of trypsin, lysozyme was recovered from the complexes by addition of poly(allyl amine), and its residual Materials activity was assessed. In fact, poly(allyl amine) can ionically interact with A-B copolymers, inducing complex Sodium All phosphate dibasic (99.9 %), sodium dehydrate (≥99.0 %), phosphoric acid, disassembly. polymers, especially longer L30phosphate and M90 monobasic copolymers, protected lysozyme against enzymatic micrococcus lisodeikticus lyophilisedofcells, trypsinlysozyme from bovine pancreas, poly(allyl amine) solution 17,000 degradation; whereas in the absence polymer, activity dropped to less than 40 %(Mw: of its initialg value -1 ), L-leucine and D-(+)-trehalose dihydrate were purchased from Sigma Aldrich. Lysozyme molecular biology (Figuremol 2-B). grade and BCA protein assay kit were supplied by AppliChem and ThermoFisher Technologies, respectively. Preparation of polymer-protein nanocomplexes

****

B

N o rm a liz e d e n z y m a tic a c tiv ity (% )

The copolymers used in this study possessed linear A-B or miktoarm A-B3 structures where the length of the **** 120 20 hydrophilic A block was kept constant and that of protein-binding B1arms was systemically varied. The number of monomer units “b” in B block arms were either 10 or 30, to give a library of Linear A-b10 (L10) and A-b30 (L30); 100 100 and Miktoarm A-m-(b10)3 (M30) and A-m-(b30)3 (M90) copolymers. The copolymers and the lysozyme model 0 protein 8were mixed in phosphate buffer (PB) (10 mM, pH 7.4) at relative 8 0 molar charge ratios of 2.5 (ratio between the overall charge of the protein and charged “b” units of the copolymers). 60 40 20

40 20

0 9 M

0 3 M

0

0 1

3 L

zy L

y

L

s

y

o

s

L

m

9

e

0

0

M

0 3 M

0 3 L

1 L

m

60

o

zy

0

0

e

N o rm a liz e d e n z y m a tic a c tiv ity (% )

A

Figure 2. (A) Lysozyme activity after incubation of the dry powders in PB and (B) lysozyme activity after incubation of the dry powders in presence of trypsin for 3 h followed by the addition of poly(allyl amine) for 1 h. All formulations contained 6.6 % (w/w) phosphate buffer salts. All experiments were carried out in triplicate (n=3) and data are presented as mean ± SD. One-way ANOVA was used for statistical analysis. **** p < 0.0001.

90


Drug Delivery to the Lungs 27, 2016 - Polymer-protein-based dry powder for efficient pulmonary protein delivery Discussion In this work, spray-dried powders containing polymer-lysozyme nanocomplexes were designed as potential model formulations for pulmonary delivery of protein therapeutics. Most of the inhalation products currently available on the market use lactose as a carrier material. However, it is not suitable for formulations that contain primary amino groups, such as peptides and proteins, because of a range of potential side-reaction occurring at the sugar [5] reducing end, e.g. Maillard reaction . In contrast, non-reducing sugars such as trehalose have been extensively [6] investigated as potential alternatives in inhalable dry powders . Due to the highly hygroscopic and aggregative nature of the amorphous form of trehalose, high concentrations of this sugar typically result in poor aerosolisation properties. In this work, leucine was therefore used as a disperser agent which allowed the production of dry powders with FPFs as high as 68 %. Protein loading into the particles was found to be very efficient, with up to 100 % of the protein added to the liquid feed solution being encapsulated. The presence of shorter polymers showed a significant improvement on protein encapsulation efficiencies from 80 to 100 % when compared to longer polymers. Importantly, high FPF was also retained, and in some cases slightly increased (from 62 % to 68 % with linear L10 copolymers) in trehalose-leucine-based dry powders. Pleasingly, the size of the nanocomplexes did not change significantly when compared to nanocomplexes before spray-drying and the polydispersity was founded to be even smaller. These results were particularly encouraging considering that very often the process of spray-drying can modify the size of nanoparticles once they are [3] redispersed . Lysozyme activity of polymer-protein nanocomplexes decreased when compared to that of the free protein, indicating that lysozyme was located in the “core� of these assemblies, where the enzyme was less accessible. In previous work (data not shown), full activity could be recovered by treating the nanoassemblies with poly(allyl amine), confirming that the low activity observed was due to decreased accessibility. Finally, the ability of polymer-protein nanocomplexes-based dry powders to protect proteins from proteolytic enzymes was evaluated. Trypsin is an enzyme which cleaves proteins specifically at the carboxyl side of the aminoacids lysine [7] and arginine. Lysozyme is particularly sensitive to trypsin due to its high content in those aminoacids . An important decrease (> 60 %) in its enzymatic activity was observed when no polymer was added to the dry powder. In contrast, lysozyme was protected from trypsin digestion when polymers were present in the formulations. Interestingly, when the longest L30 and M90 copolymers were used, virtually full retention of the protein activity was observed. Conclusion In the present study we demonstrated that polymer-protein nanocomplexes can be successfully formulated into inhalation dry powders with optimal aerodynamic properties. All polymers, and especially the longer L30 and M90 copolymers, provided a very valuable protection of proteins against proteolytic enzymes. Dry powders based on polymer-protein nanocomplexes are therefore promising formulations for efficient delivery of proteins into the lungs. References 1 2

3 4 5

6

7

Manning M C, Chou D K, Murphy B M, Payne R W, Katayama D S: Stability of protein pharmaceuticals: an update, Pharmaceut Res 2010; 27 (4): pp544-575. (a) Tsiourvas D, Sideratou Z, Sterioti N, Papadopoulos A, Nounesis G, Paleos C M: Insulin complexes with PEGylated basic oligopeptides, J Colloid Interf Sci 2012; 384 (1): pp61-72; (b) Coue G, Engbersen J F: Functionalized linear poly(amidoamine)s are efficient vectors for intracellular protein delivery, J Control Release 2011; 152 (1): pp90-98. Sham J O, Zhang Y, Finlay W H, Roa W H, Lobenberg R: Formulation and characterization of spray-dried powders containing nanoparticles for aerosol delivery to the lung, Int J Pharm 2004; 269 (2): pp457-467. Buttini F, Colombo P, Rossi A, Sonvico F, Colombo G: Particles and powders: tools of innovation for noninvasive drug administration, J Control Release 2012; 161 (2): pp693-702. Littringer E M, Mescher A, Eckhard S, SchrĂśttner H, Langes C, Fries M, Griesser U, Walzel P, Urbanetz N A: Spray drying of mannitol as a drug carrier - The impact of process parameters on product properties, Dry Technol 2012; 30 (1): pp114-124. Maury M, Murphy K, Kumar S, Mauerer A, Lee G: Spray-drying of proteins: effects of sorbitol and trehalose on aggregation and FT-IR amide I spectrum of an immunoglobulin G, Eur J Pharm Biopharm 2005; 59 (2): pp251-261. Canfield R E: Peptides derived from tryptic digestion of egg white lysozyme, J Biol Chem 1963; 238: pp26912697.

91


Drug Delivery to the Lungs 27, 2016 - Kev Dhaliwal Multiplexed optical molecular sensing and imaging in the lungs Kev Dhaliwal, MBChB (Hons), BSc (Hons), PhD, FRCP Senior Clinical Lecturer in Pulmonary Molecular Imaging, MRC Centre for Inflammation Research in the Queen’s Medical Research Institute (QMRI) and Consultant Physician in Respiratory Medicine.

Summary

The talk will describe the emerging technology of Molecular Alveoscopy that is being developed at the University of Edinburgh to provide immediate bedside sensing and imaging of the distal alveoli in humans. The Proteus Project ( www.proteus.ac.uk) is the UK’s largest biophotonics project and is focussed on developing disruptive technologies for pulmonary imaging to develop in vivo in situ optical molecular pathology approaches . The technology is being developed to measure inflammation, bacterial presence and physiological parameters such as tissue oxygen, pH, redox in the alveolar space alongside sampling distal airway fluids. The goal is to develop alveolar telemetry which will be of widespread utility especially in critical care but also in experimental medicine.

92


DDL27 - 2016 POSTERS Page Nos Characterisation of jet-milled and spray dried isoniazid for pulmonary administration

102

I. Sibum, F. Grasmeijer, P. Hagedoorn & H.W. Frijlink Department of Pharmaceutical Technology and Biopharmacy, Ant. Deusinglaan 1, Groningen, 9713 AV, the Netherlands Understanding DPI blends behaviour during high speed capsule filling

106

S. Abdrabo*, M. Fridez, T. Eggimann & R. Müller-Walz Skyepharma AG, a member of the Vectura Group of companies, Eptingerstasse 61, 4132 Muttenz, Switzerland Nano-in-microparticle powders for mucosal vaccination – understanding the particle forming process

110

Judith Heidland & Regina Scherließ Department of Pharmaceutics and Biopharmaceutics, Kiel University, Grasweg 9a, 24118 Kiel, Germany Impact of different capsules for dry powder inhalers on the aerodynamic performances of formoterol-based binary and ternary blends 1

1

2

Nathalie Wauthoz , Ismaël Hennia , Susana Ecenarro & Karim Amighi

114

1

1

Laboratory of Pharmaceutics and Biopharmaceutics, Université libre de Bruxelles (ULB), Boulevard du Triomphe CP207, Brussels 1050, Belgium 2 Qualicaps Europe S.A.U., Avenida Monte Valdelatas 4, Alcobendas 28108, Spain The surface energy effect of co-spray-dried mannitol with polyethylene glycol on the aerosolization performance in a dry powder inhalation formulation

118

Makoto Kamada, Mitsuhide Tanimoto, Atsutoshi Ito, Michiko Kumon, Kazuhiro Inoue, Shuichi Yada & Hidemi Minami Formulation Technology Research Laboratories, Daiichi Sankyo Co., Ltd., 1-2-58, Hiromachi, Shinagawa-ku, Tokyo 140-8710, Japan Impact of particle engineering on the processability and aerosolization performance of DPI formulations 1

1

1

1,2

S Zellnitz , E Faulhammer , T Wutscher , J G Khinast , A Paudel 1 2

122

1

Research Center Pharmaceutical Engineering GmbH, Inffeldgasse 13, Graz, 8010, Austria Institute for Process and Particle Engineering, Graz University of Technology, Inffeldgasse 13, Graz, 8010, Austria

Optimization of Inhalable Spray Dried Powder Formulations of Small Nucleic Acids with Mannitol and Leucine

127

Michael Chow, Hinson Lin & Jenny Lam Department of Pharmacology & Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, 21 Sassoon Road, Pokfulam, Hong Kong Optimization of Supercritical-CO2 Assisted Spray Drying for the Production of Inhalable composite particles 1,2

1

Cláudia Moura , Eunice Costa & Ana Aguiar-Ricardo 1 2

2

R&D, Hovione Farmaciencia SA, Loures, 2674-506, Portugal LAQV-REQUIMTE, Departamento de Química, Universidade NOVA de Lisboa, Caparica, 2829-516, Portugal

93

131


Impact of Spray Drying on Superoxide Dismutase Activity in Composite Systems with Optimal Aerodynamic Performance for Dry Powder Inhalers 1,2

1

1

1

Diana A. Fernandes , Raquel Barros , Cláudia Moura , Eunice Costa & Maria L. Corvo

135

2

1

Hovione SA, Loures, 2074-506, Portugal Research Institute for Medicines (iMed.ULisboa), Faculdade de Farmácia, Universidade de Lisboa, Lisboa, 1649-003, Portugal 2

®

Parteck M DPI – a novel mannitol as carrier in dry powder inhalation formulations 1

2

Nancy Rhein , Gudrun Birk , Regina Scherließ

139

1

1Department of Pharmaceutics and Biopharmaceutics, Kiel University, Grasweg 9a, 24118 Kiel, Germany 2Merck KGaA, Frankfurter Straße 250, 64293 Darmstadt, Germany

143

Dry powder intended for pulmonary delivery: Comparison between PGA-co-PDL and chitosan Nano Composite Microparticles 1

2

3

3

Valeria Carini , Ayca Y. Pekoz , Gillian A. Hutcheon , Imran Y. Saleem , Adel A. Mohamed

3

1

School of Pharmacy, Unicam, Camerino University, Italy. School of Pharmacy, Faculty of Pharmacy Istanbul University, Turkey 3 School of Pharmacy and Biomolecular Sciences, Liverpool John Moores University, Liverpool, L3 3AF, UK. 2

146

Detachment of various spray dried drugs from engineered mannitol carrier particles 1

2

2

3

4

Mathias Mönckedieck , Jens Kamplade , Peter Walzel , Hartwig Steckel , Nora Urbanetz & Regina Scherließ

1

1

Department of Pharmaceutics and Biopharmaceutics, Kiel University, Grasweg 9a, 24118 Kiel, Germany Technical University Dortmund, Emil-Figge Str. 68, 44227 Dortmund, Germany 3 Deva Holding AS, Halkalı Merkez Mah. Basın Ekspres Caddesi, Istanbul, Turkey 4 Daiichi Sankyo, Luitpoldstr. 1, 85276 Pfaffenhofen, Germany 2

Development and characterization of miconazole-loaded lipid nanoparticles against pulmonary mycoses

150

1

Zsófia Edit Pápay , Eszter Gulyás, Petra Füredi & István Antal Semmelweis University, Hőgyes Endre Street 7, Budapest, 1092, Hungary Synthetic KL4 peptide as new carrier of siRNA therapeutics for pulmonary delivery

154

Yingshan Qiu, Michael Y.T. Chow & Jenny K.W. Lam Department of Pharmacology & Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, 21 Sassoon Road, Pokfulam, Hong Kong Dextran Sulfate Microparticles Encapsulating Isoniazid and/or Rifabutin as Carriers for Pulmonary Tuberculosis Therapy

158

Flávia Musacchio and Ana Grenha Center for Biomedical Research (CBMR), Faculty of Sciences and Technology, University of Algarve, Faro, 8005-139, Portugal Center for Marine Sciences (CCMar), University of Algarve, Faro, 8005-139, Portugal Inhalable chitosan microparticles as tools in tuberculosis therapy 1,2

3

LC Cunha , AM Rosa da Costa , A Grenha

1,2

1

CBMR – Centre for Biomedical Research Drug Delivery Laboratory, University of Algarve, Campus de Gambelas, Faro, 8005-139, Portugal 2 CCMAR – Centre of Marine Sciences, University of Algarve, Campus de Gambelas, Faro, 8005-139, Portugal 3 Algarve Chemistry Research Center (CIQA) and Department of Chemistry and Pharmacy, Faculty of Sciences and Technology, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugal

94

162


166

Preparation of hybrid silver/ciprofloxacin nanoparticles for pulmonary drug delivery 1

Hisham Al-Obaidi and Mridul Majumder 1 2

The School of Pharmacy, University of Reading, Reading RG6 6AD, UK M2M Pharmaceuticals Ltd, Science & Technology Centre, Earley Gate, Reading RG6 6BZ, UK

170

Chondroitin sulphate microparticles for tuberculosis treatment: a way to target macrophages Susana Rodrigues,

1, 2

3

Ana M Costa & Ana Grenha

1, 2

1

CBMR – Centre for Biomedical Research Drug Delivery Laboratory, University of Algarve, Campus de Gambelas, Faro, 8005-139, Portugal 2 CCMAR – Centre of Marine Sciences, University of Algarve, Campus de Gambelas, Faro, 8005-139, Portugal 3 Algarve Chemistry Research Center (CIQA) and Department of Chemistry and Pharmacy, Faculty of Sciences and Technology, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugal Optimisation of phorbol myristate acetate (PMA) mediated differentiation of U937 human lung monocytes to alveolar like macrophages

174

A Martin, D Murnane, DYS Chau, MB Brown, V Hutter Department of Pharmacy, School of Life and Medical Sciences, University of Hertfordshire, AL10 9AB

178

Development of novel efficient rodent nose-only inhalation exposure systems Mikael Brülls 1

1

AstraZeneca R&D, AstraZeneca R&D Gothenburg, SE-431 83 Mölndal, Sweden

182

Chitosan nanocarrier systems for delivery of pneumococcal vaccine via nebulization 1,4

2

2

2

1

Nicoletta Zallocco , Eliane N. Miyaji , Viviane M. Goncalves , Douglas B. Figueiredo , Imran Y. Saleem , 3 Ronan MacLoughlin 1

School of Pharmacy & Biomolecular Sciences, Liverpool John Moores University, Liverpool, UK.

2 3

Centro de Biotecnologia, Instituto Butantan, Sao Paulo, Brazil.

Aerogen Ltd. Galway Business Park, Dangan, Galway, Ireland. 4 School of Pharmacy, University of Camerino, Camerino, Italy.

Enrichment of the surface of spray-dried powder particles with a hydrophobic material to improve their aerosolization 1

1

2

3

Mohammad AM Momin , Ian G Tucker , Colin Doyle , John Denman , & Shyamal C Das

186

1

1

New Zealand's National School of Pharmacy, University of Otago, Dunedin 9054, Dunedin, New Zealand 3 The University of Auckland, 20 Symonds Street, Auckland, New Zealand, Future Industries Institute, University of South Australia, Mawson Lakes, SA 5095, Australia 2

Comparative Nanotoxicology of Novel Polymeric Nanocarriers with Different Surface Charge against Human Lung Epithelial Cells

190

Nashwa Osman, Darren Sexton, Gillian Hutcheon, and Imran Saleem School of Pharmacy & Biomolecular Sciences, Liverpool John Moores University, Liverpool, L3 3AF, UK. The influence of lung surfactant, 1,2-dipalmitoyl-sn-glycero-3-phosphatidylcholine (DPPC) on the aerosolization of two anti-tubercular drugs, pyrazinamide and moxifloxacin Shyamal C. Das1, Bhamini Rangnekar & Basanth Babu Eedara

New Zealand's National School of Pharmacy, University of Otago, Dunedin 9054, New Zealand

95

194


198

Preclinical sheep models for pulmonary disease and drug delivery Rob Bischof The Ritchie Centre, Hudson Institute of Medical Research, Clayton VIC 3168, AUSTRALIA

201

A Study on the Drug Deposition Mechanisms of Surface-treated pMDI Canisters 1

1

2

Athina Skemperi , David Worrall Gary Critchlow & Phil Jinks

3

1

Chemistry Department, Loughborough University, Loughborough, Leicestershire, UK Materials Department, Loughborough University, Loughborough, Leicestershire, UK 3 3M Drug Delivery Systems, Loughborough, Leicestershire, UK 2

Generation of Respirable Particles from Surfactant Suspensions and Viscous Solutions at High Dose Rates

205

Donovan B. Yeates & Xin Heng KAER Biotherapeutics, 926 S. Andreasen Drive, Escondido, California 92029, USA Nebuliser Device Issues To Be Improved For The Pulmonary Administration Of Nano Encapsulation Delivery 1

2

1

A. A. Desta , K. C. Carter , T. Gourlay &A. B. Mullen

209

2

1

Biomedical Engineering, 106 Rottenrow East, Glasgow, G4 0NW, United Kingdom Strathclyde Institute of Pharmacy and Biomedical Sciences, 161 Cathedral Street, Glasgow, G4 0RE, United Kingdom 2

Opt2Fill™Dispersible Tablet – A Novel Method for the Manufacture of pMDIs

213

Cuong Tran, Chen Zheng, Simon Warren & Glyn Taylor i2c Pharmaceutical Services, Cardiff Medicentre, Cardiff, CF14 4UJ, UK.

217

Droplet Sizes of Electronic Cigarette Aerosols 1

2

Philip Chi Lip Kwok , Philippe Rogueda & Lu Hou

1

1

Department of Pharmacology and Pharmacy, Li Ka Shing Faculty of Medicine, The University of Hong Kong, 21 Sassoon Road, Pokfulam, Hong Kong SAR Aedestra Limited, 11/F, Rykadan Capital Tower, 135-137 Hoi Bun Road, Kwun Tong, Kowloon, Hong Kong SAR

2

221

A Study of Factors Affecting Nucleation and Bubble Growth in Pressurised Metered Dose Inhalers Nazli Nezami & Hendrik Versteeg Loughborough University, Wolfson School of Mechanical, Electrical and Manufacturing Engineering, Ashby Road, Loughborough, LE11 3TU, UK Leucine Coated Inhalable Insulin and Thymopentin Peptide Powders Produced by Aerosol Flow Reactor Method 1

1

2

2

1

Nurcin Ugur , Ville Vartiainen , Luis M. Bimbo , Jouni Hirvonen , Esko I. Kauppinen and Janne Raula 1 2

225

1

Department of Applied Physics, Aalto University School of Science, Espoo, 00076, Finland Division of Pharmaceutical Chemistry and Technology, Faculty of Pharmacy, University of Helsinki, 00014, Finland

Application of Effective Data Analysis (EDA) and Abbreviated Impaction Method (AIM) in the Metered Dose Inhaler (MDI) Product Screening Lei Mao, Alexandra Hughes & Carole Evans Catalent Pharma Solutions, 160N Pharma Drive, Morrisville, North Carolina, 27560, USA

96

229


233

Development of an inhaled ion-paired salbutamol formulation 1

1

2

2

1

Bridie Dutton , Arcadia Woods , Robyn Sadler , Nathalie Fa , Ben Forbes & Stuart Jones 1 2

1

Institute of Pharmaceutical Science, King’s College London, 150 Stamford Street, London, SE1 9NH, UK GlaxoSmithKline, Park Road, Ware, Hertfordshire, SG12 0DP, UK

The continued development of a cyclic olefin copolymer elastomer sealing system for metered dose inhalers

237

Segolene Sarrailh, Chris Baron, Gerallt Williams Aptar Pharma, Le Vaudreuil, 27100, France Investigation Into The Effects Of Valve Pressure On Total Delivered Dose (TDD) From Vhcs 1

241

1

Scott Courtney , PhD; Benjamin Pratt , B.App.Sci. 1

Medical Developments International, Research & Development Dept., 4 Caribbean Drive, Scoresby, Melbourne, AUSTRALIA Comparison Between Pharmacopeial Testing and Testing Based on Mixing Inlet Lung Simulator (MILS) Methodology with Application to Nebuliser Testing: Part 1 – Pilot Study to Select Most Feasible MILS Method 1

1

2*

Mårten Svensson , Elna Berg , Jolyon Mitchell & Dennis Sandell 1 2

3

245

3

Emmace Consulting AB, Lund, SE-22381, Sweden Jolyon Mitchell Inhaler Consulting Services Inc., London, N6H 2R1, Canada (* corresponding author)

S5 Consulting, Blentarp, SE-27562, Sweden

Comparison Between Pharmacopeial Testing and Testing Based on Mixing Inlet Methodology with Application to Nebuliser Testing: Part 2 – Main Study 1

1

2*

Mårten Svensson , Elna Berg , Jolyon Mitchell & Dennis Sandell

249

3

1

Emmace Consulting AB, Lund, SE-22381, Sweden Jolyon Mitchell Inhaler Consulting Services Inc., London, N6H 2R1, Canada (* corresponding author) 3 S5 Consulting, Blentarp, SE-27562, Sweden 2

Comparison of bronchoconstrictor responses to AMP and adenosine in sensitised guinea pigs ® using the PreciseInhale system for generation and administration of dry powder aerosols

253

A.J. Lexmond, S. Keir, C.P. Page, B. Forbes King’s College London, Institute of Pharmaceutical Science, 150 Stamford Street, London, SE1 9NH, UK Local and systemic pharmacokinetic evaluation of immediate-release and controlled-release cisplatin dry powders for inhalation against lung cancer

257

Vincent Levet, Rémi Rosière, Karim Amighi & Nathalie Wauthoz Laboratory of Pharmaceutics and Biopharmaceutics, Faculty of Pharmacy, Université libre de Bruxelles (ULB), CP 207, Boulevard du Triomphe, 1050 Brussels, Belgium

261

Ensuring the Consistency of Performance of Mesh Nebulizers R. H M Hatley, L. E A Hardaker, A. P Metcalf, J. Parker, F. Quadrelli & J. N Pritchard Respironics Respiratory Drug Delivery (UK) Ltd, a business of Philips Electronics UK Limited, Chichester, West Sussex, UK.

97


264

Drug Loading of Human Albumin Nanocarriers for Inhaled Anti-Tuberculosis Therapy 1

1

2

2

3

1,2

Ayasha Patel , Arcadia Woods , Peter Imming , Adrian Richter , Nick Childerhouse , Lea Ann Dailey 1 Forbes

& Ben

1

Institute of Pharmaceutical Science, King’s College London, 150 Stamford Street, London SE1 9NH, United Kingdom Department of Pharmacy, Martin Luther University of Halle-Wittenberg, Wolfgang-Langenbeck-Str.4, 06120 Halle, Germany 3 Vectura Group plc, One Prospect West, Chippenham, Wiltshire, SN14 6FH, United Kingdom 2

Development and Characterization of Nanocrystal-embedded Microparticles for Pulmonary Delivery of Budesonide 1

1

1

Dongmei Cun , Tingting Liu , Meihua Han , Mingshi Yang

268

1.2

1

Department of Pharmaceutics, School of Pharmacy, Shenyang Pharmaceutical Univeristy, Wenhua Road No. 103, Shenhe District, Shenyang, 110023, China 2 Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen , Universitetsparken 2, DK-2100 Copenhagen, Denmark Assessment of Delivered Dose Rate from 2 Mesh Nebulizers tested with an Adult Breathing Pattern and 2 Drugs

272

L. Slator, J. Parker, L. Hardaker & R. Potter Respironics Respiratory Drug Delivery (UK) Ltd, a business of Philips Electronics UK Limited, Chichester, West Sussex, UK ®

Effect of inhalation manoeuvre parameters on the dose emitted from Onbrez Breezhaler using inhalation profiles of patients with chronic obstructive pulmonary disease (COPD)

276

M. Abadelah1, H. Chrystyn2, H.Larhrib1 1

Department of Pharmacy and Pharmaceutical Science, University of Huddersfield, Queensgate, Huddersfield, HD1 3DH, UK 2 Inhalation Consultancy Ltd, Yeadon, Leeds, LS19 7SP, United Kingdom Development of dry powder formulations combining both a biologic therapeutic entity and a small molecule drug substance

280

Amy Worle & Harriet Bridgwater Vectura Group plc, Vectura House, Bumpers Way, Chippenham, Wiltshire, SN14 6FH Does glycerol interact with dipalmitoylphosphatidylcholine membranes?

284

Wachirun Terakosolphan1, Precious Akhuemokhan1, Richard Harvey2 & Ben Forbes1 1 2

Institute of Pharmaceutical Sciences, King’s College London, London, SE1 9NH, UK Institute of Pharmacy, Martin-Luther-University Halle-Wittenberg, Halle (Saale), 06099, Germany

Breathing Simulators: One step closer to representative deposition profiles? 1

1

1

Maria Palha , Sofia Silva , Isabel S. Lopes & Eunice Costa 1

1

Hovione FarmaCiencia SA, Sete Casas, 2674-506 Loures, Portugal

Further Investigation of the Maldistribution of Aerosol Deposits on the Upper Stages of the Andersen 8-Stage Non-Viable Cascade Impactor (ACI): Evidence for Pre-Classification of Incoming Aerosol 1

Daryl L. Roberts and Jolyon P. Mitchell 1 2

288

2

MSP Corporation, 5910 Rice Creek Parkway, Suite 300, Shoreview, Minnesota 55126, USA Jolyon Mitchell Inhaler Consulting Services Inc., 1154 St. Anthony Road, London, Ontario, N6H 2R1, Canada

98

292


Ring Opening Metathesis Polymerization of a ciprofloxacin-conjugated copolymer for pulmonary intracellular antibiotic delivery 1

1

2

296

1

Nina Warner , Maryann Zhao , Daniel M. Ratner , & Daniel J. O'Leary 1 2

Department of Chemistry, Pomona College, Claremont, CA 91711, USA Department of Bioengineering, University of Washington, Seattle, WA 98195, USA

300

Comparison of Aerosol Particle Size Measured by Two Methods Using Three Brands of Mesh Nebulizer 1

1

1

2

1

L. Slator , Y. Degtyareva , L. Hardaker , D. von Hollen & J. Pritchard 1

Respironics Respiratory Drug Delivery (UK) Ltd, a business of Philips Electronics UK Limited, Chichester, West Sussex, UK 2 Respironics, Inc., a Philips Healthcare Company, Murrysville, PA, USA

304

Polyamine ion-pairs to target drugs to the lung 1

1

1

Zarif Mohamed Sofian , Arcadia Woods , Paul G. Royall & Stuart A. Jones 1

1

Institute of Pharmaceutical Science, King’s College London, 150 Stamford Street, London, SE1 9NH, UK

Paddle over disk as a dissolution test for orally inhaled drugs: discriminating composite from carrier-based formulations 1,2

1

1

2

Beatriz Fernandes , Filipa M. Maia , A. Mafalda Paiva , Maria Luisa Corvo & Eunice Costa

308

1

1

Hovione FarmaCiencia SA, Sete Casas, Loures, 2674-506, Portugal iMed.ULisboa, Faculdade de Farmácia, Universidade de Lisboa, Avenida Prof. Gama Pinto, Lisboa, 1649-003, Portugal 2

Aerosol Performance and Stability Characteristics of Spray-Dried Tobramycin Excipient Enhanced Growth Inhalation Powder Formulations 1

1

3

Xiangyin Wei , Anubhav Kaviratna , Ruba S. Darweesh , P. Worth Longest

1,2

and Michael Hindle

312

1

1

Department of Pharmaceutics, Virginia Commonwealth University, Richmond, VA, 23298, USA Department of Mechanical and Nuclear Engineering, Virginia Commonwealth University, Richmond, VA, 23284, USA 3 Department of Pharmaceutical Technology, Jordan University of Science and Technology, Irbid, 22110, Jordan 2

316

A deeper insight into the impact of chemical surface properties on inhalation performance 1

Niklas Renner , Hartwig Steckel², Nora A. Urbanetz³ & Regina Scherließ

1

1

Department of Pharmaceutics and Biopharmaceutics, Kiel University, Grasweg 9a, 24118 Kiel, Germany ²Deva Holding A.S., Istanbul, Turkey; ³Daiichi Sankyo, Tokyo, Japan 60 Years of the MDI – A History of Innovation 1

Alex Slowey & Steve Stein 1 2

320

2

3M Drug Delivery Systems, Morley Street, Loughborough, Leicestershire, LE11 1EP, UK. 3M Drug Delivery Systems,3M Center,St. Paul, MN 55144-1000, USA.

324

Performance Indicating Acoustic Emission Measurements on a Dry Powder Inhaler 1

1

1

2

1

Lars Karlsson , Lisa Holmstén , Mats Josefson , Roland Greguletz , Kyrre Thalberg & Staffan Folestad 1 2

1

Pharmaceutical Technology & Development, AstraZeneca R&D Gothenburg, S-43183 Mölndal, Sweden Technology Department, AstraZeneca Sofotec, D-61352 Bad Homburg, Germany ®

Effect of using bio-relevant media in the DissolvIt system to measure dissolution of fluticasone propionate from Flixotide 50 µg Evohaler 1

2

1

1

2

2

Mireille Hassoun , Maria Malmlöf , Abhinav Kumar , Sukhi Bansal , Mattias Nowenwik , Per Gerde , Ben 1 Forbes 1 2

Institute of Pharmaceutical Science, King’s College London, SE1 9NH, UK Inhalation Sciences Sweden AB, Stockholm, Sweden

99

328


Solubility of fluticasone propionate and beclomethasone dipropionate in simulated lung lining fluids 1

1

1

Wachirun Terakosolphan , Mireille Hassoun , Abhinav Kumar & Ben Forbes 1

333

1

King’s College London, Institute of Pharmaceutical Science, London, SE1 9NH, UK

Using 3D printed standards to isolate the effect of surface morphology and surface chemistry in DPIs 1

1

2

3

337

4

Georgina E. Marsh , Morgan R. Alexander , Ricky D. Wildman , Matt J Bunker , Mark Nicholas & Clive J. 1 Roberts 1

School of Pharmacy, The University of Nottingham, Nottingham, UK Department of Chemical and Environmental Engineering, The University of Nottingham, Nottingham, UK 3 AstraZeneca R&D, Macclesfield, UK 4 AstraZeneca R&D, Mölndal, Sweden 2

Modelling Drug Entrainment in a Dry Powder Inhaler: Benchmarking and Sensitivity Analysis of a Multiphase CFD Approach 1

1

Thomas Kopsch , Digby Symons & Darragh Murnane 1 2

341

2

University of Cambridge, Trumpington Street, CB2 1PZ, UK University of Hertfordshire, College Lane, AL10 9AB, UK

Tracheal Inhalation Comparison across Three Different Inhalation Methods - Normal Adult vs. an Asthmatic Adult

345

Cathy McKenna, Andrew O’Sullivan, Patrick Kelly, Louise Sweeney, Conor Duffy, Ronan MacLoughlin Aerogen, IDA Business Park, Dangan, Galway, Ireland A New Size Facemask for Use with a Valved Holding Chamber (VHC) by Adults with Smaller Facial Features 1

1

1

1

Neritan Alizoti , Noel Gulka , Jason A Suggett , Robert Morton , Jolyon P. Mitchell 1 2

348

2

Trudell Medical International, 725 Third Street, London, Ontario, N5V 5G4, Canada Jolyon Mitchell Inhaler Consulting Services Inc., 1154 St. Anthony Road, London, Ontario, N6H 2R1, Canada

Can Aerosol Therapy Keep Pace with Innovations in Patient Care? A Review.

352

Ronan Mac Loughlin Aerogen, IDA Business Park, Dangan, Galway, IRELAND Effect of Nebulizer Type and Position on Aerosol Drug Delivery during Support Mechanical Ventilation and Spontaneously Breathing for Tracheostomized Adult Patients

356

Patrick M. Kelly, Andrew O’Sullivan, Cathy McKenna, Louise Sweeney & Ronan MacLoughlin Aerogen, Galway Business Park, Dangan, Galway, Ireland

359

Numerical Simulation of Salbutamol Deposition in VHC Devices 1

2

3

R F Oliveira , S F C F Teixeira , H M Cabral-Marques & J C F Teixeira 1 2 3

MEtRiCS R&D Centre, University of Minho, 4800-058 Guimarães, Portugal ALGORITMI R&D Centre, University of Minho, 4800-058 Guimarães, Portugal iMed.ULisboa R&D Centre, Universidade de Lisboa, 1649-003 Lisboa, Portugal

363

A Fresh Look at Designing Respiratory Health Devices 1

1

Daniel P. Jenkins , Chris Langley & Paul Draper 1

1

1

DCA Design International, 19 Church Street, Warwick, CV34 4AB, UK

100


101


Drug Delivery to the Lungs 27, 2016 – I. Sibum et al. Characterisation of jet-milled and spray dried isoniazid for pulmonary administration I. Sibum, F. Grasmeijer, P. Hagedoorn & H.W. Frijlink Department of Pharmaceutical Technology and Biopharmacy, Ant. Deusinglaan 1, Groningen, 9713 AV, the Netherlands Summary The aim of this study was to develop a dry powder isoniazid formulation with no or a limited amount of excipients for pulmonary administration. Milled isoniazid showed an excellent particle size distribution for inhalation, however dispersion was poor. In 78% of the dispersion measurements the inhaler blocked, retaining most of its dose. Pure spray dried isoniazid yielded particles too large for pulmonary delivery, but the addition of 5% of L-leucine resulted in spray dried particles of inhalable size. DSC data showed complete crystallinity for all samples, while TGA analysis showed that isoniazid sublimates around 100°C. SEM imaging showed that pure jet milled and spray dried isoniazid particles fused together. Isoniazid spray dried with L-leucine resulted in spherical particles with no fusion visible. The most likely explanation for particle fusion is that isoniazid crystalizes, resulting in solid bridge formation. L-leucine however, forms a coating around isoniazid particles, thereby preventing this phenomenon. Further experiments are needed to show why isoniazid fuses together in the jet mill. A possible explanation is that some isoniazid sublimates due to heat generation during particle collisions, and causes solid bridge formation between particles when it ripens. Further experiments have to show whether isoniazid co-spray dried with Lleucine disperses efficiently and is stable over time. Introduction The World Health Organization (WHO) estimates that one third of the world population is (latently) infected with [1] tuberculosis (TB). Isoniazid is an antibiotic used in the first-line treatment of TB. However, in 15% of all TB [2] cases worldwide resistance to isoniazid is encountered. Thus far only oral and injectable forms are available. Pulmonary administration of isoniazid can provide significant advantages over oral and parenteral administration. Inhaled drug is administered directly to the target area (port of entry), which means that with the same dose [3] administered a higher local concentration can be achieved, eradicating bacteria considered resistant. Since [1] most TB cases are in third world countries, a stable and cheap formulation is needed. A dry powder combined [3] with a cheap inhaler would fit these requirements. In the scientific literature, experimental isoniazid formulations consist either of drug loaded liposomes which only contain 10% drug, or of a combination formulation in which [4]–[6] These formulations are of limited use since anti-TB drugs have isoniazid only makes up 12.5% of the powder. [7] to be given in high doses, as recommended by the WHO. For these reasons we aimed to develop an isoniazid powder formulation suitable for pulmonary delivery with no or only a limited amount of excipients. Materials and Methods Isoniazid and L-leucine used during spray drying were purchased from Sigma Aldrich (St. Louis, United States). Isoniazid used during jet milling was provided by Fluka (St. Louis, United States). Spray drying was performed using a B-290 mini spray dryer supplied by Büchi (Flawil, Switzerland). A range of parameters were investigated, as Concentration Inlet Feed Atomizing Atomizing shown in Table 1. All samples were isoniazid temperature rate air flow air flow produced in duplicate. The first sample (mg/ml) (°C) (ml/min) (mm) (%) (50 mg/mL concentration and 60°C 50 60 2.5 50 100 inlet) was also spray dried with 5% 50 40 1 50 100 leucine w/w. Samples were always 50 160 2.5 50 100 prepared in 20 mL demineralised water before spray drying. Before 50 160 12.5 50 100 each run the outlet temperature was 50 20-30 0.1 50 100 stabilised by spraying demineralised 25 60 2.5 50 100 water in the system. 75 60 2.5 50 100 Table 1 - spray drying conditions with the B-290 mini spray dryer. All samples were produced twice.

Low pressure Medium pressure High pressure

Nozzle pressure (bar) 4 5 7

Mill pressure (bar) 0.5 2 2

Milling was performed using a 50 AS spiral jet mill supplied by Alpine (Augsburg, Germany). The mill was equipped with a 0.8 mm nozzle. Table 2 shows the three different pressure settings tested, the milling gas used was nitrogen.

Table 2 - jet milling conditions with the 50 AS spiral jet mill. All samples were produced twice.

102


Drug Delivery to the Lungs 27, 2016 - Characterisation of jet-milled and spray dried isoniazid for pulmonary administration Particle sizes were measured using a HELOS BF diffraction unit with a RODOS dry disperser, both provided by Sympatec (Clausthal-Zellerfeld, Germany). The HELOS diffraction unit was equipped with an R3 lens, resulting in a measurement range of 0.1 μm-175 μm. Powders were dispersed with the RODOS at 0.5, 3 and 5 bars. All measurements were performed two times for the milled samples and three times for the spray dried samples. The average with standard deviation was then calculated. Dispersion measurements were performed using the HELOS diffraction unit equipped with the INHALER 2000 attachment, also provided by Sympatec (Clausthal-Zellerfeld, Germany). The Twincer® dry powder inhaler (DPI), provided by PureIMS (Roden, the Netherlands), was used as test inhaler. Measurements were performed with a pressure drop of 2, 4 and 6 kPa and with a dose of 25 mg. All samples were measured twice. Scanning electron microscopy (SEM) images were obtained using a JSM 6301-F microscope provided by JEOL (Tokyo, Japan). Samples were placed on carbon tape and coated with 10 nm of gold using the JFC-1300 Auto Fine Coater, also provided by JEOL (Tokyo, Japan), while purged with argon gas. Imaging itself took place under high vacuum, with an acceleration voltage of 10kV, a working distance of 10 mm and a spot size of 25. The secondary electron detector was used. Differential scanning calorimetry (DSC) was performed with the Q2000 supplied by TA instruments (New Castle, United States). A sample of 1-4 mg was placed in a Tzero pan and hermetically sealed. Pans were loaded into the DSC at 20°C and a ramp from 20°C until 210°C was made at a rate of 20°C /min. Samples were always measured the same day as they were spray dried or milled, each sample was measured twice. Thermogravimetric analysis (TGA) was performed on the isoniazid starting material for spray drying. A sample of 6-8 mg was placed on a platinum pan. The sample was heated at 80°C for 4 hours after which the temperature was increased with 20°C and again the temperature was kept stable for 4 hours. This process was repeated until 140°C was reached. Non-cumulative weight loss percentage over these 4 hour periods was then calculated. Results In Table 3 the laser diffraction data of the jet milled isoniazid samples is given. The particle sizes are consistently smaller compared to the spray dried samples. Furthermore, the higher the pressure the smaller the particles, as is to be expected. Low pressure Medium pressure High pressure

X10 (μm) 1.59 ± 0.01 1.26 ± 0.42 1.11 ± 0.26

X50 (μm) 2.91 ± 0.06 2.50 ± 0.58 2.22 ± 0.26

X90 (μm) 5.34 ± 0.63 4.57 ± 1.19 3.82 ± 0.28

Table 3 - the jet milled isoniazid laser diffraction data. Depicted the X10, X50 and X90 at 3 bars (average ± SD, n = 2).

Table 4 shows the amount of ‘blockages’ (i.e. where agglomerates in the formulation cause almost complete inhaler retention) during the dispersion measurements. At a pressure drop of 2 kPa all measurements showed blocked Twincers®, at 4 kPa the situation improves for the low and high pressure milled samples with 1 blockage for every 2 measurements. However, at 2 kPa 4 kPa 6 kPa 6 kPa these samples show again Low pressure 2/2 1/2 2/2 blockages in all measurements except Medium pressure 2/2 2/2 0/2 for the medium pressure sample, there High pressure no blocked Twincers® were seen. 2/2 1/2 2/2 Table 4 - amount of measurements that showed blockages in the Twincer® dpi during dispersion measurements (n=2).

Concentration isoniazid (mg/ml) 50 50 50 50 50 25 75 50 + Leucine

Inlet temperature (°C) 60 40 160 160 20-30 60 60 60

Feed rate (ml/min) 2.5 1 2.5 12.5 0.1 2.5 2.5 2.5

Outlet temperature (°C) 37 31 94 57 20-27 37 37 37

X10 (μm)

X50 (μm)

X90 (μm)

3.09 ± 0.29 2.93 ± 0.15 3.39 ± 0.56 3.42 ± 0.17 2.52 ± 0.37 2.72 ± 0.16 3.66 ± 0.27 0.99 ± 0.06

10.95 ± 0.16 9.68 ± 0.93 20.32 ± 4.18 16.06 ± 3.85 9.61 ± 1.46 9.22 ± 0.20 14.08 ± 0.92 2.57 ± 0.13

30.90 ± 1.62 23.27 ± 3.20 127 ± 11.53 86.96 ± 31.91 30.98 ± 8.51 24.54 ± 0.83 44.61 ± 6.19 5.08 ± 0.22

Table 5 - the spray drying settings used in the B-290 spray drier and the resulting X10, X50 and X90 as measured with laser diffraction at 3 bars (average ± SD, n = 2). Atomizing air flow was 50 mm while the aspirator flow was 100% for all samples.

103


Drug Delivery to the Lungs 27, 2016 – I. Sibum et al. Table 5 shows the different spray drying settings tested. The settings that have not been changed are the aspirator flow and the atomizing air flow, these were 100% and 50 mm respectively. Particle size seems to be mostly affected by the isoniazid concentration in the feed solution and the inlet temperature, while the feed rate has a negligible effect. Only with Lleucine particles small enough for pulmonary administration were obtained. Figure 1 shows representative DSC data from the jet milled, spray dried and spray dried with 5% L-leucine samples. All samples show a sharp melting peak at 171°C except for the samples containing L-leucine, which have a broader melting peak at 170°C. All samples are completely crystalline.

Figure 1 - representative DSC data from the jet milled and spray dried samples. All samples were measured twice.

Temperature (°C) 80 100 120 140

Mass loss (°C) 0.13 1.13 8.18 45.10

In Table 6 the TGA data is shown. At 80°C no mass loss is seen with a decrease over a 4-hour period of only 0.13%. However, at 100°C isoniazid starts to show signs of sublimation, with a decrease of 1.13% in mass. As expected, the rate of sublimation increases with increasing temperature. At the highest temperature tested, 140°C, 45.10% of the sample has sublimated in a 4-hour period.

Table 6 - the TGA data showing the temperature and the mass loss at that temperature. Mass loss was measured over a 4 hour period.

Scanning electron microscopy images of the low and high pressure mill samples are depicted in Figure 2, together with the 50 mg/mL 60°C inlet spray dry samples with and without 5% L-leucine. The two jet milled samples unexpectedly showed highly smoothed particles. Furthermore, the high pressure sample shows this to a greater extent than the low pressure one, the particles look almost completely spherical and some particles seem molten together. This ’fused’ appearance is even more pronounced for the pure isoniazid spray dried sample in Figure 2C. One large particle is visible seemingly consisting of multiple particles fused together. The spray dried sample that contains 5% L-leucine consists of smaller particles compared to the formulation of pure isoniazid. These particles are spherical, which is expected from spray dried powders. The SEM images qualitatively corroborate the laser diffraction data shown in Table 1.

Figure 2, Scanning electron microscopy images of the spray dried and jet milled samples. A shows a jet milled sample at low pressure, while B shows a jet milled sample at high pressure. C shows the 50mg/ml 60 inlet sample, while D shows the sample containing 5% leucine. Magnification 5000 x.

104


Drug Delivery to the Lungs 27, 2016 - Characterisation of jet-milled and spray dried isoniazid for pulmonary administration Discussion Milled isoniazid showed an appropriate particle size for pulmonary administration, with an X50 between 2.91 ± 0.06 μm and 2.22 ± 0.26 μm for all jet milled samples. However, milled isoniazid blocked the DPI during most dispersion measurements (Table 4). It is unclear how these particles gain their molten morphology. This seems to be related to milling pressure, and could result from mechanical stress or the associated increase in temperature locally on the particles surfaces. The TGA data in Table 6 showed that isoniazid sublimates at relatively low temperatures. A possible explanation is that the local increase in temperature sublimates some of the isoniazid, when it cools down the isoniazid deposits on the surface of agglomerates fusing these particles together. However, the binding force seems to be weak since the laser diffraction is able to disperse this powder into primary particles. Further experiments are needed to corroborate this theory. It would be interesting to see what the effect of excipients would be that limit the local increase in temperature, such as a lubricant like magnesium stearate. As is visible in Table 5 the spray dried samples with pure isoniazid all have large particle sizes and are not suitable for pulmonary administration. Unexpectedly it is the inlet temperature which has the biggest influence of the parameters tested, which increases the X50 of 10.95 ± 0.16 μm at 60°C to 20.32 ± 4.18 μm at 160°C inlet temperature. These particles are larger than the droplets generated in the nozzle, which are somewhere between 8 and 9 μm. The most likely explanation for this remarkable observation is that isoniazid crystalizes in the collection vial fusing different particles together. This seems to be corroborated by DSC data (Figure 1), which are indicative of complete crystallinity, and the SEM image in Figure 2C, where different particles can be distinguished that have fused together. Isoniazid + 5% L-leucine seems to result in a particle size distribution suitable for inhalation, with an X50 of 2.57 ± 0.13 μm and an X90 of 5.08 ± 0.22 μm. The most likely explanation is that during the spray drying process L[8] leucine enriched at the droplet surface and formed a coating, which L-leucine has been known to do. This coating may have prevented the isoniazid cores from interacting with each other and thus prevented the fusion between multiple particles. Furthermore, no fusion between individual particles is visible. Conclusion Isoniazid displays some strange behaviour which may be heat related. Dispersibility of pure jet milled isoniazid is poor and pure spray dried isoniazid has an unsuitable particle size distribution. However, with the use of 5% Lleucine it is possible to improve the particle size distribution of spray dried material to an acceptable level. Further tests are needed to show if this obtained formulation is dispersible and stable over time. References 1 2 3 4 5 6 7 8

World health organisation, “Global tuberculosis report 2015,” 2015. C. M. Yuen, H. E. Jenkins, C. A. Rodriguez, S. Keshavjee, and M. C. Becerra, “Global and Regional Burden of Isoniazid-Resistant Tuberculosis.,” Pediatrics, vol. 136, no. 1, pp. e50–9, 2015. M. Hoppentocht, P. Hagedoorn, H. W. Frijlink, and A. H. De Boer, “Developments and strategies for inhaled antibiotic drugs in tuberculosis therapy: A critical evaluation,” Eur. J. Pharm. Biopharm., vol. 86, no. 1, pp. 23–30, 2014. J. G. Y. Chan, H. K. Chan, C. A. Prestidge, J. A. Denman, P. M. Young, and D. Traini, “A novel dry powder inhalable formulation incorporating three first-line anti-tubercular antibiotics,” Eur. J. Pharm. Biopharm., vol. 83, no. 2, pp. 285–292, 2013. W. Rojanarat, N. Changsan, E. Tawithong, S. Pinsuwan, H. K. Chan, and T. Srichana, “Isoniazid proliposome powders for inhalation-preparation, characterization and cell culture studies,” Int. J. Mol. Sci., vol. 12, no. 7, pp. 4414–4434, 2011. A. Bhardwaj, L. Kumar, R. K. Narang, and R. S. R. Murthy, “Development and characterization of ligandappended liposomes for multiple drug therapy for pulmonary tuberculosis.,” Artif. Cells. Blood Substit. Immobil. Biotechnol., vol. 142001, no. May, pp. 1–8, 2012. T. Parumasivam, R. Y. K. Chang, S. Abdelghany, T. T. Ye, W. J. Britton, and H.-K. Chan, “Dry powder inhalable formulations for anti-tubercular therapy,” Adv. Drug Deliv. Rev., vol. 102, pp. 83–101, 2016. L. Li, S. Sun, T. Parumasivam, J. A. Denman, T. Gengenbach, P. Tang, S. Mao, and H.-K. Chan, “lLeucine as an excipient against moisture on in vitro aerosolization performances of highly hygroscopic spray-dried powders.,” Eur. J. Pharm. Biopharm., vol. 102, pp. 132–141, 2016.

105


Drug Delivery to the Lungs 27, 2016 – S. Abdrabo* et al. Understanding DPI blends behaviour during high speed capsule filling S. Abdrabo*, M. Fridez, T. Eggimann & R. Müller-Walz Skyepharma AG, a member of the Vectura Group of companies, Eptingerstasse 61, 4132 Muttenz, Switzerland * Corresponding author : sameh.abdrabo@vectura.com

Summary Introduction: It is important to investigate the particle-particle interactions taking place in DPI formulations. These are cohesive (drug-drug), and adhesive (drug-carrier) interactions that play an important role in DPI performance during downstream processing such as capsule filling. Understanding such interactions helps in predicting problems such as aggregation or powder segregation, and controlling drug product quality. Aim: Understanding interparticulate interactions to predict DPI behaviour during high speed capsule filling for ensuring uniformity of weight and content. Materials and Methods: Three different powder blends of either a bronchodilator (X) or an ICS, Inhaled Corticosteroid (Y) or both (X plus Y) were manufactured with inhalation grade lactose as a carrier. All blends were characterized by dynamic testing (Basic Flowability Energy and Specific Energy), bulk measurements (Permeability and Compressibility), shear testing (Cohesion/Adhesion). Capsules of the three blends were filled by a high speed capsule machine dosator type. Fill weights and content uniformity were assessed. Results and Discussion: Capsule fill weight achieved was consistent for all blends, however slightly different content uniformity at higher filling speeds was found, which can be related to different drug-drug particle interactions and drug-carrier particle interactions. Conclusion: Similar DPI blends could react differently depending on the particle-particle interactions, therefore applying dynamic and bulk measurements as a characterization tool is valuable in understanding the behaviour of these blends in downstream process steps, and selecting the appropriate processing conditions accordingly to ensure that quality and robustness are built-in. Introduction Powders are considered a dispersion of solids in air with highly complex interparticulate interactions. In dry powder blends for inhalation, an active drug should be less cohesive and optimally adhesive to the carrier (usually inhalation grade lactose monohydrate) to allow both weight and content uniformity during DPI capsule filling, and also later [1] during aerosolization from the device . It is usually difficult to investigate the effect of individual physical properties such as surface free energy, mechanical interlocking, particle size, shape, surface roughness, particle deformation, and [2,3] . Measurement techniques relative humidity on powder performance during processing or aerosolization one at a time such as dynamic testing (Basic Flowability Energy and Specific Energy), bulk measurements (Permeability and Compressibility), and shear testing (Cohesion/Adhesion) can help to put these impactful physicochemical parts of the puzzle together to enable a better understanding about powder performance in downstream processing steps and during drug [4,5] . The aim of this study was to investigate the effect of powder physical properties on capsule filling uniformity by delivery a dosator type, high speed capsule filling machine. By investigating the physical properties, it will be feasible to apply the appropriate conditions in capsule filling of the DPI powder blends, to avoid problems such as aggregation, segregation, poor flow, adhesion, and cohesion in downstream processing. Measuring such physical properties should be routine for DPI powder blends characterization in line of QbD principles which require the elucidation of critical material attributes. Experimental Methods Manufacture of powder blends Three powder blends were prepared at 2kg scale with lactose monohydrate as a carrier, one blend was a combination of a bronchodilator (X) and an inhaled corticosteroid ICS (Y), and the other two were mono product blends of the respective APIs (X) and (Y), respectively at the same individual API dose concentration as the combination blend. Dose concentration of X was 27 times higher than dose concentration of Y. Capsule Filling Capsule filling was performed with the continuous motion Planeta capsule filling machine (MG2; Italy), fitted with 4 dosators, and enabled 100% net weight check (MultiNett system), and statistical gravimetric net weight control. A high-speed capsule filling process for cohesive powder blends was developed using this pilot scale equipment which is easily and fully scalable to commercial batch sizes. The target fill weight was set to 25 mg/capsule with automatic rejection limits at ±5% (23.75 – 26.25 mg), and automatic adjustment of the dosator chamber to reach the target fill weight. Two different sets of dosators with an internal diameter of 2.8 mm and 3.4 mm were used. The machine speed was varied from 6’000 to 12’000 capsules/hour (25-50 rpm), with a run time of 1 hour. Stratified sampling of the capsules was performed throughout the filling process per batch, and capsule API assay was analysed by UPLC on these samples, to assess a potential effect of high speed capsule filling on API content and content uniformity. Powder blends characterization

106


automatic rejection limits at ±5% (23.75 – 26.25 mg), and automatic adjustment of the dosator chamber to reach the target fill weight. Two different sets of dosators with an internal diameter of 2.8 mm and 3.4 mm were used. The machine speed was varied from 6’000 to 12’000 capsules/hour (25-50 rpm), with a run time of 1 hour. Stratified sampling of the capsules was performed throughout the filling process per batch, and capsule API assay was UPLC on these samples, to assess aDPI potential of high speed capsule on API content Druganalysed Delivery by to the Lungs 27, 2016 - Understanding blendseffect behaviour during high speedfilling capsule filling and content uniformity. Drug Delivery to the Lungs 27, 2016 - Understanding DPI blends behaviour during high speed capsule filling Powder characterization The threeblends powder blends were tested for blend content uniformity and characterized by FT4 Powder Rheometer (freemantechnology; UK) in terms of the following (samples were measured in duplicate): The three powder blends were tested for blend content uniformity and characterized by FT4 Powder Rheometer  Dynamic Testing: Flowability Energy,(samples Stability were Index,measured Flow RateinIndex, Specific Energy, Conditioned (freemantechnology; UK) Basic in terms of the following duplicate): Bulk Density, and Aeration Ratio / Aerated Energy  Dynamic Testing: Basic Flowabilityand Energy, Stability Index, Flow Rate Index, Specific Energy, Conditioned Bulk Measurements: Permeability Compressibility Bulk Density, Aeration Ratio / Aerated Energy Yield Strength, Flow Function, Wall Friction Angle  Shear Testing:and Shear Stress, Cohesion, Unconfined  Bulk Measurements: Permeability and Compressibility  Shear Testing: Shear Stress, Cohesion, Unconfined Yield Strength, Flow Function, Wall Friction Angle Results

Powder blends characterization Results All blendsblends met the acceptance criteria of the content uniformity test (API Assay ≥ 95%; RSD ≤ 3%). Table 1 shows Powder characterization the results of dynamic, bulk property, and shear measurements. All blends met the acceptance criteria of the content uniformity test (API Assay ≥ 95%; RSD ≤ 3%). Table 1 shows the results of dynamic, bulk property, and shear Blend measurements. Y Blend XY Blend X Dynamic Measurements Average± SD (n=2) Blend Y Blend XY Blend X BasicDynamic Flowability Energy, BFE (mJ) 79.2(±1.6) 123(±4.5) 103(±6.1) Measurements Average± SD (n=2) Stability Index, SI 1.1(±0.01) 1.3(±0.01) 1.3(±0.04) Basic Flowability Energy, BFE (mJ) 79.2(±1.6) 123(±4.5) 103(±6.1) Flow Rate Index, FRI 1.9(±0.002) 5.3(±0.31) 3.3(±0.006) Stability Index, SI 1.1(±0.01) 1.3(±0.01) 1.3(±0.04) Specific Energy, SE (mJ/g) 4.37(±0.05) 10.3(±0.32) 9.1(±0.42) Flow Rate Index, FRI 1.9(±0.002) 5.3(±0.31) 3.3(±0.006) Conditioned Bulk Density, CBD (g/ml) 0.803(±0.006) 0.675(±0.01) 0.697(±0.01) Specific Energy, SE (mJ/g) 4.37(±0.05) 10.3(±0.32) 9.1(±0.42) Aeration Ratio, AR5 3.7(±0.2) 2.2(±0.2) 2.2(±0.03) Conditioned Bulk Density, CBD (g/ml) 0.803(±0.006) 0.675(±0.01) 0.697(±0.01) Aeration Energy, AE5 (mJ) 20.1(±1.8) 36.5(±2.2) 30.8(±0.27) Aeration Ratio, AR5 3.7(±0.2) 2.2(±0.2) 2.2(±0.03) Bulk Property Measurements Blend Y Blend XY Blend X Aeration Energy, AE5 (mJ) 20.1(±1.8) 36.5(±2.2) 30.8(±0.27) Pressure Drop, PD15,2 (mbar) 17.3(±0.22) 37.1(±0.33) 32.0(±0.5) Bulk Property Measurements Blend Y Blend XY Blend X Permeability, k15,2 x109 (cm2) 3.4(±0.01) 1.4(±0.02) 1.6(±0.04) Pressure Drop, PD15,2 (mbar) 17.3(±0.22) 37.1(±0.33) 32.0(±0.5) Compressibility, CPS15 (%) 16.8(±0.3) 26.8(±0.2) 25.2(±0.6) Permeability, k15,2 x109 (cm2) 3.4(±0.01) 1.4(±0.02) 1.6(±0.04) Shear Measurements Blend Y Blend XY Blend X Compressibility, CPS15 (%) 16.8(±0.3) 26.8(±0.2) 25.2(±0.6) Shear Stress, Ʈ7,9 (kPa) 4.1(±0.11) 6.1(±0.01) 6.0(±0.2) Shear Measurements Blend Y Blend XY Blend X Shear Stress, Ʈ3,9 (kPa) 2.1(±0.006) 3.5(±0.05) 3.4(±0.2) Shear Stress, Ʈ7,9 (kPa) 4.1(±0.11) 6.1(±0.01) 6.0(±0.2) Flow Function, FF 6.7 3.0 3.0 Shear Stress, Ʈ3,9 (kPa) 2.1(±0.006) 3.5(±0.05) 3.4(±0.2) Wall Friction Angle, WFA1.2µm (°) 19.5 26.4 23.3 Flowblends Function, FF 6.7 Rheometer 3.0 3.0 Table 1 - Powder characterization by FT4 Powder Wall Friction Angle, WFA1.2µm (°) 19.5 26.4 23.3 Table 1 - Powder blends characterization by FT4 Powder Rheometer

Figure 1 – Basic Flowability Energy and Specific Energy of the DPI blends Figure 1 – Basic Flowability Energy and Specific Energy of the DPI blends

Figure 2 – Permeability and Compressibility of the DPI blends Figure 2 – Permeability and Compressibility of the DPI blends

107


Drug Delivery to the Lungs 27, 2016 – S. Abdrabo* et al. As shown in Table 1 and Figures 1-3, the blend comprising Y differs significantly (p<0.05) from the powder blends containing drug substance X either alone or in combination with Y in terms of Basic Flowability Energy BFE, Specific Energy SE, and Flow Rate Index FRI, Aerated Energy AE, Compressibility, Permeability and Shear Stress values.

Figure 3 – Shear cell analysis of the DPI blends

Capsule Filling Table 2 shows capsule filling results in terms of fill weight accuracy. Overall, all powder blends showed good capsule filling performance in terms of fill weight and RSD. The larger dosators (diameter 3.4 mm) were used only once and resulted in a higher variability in fill weight compared to the smaller dosators with 2.8mm diameter (batches 1B to 1A). For all three blends, slightly smaller fill weight variability was seen for the batches filled at higher speed. All fill weight distributions showed a Normal distribution. Batch Active ingredient Dosator diameter [mm] Filling speed [caps/hr] Conforming capsules filled Non-conforming capsules Non-conforming capsules [%] Fill weight average [mg] Fill weight RSD [%]

1A

1B

3A

3B

4A

4B

5A

5B

Y

Y

Y

Y

X

X

XY

XY

2.8

3.4

2.8

2.8

2.8

2.8

2.8

2.8

12,000

12,000

6,000

10,000

6,000

10,000

6,000

10,000

9,974

8,509

5,838

9,539

4,073

9,626

4,070

9,638

22

798

115

19

287

83

135

182

0.22

9.38

1.97

0.20

7.05

0.86

3.32

1.89

24.98

25.04

24.80

24.96

25.02

25.02

25.03

25.03

1.57

2.15

1.85

1.65

2.17

1.69

2.09

1.83

Table 2 – Capsule filling parameters and results. The fill weights shown were measured from the 100% capacitive weight control system (MultiNett). Rejection limits were set at ±5% (23.75 – 26.25mg)

The API content in the capsules was analysed for batches 3, 4 and 5. For batch 3 (single API, Y), very stable assay results were obtained, the initial higher value for batch 3A can be linked to a higher fill weight at the start of filling. Batch 4 (single API, X), showed more variability in the results, especially at a speed of 10’000 capsules/hour (time 0 and 60 minutes), see Figure 4. For the combination batch (batch 5), slightly variable assays were obtained for both actives, see Figure 5.

Figure 4 – Capsule content results of the single API batches filled with the 2.8mm Ø dosators at different speeds

108


Drug Delivery to the Lungs 27, 2016 - Understanding DPI blends behaviour during high speed capsule filling

Figure 5 – Capsule content results of the combination batch filled with the 2.8mm Ă˜ dosators at different speeds

Discussion Blend Y showed a significantly lower Specific Energy (SE) (p<0.05) indicating less mechanical interlocking and friction, Figure 1; the lower Aerated Energy (AE), (p<0.05), in comparison to blends containing X (Table 1) indicates that this blend has the lowest cohesion, hence better flow resulting in more uniform capsule filling. The lower Compressibility and Permeability values (p<0.05), Figure 2; also indicate a denser particle packing, a characteristic of free-flowing powders. Also, blend Y showed a lower shear stress value in comparison to blends containing X (p<0.05), Figure 3, confirming the better flowability of the Y blend. On the other hand, the combination blend containing X and Y showed the highest Basic Flowability Energy (BFE), Figure 1, and the highest Flow Rate Index, as shown in Table 1 indicating the highest resistance to flow, which is characteristic for cohesive powders. Moreover, as shown in Figure 2, this blend together with the mono blend containing X had the highest Pressure Drop across powder bed i.e. the least permeability, and the highest Wall Friction Angle (WFA), Table 1, which further confirms the cohesiveness of blends containing X. It can be assumed that active X having a higher concentration in the blend; it dominates the properties of the blend. All these results explain why blends containing X as mono blend or in combination showed larger variability in capsule content (Figures 4, 5). Conclusion Dynamic, bulk property, and shear measurements using the FT4 Powder Rheometer correlated well with the results from the capsule filling process, and can be used as a predictive tool in the formulation screening of blends for high speed capsule filing. The results showed clearly that the cohesiveness of blends containing the active drug X resulted in slightly more variable filling into capsules mainly at higher speeds, while a more uniform content was found for the powder blend with drug Y due to better flow. Acknowledgment Vectura would like to acknowledge freemantechnology (Gloucestershire, UK) for powder blend characterization using FT4 Powder Rheometer. 1

Le V N P, Hoang Thi T H, Robins E, Flament M P: Dry Powder Inhalers: Study of the Parameters Influencing Adhesion and Dispersion of Fluticasone Propionate, AAPS PharmSciTech 2012; 13: pp 477-484.

2

Tan B M J, Chan LW, Heng P W S: Improving Dry Powder Inhaler Performance by Surface Roughening of Lactose Carrier Particles, Pharm Res 2016; 33: pp 1923-1935. 3

Depasquale r, Lee S L, Saluja B,Shur J, Price R: The Influence of Secondary Processing on the Structural Relaxation Dynamics of Fluticasone Propionate, AAPS PharmSciTech 2015; 16: pp 589-599.

4

Osorio JG, Muzzio FJ: Effects of powder flow properties on capsule filling weight uniformity, Drug Dev Ind Pharm 2013 Sep;39(9): pp1464-1475. 5

Freeman T, Moolchandani V, Hoag S W, Fu X: Capsule Filling Performance of Powdered Formulations in Relation to Flow Characteristics. In C Y Wu, and W Ge (eds.): Particulate Materials: Synthesis, Characterization, Processing and Modelling; 2012.

109


Drug Delivery to the Lungs 27, 2016 - Judith Heidland et al. Nano-in-microparticle powders for mucosal vaccination – understanding the particle forming process Judith Heidland & Regina Scherließ Department of Pharmaceutics and Biopharmaceutics, Kiel University, Grasweg 9a, 24118 Kiel, Germany Summary Nanoparticulate antigen encapsulation to induce a local immune response upon nasal application is a promising alternative to the commonly used parenteral vaccination. In this study, the particle forming process of such nanoparticles, consisting of chitosan and a low viscosity sodium carboxymethylcellulose sodium salt – produced via ionic gelation, is illustrated. Ovalbumin as model antigen was incorporated into the particles for some experiments. Different qualities of chitosan (varying degree of deacetylation as well as molecular weight) were investigated regarding their influence on the resulting particle sizes, showing an increase of size for qualities with a higher molecular weight. For selected qualities a closer look into the particle forming process was taken to 1 improve its understanding. For this, gel permeation chromatography and H-NMR experiments were performed to assess whether changes in the chitosan regarding molecular weight or degree of deacetylation occur during particle formation, to get an idea about the “reactive share” of the chitosan used. To gain further information regarding the binding mechanisms between the components isothermal titration calorimetry experiments were performed finding differences between the different chitosan qualities. Moreover, a different counterion (sodium deoxycholate) was utilised to detect potential differences depending on the counterion (large carboxymethylcellulose sodium salt, approximately 90 kDa vs. small sodium deoxycholate, 0.4 kDa) regarding the particle size as well as the zeta potential. Nanoparticles can be incorporated in a microparticulate stabilising matrix to allow respiratory application and deposition.

Introduction [1]

Mucosal vaccination via the respiratory tract holds promise and has been under research for some time . Both, the nasal and the pulmonary mucosa, are easily accessible and comprise many immune cells, making it an especially attractive application route. Furthermore, the respiratory tract is the natural site of entrance for many pathogens and therefore an even more promising target for vaccination. Still, there are challenges associated with this as well, i.e. a sufficient dwell time on the mucosa as well as a sufficient uptake into and activation of antigen presenting cells. Antigens encapsulated into particulate systems are able to induce a local immune response in addition to a [1] systemic one . Moreover, particulate systems have shown superior immune effects compared to dissolved antigen. The uptake furthermore depends on the particle size. Those ranging from 250 to 400 nm are most [2] favourable . Therefore, in this project a nano-in-microparticle-formulation (NiM) has been chosen as vehicle because the nanoparticle ensures the uptake into cells whereas the microparticle (produced via spray drying of the nanosuspension with mannitol as matrix) secures the deposition on the mucosa (10-50 µm particles for [3] deposition on the nasal mucosa and particles smaller than 5 µm for pulmonary deposition). In this study the focus lies on the understanding of the nanoparticle forming process. Hence, nanoparticles based on chitosan and two different counterions (large carboxymethylcellulose sodium salt and small sodium deoxycholate) have been produced and analysed.

Materials and Methods Preparation of nanoparticles Different chitosan qualities (with varying degree of deacetylation (DDA) and molecular weight) obtained from Heppe Medical Chitosan, Germany, were used. A low viscosity carboxymethylcellulose sodium salt (CMC, approximately 90 kDa, Sigma-Aldrich, USA) and sodium deoxycholate (DOC, 0.4 kDa, Carl Roth GmbH + Co. KG, Germany) served as counterions. Ovalbumin (OVA, Sigma-Aldrich, USA) was used as a model antigen. Nanoparticles were formed by ionic gelation of the positively charged chitosan with the negatively charged counterion (CMC or DOC). Chitosan was dissolved in acetic acid (2 % (w/v) for particle formation with CMC and 1 % (w/v) with DOC, respectively) to a concentration of 0.1 %. CMC or DOC was dissolved in ultrapure water (Direct-Q 3 UV, Merck Millipore, Germany) to the same concentration. Nanoparticles formed spontaneously upon mixing of these two components utilising a magnetic stirrer.

110


Drug Delivery to the Lungs 27, 2016 - Nano-in-microparticle powders for mucosal vaccination – understanding the particle forming process Size and zeta potential measurements Size and zeta potential of the nanoparticles were determined with the Zetasizer Nano ZS (Malvern Instruments, UK). Size measurements (utilising dynamic light scattering) were performed with undiluted nanosuspensions whereas the zeta potential (utilising Laser Doppler anemometry) was measured with a 1:2 diluted suspension. Each measurement was performed in triplicate, results are given as mean. Determination of molecular weight Gel permeation chromatography (GPC) served to determine the molar mass using a PL-GPC 50 Plus (Polymer Laboratories, USA). The system contains an integrated degasser and a differential refractive index detector (Polymer Laboratories, Varian Inc., USA). A MALS detector (Mini DAWN Tristar, Wyatt) was utilised for the [4] determination of molar mass, assuming a refractive index increment of 0.185 mL/g . For the separation by size, a PL aquagel OH Guard 8 μm precolumn followed by a PL aquagel OH 40 8 μm column and one PL aquagel OH 30 8 μm column (Agilent Technologies, Santa Clara, USA) in series were used. Samples (3 mg/mL in acetic acid 1 %) were eluted with 0.25 M NaNO3 und 0.01 M NaH2PO4 in water (pH 2.3) at a flow rate of 1.0 mL/min and the column temperature was kept at 35 °C. Determination of DDA 1

H-NMR was utilised to assess the DDA of chitosan. The experiment was performed modifying the method of [5] Lavertu et al. . Samples were dissolved in 2 %CD3COOD (Carl Roth GmbH + Co. KG, Germany) to a concentration of 5 mg/mL. The measurements were performed with a BRUKER Avance III 300 MHz spectrometer (Bruker Corporation, USA) at 80 °C. The DDA was determined using Equation 1: Equation 1 Determination of binding characteristics

Typically, isothermal titration calorimetry (ITC) is used to characterise the binding affinities between proteins and ligands. In our study we used the MicroCal VP-ITC (Malvern Instruments, UK) to gain knowledge about the interaction between chitosan and CMC. For this chitosan and CMC were dissolved in 1 % acetic acid, chitosan to a concentration of 0.05 % in the cell and CMC to 0.9 % in the syringe (=injectant). For some experiments OVA was added to the reaction, either in the syringe (0.44 %) or in the cell (0.05 %). 35 injections (8 µL over 16 s, spaced by 200 s) were made per experiment. Stirring speed was set at 307 rpm. Measurements were performed at 25 °C.

Results and Discussion Particle sizes obtained when using different qualities of chitosan for the process are shown in Figure 1. Size can be tuned by using different molecular weight chitosan, whereas the DDA does not influence the particle size. Further, particles get slightly smaller upon the incorporation of antigen compared to unloaded particles (data not shown). As target particle size is between 250 and 400 nm to ensure efficient uptake by immune competent cells, the molecular weight should be kept below 150 kDa.

Figure 1 – Resulting nanoparticle sizes utilising different chitosan qualities (first set of numbers represents the DDA in percent, second set the viscosity in mPas (1 % solution in acetic acid 1 %, 20 °C); for the chitosan 90/10 three different batches (a-c) have been utilised). On the left the different chitosans have been ordered according to their molecular weight, on the right according to their DDA. (n=3, error bars=sd)

111


Drug Delivery to the Lungs 27, 2016 - Judith Heidland et al. During nanoparticle processing it became obvious that the polymer does not completely react to nanoparticles. To better understand the particle forming process, DDA and molar mass were determined before and after particle formation. For these experiments it was necessary to obtain unreacted chitosan after particle formation; consequently, a nanosuspension was centrifuged to separate the nanoparticles from the supernatant. The excess chitosan in the supernatant was precipitated using 1 N NaOH prior to repeated centrifugation. The chitosan was dried for 48 h in a vacuum drier subsequently. 1

Figure 2 shows the results for the H-NMR as well as the GPC experiments. DDA and molar mass were lower after particle formation compared to the starting substances. To conclude, the “more reactive chitosan” was found to have both a higher molecular weight and a higher DDA (meaning more free amino groups). Nevertheless, no separate fractions of molecular weight could be found for the examined chitosan qualities. The respective peak simply shifts slightly after particle formation. Hence, it was not possible to identify the “reactive part” of the chitosan to work with this exclusively. The same is true for the DDA, a “reactive part” could not be separated from the rest.

Figure 2 – Left: Results from 1H-NMR experiments showing the DDA values for the different chitosan qualities before and after particle formation. Error bars show standard deviation. Right: Resulting molar mass values for the different chitosan qualities before and after particle formation.

Figure 3 – ITC results. Left: Binding curves for different chitosan qualities reacting with CMC showing a difference in binding affinities between varying DDA and no difference for different molecular weights. Right: Differences between particle formation including or not including OVA in the cell displaying that more CMC is necessary for the reaction if additional cations (OVA) are present. The concentration of the OVA is equal to the concentration of chitosan but not taken into account on the axis.

Isothermal titration calorimetry experiments showed that no reaction took place between chitosan and OVA which was expected because both polymers have a positive charge in acidic medium. Further experiments (Figure 3) showed differences in the particle formation process: Results suggest that a higher DDA led to more chitosan reacting with the CMC as well as OVA in the chitosan phase resulted in an increased amount of reacting CMC. Molecular weight however does not influence the reaction kinetics. Problem with these experiments was the titration to the “endpoint” – in the case of chitosan nanoparticles this meant agglomeration as the nanosuspension is only stabilised electrostatically. Consequently, if all the positive charges of the chitosan are neutralised, agglomeration occurs. Formulations with [6] a zeta potential larger than +30 mV (or smaller than -30 mV) are considered stable . Figure 4 shows the results of the titration of CMC to chitosan 90/10 – either containing or not containing OVA. It could be shown clearly that at a certain ratio of the two ions (chitosan and CMC) z-average increases dramatically. At the same time, zeta potential drops below 30 mV indicating loss of stabilisation and agglomeration of nanoparticles. In this experiment

112


Drug Delivery to the Lungs 27, 2016 - Nano-in-microparticle powders for mucosal vaccination – understanding the particle forming process no apparent difference in zeta potential was observed irrespective of the formulation including OVA or not. Accordingly, the difference observed using ITC is probably rather small and not relevant for the particle formation on a larger scale.

Figure 4 – Left: Resulting nanoparticle sizes from titration experiment with either placebo- or OVA-containing NP. Right: Resulting zeta potentials for the nanosuspensions.

These findings were compared to particles formed with chitosan and DOC showing (Figure 4) a difference in particle formation between the two counterions. The smaller DOC formed smaller particles with a higher zeta potential most likely due to a lower charge density. Further experiments need to be performed to verify this hypothesis.

Figure 5 – Differences in particle size (left) and zeta potential (right) of nanosuspension formed with CMC and DOC, respectively.

Conclusion and Outlook The particle forming process for chitosan:CMC nanoparticles could be elucidated with the chosen methods. The nanoparticle size could be tuned by the molecular weight of chitosan, an increasing molecular weight resulted in larger particles. The experiments examining molecular weight and DDA of the chitosan before and after particle formation showed that these parameters change during the process and therefore it is not simple to determine the amount of chitosan reacted as most quantification methods depend on the DDA. The ITC experiments further showed an influence of the DDA on the particle forming process as a higher DDA resulted in a more intense reaction (in terms of energy and binding ratio). For the counterion both the type as well as the amount influences particle size and zeta potential. Future work will be focussing on the protein encapsulation efficiency for different formulations as well as their efficiency regarding uptake into and activation of immunocompetent cells.

Acknowledgements The authors would like to thank Jeroen Mesters for conducting the ITC measurements. References 1 Holmgren J, Czerkinsky C: Mucosal immunity and vaccines, Nat Med 2005; 11 (4 Suppl): pp 45-53 2 Rietscher R, Schröder M, Janke J, Czaplewska J, Gottschaldt M, Scherließ R, Hanefeld A, Schubert U S, Schneider M, Knolle P A, Lehr C: Antigen delivery via hydrophilic PEG-b-PAGE-b-PLGA nanoparticles boosts vaccination induced T cell immunity, Eur J Pharm Biopharm 2016; 102: pp 20-31 3 Garmise R J, Hickey A: Dry powder nasal vaccines as an alternative to needle-based delivery, Crit Rev Ther Drug Carrier Syst 2009; 26:pp 1-27 4 Sorlier P, Rochas C, Morfin I, Viton C, Domard A: Light Scattering Studies of the Solution Properties of Chitosans of Varying Degrees of Acetylation, Biomacromolecules 2003; 4: pp 1034-1040 5 Lavertu M, Xia Z, Serrequ A N, Berrada M, Rodrigues A, Wang D, Buschmann M D, Gupta A: A validated 1H-NMR method for the determination of the degree of deacetylation of chitosan, J Pharm Biomed Anal 2003; 32: pp 1149-1158 6 Zetasizer Nano Series User Manual, Malvern Instruments, UK, 2013

113


Drug Delivery to the Lungs 27, 2016 – Nathalie Wauthoz et al. Impact of different capsules for dry powder inhalers on the aerodynamic performances of formoterol-based binary and ternary blends 1

1

2

Nathalie Wauthoz , Ismaël Hennia , Susana Ecenarro & Karim Amighi 1

1

Laboratory of Pharmaceutics and Biopharmaceutics, Université libre de Bruxelles (ULB), Boulevard du Triomphe CP207, Brussels 1050, Belgium 2 Qualicaps Europe S.A.U., Avenida Monte Valdelatas 4, Alcobendas 28108, Spain

Summary Background: Independently of the patient, the aerodynamic performance of a dry powder for inhalation depends on the formulation and the dry powder inhaler (DPI). In the case of capsule-based DPIs, the capsule also plays a role, not only in the packaging of the formulation, but also in the powder aerosolisation and the dispersion of the micronised drug from the carrier during inhalation. Therefore, the capsule is an important parameter. However, few studies have been conducted on the impact of the kind of capsule on aerodynamic performance. Methods: This study evaluated TM ® ® the impact of different capsules (Quali-G and Quali-V -I from Qualicaps , and hard gelatin capsules for DPIs, ® ® ® Vcaps and Vcaps Plus from Capsugel ) on the delivered doses (DDs) and fine particle doses (FPDs) of formoterol® based binary and ternary blends using the Axahaler DPI. Results: Similar trends were observed for the DD, FPD and formoterol capsule retention for both dry powder formulations contained in the different capsules. The highest DDs and FPDs and the lowest formoterol capsule retention were observed with hypromellose capsules such as QualiV-I and Vcaps, without significant differences between these capsules (p ˃ 0.05, one-way ANOVA with NewmanKeuls post-hoc test) for both dry powders. The ternary blend showed higher and less sensitive FPDs to the kind of capsules used than those observed for the binary blend. Conclusion: The choice of the kind of capsule used to package the dry powder has an influence on its aerodynamic performance, with the best results obtained from ® ® hypromellose capsules (Quali-V -I and Vcaps ) for the binary and ternary mixtures used in this study. Introduction Independently of the patient, the aerodynamic performance of a dry powder for inhalation depends on the formulation and the dry powder inhaler (DPI). In the case of capsule-based DPIs, the capsule also plays a role, not only in the packaging of the formulation, but also in the powder aerosolisation and the dispersion of the micronised drug from the [1] carrier during inhalation . Therefore, the capsule is an important parameter. However, few studies have been conducted on the impact of the kind of capsule on dry powder aerodynamic performance. The aim of this work was to evaluate the impact on the aerodynamic performance of two dry powder formulations for ® inhalation based on formoterol by using different capsules in the Axahaler capsule-based DPI. The work evaluated TM ® different capsules used in, though not necessarily marketed for, the inhalation field: Quali-G and Quali-V -I from ® Qualicaps for gelatin and hypromellose (HPMC) capsules, respectively, and hard gelatin capsules for DPIs (HCG), ® ® ® Vcaps and Vcaps Plus from Capsugel for gelatin, HPMC and second-generation HPMC capsules, respectively. To [2] be representative of the kind of dry powder found on the market for low-content drug formulations , two dry powders for inhalation were produced using the same micronised formoterol, but with different lactoses for inhalation: milled lactose presenting a broad particle size distribution (PSD) to use in a binary mixture, and sieved lactose presenting a narrow PSD with the addition of 10% of fine lactose to use in a ternary mixture.

Experimental Methods Materials Milled lactose carrier with a broad PSD characterised by a d50 of 46 µm and a span of 3.13 (Respitose ML001), sieved lactose carrier with a narrow PSD characterised by a d50 of 113 µm and a span of 1.11 (Respitose SV010) and fine lactose characterised by a d50 of 3 µm and a d90 of 7 µm (Lactohale LH300) were kindly donated by DFE Pharma (Goch, Germany). Micronised formoterol fumarate dihydrate characterised by a d50 of 1.5 µm and by 99.81% < 5 µm (formoterol) was purchased from Chemo (Madrid, Spain). Size 3 capsules for inhalation were kindly donated by Qualicaps (Quali-G and Quali-V-I for gelatin and HPMC capsules, respectively) and Capsugel (HCG DPI, Vcaps, VcapsPlus for gelatin, HPMC, and second-generation HPMC, respectively). Low resistance Axahaler capsule-based DPIs were kindly donated by SMB (Brussels, Belgium). Potassium phosphate and HPLC-grade acetonitrile were purchased from Merck (Darmstadt, Germany) and HPLC-grade methanol from VWR (Fontenay sous Bois, France).

114


Drug Delivery to the Lungs 27, 2016 - Impact of different capsules for dry powder inhalers on the aerodynamic performances of formoterol-based binary and ternary blends Dry powders Dry powder A (binary mixture) was made by blending Respitose ML001 and formoterol (0.05% w/w) in a plastic vessel filled at most to 40% of its inner volume. The blending was performed using a laboratory-scale threedimensional motion mixer, the Turbula 2C (Bachofen AG, Switzerland), as follows. First, 12 mg of formoterol was set in with 12 g of carrier using the so-called sandwich method and all the powder was sieved using a 224 µm sieve. The mixture was then blended in the Turbula 2C for 5 minutes. The obtained pre-mix was then set in sandwich with an additional quantity of 12 g of carrier and all the powder was sieved using a 224 µm sieve and blended in the Turbula 2C for 10 minutes. Then, the blend was sieved again through the 224 µm sieve before being blended in the Turbula 2C for 10 minutes. The mixing speed was set at 46.2 rpm. Dry powder B (ternary mixture) was made by blending Respitose SV010, Lactohale 300 and formoterol (0.05% w/w) in a plastic vessel filled at most to 40% of its inner volume. The blending was performed using a Turbula 2C, as follows. First, 12 mg of formoterol was set in with 2.4 g of coarse carrier and 2.4 g of fine lactose using the so-called sandwich method and all the powder was sieved using a 224 µm sieve. The mixture was then blended in the Turbula 2C for 10 minutes. The obtained pre-mix was then sieved with 19.2 g of coarse carrier using a 224 µm sieve and blended in the Turbula 2C for 15 minutes. Then, the blend was sieved again through the 224 µm sieve before being blended in the Turbula 2C for 10 minutes. The mixing speed was set at 46.2 rpm. Uniformity of drug content Uniformity of drug content of both dry powder blends was assessed using test B for uniformity of content of single[3] dose preparations in European Pharmacopeia 8.0 . This test consists of determining the individual formoterol content of 10 dosage units. Briefly, about 24 ± 1 mg of each blend was weighed accurately (to 0.01 mg precision) in a 50-mL volumetric flask that was then filled with HPLC dilution phase. The formoterol determination was performed using a validated analytical method. Packaging and storage Amounts of 24 ± 1 mg of each blend were weighed in 50 capsules of each kind. The filled capsules were stored in a plastic container for a minimum of three weeks at 20°C and 50% RH, which corresponds to the usual environmental conditions used by firms producing dry powders for inhalation. Uniformity of Delivered Doses (DDs) The uniformity of DDs of both dry powders packaged in the different kinds of capsule was determined using a dosage unit sampling apparatus (Copley, Scientific Limited, Nottingham, UK) containing a 47 mm glass fiber filter (PALL ® CORPORATION , USA) on 10 dosage units. A flow rate of 100 L/min for 2.4 seconds was adopted, as recommended [4] with this low-resistance device by European Pharmacopeia 8.0. . Then, each DD was recovered from the dosage unit sampling apparatus with the dilution phase in a 50-mL volumetric flask, sonicated for 20 min and filtered on a 25 mm glass microfiber filter (GE Healthcare Life Science, UK). Fine particle dose (FPD) and capsule retention The aerodynamic behaviour of both dry powders packaged in the different capsules was determined using a Next Generation Impactor (NGI, apparatus E) (Copley Scientific Limited, Nottingham, UK) with a pre-separator and uncoated plates. A new Axahaler was used for each kind of capsule filled with each blend. A flow rate of 100 L/min for [4,5] . 2.4 seconds was adopted, as recommended with this low-resistance device by European Pharmacopeia 8.0. Three independent tests were performed on the capsules. For each test, 10 capsules used in DPIs were pre-filled with 24 ± 1 mg of the blend and stored as mentioned in “Packaging and storage”. The quantity of formoterol deposited at each level, i.e. capsules, device, induction port, pre-separator, stages 1-7 and micro-orifice collector, was recovered using an HPLC dilution phase. The quantity of formoterol deposited was determined using a validated analytical method. The FPD was determined using the Copley inhaler testing data analysis software (Copley Scientific Limited, Nottingham, UK). Analytical method The determination of formoterol was performed using a chromatographic system (HP 1200 series, Agilent Technologies, Brussels, Belgium) equipped with a binary pump, an auto-sampler and a diode array detector. The separations were performed on a LiChrospher 60 RP Select B (125 × 4 mm, 5 µm) column (Merck, Darmstadt, Germany) equipped with a pre-column (Merck, Darmstadt, Germany). The mobile phase consisted of acetonitrile and phosphate buffer 0.01M, pH 2.7 (25:75 v/v) at 1.0 mL/min. The quantification was performed at 245 nm. The standard curve was linear in the 25-1000 ng/mL range. The samples were recovered and diluted with a dilution phase

115


Drug Delivery to the Lungs 27, 2016 – Nathalie Wauthoz et al. consisting of methanol and milliQ water (25:75 v/v). The volume injected was 200 µL using a 400 µL extension loop (Agilent, Brussels, Belgium). The temperature was set at 30°C and the analysis time was 6 min. Data analysis Statistical comparisons were made using a one-way analysis of variance (ANOVA), having checked the homoscedasticity of variances. When homoscedasticity of variances was not encountered, a non-parametric Kruskal Wallis test was performed. When these multi-group tests were significant, post-hoc tests (Newman Keuls or Dunn’s procedure) were used to avoid multiple comparison effects when comparing the group pairs of interest for parametric or non-parametric tests, respectively. Results and Discussion The formoterol-based blends complied with the test for uniformity of content of single-dose preparations (Test B) in European Pharmacopeia 8.0. This was tested on 10 dosage units, as no more than one individual content was outside the limits of 85% and 115% of the average content, and none was outside the limits of 75% and 125% of the average content. The formoterol contents were 12.0  0.1 µg (CV% of 1.2%) and 11.6  0.3 µg (CV% of 2.8%) per 24 mg of the dry powders A and B, respectively. The blends were considered homogeneous because the CV% of the [2] drug content was below 5% . DDs (i.e. the formoterol dose emitted from the capsule and the device), FPDs (i.e. the formoterol dose presenting an aerodynamic diameter ≤ 5 µm and able theoretically to deposit into the lungs) and formoterol capsule retention (i.e. formoterol dose recovered in the capsule after the dose delivery) were evaluated for dry powders A and B packaged in different kinds of capsules (Quali-G and HCG for gelatin capsules; Quali-V-I and Vcaps for HPMC capsules; and VcapsPlus for second-generation HPMC), as shown in Figure 1. Similar trends were observed for the DD, FPD and formoterol capsule retention for both dry powders packaged in the different capsules. The highest DD and FPD and the lowest formoterol capsule retention were observed with HPMC capsules such as Quali-V-I and Vcaps without significant differences between the HPMC capsules (p ˃ 0.05, one-way ANOVA with Newman-Keuls post-hoc test) for both dry powders. More specifically for dry powder A, significantly higher FPDs (p < 0.01, one-way ANOVA with Newman-Keuls post-hoc test) and lower capsule retentions (p < 0.001, one-way ANOVA with Newman-Keuls post-hoc test) were observed for HPMC capsules (Quali-V-I, Vcaps) in comparison with gelatin capsules (Quali-G and HCG DPI) and secondgeneration HPMC (VcapsPlus). Dry powder B, made with a ternary mixture based on sieved lactose and fine lactose, showed higher FPDs than those observed for dry powder A, made with a binary mixture based on milled lactose (4.15 ± 0.1 µg to 5.15 ± 0.33 µg vs 2.6 ± 0.1 µg to 3.71 ± 0.2 µg, respectively) and was also less sensitive to the kind of capsule used (less significant differences between the different kind of capsules than observed for dry powder A). However, the formoterol capsule retention was higher with the dry powder B than with dry powder A (1.1 ± 0.2% to 4.3 ± 1.2% vs 0.0 ± 0.0% to 2.0 ±.5%, respectively). This higher capsule retention did not impact the FPDs or the DDs which were slightly higher than those from dry powder A (8.8 ± 0.6 µg to 9.5 ± 0.5 µg vs 8.4 ± 0.2 µg to 8.98 ± 0.5 µg, respectively).

116


Drug Delivery to the Lungs 27, 2016 - Impact of different capsules for dry powder inhalers on the aerodynamic performances of formoterol-based binary and ternary blends

Figure 1 - (A) Uniformity of delivered doses (µg), (B) Fine particle doses (µg) or (C) formoterol retention (%) in the capsule for dry powders A (binary mixture) and B (ternary mixture), packaged in different capsules: hard gelatin TM ® ® ® capsules for DPI (HGC), Quali-G (QG), Vcaps (VC), Vcaps Plus (VC+) or Quali-V -I (QV I). The tests were ® performed using the Axahaler capsule-based DPI at 100 L/min for 2.4 sec. Conclusion Similar trends were observed for the DD, the FPD and formoterol capsule retention for both dry powder formulations packaged in the different capsules. The best results were obtained with HPMC (Quali-V-I and Vcaps) for the binary and ternary mixtures used in this study. Therefore, the choice of the kind of capsule used to package the dry powder formulations has an influence on the DDs and FPDs. Further investigations are needed to better understand the differences between the different capsules. References [1] Coates MS, Fletcher DF, Chan HK, Raper JA: The role of capsule on the performance of a dry powder inhaler using computational and experimental analyses, Pharm Res 2005; 22: pp 923-932. [2] Pilcer G, Wauthoz N, Amighi K: Lactose characteristics and the generation of the aerosol, Adv Drug Deliv Rev 2012; 64: pp 233-256. [3] European pharmacopeia 8.0 (2008): Uniformity of content of single-dose preparations (Test B): pp 295. [4] European pharmacopeia 8.0 (2012): Preparation for inhalation pp 800-805. [5] European pharmacopeia 8.0 (2008): Preparation for inhalation: aerodynamic assessment of fine particles: pp 309320.

117


Drug Delivery to the Lungs 27, 2016 – Makoto Kamada et al. The surface energy effect of co-spray-dried mannitol with polyethylene glycol on the aerosolization performance in a dry powder inhalation formulation Makoto Kamada, Mitsuhide Tanimoto, Atsutoshi Ito, Michiko Kumon, Kazuhiro Inoue, Shuichi Yada & Hidemi Minami Formulation Technology Research Laboratories, Daiichi Sankyo Co., Ltd., 1-2-58, Hiromachi, Shinagawa-ku, Tokyo 140-8710, Japan Summary The objective of this study is to clearly assess the surface energy effect of co-spray-dried mannitol (co-SDM) particles with polyethylene glycol (PEG) on the aerosolization performance in a dry powder inhalation (DPI) formulation. From the results of in vitro deposition studies in the model DPI formulation containing salbutamol sulphate (SS), the effect of PEG molecular weight (MW) and formulated amount of PEG in the co-SDM on the aerosolization performance was evaluated. The fine particle fraction (FPF) of SS in the formulation containing coSDM with PEG 20000 was superior to that in the formulation containing spray-dried mannitol (SDM) without PEG. On the other hand, the FPFs of SS were found to be influenced by the PEG ratio and MW of PEG used in the formulation of co-SDM with PEG. The surface energies of co-SDMs with PEG were measured by an IGC technique. It was considered that the lower dispersive component of the surface energy of co-SDM may contribute to a higher FPF of SS when using PEG with a high MW. Moreover, the elevation of the basic (electron donor) energy might result in the decrease of FPF of SS when using PEG with low MWs. The elevation of the basic (electron donor) energy of a co-SDM would relate to both/either the difference in the adhesion force to the SS and/or the difference in the hygroscopicity of each PEG. Introduction In our previous research, a manufacturing method of producing novel mannitol carrier particles by co-spray drying [1] with polyethylene glycol (PEG) was developed . For evaluation of the properties of resultant co-spray-dried mannitol with PEG (co-SDM) particles as carrier particles, model dry powder inhalation (DPI) formulations were prepared. Salbutamol sulfate (SS) was used as the model drug substance, because SS is widely used as a model [2] compound for DPI formulations . The aerosolization performance of SS in the model formulations was evaluated by using an Andersen cascade impactor (ACI). From the results of in vitro deposition studies, the formulation using co-SDM with 2.5% PEG 4000 and co-SDM with 5% PEG 20000 showed a superior aerosolization performance of SS when compared to a commercial mannitol and its corresponding spray-dried mannitol without [3] PEG (SDM) as carrier particles . Therefore, employing co-SDM seemed to be more preferable for the DPI carrier particle than using the commercial mannitol alone or SDM. However, significant differences were found in the aerosolization performance of SS dependent on the molecular weight (MW) and the formulated amount of PEG in the co-SDMs. The quality of DPI formulation, such as aerosolization performance, depends on the physical properties of not only the drug substance but also the carrier particles. For a preferable aerosolization performance of a drug, the characteristics of carrier particles must be well controlled in terms of size, shape, surface roughness, surface [4] energy, etc . In this study, we focused upon the surface adhesion force between the carrier particles and drug particles. Because there are numbers of studies focused on the surface adhesion force between the carrier [5, 6] . particles and drug substances in DPI formulation For the evaluation of the surface adhesion force, the surface energies of carrier particles were evaluated by using an inverse gas chromatography (IGC) technique. The surface energy was known to be one of the important [5, 6] . IGC can selectively detect the dispersive factors in managing the adhesion forces on the particle surface [5, 7] components and the specific (non-dispersive) components of the surface energies of solid materials . These two components are well known to indicate the polarity of the surface of measured materials. The specific interactions can be also classified as either electron donor or electron acceptor type interactions based on the [5, 7] . In fact, IGC is widely used to evaluate the effect of the acid/base approach using different polar probe gases [5-7] . surface energy of carrier particles on the aerosolization performance of drug substances in DPI formulations Therefore, IGC is considered preferable for evaluating the adhesion force between carrier particles and drug substances. The objective of this study is to clearly assess the effects of the formulation parameters (PEG molecular weight and formulated amount in the co-SDM) on the aerosolization performance. The surface energy of each carrier particle, including co-SDMs and SDMs, was evaluated. From the viewpoint of the surface energy, the relationship between the surface adhesion force and the aerosolization performance in each of the formulations was evaluated.

118


Drug Delivery to the Lungs 27, 2016 - The surface energy effect of co-spray-dried mannitol with polyethylene glycol on the aerosolization performance in a dry powder inhalation formulation Materials and Methods Mannitol (β-mannitol) was purchased from Merck Co., Ltd. PEG 400 and PEG 4000 were purchased from Wako Pure Chemicals Industries, Ltd. PEG 20000 was purchased from NOF Corporation. Salbutamol sulphate (SS),, which was used as the model drug for the in vitro deposition studies, was purchased from Tokyo Chemical Industry Co., Ltd. Hypromellose (HPMC) capsules (size 3) were purchased from Qualicaps Co., Ltd. Two SDMs (SDM-1 and SDM-2) were manufactured by spray drying the mannitol aqueous solution using a spray dryer CL-8i (Ohkawara Kakohki Co., Ltd.) equipped with a rotation disk. Co-SDMs were manufactured by spray drying the mannitol aqueous solution with several amounts of each PEG (2.5, 5 and 10% to mannitol weight) using a spray dryer GS-31 (Yamato Scientific Co., Ltd.) equipped with a two-fluid nozzle. The particle size of the mannitol carrier particles was measured with a laser diffraction particle size analyzer HELOS/RODOS system (Sympatec GmbH). In this research, the volume mean diameter (VMD) was used as the representative particle size for each sample. The shapes of carrier particles were observed by using a scanning electron microscope VE-7800 (Keyence Corporation). The model formulations for deposition studies were prepared as follows. SS was milled by a Turbo Counter Jet Mill TJ 60 (Turbo Kogyo Co., Ltd.) for reduction of the particle size. The VMD of micronized SS was confirmed as 2.23 μm. SDMs and Co-SDMs were also sieved by 100 mesh screen to reduce the particle size difference of each carrier particle. Either sieved SDM or co-SDM and milled SS were blended at a constant ratio of mannitol to SS, 67.5:1 (w/w) by a Turbula mixer T2F (Willy A. Bachofen AG Maschinenfabrik). After blending, each formulation was manually filled into HPMC capsules with 34.25 mg ± 1.50 mg of the formulation powder. The deposition profiles of model formulations aerosolized with a Handihaler (Boehringer Ingelheim GmbH) were assessed by an Andersen Cascade Impactor (Copley Scientific Ltd.). Ten capsules were continuously tested under a 28.3 L/min air flow condition. Fine particle dose (FPD) was defined as the sum of either SS or mannitol mass collected below stage 2, with a 50% cut-off diameter of aerodynamic particle size of 4.7 μm. The emitted dose (ED) was defined as the total mass of drug or mannitol emitted from the capsules. Fine particle fraction (FPF) was calculated as the percentage of FPD compared to ED. Surface energy analyses were conducted using an inverse gas chromatography system (Surface Measurement Systems Ltd.) equipped with a flame ionization detector. The experiment was carried out at 303 K and 0% RH. The gas flow rate was 10 mL/min and helium gas was used as the carrier gas. Methane was used for the inert reference; n-decane, n-nonane, n-octane and n-heptane were used to determine the alkane line, and chloroform d and ethyl acetate were employed as polar probes. The dispersive component of the surface energy (γ ) was [5, 7, 8] . The acidic (electron acceptor) calculated based on the retention volume of a series of injected n-alkanes + energy (γ ) and the basic (electron donor) energy (γ ) were calculated based on the retention volumes of ethyl [5, 8] sp . The specific component of surface energy (γ ) was calculated using the acetate and chloroform, respectively + - [5, 8] . acidic energy (γ ) and the basic energy (γ ) Results and Discussion The SDM particle sizes were 24.4 μm (SDM-1) and 47.9 μm (SDM-2). Particle sizes of all co-SDMs were ranged between 27.9 μm and 47.3 μm. As shown in Figure 1, the FPFs of SS when SDM-1 and SDM-2 were used were 13.5% and 14.1%, respectively. From the statistical evaluation, there was no significant difference in FPFs of SS between the model formulations prepared with both SDMs (p > 0.05), so that the particle size of carrier particles had little effect on the aerosolization performance of drug substance within the range of 24.4 μm to 47.9 μm in this system. The particle shape of SDMs showed smooth, spherical forms, whereas the appearances of all of the coSDMs were irregular forms with corrugated surfaces. From a morphological approach, no significant difference was found by changing the PEG molecular weight and formulated amount. It was found that the FPF of SS in the model formulation depended on the MW of PEG when co-SDMs were used as carrier particles (Figure 1). The FPF of SS when co-SDM with PEG 20000 was used was higher compared to that of the two cases of SDM, regardless of the PEG ratio. In contrast, the FPF of SS decreased along with the formulated amount of PEG when co-SDM with PEG 400 or with PEG 4000 was used. When co-SDM with 2.5% PEG 400 or co-SDM with 2.5% PEG 4000 was used as a carrier particle, the FPF of SS was higher compared to that when SDM was used. However, the FPF of SS displayed an extreme decrease in correlation with the increasing PEG ratio. It was found that the PEG property was important for the aerosolization performance of the model formulation. The aerosolization performance of each carrier particle itself was also measured by ACI. On the induction port and preseparator, 85% to 95% of the loaded weights of mannitol carrier particles were detected. The FPF of mannitol was less than 1%. The deposition profiles of each carrier particle were almost exactly the same. This result showed that most of each mannitol carrier particle was captured before the preseparator. Hence, SS was indicated to separate from mannitol carrier particles when aerosolized and the SS separated from the mannitol carrier particles existed below the lower stage.

119


Drug Delivery to the Lungs 27, 2016 – Makoto Kamada et al.

FPF (%, <4.7μm)

35 It was considered that the difference of 30 the surface adhesion force between 25 carrier particle and drug substance was 20 related to the variety of the FPF of SS 15 when co-SDM was used as a carrier 10 particle. Therefore, the surface energy of 5 mannitol carrier particles was measured as an influencing property affecting the 0 adhesion by IGC. The dispersive d component (γ ) and the specific sp component (γ ) of surface energy were assessed among mannitol carrier particles. It was reported that a linear inverse relationship between the dispersive component of the surface Figure 1 - FPF of salbutamol sulfate in carrier-based energy of carrier particles and the FPFs formulations using mannitol carrier particles. of the drug substance was found by using (Mean ± SD, n = 3) [5] d the IGC technique . The γ of co-SDMs d had a lower energy than that of SDM (Figure 2A). It was considered that the lower γ of a co-SDM with PEG d 20000 may contribute to the higher FPF of SS (Figure 2A). But only this result of γ cannot explain well enough d the decreasing FPF of SS when co-SDM with PEG 400 or with PEG 4000 was used. Although the γ of co-SDM with PEG 400 or PEG 4000 had a lower energy compared to that of SDM, the FPF of SS when the PEG ratio was 5% or 10% was lower than that when SDM was used as a carrier particle (Figure 2A). Figure 2B shows the sp sp relationship between the FPF of SS and the γ of each carrier particle. The γ of co-SDM with PEG 20000 was sp lower than that of SDM. In contrast, the γ of co-SDM with PEG 400 and with 10% PEG 4000 was higher than sp that of SDM. Moreover, the γ of co-SDM with PEG increased dependent on the increasing PEG ratio. The FPF sp of SS tended to decrease along with the increase of the γ of co-SDMs and SDMs. From these results, it was d sp suspected that not only the γ of each carrier particle, but also its γ affected to the aerosolization performance of the DPI formulation. 35

FPF (%, <4.7μm)

30

2.5%

25

(B)

2.5%

2.5%

5%

10%

20 15

5%

10 10%

5 0

0

5% 20

35

10%

2.5%

2.5%

30 FPF (%, <4.7μm)

(A)

2.5%

25

5%

10%

20 15

5%

10 10%

5

40 γd (mJ/m2)

0

60

0

20

40 60 γsp (mJ/m2)

5% 80

SDM without PEG

SDM with PEG400

SDM without PEG

SDM with PEG400

SDM with PEG4000

SDM with PEG20000

SDM with PEG4000

SDM with PEG20000

10% 100

d

Figure 2 – The relationship between FPF of SS and dispersive component, γ (A) and specific sp component, γ (B) of surface energy of mannitol carrier particles. Data labels show PEG ratio. +

-

The acidic (electron acceptor) energy (γ ) and the basic (electron donor) energy (γ ) were compared among + mannitol carrier particles in order to evaluate the differences of specific energy in detail (Figure 3). The γ was + higher than the γ in each co-SDM with PEG 20000. In contrast, the γ was higher than the γ in each co-SDM with PEG 400. Additionally, the γ of co-SDMs increased with increasing PEG ratio and the ratio of the γ compared to + - + the γ (γ /γ ) also increased with increasing the PEG ratio in co-SDMs with PEG 400. As shown in Figure 4, the FPF of SS decreased with the increase of γ in co-SDM for each PEG system. Hence, it was implied that the electron donor function affected the adhesion between the mannitol carrier particle and SS. This elevation of γ with increasing PEG ratio could be attributed to the hydroxyl group on the terminal of the carbon chain of PEG [9] due to the higher hydroxyl value of PEG with lower MW . There seemed to be two possible mechanisms that could explain the increased adhesion between mannitol carrier particles and SS when the γ increases. SS has a secondary amino group that behaves as a cation and several hydroxyl groups. These functional groups can work as electron acceptors. Consequently, it was considered that the surface regions of the carrier particles with high electron donor energy attached to the functional groups of SS that work as electron acceptors. Adhesion of water is also considered to be an important [9] issue because PEG has high viscosity and high hygroscopicity . Increasing viscosity on the carrier particle surface may occur due to the hygroscopicity of PEG. From the water absorption profile, hygroscopicity was found [10] to be increased due to the ratio of PEG in the co-SDM . The elevation of the basic (electron donor) energy with the increasing PEG ratio may also imply an increasing hygroscopicity. It is anticipated that the altering of the surface adhesion force of carrier particles induced based on these expected mechanisms will result in a variety of

120


Drug Delivery to the Lungs 27, 2016 - The surface energy effect of co-spray-dried mannitol with polyethylene glycol on the aerosolization performance in a dry powder inhalation formulation FPFs for the drug substance. The details of this mechanism that alters the aerosolization performance of SS will be clarified in our future work. 35 30

50

FPF (%, <4.7μm)

Surface Energy (mJ/m2)

60 40 30 20 10 0

2.5%

2.5% 2.5% 5%

25

10%

20 15

5%

10 10%

5 0

γ+

γ+

Figure 3 - Acidic (electron acceptor) energy, γ and basic (electron donor) energy, γ of mannitol carrier particles.

5%

0

20

40 γ- (mJ/m2)

10%

60

SDM without PEG

SDM with PEG400

SDM with PEG4000

SDM with PEG20000

Figure 4 - The relationship between FPF of SS and basic (electron donor) energy, γ of mannitol carrier particles. Data labels show PEG ratio.

Conclusion The FPF of SS in the formulation containing co-SDM with PEG 20000 was superior to that in the formulation containing SDM. It was considered that the lower dispersive component of the surface energy of co-SDM may contribute to the higher FPF of SS when using PEG with a high MW. On the other hand, the FPFs of SS were found to be influenced by the formulation of the co-SDM, such as the ratio and molecular weight. It was considered that the elevation of the basic (electron donor) energy resulted in the decrease of FPF of SS when using PEG with low MWs. The elevation of the basic (electron donor) energy of a co-SDM would relate to both/either the difference in the adhesion force to the SS and/or the difference in the hygroscopicity of each PEG. References 1

2 3

4 5 6

7

8

9 10

Ito A, Tanimoto M, Yano H, Kumon M, Inoue K, Yada S, Wakiyama N: The effect of co-spray drying with polyethylene glycol on the polymorphic state of mannitol carrier particles for dry powder inhalation. (Abstract). Presented at: Drug Delivery to the Lungs 24. Scotland, UK, December 11–13, 2013; J Aerosol Med Pulm Drug Deliv. 27:4; A1–A27. Shariare M H, York M de M P: Effect of crystallisation conditions and feedstock morphology on the aerosolization performance of micronised salbutamol sulphate, Int J Pharm 2011; 415: pp62-72. Tanimoto M, Ito A, Kumon M, Inoue K, Yada S, Minami H: Co-spray-dried mannitol with polyethylene glycol (PEG) for dry powder inhalation -The effect of PEG on the aerosolization performance and the physical stability. (Abstract). Presented at: Drug Delivery to the Lungs 26. Scotland, UK, December 9–11, 2015; J Aerosol Med Pulm Drug Deliv. 29:3; A1–A25. Timsina M P, Martin G P, Marriott C, Ganderton D, Yianneskis M: Drug delivery to the respiratory tract using dry powder inhalers, Int J Pharm 1994; 101: pp1-13. Traini D, Young P M: The influence of lactose pseudopolymorphic form on salbutamol sulfate-lactose interactions in DPI formulations, Drug Dev Ind Pharm 2008; 34: pp992-1001. Kumon M, Machida S, Suzuki M, Kusai A, Yonemochi E, Terada K: Application and mechanism of inhalation profile improvement of DPI formulations by mechanofusion with magnesium stearate, Chem Pharm Bull 2008; 56: pp617-625. Hickey A J, Mansour H M, Telko M J, Xu Z, Smyth H D C, Mulder T, Mclean R, Langridge J, Papadopoulos D: Physical characterization of component particles included in dry powder inhalers. I. Strategy review and static characteristics, J Pharm Sci 2007; 96: pp1282-1301. Hasegawa S, Furuyama N, Yada S, Hamaura T, Kusai A, Yonemochi E, Terada K: Effect of physical properties of troglitazone crystal on the molecular interaction with PVP during heating, Int J Pharm 2007; 336: pp82-89. Rowe R C, Sheskey P J, Quinn M E (eds): Handbook of pharmaceutical excipients 6th edition. Pharmaceutical Press, London; pp517-522, 2009. Kumon M, Tanimoto M, Ito A, Yano H, Inoue K, Yada S, Minami H: Co-spray-dried mannitol with polyethylene glycol for dry powder inhalation: Factors affecting the polymorphism and stability. (Abstract). Presented at: Drug Delivery to the Lungs 25. Scotland, UK, December 10–12, 2014; J Aerosol Med Pulm Drug Deliv. 28:4; A1–A25.

121


Drug Delivery to the Lungs 27, 2016 – S Zellnitz et al. Impact of particle engineering on the processability and aerosolization performance of DPI formulations 1

1

1

1,2

S Zellnitz , E Faulhammer , T Wutscher , J G Khinast , A Paudel 1

1

Research Center Pharmaceutical Engineering GmbH, Inffeldgasse 13, Graz, 8010, Austria 2 Institute for Process and Particle Engineering, Graz University of Technology, Inffeldgasse 13, Graz, 8010, Austria

Summary Active pharmaceutical ingredient (API) particles administered via dry powder inhalers (DPIs) must exhibit an aerodynamic diameter of 1µm - 5µm. Particles of such a small size are rather cohesive and show poor flow properties. This is challenging for the dose uniformity, as a good flowing powder is required to guarantee uniform doses for example during capsule filling or dose metering within a reservoir inhaler. To overcome this problem, carrier based formulations where the small API particles are attached to larger carrier particles with adequate flowability have been invented. A crucial step during inhalation is the detachment of the API from the carrier. Only detached API particles are able to reach the lung. The detachment process is governed by API as well as carrier properties. Therefore, within this study the effect of particle engineering on capsule filling performance represented by capsule fill weight and weight variability and on aerosolization performance represented by the fine particle fraction (FPF) and the emitted dose (ED) were investigated. Inhalation grade lactose was used as carrier material as received and after wet surface processing. Spray dried and jet milled salbutamol sulphate served as model API. Results showed that compared to API engineering carrier engineering had a positive effect on weight variability. Moreover, the use of spray dried API particles overall decreased the FPF. Engineered carrier particles showed improved FPF but only in combination with jet milled API. The highest FPF (≈33%) was obtained for jet milled API in combination with engineered lactose carrier particles. Introduction Formulation development for dry powder inhalers (DPIs) is challenging as in order to reach the deep lung active pharmaceutical ingredient (API) particles must exhibit an aerodynamic diameter of 1µm - 5µm. Particles of such a small size are typically rather cohesive and show poor flow properties. This however is challenging during the formulation and dosing process where a good flowing powder is required to guarantee uniform doses for example during capsule filling or dose metering within a reservoir inhaler. To overcome this challenge carrier based formulations where the small API particles are attached to a larger carrier particle (usually 50µm – 200µm) with adequate flowability have been developed. A crucial step during inhalation is the detachment of the API from the carrier. Only detached single API particles are able to reach the lung where they are supposed to cause a therapeutic effect. This detachment process is impacted by API as well as carrier properties and many studies have focused on either carrier or API engineering in order to tailor drug detachment from the carrier [1]. There is a complex interplay between particle engineering, particle properties, their processability and aerodynamic performance. Most studies focus on the effect of particle engineering on the aerodynamic performance but also the effect on processability must not be neglected. Therefore, the present study evaluates the effect of particle engineering on the processability during capsule filling as well as on the aerodynamic performance. This study presents a follow up study to work that was shown at last year’s DDL, where also the influence of carrier surface processing on capsule fill performance and aerosolization performance was analyzed [2]. This study should complement the data already presented by investigating a different carrier engineering technique. The obtained data shall help in a more comprehensive understanding of the interplay between carrier morphology and carrier type, drug detachment and capsule filling efficiency. The final goal would be to collect these data for various different carrier materials to generate a platform technology encompassing a detailed understanding on how different engineering techniques affect particle characteristics and further capsule fill performance and aerodynamic performance. Ideally, this platform should allow the preselection of matching API carrier pairs and process settings based on particle characteristics. Materials and methods API engineering Salbutamol sulphate (SS) (Selectchemie, Zurich, Switzerland) was chosen as a model API in the present study. To generate inhalable sized particles salbutamol sulphate was micronized using spray drying (Nano Spray Dryer B-90 (Buechi Labortechnik AG, Flawil, Switzerland) and jet milling (Spiral Jet Mill 50 AS, Hosokawa Alpine AG, Augsburg, Germany). Spray drying conditions were chosen according to our previous work [3]. Air jet milling was done at an injection pressure of 6 bar and a pressure inside the micronizer chamber of 3 bar. Carrier engineering Besides engineered API particles, also engineered carrier particles were used. Carrier engineering was performed through wet surface processing of α-lactose monohydrate, with the intention to on the one hand remove fine particles adhering to the bulk carrier material and on the other hand to modify the carrier surface topography [4].

122


Drug Delivery to the Lungs 27, 2016 – Impact of particle engineering on the processability and aerosolization performance of DPI formulations Preparation of adhesive mixtures For evaluating the performance of engineered materials, blends of spray dried (SDSS) and jet milled (JMSS) salbutamol with lactose raw material (LAC R) and engineered lactose (LAC E) were prepared. In total four adhesive mixtures with 2% API content were manufactured in a tumble blender TC2 (Willy A. Bachofen Maschinenfabrik, Muttenz, Switzerland). Mixing parameters were chosen as follows, 60min at 60rpm. Powder characterization Both, raw and engineered carrier as well as the differently prepared APIs were extensively characterized. Particle size was measured via laser diffraction (HELOS, Sympatec, Clausthal-Zellerfeld, Germany). Particle morphology was measured via scanning electron microscopy ((SEM), Zeiss Ultra 55, Zeiss, Oberkochen, Germany). Solid state was analyzed via differential scanning calorimetry ((DSC), 204F1 Phoenix®, Netzsch GmbH, Selb, Germany) and small- and wide angle x – ray (SWAXS) scattering (S3 – MICRO camera (formerly Hecus X ray systems, Graz, now Bruker AXS, Karlsruhe, Germany)). Particle characteristics were used to evaluate the effect of processing by comparing the different particulate properties. For all adhesive mixtures SEM images (SEM, Zeiss Ultra 55, Zeiss, Oberkochen, Germany) were taken to visualize the distribution of API on the carrier surface. Moreover, adhesive mixtures were characterized in terms of flowability. Therefore, bulk density (BD) and tapped density (TD) were determined with a PT-TD200 (Pharma Test Apparatebau AG, Hainburg, Germany) according to a standardized method described in the United States Pharmacopeia (USP 2011). From these results the Carr´s Index or Compressibility Index (CI) was calculated upon the volumes of BD and TD and Houser Ratio (HR) calculated. Capsule filing Capsule filling was performed with two different process setting on a dosator nozzle capsule filling machine (Labby, MG2, Bologna, Italy) with a target fill weight of 20mg to 25mg at a filling rate of 2500 capsules per hour (cph). Setting 1: 3.4mm dosator, 2.5mm dosing chamber, 5mm powder layer Setting 2: 3.4mm dosator, 2.5mm dosing chamber, 10mm powder layer Blend uniformity during filling was checked by sampling filled capsules during filling. Therefore 25 filled capsules were collected after 5 and 10 minutes from the beginning of capsule filling. The API content in each capsule was evaluated by dissolved the capsule content in 20ml of buffer (diluted acetic acid solution (pH=3)) and quantifying the salbutamol content using a validated HPLC method. The homogeneity of the mixtures is expressed by the relative standard deviation (RSD) of the mean sample drug content. Aerodynamic assessment of particles To assess the impact of API engineering as well as carrier engineering on the in vitro performance, lung deposition experiments were carried out with a next generation impactor (NGI, Copley Scientific, Nottingham, United Kingdom). For each NGI experiment, three capsules were discharged via the Aerolizer®/Cyclohaler® and the salbutamol content in each part of the impactor quantified using a validated HPLC method. For all experiments the emitted dose (ED) and the fine particle fraction (FPF) were calculated based on the specification of the European pharmacopoeia. Results and discussion Powder characteristics API processing significantly affects the final product characteristics. Spray drying resulted in spherical amorphous particles whereas by jet milling crystalline needle shaped particles were generated. Further, the engineered API particles differ in terms of size. (Characteristic diameters of spray dried particles x10=0.45µm, x50=3.07µm and x90=6.73µm, characteristic diameters of jet milled particles: x10=0.52µm, x50=1.99µm and x90=5.03µm). Carrier engineering via wet surface processing resulted in smoother particles with less fine lactose attached compared to the raw material. This was visualized via SEM images and reflected via particle size measurements by a slightly reduced particle size for the engineered carrier material (Characteristic diameters of the engineered carriers particles: x10=71.4µm, x50=179.04µm and x90=313.93µm; characteristic diameters of the raw material: x10=95.27µm, x50=216.03µm and x90=354.7µm). Moreover, SEM images revealed that wet surface processing introduced shallow cavities on the surface. Solid state analysis revealed that during the wet surface processing the surface was solvated and during drying recrystallized in form of small anhydrous lactose particles [4, 5]. As a measure for the flowability of the different adhesive mixtures, the CI and the HR were evaluated. Results in Figure 1 show that carrier engineering lowered the CI and the HR indicating a better flowability of the mixture. This is true for mixtures of engineered carrier and both spray dried and jet milled API. The use of engineered API in the blend preparation seemed not to affect the flowability of the mixtures what is most likely due to the small amount of API present in the mixtures.

123


Powder characteristics API processing significantly affects the final product characteristics. Spray drying resulted in spherical amorphous Drug Delivery to the Lungs 27, 2016 – S Zellnitz et al.

particles whereas by jet milling crystalline needle shaped particles were generated. Further, the engineered API particles differ in terms of size. (Characteristic diameters of spray dried particles x10=0.45µm, x50=3.07µm and x90=6.73µm, characteristic diameters of jet milled particles: x10=0.52µm, x50=1.99µm and x90=5.03µm).

Carrier engineering via wet surface processing resulted in smoother particles with less fine lactose attached compared to the raw material. This was visualized via SEM images and reflected via particle size measurements by a slightly reduced particle size for the engineered carrier material (Characteristic diameters of the engineered carriers particles: x10=71.4µm, x50=179.04µm and x90=313.93µm; characteristic diameters of the raw material: x10=95.27µm, x50=216.03µm and x90=354.7µm). Moreover, SEM images revealed that wet surface processing introduced shallow cavities on the surface. Solid state analysis revealed that during the wet surface processing the surface was solvated and during drying recrystallized in form of small anhydrous lactose particles [4, 5]. As a measure for the flowability of the different adhesive mixtures, the CI and the HR were evaluated. Results in Figure 1 show that carrier engineering lowered the CI and the HR indicating a better flowability of the mixture. This is true for mixtures of engineered carrier and both spray dried and jet milled API. The use of engineered API Drug to the Lungs 27,not 2016 – S Zellnitz et al. of the mixtures what is most likely due to the small in the Delivery blend preparation seemed to affect the flowability amount of API present in the mixtures. Figure 1. CI and HR for all adhesive mixtures, calculated from bulk and tapped density measurements (n=3)

Processability – Capsule filling performance Blend uniformity during filling was evaluated after 5 and 10 minutes from the beginning of capsule filling. A decrease in blend uniformity values represents a more homogenous distribution of API within the mixture. Overall results in table 1 show that the blend uniformity is good for all formulations independent of what carrier and API combination is used and what process settings are chosen. Comparing the two different process settings one can see that for settings 2 the blend uniformity improves with increasing capsule filling time. Whereas for process settings 1 this effect is less pronounced. Having a look on the impact of particle engineering, values of mixture homogeneity show a different behavior for jet milled and spray dried blends. The use of engineered carrier shows improved blend uniformity but only when combined with spray dried API. This is true for both process settings and sampling time steps. For the micronized blends the use of engineered lactose seems to reduce the blend uniformity except for a pbh of 10mm after 10 minutes where the blend uniformity improves. These results show that wet surface processing of the lactose carrier leads to more uniform mixtures when formulated with spray Figure dried API.1. CI and HR for all adhesive mixtures, calculated from bulk and tapped density measurements (n=3) Table 1. Mixture homogeneity process at different powder bed heights. Samples were taken from 5 to 10 Processability – Capsuleduring filling filling performance minutes and from 10 to 15 minutes (n=25, RSD from mean drug content)

Blend uniformity during filling was evaluated after 5 and 10 minutes from the beginning of capsule filling. A Blend uniformity - Setting 1 [%] Blend uniformity - Setting 2 [%] decrease in blend uniformity values represents a more homogenous distribution of API within the mixture. Overall Min 5-10 Min 10-15 Min 5-10 10-15 results in table 1 show that the blend uniformity is good for all formulations independent Min of what carrier and API combination is used and what process settings are chosen. Comparing the two different process JMSS+LAC_R 1.44 0.98 2.96 1.35 settings one can see that for settings 2 the blend uniformity improves with increasing capsule filling time. Whereas for process JMSS+LAC_E 0.93values of mixture settings 1 this effect is less 1.75 pronounced. Having 1.62 a look on the impact3.80 of particle engineering, homogeneity show a different2.11 behavior for jet milled and spray dried blends. SDSS+LAC_R 1.33 3.90 The use of engineered 1.81 carrier shows improved blend uniformity but only when combined with spray dried API. This is true for both process settings and SDSS+LAC_E 1.07 0.66 3.32 1.48 sampling time steps. For the micronized blends the use of engineered lactose seems to reduce the blend uniformity except for a pbh of 10mm after 10 minutes where the blend uniformity improves. These results show that surfaceofprocessing of the lactose carrier leads to more formulated spray For thewet evaluation capsule filling performance two parameters wereuniform chosen,mixtures the targetwhen fill weight and thewith weight dried API. variability expressed by the relative standard deviation (RSD). Results in table 2 show that the intended fill weight of 25mg per capsule was achieved for all blends within the given limits of accuracy of a RSD below 5 %. Table 1. Mixture homogeneity during heights. Samples taken from 5 to 10 However, when having a closer look filling on theprocess values,atitdifferent can be powder clearly bed seen that the weightwere variability is reduced minutes from 10 tolactose 15 minutes (n=25,(lower RSD from drugThis content) when usingand engineered particles RSDmean values). effect is true for both process settings but more pronounced for settings 1. Moreover, when applying process settings 1 the use of engineered carriers led to Blend uniformity - Setting 1 [%] Blend uniformity - Setting 2 [%] higher fill weights compared to raw lactose carriers.

Min 5-10 1.44 Setting 1 [mg] 1.75

Table 2. Fill weight and fill weight variability (RSD)

JMSS+LAC_R JMSS+LAC_E SDSS+LAC_R JMSS+LAC_R SDSS+LAC_E JMSS+LAC_E

2.11 19.74 1.07 24.07

Min 10-15 0.98 1.62RSD [%] 1.33 6.57 0.66 2.37

Min 5-10 2.96 Setting 2 [mg] 3.80 3.90 27.22 3.32 25.08

Min 10-15 1.35 0.93RSD [%] 1.81 2.00 1.48 1.97

20.85 4.19 were chosen,27.19 2.33 For theSDSS+LAC_R evaluation of capsule filling performance two parameters the target fill weight and the weight variability expressed by the relative standard deviation (RSD). Results in table 2 show that the intended fill weight SDSS+LAC_E 24.74 2.21 27.33 2.00 of 25mg per capsule was achieved for all blends within the given limits of accuracy of a RSD below 5 %. However, when having a closer look on the values, it can be clearly seen that the weight variability is reduced when using engineered lactose particles (lower RSD values). This effect is true for both process settings but more pronounced for settings 1. Moreover, wheneffect applying process settingswhereas 1 the use engineered carriersnotled Concluding, carrier engineering had a positive on weight variability API of engineering seemed to to higher weightsfilling compared to raw lactose carriers. affect thefill capsule performance. This could be explained by the amount of API that is relatively small in the mixtures. The positive effect of engineered lactose on the capsule filling performance can be related to the Table 2. Fill weight and fill weight variability (RSD) increased flowability of adhesive mixtures containing engineered lactose particles (Figure 1). Setting 1 [mg]

124 RSD [%]

Setting 2 [mg]

RSD [%]

JMSS+LAC_R

19.74

6.57

27.22

2.00

JMSS+LAC_E

24.07

2.37

25.08

1.97


For the evaluation of capsule filling performance two parameters were chosen, the target fill weight and the weight variability expressed by the relative standard deviation (RSD). Results in table 2 show that the intended fill weight of 25mg per capsule was achieved for all blends within the given limits of accuracy of a RSD below 5 %. Drug Delivery the Lungs 27,look 2016on– the Impact of particle thetheprocessability andis reduced However, whento having a closer values, it can beengineering clearly seenon that weight variability when using engineered lactoseof particles (lower RSD values). This effect is true for both process settings but more aerosolization performance DPI formulations pronounced for settings 1. Moreover, when applying process settings 1 the use of engineered carriers led to higher fill weights compared to performance raw lactose carriers. Assessment of aerodynamic Table 2. Fill weight and fill weight variability For the evaluation of the performance of (RSD) the different mixtures, the fine particle fraction (FPF) and the emitted

dose (ED) were chosen. Results in table 3 overall show that mixtures comprising of spray dried API and either 1 [mg] RSD [%]to adhesiveSetting 2 [mg] [%]API raw or engineered lactose show Setting much lower FPFs compared mixtures containing jetRSD milled particles. Thus, API engineering has a larger effect on the aerodynamic performance than carrier engineering. 6.57 using jet milled API 27.22 2.00 The use JMSS+LAC_R of engineered lactose particles19.74 increase the FPF when in combination with process setting 1JMSS+LAC_E and had no effect at process settings 2. By contrast, when spray dried API is use the FPF decreases 24.07 2.37 25.08 1.97 for the engineered carrier material. The highest FPF was obtained for jet milled API in combination with engineered lactose particles at process settings 20.85 1, indicating this as4.19 most favorable API-carrier 27.19 combination within 2.33 the SDSS+LAC_R presents study. For both process settings the same trends for FPF can be observed, so the different process 2.21 27.33 2.00 SDSS+LAC_E settings do not affect the aerosolization24.74 performance.

The ED is within the same range for all adhesive mixtures. Although there is a slight tendency that the use of spray dried API increases the ED, especially at effect process 1. Concluding, carrier engineering had a positive onsettings weight variability whereas API engineering seemed not to affect the capsule filling performance. This could be explained by the amount of API that is relatively small in the Table 3. In The Vitro positive performance: fraction (FPF) and Emitted dose filling (ED) performance can be related to the mixtures. effectFine of particle engineered lactose on the capsule increased flowability of adhesive mixtures containing engineered lactose particles (Figure 1). Samples 2 Drug Delivery to the Lungs 27, 2016 –Setting Impact1 of particle engineering on the Setting processability and FPFof[%] ED [µg] aerosolization performance DPI formulations

FPF [%]

ED [µg]

829.88 1131.95 JMSS+LAC_R 19.20 22.41 Assessment of aerodynamic performance 939.91 1077.56 JMSS+LAC_E 32.53 22.63 For the evaluation of the performance of the different mixtures, the fine particle fraction (FPF) and the emitted SDSS+LAC_R 1035.02 6.49 5.07 dose (ED) were chosen. Results in table 3 overall show that mixtures comprising of spray1114.91 dried API and either raw orSDSS+LAC_E engineered lactose show much lower FPFs compared to adhesive mixtures containing jet milled API 1261.05 1162.72 2.39 3.35 particles. Thus, API engineering has a larger effect on the aerodynamic performance than carrier engineering. The use of engineered lactose particles increase the FPF when using jet milled API in combination with process setting 1 and had no effect at process settings 2. By contrast, when spray dried API is use the FPF decreases for Concluding, results showed that drug detachment and aerodynamic performance is governed by the API particles the engineered carrier material. The highest FPF was obtained for jet milled API in combination with engineered used. Although carrier engineering positively affected capsule filling no benefit on the aerosolization performance lactose particles at process settings 1, indicating this as most favorable API-carrier combination within the could be observed. This could be explained by particle characteristics that were more distinct for the differently presents study. For both process settings the same trends for FPF can be observed, so the different process prepared API particles than for raw and engineered carrier particles. API engineering resulted in particles with settings do not affect the aerosolization performance. different size shape and solid-state whereas carrier engineering manly changed surface topography. The ED is within the same range for all adhesive mixtures. Although there is a slight tendency that the use of Conclusions spray dried API increases the ED, especially at process settings 1. The obtained data are highly useful, to improve the understanding of the relationship between carrier morphology on capsule filling efficiency and aerosolization performance. Moreover, these data help to approach the platform technology with the final goal to preselect matching API carrier pairs and process settings Samples Setting 1 Setting 2 based on particle characteristics. FPF [%] ED [µg] FPF [%] ED [µg]

Table 3. In Vitro performance: Fine particle fraction (FPF) and Emitted dose (ED)

References JMSS+LAC_R

19.20

829.88

22.41

1131.95

JMSS+LAC_E

32.53

939.91

22.63

1077.56

1

Chow A H L, Tong H H Y, Chattopadhyay P, Shekunov P Y: Particle Engineering Delivery; Pharm. Res. SDSS+LAC_R 1035.02 1114.91 6.49 5.07 for Pulmonary Drug 2007; 24 (3): pp 411-437.

SDSS+LAC_E

2.39

1261.05

3.35

1162.72

2

Zellnitz S, Faulhammer E, Wahl V, Khinast J G, Paudel A: Carrier-based Dry Powder Inhalation: Impact of Carrier Modification on Capsule Filling Processability and in vitro Aerodynamic Performance – Part I (Abstract). Presented at: Drug Concluding, showed drug detachment andScotland, aerodynamic performance governed API particles Delivery toresults the Lung (DDL) that 27 Conference, Edinburgh, December 9-11, 2015;isDDL27 bookby of the abstracts pp 187 used. Although carrier engineering positively affected capsule filling no benefit on the aerosolization performance 190.

could be observed. This could be explained by particle characteristics that were more distinct for the differently 3 prepared APIE particles for raw and engineered carrier particles. engineering particlessulfate with Littringer M, Zellnitzthan S, Hammernik K, Adamer V, Friedl H, Urbanetz N API A: Spray drying of resulted aqueous in salbutamol different size shape and solid-state whereas carrier engineering manly changed surface topography. solutions using the Nano Spray Dryer B-90 – The impact of process parameters on particle size. Drying Technol. 2013; 31: pp 1346–1353.

4 Conclusions Wahl V: Modification of dry powder inhaler carrier surfaces by wet decantation; Master thesis, Graz University of Technology,

Graz, Austria. 2014.

5 The obtained dataS,are highlyG: useful, to improve the understanding ofbehavior the relationship morphology Garnier S, Petit Coquerel Dehydratio mechanism and crystallization of lactose.between J. Therm.carrier Anal. Calorim. 2002; on capsule filling efficiency and aerosolization performance. Moreover, these data help to approach the platform 68: pp 486-502. technology with the final goal to preselect matching API carrier pairs and process settings based on particle characteristics.

References 1

Chow A H L, Tong H H Y, Chattopadhyay P, Shekunov P Y: Particle Engineering for Pulmonary Drug Delivery; Pharm. Res. 2007; 24 (3): pp 411-437.

2

Zellnitz S, Faulhammer E, Wahl V, Khinast J G, Paudel A: Carrier-based Dry Powder Inhalation: Impact of Carrier Modification on Capsule Filling Processability and in vitro Aerodynamic Performance – Part I (Abstract). Presented at: Drug Delivery to the Lung (DDL) 27 Conference, Edinburgh, Scotland, December 9-11, 2015; DDL27 book of abstracts pp 187 190.

3

125


Conclusions The are Lungs highly useful, to improve the understanding of the relationship between carrier morphology Drugobtained Deliverydata to the 27, 2016 – S Zellnitz et al. on capsule filling efficiency and aerosolization performance. Moreover, these data help to approach the platform technology with the final goal to preselect matching API carrier pairs and process settings based on particle characteristics. Impact of particle engineering on the processability and aerosolization performance of DPI formulations References 1

1

1

1

1

1,2

S Zellnitz , E Faulhammer , T Wutscher , J G Khinast , A Paudel

1

Research Center Pharmaceutical Engineering GmbH, Inffeldgasse 13, Graz, 8010, Austria

Chow A H L,2Tong H H Y, Chattopadhyay P, Shekunov P Y: Particle Engineering for Pulmonary Drug Delivery; Pharm. Res. Institute for Process and Particle Engineering, Graz University of Technology, Inffeldgasse 13, 2007; 24 (3): pp 411-437.

Graz, 8010, Austria

2

Zellnitz S, Faulhammer E, Wahl V, Khinast J G, Paudel A: Carrier-based Dry Powder Inhalation: Impact of Carrier Summary

Modification on Capsule Filling Processability and in vitro Aerodynamic Performance – Part I (Abstract). Presented at: Drug Delivery to the Lung (DDL) 27 Conference, Edinburgh, Scotland, December 9-11, 2015; DDL27(DPIs) book ofmust abstracts pp 187 Active pharmaceutical ingredient (API) particles administered via dry powder inhalers exhibit an 190. aerodynamic diameter of 1µm - 5µm. Particles of such a small size are rather cohesive and show poor flow

properties. This is challenging for the dose uniformity, as a good flowing powder is required to guarantee uniform Littringer E M, Zellnitz S, capsule Hammernik K, Adamer Friedl H,within Urbanetz N A: Spray dryingTo of overcome aqueous salbutamol sulfate doses for example during filling or dose V, metering a reservoir inhaler. this problem, solutions using the Nano Spray Dryer – The of process parameters on particle size. Drying Technol. 2013; 31: carrier based formulations where theB-90 small API impact particles are attached to larger carrier particles with adequate pp 1346–1353. flowability have been invented. A crucial step during inhalation is the detachment of the API from the carrier. Only 4 detached particlesofare to inhaler reach carrier the lung. The detachment process is governed by API as well as carrier Wahl V: API Modification dry able powder surfaces by wet decantation; Master thesis, Graz University of Technology, properties. Therefore, Graz, Austria. 2014. within this study the effect of particle engineering on capsule filling performance 5 represented by capsule fill weight and weight variability and on aerosolization performance represented by 2002; the Garnier S, Petit S, Coquerel G: Dehydratio mechanism and crystallization behavior of lactose. J. Therm. Anal. Calorim. fine68: particle fraction (FPF) and the emitted dose (ED) were investigated. Inhalation grade lactose was used as pp 486-502. carrier material as received and after wet surface processing. Spray dried and jet milled salbutamol sulphate served as model API. Results showed that compared to API engineering carrier engineering had a positive effect on weight variability. Moreover, the use of spray dried API particles overall decreased the FPF. Engineered carrier particles showed improved FPF but only in combination with jet milled API. The highest FPF (≈33%) was obtained for jet milled API in combination with engineered lactose carrier particles. 3

Introduction Formulation development for dry powder inhalers (DPIs) is challenging as in order to reach the deep lung active pharmaceutical ingredient (API) particles must exhibit an aerodynamic diameter of 1µm - 5µm. Particles of such a small size are typically rather cohesive and show poor flow properties. This however is challenging during the formulation and dosing process where a good flowing powder is required to guarantee uniform doses for example during capsule filling or dose metering within a reservoir inhaler. To overcome this challenge carrier based formulations where the small API particles are attached to a larger carrier particle (usually 50µm – 200µm) with adequate flowability have been developed. A crucial step during inhalation is the detachment of the API from the carrier. Only detached single API particles are able to reach the lung where they are supposed to cause a therapeutic effect. This detachment process is impacted by API as well as carrier properties and many studies have focused on either carrier or API engineering in order to tailor drug detachment from the carrier [1]. There is a complex interplay between particle engineering, particle properties, their processability and aerodynamic performance. Most studies focus on the effect of particle engineering on the aerodynamic performance but also the effect on processability must not be neglected. Therefore, the present study evaluates the effect of particle engineering on the processability during capsule filling as well as on the aerodynamic performance. This study presents a follow up study to work that was shown at last year’s DDL, where also the influence of carrier surface processing on capsule fill performance and aerosolization performance was analyzed [2]. This study should complement the data already presented by investigating a different carrier engineering technique. The obtained data shall help in a more comprehensive understanding of the interplay between carrier morphology and carrier type, drug detachment and capsule filling efficiency. The final goal would be to collect these data for various different carrier materials to generate a platform technology encompassing a detailed understanding on how different engineering techniques affect particle characteristics and further capsule fill performance and aerodynamic performance. Ideally, this platform should allow the preselection of matching API carrier pairs and process settings based on particle characteristics. Materials and methods API engineering Salbutamol sulphate (SS) (Selectchemie, Zurich, Switzerland) was chosen as a model API in the present study. To generate inhalable sized particles salbutamol sulphate was micronized using spray drying (Nano Spray Dryer B-90 (Buechi Labortechnik AG, Flawil, Switzerland) and jet milling (Spiral Jet Mill 50 AS, Hosokawa Alpine AG, Augsburg, Germany). Spray drying conditions were chosen according to our previous work [3]. Air jet milling was done at an injection pressure of 6 bar and a pressure inside the micronizer chamber of 3 bar. Carrier engineering Besides engineered API particles, also engineered carrier particles were used. Carrier engineering was performed through wet surface processing of α-lactose monohydrate, with the intention to on the one hand remove fine particles adhering to the bulk carrier material and on the other hand to modify the carrier surface topography [4].

126


Drug Delivery to the Lungs 27, 2016 – Michael Chow et al. Optimization of Inhalable Spray Dried Powder Formulations of Small Nucleic Acids with Mannitol and Leucine Michael Chow, Hinson Lin & Jenny Lam Department of Pharmacology & Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, 21 Sassoon Road, Pokfulam, Hong Kong Summary Several studies have demonstrated that the pulmonary delivery of small nucleic acids could be used in the management of various lung diseases and infections. Most formulations used in those studies are liquid aerosol which are not desirable for macromolecules like nucleic acids and proteins that are prone to hydrolysis and enzymatic degradation. A stable solid formulation of nucleic acid for inhalation is in demand. This study focused on formulating small nucleic acids (herring sperm DNA, with size below 50 base pairs) into inhalable dry powders by spray drying using mannitol and L-leucine as the excipients. Formulations of different compositions were prepared and their in vitro aerodynamic performances were evaluated using the Next Generation Impactor. The addition of L-leucine, as a dispersion enhancer, increased the fine particle fraction (FPF) by about 50% compared to those without L-leucine. At 0.75% w/w of DNA, powders with excipients of mannitol: L-leucine at 8:2 or 5:5 achieved a FPF of 62%. When the amount of DNA was increased to 2% w/w, formulation with mannitol: L-leucine equalled 8:2 gave the highest FPF at 58%. Introduction Using small therapeutic nucleic acids for the management of diseases has become a popular direction for future [1] medicine development. They can act on targets that are not accessible by traditional small molecules . Several clinical studies, noticeable the use of small interfering RNA (siRNA) for the treatment of respiratory syncytial virus [2] [3] (RSV) infections , as well as other in vivo studies involving the delivery of siRNA to the lungs , have successfully demonstrated the therapeutic potential of these nucleic acids. On the contrary, the development of inhalable nucleic acids formulations, particularly in dry powder forms, is lacking. Among the few techniques that [4] could be exploited to prepare nucleic acid dry powders , spray drying stands out as it offers versatility that allows formulation scientists to design and engineer particles with properties that fit particular purposes, and its possibility to scale up to industrial manufacture. In the current study, we aimed to design and evaluate spray dried powders of small nucleic acids using mannitol and L-leucine as bulking agent and dispersion enhancer, respectively. It has been shown in previous studies that the addition of amino acids such as leucine or arginine could substantially improve the aerosolization performance [5, 6] . Herring sperm DNA (HSDNA; with size <50 base pairs) was chosen as a model of spray dried powders nucleic acid as its size is comparable to other therapeutic nucleic acids such as siRNA and microRNA (miRNA), and the ease of supply. We anticipated that the result of this study would provide valuable insights regarding the direction and limitations in developing inhalable formulations for small nucleic acids. Experimental Methods and Materials Aqueous solutions of mannitol (Pearlitol 160C; Roquette, Lestrem, France) and L-leucine (J&K Chemical Ltd; Shanghai, China) was prepared. Crude HSDNA (Sigma; Poole, UK) was purified by ethanol precipitation before use. The concentration and the purity of the purified HSDNA solution were determined using an UV spectrometer. HSDNA solution with satisfactory purity would give a ratio of absorbance at 260nm and 280 nm of around 1.8. Solutions containing mannitol, L-leucine and purified HSDNA were prepared according to the formulations as shown in Table 1. These solutions were spray dried with a laboratory scale Büchi B-290 spray dryer (Büchi Labortechnik AG, Postfach, Switzerland) configured as closed loop and operated in suction mode. Other operating parameters were summarized in Table 2. Powders prepared were stored in glass vials inside a desiccator at room temperature until further analysis.

127


Drug Delivery to the Lungs 27, 2016 - Optimization of Inhalable Spray Dried Powder Formulations of Small Nucleic Acids with Mannitol and Leucine Formulation

Mannitol: L-leucine (w/w)

M10L0 M8L2 M5L5 0.75%-M10L0 0.75%-M8L2 0.75%-M5L5 2%-M10L0 2%-M8L2 2%-M5L5

Mannitol only 8:2 5:5 Mannitol only 8:2 5:5 Mannitol only 8:2 5:5

Mannitol 100 80 50 99.25 79.4 49.6 98 78.4 49

Percentage by mass L-leucine 0 20 50 0 19.85 49.6 0 19.6 49

Table 1 – Compositions of different spray dried formulations containing mannitol, L-leucine and HSDNA

Parameters Values

Inlet temperature 80°C

Feed rate 1.4 ml/min

Feed volume 25 ml

Feed concentration 1.5% w/v

Table 2 – Operating parameters of the spray dryer for producing HSDNA dry powders

HSDNA 0 0 0 0.75 0.75 0.75 2 2 2

Atomization

Aspiration

742 L/h

35 m /h

3

The in vitro aerodynamic performance of the powders were evaluated using the Next Generation Impactor (NGI) ® in accordance to the protocol as stated in the British Pharmacopoeia 2016 (Apparatus E). A Breezhaler (Novartis Pharmaceuticals (HK), Hong Kong) was selected as the model inhaler for its capsule-based design and low resistance. The resultant flow rate corresponding to a 4kPa pressure drop was 100 L/min. For each formulation an average of 8mg of powders were loaded into size 3 hydroxypropyl methylcellulose (HPMC) capsules (Capsugel, West Ryde, NSW, Australia). The trays of every impactor stage were coated with a thin layer of silicon grease (Slipicone; DC Products, Waverley, VIC, Australia) prior to dispersion to reduce particle bounce. To quantify the amount of powder deposited on each stage after dispersion, 4 ml of water was used to rinse and dissolve the collected powders. Each measurement used one capsule and the measurements were carried out in triplicate for each formulation. An HPLC system with a refractive index detector installed (Agilent Technologies; California, US) was used to quantify the amount of mannitol in each stage with ultrapure water as the mobile phase. The Agilent Hi-Plex Ca column (7.7 × 300mm, 8µm) was used. Recovered dose was calculated by the sum of mannitol quantified from all stages by the HPLC. Fine particle fraction (FPF) was defined as the fraction of powders with aerodynamic diameter under 5µm, with respect to the recovered dose. Surface morphology and surface atomic analysis of selected formulations were observed by scanning electron microscopy (SEM) (Hitachi S-4800 FEG Scanning Electron Microscope; Hitachi, Tokyo, Japan) and analyzed by x-ray photoelectron spectroscopy, respectively. Results and Discussions Mannitol was used as the bulking agent since it is biocompatible and has shown crystallinity upon spray drying conferring stability. L-leucine has been widely used to improve the dispersibility of particles and their aerodynamic properties. In the current spray drying conditions, the outlet temperature, which referred to the maximum temperature experienced by the particles, was 50°C. The production yield was satisfactory, ranging from 58.5% to 75.9%. The FPF of all formulations was shown in Figure 1. Powders without DNA were included for comparison. Among formulations without L-leucine (with mannitol only, or mannitol with DNA), the FPF remained similar at about 38% when the mass percentage of DNA increased from null to 2%. When L-leucine was added to the formulation, a substantial increase in FPF of about 50% was seen at 0.75% DNA, from 37% to 62% for both 0.75%-M8L2 and 0.75%-M5L5, and the improvement persisted when DNA composition was further increased to 2%. Interestingly, M8L2 demonstrated a poorer FPF than M10L0, suggesting a potential interaction between HSDNA and L-leucine may exist during the course of spray drying, which require further investigation. The optimal amount of L-leucine is also remained to be determined.

128


Drug Delivery to the Lungs 27, 2016 – Michael Chow et al.

F in e P a r t ic le F r a c t io n

P e r c e n ta g e r e la tiv e to r e c o v e r e d d o s e

75

60

45

30

15

0 No D NA M a n n ito l o n ly

0 .7 5 % D N A

2% D NA

M a n n ito l : L -le u c in e

M a n n ito l : L -le u c in e

8 :2

5 :5

Figure 1 – Fine particle fraction expressed as percentage relative to the recovered dose

The surface morphology of powders after the addition of L-leucine was studied using SEM. Figure 2 shows the morphology of formulations M10L0, M8L2 and M5L5. Spray dried mannitol gave a spherical particle with a smooth surface. With L-leucine incorporated into the formulation, the particle changed. Particles of M5L5 formulation showed corrugated surface and irregular shape. These features could reduce inter-particle interactions and promote deagglomeration. It has been suggested that L-leucine, being more hydrophobic and less soluble in water than mannitol, would preferentially stay at the surface of the particle during spray drying, [7] possibly forming an outer shell that encapsulates the particles . Surface atomic analysis was thus carried out using X-ray photoelectric spectrometry. The surface atomic concentration of mannitol, L-leucine and M5L5 was determined, as shown in Table 3. Concentration of nitrogen atoms could be used to estimate the amount of Lleucine on the surface. The measured surface concentration of carbon, oxygen and nitrogen atoms resembled that of the L-leucine, suggesting a surface enrichment of L-leucine.

Figure 2 – SEM images of M10L0, M8L2 and M5L5 at 10K and 20K magnification. Scale bar represents 5µm and 2µm at 10K and 20K magnification, respectively

129


Drug Delivery to the Lungs 27, 2016 - Optimization of Inhalable Spray Dried Powder Formulations of Small Nucleic Acids with Mannitol and Leucine HSDNA was selected as a model of small nucleic acids for the purpose of formulation engineering. It is a mixture of nucleic acids of different sizes under 50 base pairs. As a result, structure integrity of HSDNA after spray drying would be difficult to evaluate. Nonetheless, when nucleic acids with well-defined characteristics such as siRNA or miRNA are formulated, the structural integrity after spray drying should be investigated by means of gel retardation assay or liquid chromatography. Percentage mass concentration of surface atoms Mannitol L-leucine M5L5

Carbon 57.2 64.0 62.0

Measured values Nitrogen Oxygen 0.0 52.8 11.8 24.3 10.3 27.7

Table 3 – Surface analysis of M5L5 using X-ray photoelectron spectroscopy

Theoretical values Carbon Nitrogen Oxygen 42.9 0 57.1 61.0 11.9 27.1 55.6 5.9 38.5

Conclusions Dry powder formulations of small nucleic acids were prepared by spray drying, with mannitol and L-leucine used as bulking agent and dispersion enhancer respectively. The incorporation of L-leucine in the formulation could enhance the aerodynamic performance. These results were promising and served to offer a possibility to bridge investigational therapies into real clinical practice. Further studies involving small nucleic acids with therapeutic potential in suitable animal model will be carried out. References 1. 2. 3. 4. 5. 6.

7.

Daka, A. and D. Peer, RNAi-based nanomedicines for targeted personalized therapy. Adv Drug Deliv Rev, 2012. 64(13): p. 1508-21. DeVincenzo, J., R. Lambkin-Williams, T. Wilkinson, J. Cehelsky, S. Nochur, E. Walsh, R. Meyers, J. Gollob, and A. Vaishnaw, A randomized, double-blind, placebo-controlled study of an RNAi-based therapy directed against respiratory syncytial virus. Proc Natl Acad Sci U S A, 2010. 107(19): p. 8800-5. Rosas-Taraco, A.G., D.M. Higgins, J. Sanchez-Campillo, E.J. Lee, I.M. Orme, and M. GonzalezJuarrero, Intrapulmonary delivery of XCL1-targeting small interfering RNA in mice chronically infected with Mycobacterium tuberculosis. Am J Respir Cell Mol Biol, 2009. 41(2): p. 136-45. Chow, M.Y. and J.K. Lam, Dry Powder Formulation of Plasmid DNA and siRNA for Inhalation. Curr Pharm Des, 2015. 21(27): p. 3854-66. Li, H.Y., P.C. Seville, I.J. Williamson, and J.C. Birchall, The use of amino acids to enhance the aerosolisation of spray-dried powders for pulmonary gene therapy. J Gene Med, 2005. 7(3): p. 343-53. Sou, T., L.M. Kaminskas, T.H. Nguyen, R. Carlberg, M.P. McIntosh, and D.A. Morton, The effect of amino acid excipients on morphology and solid-state properties of multi-component spray-dried formulations for pulmonary delivery of biomacromolecules. Eur J Pharm Biopharm, 2013. 83(2): p. 23443. Vehring, R., Pharmaceutical particle engineering via spray drying. Pharm Res, 2008. 25(5): p. 999-1022.

130


Drug Delivery to the Lungs 27, 2016 - Cláudia Moura et al. Optimization of Supercritical-CO2 Assisted Spray Drying for the Production of Inhalable composite particles 1,2

1

Cláudia Moura , Eunice Costa & Ana Aguiar-Ricardo

2

1

2

R&D, Hovione Farmaciencia SA, Loures, 2674-506, Portugal LAQV-REQUIMTE, Departamento de Química, Universidade NOVA de Lisboa, Caparica, 2829-516, Portugal

Summary The main goal of this work was to assess if the Supercritical CO2-Assisted Spray Drying (SASD) technology is suitable for the production of inhalable composite particles with enhanced properties and improved in-vitro aerodynamic performance while maintaining a high process throughput and yield when comparing to other standard particle engineering technologies. For that purpose, a systematic QbD approach using the design of experiments (DoE) tool, followed by a statistical analysis to predict the powder fine particle fraction (FPF), were implemented. An established formulation composition previously investigated and optimized (trehalose/leucine at 80/20% (w/w) solubilized in a water/ethanol 80/20% w/w mixture) was used to produce composite particles with the SASD apparatus, using a full-factorial design to assess the impact of the static mixer pressure (P_sat), inlet drying gas temperature (T_in) and feed flowrate (F_feed) on the powder properties and on the in-vitro aerodynamic performance by a gravimetric Andersen Cascade Impactor (ACI) using a Plastiape RS01 at 60 L/min, 4 kPa. The powders produced using the SASD apparatus presented yields up to 70% (batch size of 11 grams) while enabling the optimization of the overall throughput, the powder properties and the in-vitro aerodynamic performance. Improved in-vitro aerodynamic performance was driven by the successful manipulation of the powder aerodynamic particle size (aPS), which is mainly dictated by the particle size and density by decreasing the feed flowrate (F_feed) and increasing the inlet drying temperature (T_in), reaching FPFED(<5µm) up to 86%. Introduction The pharmaceutical interest to deliver drugs to the lungs is increasing, either to treat local diseases such as asthma, cystic fibrosis, lung cancer, amongst others, or for systemic drug delivery, such as insulin for diabetes [1][2] . Drug delivery to the lungs presents several advantages such as the rapid onset of drug action and improved bioavailability by avoiding first pass metabolism, which reduces the total drug load required and minimizes the [1] potential for adverse side effects, leading to a more efficient therapy for the patients . The new strategies of carrier-free inhaled drug delivery are possible due to advances in the particle engineering field, in which Spray-drying (SD) is one of the most well-known and enabling technologies. However, due to the environmental deterioration, the exploration and use of alternative greener technologies should be encouraged. For this reason, the production of composite particles using supercritical fluids, namely, Supercritical-CO2 Assisted Spray Drying (SASD) was investigated, using supercritical carbon dioxide (scCO2) as a cosolvent to assist the atomization process, minimizing the use of solvents. scCO 2 was selected due to its mild critical [3] conditions (31.4 ºC, 7.4 MPa): it is inert, safe, non-toxic, recyclable, inexpensive and environmental friendly . The use of scCO2 as a cosolvent minimizes the use of organic solvents, decreases the drying temperature required for the solvents evaporation, being a more eco and energy-friendly approach, does not required additional gas for the atomization step, which reduces the consumption of gases and potentially reduces the shear stress induced during the atomization step that may cause protein denaturation. Furthermore, the possibility of using lower drying temperatures is particularly suitable for processing labile drugs such as biologics (e.g. proteins). Additionally, it is a highly versatile technology, able to work with water or organic solvents, enabling the handling of both hydrophobic and hydrophilic drugs, while generating powders with enhanced aerosolization properties. A schematic representation of the SASD technology is presented in Figure 1 (left). The feed solution composed of the solute and solvent or solvents mixture, and the scCO2 feed are independently pumped and mixed inside a heated static mixer to promote saturation of the mixture with scCO 2, by adding scCO2 in excess. The resulting mixture is atomized inside the drying chamber through a pressure nozzle. The formed droplets are then dried by a hot drying gas stream to evaporate the solvents, forming dried particles. The drying gas used was air. The particles were then collected by a high efficiency cyclone. The main difference between the SASD and SD technology is the atomization step. In the SD technology the atomization step has only one phase: the liquid disperses and forms droplets that are then dried due to the solvent evaporation, forming particles. In the SASD technology, the liquid is dispersed forming the primary droplets; then the scCO2 inside the primary droplets [4] expands and forms smaller secondary droplets that will form particles upon solvent evaporation . The main goal of this work was to optimize the SASD technology process parameters for the production of composite particles with enhanced properties and in-vitro aerodynamic performance using the DoE tool under QbD principles. A formulation composed of trehalose and leucine dissolved in a water/ethanol system was used.

131


Drug Delivery to the Lungs 27, 2016 - Optimization of Supercritical-CO2 Assisted Spray Drying for the Production of Inhalable composite particles Experimental methods Materials: Trehalose dihydrate (Tre) and L-Leucine (Leu) were purchased from Merck KGaA (Darmstadt, Germany). For the production of inhalation powders using the SASD technology, the following process conditions were used: i) Formulation: trehalose/leucine 80/20% (w/w), dissolved in a 80/20% (w/w) water/ethanol mixture. ii) SASD process parameters: the drying gas flowrate (F_drying), the scCO2 flowrate and the scCO2 heating bath and static mixer temperatures were set constant. The DoE performed was a full factorial design (8 points plus 2 central points to assess the reproducibility) and the process parameters investigated are presented in Figure 1 (right) below, being: the static mixer pressure (P_sat), the drying gas inlet temperature (T_in) and the liquid feed flowrate (F_feed). The main goal was to determine the most critical parameters. The powders generated were characterized in terms of their morphology using scanning electron microscopy (SEM), particle size distribution by laser diffraction (Sympatec), bulk density (BD) using a simple graduate beaker method, thermal behavior by modulated differential scanning calorimetry (mDSC), crystallinity by X-ray powder diffraction (XRPD), water content by Karl-Fischer (KF) and ethanol content using a gas chromatography method (GC). The aerodynamic performance was assessed on an 8 stage gravimetric Andersen cascade impactor (ACI8) with 20 mg of formulation filled in HPMC size 3 capsules using a Plastiape HR model 7 at 60 L/min, for a 4 kPa pressure drop. The powder fine particle fraction (FPF) was measured below 5 µm and relative to the capsule emitted dose (FPFED(<5µm)). The statistical analysis used to quantify the impact of the input parameters (P_sat, T_in, F_feed) on the output parameter (FPF) was performed using SIMCA v13.0.3.0 software from Umetrics.

Figure 1 – (left) Schematic representation of the SASD technology with special focus on the atomization phase; (right) scheme representing the DoE and tests performed and the process parameters studied.

Results & Discussion The powders process conditions and the analytical results are presented below in table 1. Table 1 - Physicochemical characterization of the powders produced by SASD technology, presenting the process yield, Dv50, span, BD, water and EtOH content. Process conditions Test #1 #2 #3 #4 #5 #6 #7 #8 #9 #10

F_feed

T_in

P_sat

+1 +1 -1 -1 +1 +1 -1 -1 0 0

-1 -1 -1 -1 +1 +1 +1 +1 0 0

-1 +1 -1 +1 -1 +1 -1 +1 0 0

Yield % 62.6 69.6 61.7 62.1 66.8 70.4 54.6 61.1 67.9 67.1

Dv50 µm 0.99 1.09 0.84 0.85 1.09 1.10 0.93 0.95 0.92 0.95

132

Span (-) 1.90 2.13 1.50 1.67 1.67 1.65 1.43 1.35 1.46 1.86

Powder Properties BD aPScalc water g/mL µm % w/w 0.393 0.62 3.02 3.27 0.441 0.72 0.343 0.49 2.75 2.62 0.343 0.50 1.68 0.285 0.58 1.68 0.297 0.60 1.90 0.202 0.42 1.77 0.219 0.44 2.10 0.309 0.51 2.13 0.326 0.54

EtOH ppm 2767 3190 2194 2043 1059 709 252 224 1663 1758


Drug Delivery to the Lungs 27, 2016 - Cláudia Moura et al. Process yields as high as 70% (batch size of 11 grams) were obtained. Regarding the powders PS, it was observed that the powders produced at a higher T_in and F_feed presented a larger PS than the powders produced at lower temperatures and lower F_Feed. The higher temperature possibly caused particle inflation without breakage, leading to slightly larger particles while the lower F_feed possibly produced smaller droplets because the ratio F_CO2/F_feed increased, possibly leading to a stronger secondary atomization that formed smaller droplets and then particles. The particle size span of the SASD particles is smaller than the particles produced by SD using a conventional two-fluid nozzle. Regarding the residual water and ethanol solvent content, it is possible to observe that the higher the drying temperatures (T_in and T_out) and the lower the F_Feed, the lower the residual water and ethanol content in the final powder. In addition, all powders presented a residual EtOH content well below the ICH limits (< 5000 ppm). Globally, it is possible to observe that as expected, the powders BD is inversely proportional with the powder T_in (and consequently T_out) because it is simultaneously affected by the T_in and F_feed parameters. The powders produced at the higher drying temperatures and lower F_feed values presented a lower BD since higher temperatures promote the production of hollow particles which in combination with a low F_feed, leads to the production of particles with lower water and EtOH content, which further contributes to a lower BD. The same trends were observed in previous work conducted using the Spray drying technology where higher T_in and lower F_feed produces smaller particles. According to the XRPD and DSC data (data not shown), and similarly to what was observed when spray drying this formulation, amorphous trehalose and crystalline leucine were observed in all powders. The SEM micrographs presented in Figure 2 depict the powders produced at lower and higher T_in. All powders seem similar, uniform and spherical with the powders spray dried at higher T_in presenting a slightly smoother surface, in agreement to what would be expected and was observed with conventional SD. #1

#3

#5

#7

Figure 2 – SEM micrographs of some of the powders produced by SASD technology. Magnification of 10000x. The scale presented as a white line at the bottom of the SEM images corresponds to 1 µm.

The powders were then characterized in terms of aerodynamic performance where FPF ED(< 5µm) values as high as 86% were obtained as presented in Figure 3. The MMAD values were also lower than 2.5 µm for all cases. All powders presented a high FPF (76 – 86%), which makes it harder to draw conclusions on the process parameters impact due to the very small variations. Therefore, it would not be possible to make a clear conclusion without applying statistical analysis tools. The statistical models obtained predicted well the powder properties as well as the powder FPF. The powders from test 9 and 10 presented similar FPF values which shows the good process reproducibility. As presented in Figure 4, an acceptable model was obtained, being able to predict the powder FPF based on a correlation with – [1] the aerodynamic particle size (aPS) determined based on the powder PS, density and shape factor – which 2 2 presented a R and Q of 0.658 and 0.545, respectively (A). It was observed that the main descriptors that influence aPS and therefore the FPF are the F_feed and T_in (B) where it was observed that the higher T_in and the lower F_Feed the higher the FPF. The model variability presented an average of 0 and a standard deviation (s.d.) of 1.92 (C). With this information, the design space considering T_in and F_feed was drawn in order to obtain a FPF higher than 80%, with a confidence of 95%, it would be required to operate in the green area (D). According to the model, the P_sat did not have a significant impact on the powder aPS and consequently on the FPF.

133


Drug Delivery to the Lungs 27, 2016 - Optimization of Supercritical-CO2 Assisted Spray Drying for the Production of Inhalable composite particles

Figure 3 – In-vitro aerodynamic performance results obtained for all powders produced by SASD, presenting the FPF, MMAD and GSD. A)

B)

C)

D)

Figure 4 – Statistical analysis results: A) FPF vs aPS prediction; B) Factors impacting the aPS; C) FPF model variability assessment and D) Design Space of the FPF model (yellow area: FPF > 80%; green area: FPF > 85% (to guarantee that FPF > 80% with 95% confidence).

Conclusions The SASD technology enables the production of inhalation powders with a similar solid state of the ones commonly produced using Spray drying, presenting amorphous trehalose and crystalline leucine. The trends observed regarding the impact of the process parameters on the powder properties were also observed using the Spray Drying technology. Although all powders presented a high FPF, where the maximum values reached 86%, it was observed that by manipulating F_feed and T_in, a fine control over the FPF and a good prediction model were successfully achieved: in order to increase the powder FPF, T_in should be increased and F_feed decreased. However, the process throughput and the powder FPF should be simultaneously evaluated according to each case to find the best operating range from an economic perspective. SASD technology was shown to be a suitable and scalable alternative to conventional SD for preparing inhalation formulations, particularly addressing challenges in ensuring a fine atomization, potentially reducing shear and thermal stress, and reducing the use of organic solvents. References 1

Pilcer G, Amighi K: Formulation strategy and use of excipients in pulmonary drug delivery, Int J Pharm 2010; 392:pp1-19.

2

Hoppentocht M, Hagedoorn P, Frijlink H, Boer H: Technological and practical challenges of dry powder inhalers and formulations, Adv Drug Deliv Rev 2014; 75:pp18-31.

3

Tabernero A, Martín del Valle E, Galán M: Supercritical fluids for pharmaceutical particle engineering: Methods, basic fundamentals and modelling. Chem Eng Process 2012; 60:pp9-25.

4 Reverchon: Supercritical-Assisted Atomization to produce micro and/or nanoparticles of controlled size and distribution. Ind Eng Chem Res 2002; 41:pp2405-2411.

134


Drug Delivery to the Lungs 27, 2016 - Diana A. Fernandes et al. Impact of Spray Drying on Superoxide Dismutase Activity in Composite Systems with Optimal Aerodynamic Performance for Dry Powder Inhalers 1,2

1

1

1

Diana A. Fernandes , Raquel Barros , Cláudia Moura , Eunice Costa & Maria L. Corvo

2

1

2

Hovione SA, Loures, 2074-506, Portugal Research Institute for Medicines (iMed.ULisboa), Faculdade de Farmácia, Universidade de Lisboa, Lisboa, 1649003, Portugal

Summary The aim of this study was to investigate the effect of drying process variables on the aerodynamic performance and enzymatic activity of a spray-dried model enzyme, Cu,Zn-Superoxide Dismutase (Cu,Zn-SOD). Cu,Zn-SOD is often implicated in a broad spectrum of oxidative stress related diseases, from Cystic Fibrosis to Rheumatoid Arthritis, catalyzing the breakdown of superoxide radicals and so providing the first line of defense against oxygen 3 toxicity. A 2 full factorial experiment was conducted at a constant Drying Gas Flowrate using a trehalose:leucine [1] based composite system previously optimized for dry powder inhalers . Cu,Zn-SOD:trehalose:leucine spray dried powders were successfully generated displaying Fine Particle Fraction values (FPFs) of ≈60% and Enzymatic Activity Retention (EAR) values ranging 50-80%. Aerodynamic performance of these SD powders was mainly affected by Ratom, while enzymatic activity was mainly influenced by Ratom and Ffeed. Introduction Significant advances on recombinant DNA technology, coupled with the rising capacity of large-scale production of therapeutic biologics, are paving the way towards the cumulative increase of macromolecules in the [2] pharmaceutical industry pipeline . In particular, inhaled formulations are increasingly attractive not only to treat respiratory diseases but also systemic diseases due to the large surface area available for drug absorption and the avoidance of first-pass metabolism. This ultimately translates into a rapid onset of the drug action, reducing the required drug load and minimizing adverse side effects, whilst improving patient compliance over injectables. Pulmonary delivery can be performed using three main platforms: nebulizers, pressurized metered-dose inhalers (pMDIs) and dry powder inhalers (DPIs). Biopharmaceutical formulations have already been successfully formulated in nebulizers. However, its liquid based formulations result in the early stability loss of biologics upon storage. Moreover, nebulization requires long periods of dosing due to inefficient drug deposition, leading to drug wastage and may require sterilization between uses. On the other hand, pMDIs use for biopharmaceuticals administration is relatively scarce given the exposure of these to shear stress during device actuation along with [3] the complexity of generating stable formulations in the commonly used propellants . Finally, DPIs address these shortcomings, taking advantage of more stable solid based formulations and without any need for propellants. Despite the advantages of targeting the pulmonary route through DPIs, the efficient delivery of biopharmaceuticals to the lungs still presents a significant challenge: the generation of a stable aerosol with adequate aerodynamic properties while preserving the integrity of the biologic. Hence, the particle engineering technology employed to meet this balance plays a pivotal role. Spray Drying (SD) emerges as a viable technology given its relative simplicity (one-step based process), cost effectiveness and scalability with increased control over key aerosol features that impact its aerodynamic performance such as particle size, shape, internal structure, surface, among others, by fine-tuning formulation [4] composition and process parameters . This technology is also regarded as suitable for biopharmaceuticals due to the mild temperature exposure through evaporative cooling and short residence times. In addition, the wide range of SD scales commercially available, including miniaturized set-ups with low volume requirements, render it a promising alternative when working with expensive molecules. Herein, the impact induced by SD on the aerodynamic performance and enzymatic activity retention of Cu,ZnSuperoxide Dismutase (Cu,Zn-SOD), embedded in trehalose:leucine based composite systems for DPIs, is accessed by screening three process variables - Outlet temperature, Feed and Atomization flow rates. Cu,ZnSOD is a model enzyme that is often implicated in a broad spectrum of oxidative stress related diseases, from Cystic Fibrosis to Rheumatoid Arthritis, catalyzing the breakdown of superoxide radicals and so providing the first [5] line of defense against oxygen toxicity . Materials and Methods Experimental Design Outlet temperature (Tout) together with feed (Ffeed) and atomization (Ratom) flow rates were the factors selected to study spray drying induced stress on Cu,Zn-SOD. For each process variable two levels (low and high) were set out. Tout range was defined based on a minimum suitable temperature for evaporating the working solvent system - MilliQ (double distilled deionised) water - and on Cu,Zn-SOD melting temperature, previously determined by differential scanning fluorimetry as 72°C in MilliQ water. Overall, Tout, Ffeed and Ratom ranges were adjusted in 3 agreement with a full factorial experiment 2 , as depicted in Table 1, plus an additional center point, resulting in a total of 9 experiments. Drying gas flow rate (Fdrying) and SD feed solutions composition were kept constant across the trials. All trials were executed in a lab-scale spray drier (Mini Spray Drier BUCHI model B-290).

135


Drug Delivery to the Lungs 27, 2016 - Impact of Spray Drying on Superoxide Dismutase Activity in Composite Systems with Optimal Aerodynamic Performance for Dry Powder Inhalers Table 1 – Set of Test conditions performed according to the defined design of experiments. Test 1 2 3 4 5 6 7 8 9

Tout (°C) 65

95 80

Ratom (mm) 40 60 40 60 40 60 40 60 50

-1

Ffeed (gmin ) 3 3 9 9 3 3 9 9 6

-1

Fdrying (kgh )

35

Morphology and aerodynamic performance of SD generated powders were determined to evaluate their suitability from an inhalation standpoint. To assess the impact of the tested process variables on Cu,Zn-SOD, the total protein was quantified and enzyme activity was measured for each SD run. To further understand the interaction between the studied factors (Tout, Ratom and Ffeed) on the morphology, aerodynamic performance, protein yield and enzymatic activity retention (EAR) of Cu,Zn-SOD powders, a multivariate statistical analysis based on partial least squares (PLS) regression model was performed using SIMCA v13.0.3.0 software (Umetrics). SD Feed Solutions Composition [1]

comprising The feed solutions for spray drying comprised a previously optimized excipient matrix trehalose:leucine (4:1) at 2% (w/w) of solids content in a MilliQ water based solvent system. Bovine erythrocytes Cu,Zn-SOD (Sigma, St. Louis, MO, USA) was added at 0.1% (w/w) concentration. SD Powder Characterization: The morphology of the dry powders was examined by Scanning Electron Microscopy (SEM) for particle shape and surface and Laser Diffraction (LD) for particle size distribution. The aerodynamic performance was assessed in-vitro using an 8 stage gravimetric Andersen Cascade Impactor (ACI) with 20±0.4 mg of spray-dried powder filled in HPMC size 3 capsules using a Plastiape HR model 7 at 60 L/min (for a pressure drop of 4 kPa). Cu,Zn-SOD Quantification and Enzymatic Activity Retention (EAR) For each SD run, total protein was quantified using a modified Lowry method with prior protein precipitation with [7] for trehalose and leucine removal. Cu,Zn-SOD enzymatic activity retention Trichloroacetic Acid (TCA) [5] measurements were performed using the SOD Assay Kit – WST (Sigma, St. Louis, MO, USA) . Results and Discussion Impact of process variables on Morphology and Aerodynamic Performance of Cu,Zn-SOD SD Powders The SEM micrographs of the nine SD powders were very similar among each other, presenting spherical and slightly shriveled particles, typical from a SD process, with a particle size (PS) within the inhalable size range (Figure 1). This evidences the flexibility of trehalose:leucine based composite systems to display desirable attributes from an inhalation standpoint whether prepared under mild or harsh conditions (at least considering the studied conditions). This is advantageous when dealing with biologics since lower processing temperatures can be employed, for instance, without affecting powder morphology features that potentially impact aerodynamic performance. The apparent PS observed in the SEM micrographs is also in agreement with the particle size distribution (PSD) quantitatively determined by laser diffraction (Figure 2 – A. Fine Particle Fraction expressed as the percentage particles < 5 μm for each test: error bars correspond to standard deviations based on three replicates. B. Input variables that allowed for the best PLS regression model fit (R2>0.7) and prediction (Q2>0.5; R2- Q2<0.2-0.3) for each one of the output variables.). Particle size seems to be independent of Tout (see tests 1&5, 2&6, 3&7, 4&8) and Ffeed (see tests 1&3, 2&4, 5&7, 6&8) as similar values of X50 and span ((X90 – X10)/X50) can be observed for both low and high level of these process variables. R atom is thus the predominant factor responsible for the different PS between tests with distinct values for this parameter: the higher the value of R atom the lower the X50 values. The ACI assays performed to assess the powders aerodynamic performance resulted in Fine Particle Fractions (FPF) ranging from 45-60% with a capsule emitted dose (ED) higher than 98%. FPF along with mass median aerodynamic diameter (MMAD) results followed the same trend observed for PSD being mainly dependent once again on Ratom (Figure 2) regardless of Tout and Ffeed values. The highest FPF and MMAD obtained match the test [1] conditions performed for the high level of Ratom, following the expected trend as determined in previous studies .

136


Drug Delivery to the Lungs 27, 2016 - Diana A. Fernandes et al.

Test 1

Test 2

Test 3

Test 4

Test 5

Test 6

Test 7

Test 8

Test 9

Figure 1 – SEM micrographs (x5000) for Tests 1-9. Impact of process variables on Cu,Zn-SOD Protein Yield and Enzymatic Activity Retention (EAR) To further narrow down the best conditions to spray dry Cu,Zn-SOD, its protein yield and respective enzymatic activity retention were determined for all nine tests (Table 2) with values higher than 60 and ranging from 50-80%, respectively. The protein yield observed for tests 1&5 sharing the same low level values of Ratom and Ffeed exceeded 100% while exhibiting the higher loss of activity preserving only 50-60%. This can be possibly explained by the longer residence time to which Cu,Zn-SOD was exposed causing it to aggregate/degradate, but further studies would be required to corroborate the hypothesis. When comparing tests at the same Tout and Ratom (Tests 2&4, 6&8), to a lower Ffeed corresponds a slightly higher protein yield although, despite the negligible difference, respective EAR range values do not necessarily follow the same trend, displaying instead lower EAR range values for lower Ffeed values. Tout (tests 2&6, 3&7, 4&8) does not seem to have a visible impact on these two parameters. To strengthen the previous qualitative observations, a multivariate statistical analysis was carried out to further understand the interaction between Tout, Ratom and Ffeed – inputs – on the morphology (X50), aerodynamic performance (MMAD, FPF), protein yield and EAR – outputs – of Cu,Zn-SOD powders. Figure 2B showcases 2 [7] 2 2 2 [7] which input variables allowed for the best fit (R >0.7 ) and prediction (Q >0.5; R - Q <0.2-0.3 ) of the derived PLS regression models describing each one of the output variables. PLS regression models obtained for X50, MMAD and FPF displayed the best fit and prediction when considering Ratom as input parameter. No good PLS 2 2 model could be derived for protein yield, with a R value<0.7 and Q value<0.5. On the other hand, PLS regression model obtained for EAR exhibited the best fit and prediction when this was simultaneously predicted by Ratom and Ffeed. Finally, Tout had a negligible impact on all the studied output variables. Thus, the qualitative conclusions previously withdrawn could be corroborated by the multivariate statistical analysis.

137


Drug Delivery to the Lungs 27, 2016 - Impact of Spray Drying on Superoxide Dismutase Activity in Composite Systems with Optimal Aerodynamic Performance for Dry Powder Inhalers

Figure 2 – A. Fine Particle Fraction expressed as the percentage particles < 5 μm for each test: error bars correspond to standard deviations based on three replicates. B. Input variables that allowed for the best PLS 2 2 2 2 regression model fit (R >0.7) and prediction (Q >0.5; R - Q <0.2-0.3) for each one of the output variables. Table 2 - Overview of PSD (X50), Aerodynamic Performance (MMAD, GSD), Protein Yield and Enzymatic Activity Retention (EAR) for all nine tests. X50 (μm) 2.3 1.6 2.1 1.7 2.1 1.5 2.3 1.5 1.9

Test 1 2 3 4 5 6 7 8 9

span 1.9 1.9 2.1 1.8 1.9 1.9 2.0 1.9 1.9

MMAD ( μm) 3.7 2.7 3.5 2.8 3.3 2.8 3.8 2.6 3.0

GSD 1.7 1.8 1.8 1.8 1.7 1.6 1.7 1.7 1.7

Protein Yield (%) 118 94 92 81 131 90 63 81 77

EAR (%) 50-60 65-75 65-75 70-80 50-60 65-75 70-80 75-85 70-80

Conclusions The impact induced by SD on the aerodynamic performance and enzymatic activity retention of Cu,Zn-SOD embedded in trehalose:leucine based composite systems for DPIs was accessed by screening three process variables - Tout, Ratom and Ffeed. Aerodynamic performance of these SD powders was mainly affected by Ratom while enzymatic activity was mainly influenced by Ratom and Ffeed. Further research focusing on the enzyme configuration before and after Spray Drying using other techniques like Dynamic Scanning Fluorimetry and Dynamic Light Scattering could also be carried out. It can be concluded that higher values of Ratom can be employed to successfully spray dry Cu,Zn-SOD intended for inhalation since its enzymatic activity does not seem to be greatly affected by these factors, while ensuring good aerodynamic performance. Acknowledgments We acknowledge Fundação para a Ciência e a Tecnologia for partial funding (UID/DTP/04138/2013). References 1.

Moura C, Vicente J, Palha M, Neves F, Aguiar-Ricardo A, Costa E: Screening and Optimization of Formulation and Process Parameters for the Manufacture of Inhalable Composite Particles by Spray-Drying. Journal of Aerosol Medicine and Pulmonary Drug Delivery 2015; 28, pp A20–A20.

2.

Cruz M E M, Simões S I D, Corvo M L, Martins M B F, Gaspar M M: Formulation of Nanoparticulate Drug Delivery Systems (NPDDS) for Macromolecules In: Y Pathak, D. Thassu (eds): Drug Delivery Nanoparticles: Formulation and characterization Informa Healthcare, New York; pp35-49, 2009.

3.

Berkenfeld K, Lamprecht A, McConville J T: Devices for Dry Powder Drug Delivery to the Lung AAPS PharmSciTech 2015; 16, pp 479– 490.

4.

Fernandes D A, Moura C, Campos S, Costa E, Neves F: Impact of the API Concentration on the Solid State Properties and Aerodynamic Performance of Composite Particles for DPIs. Respiratory Drug Delivery 2016; 2, pp 339–344.

5.

Peskin A V, Winterbourn C C: A microtiter plate assay for superoxide dismutase using a water-soluble tetrazolium salt (WST-1) Clinica Chimica Acta 2000; 293, pp 157–166.

6.

Corvo M L, Jorge J C S, Hof R, Cruz M E M, Crommelin D J A, Storm G: Superoxide dismutase entrapped in long-circulating liposomes: formulation design and therapeutic activity in rat adjuvante arthritis Biochimica et Biophysica Acta 2002; 1564, pp 227–236.

7.

Eriksson L, Johansson E, Kettaneh-Wold N, Wikstrӧm C, Wold S: Analysis of factorial designs In (3rd Edition): Design of Experiments Principles and Applications. Umetrics AB, Umeå, Sweden.

138


Drug Delivery to the Lungs 27, 2016 – Nancy Rhein1 et al. ®

Parteck M DPI – a novel mannitol as carrier in dry powder inhalation formulations 1

2

Nancy Rhein , Gudrun Birk , Regina Scherließ 1

1

Department of Pharmaceutics and Biopharmaceutics, Kiel University, Grasweg 9a, 24118 Kiel, Germany 2 Merck KGaA, Frankfurter Straße 250, 64293 Darmstadt, Germany

Summary Dry powder inhalation (DPI) utilising carrier-based blends is a well-known formulation strategy in local treatment of pulmonary diseases. Typically, lactose is used as carrier in these interactive powder blends. A suitable alternative to ® this established carrier material is Parteck M DPI, a particle engineered mannitol for DPI applications. In this study it is characterised with respect to its physico-chemical properties (particle size distribution, crystallinity, specific surface area, content of reducing sugars) and with respect to its performance as carrier material in interactive blends for dry powder inhalation. Blending conditions, homogeneity, fine particle fraction/dose and uniformity of dosing are ® investigated. It is shown that Parteck M DPI is a fully crystalline material with a high specific surface area which forms homogeneous blends with low dose active pharmaceutical ingredients (API) and allows constant dosing from a ® ® reservoir-based inhaler (Novolizer ). In comparison to a commercial lactose quality for inhalation Parteck M DPI exhibits good aerodynamic performance from both capsule-based and reservoir inhaler devices indicating that it is a versatile alternative carrier in DPI blends. Introduction The use of carrier-based blends in inhalation is a well-used concept. They improve powder bulk properties such as flowability, cohesion, dosing and powder homogeneity during powder handling and separation of the individual dose, but also to facilitate dispersion during the inhalation process. Hence, ideal carrier particles should be larger than the corresponding API particles, they should be able to balance between particle cohesion and adhesion, should have [1]. Traditionally, lactose is good flow characteristics, good stability and should be compatible with a wide range of APIs used as carrier material for DPI formulations. This material is mostly used as crystalline alpha lactose monohydrate, which is stable and not very susceptible to water vapor, but it is known that amorphous content can be introduced via [2] processing such as milling or spray drying . Further, it is of animal origin, which sometimes may be an issue. Mannitol as alternative to lactose has been discussed for quite some time now. Whereas 10 years ago there were still constraints with respect to the use of mannitol in inhalation as no product had been registered using mannitol as excipient, the regulatory landscape has changed since and mannitol is approved as excipient as well as active [3] material in inhalation . Furthermore the material is animal origin free and listed as GRAS (generally recognized as [4] safe) materials by the FDA . This paves the way for the use of mannitol as alternative carrier material in inhalation. However, carrier materials in inhalation have certain requirements with respect to their specifications, which are not met by commercially available mannitol qualities. These specification parameters comprise particle size distribution, particle morphology and surface area, microbiology and foreign particles, just to mention some characteristics being important for the use as carrier in dry powder inhalation. This work characterizes an engineered mannitol quality which has been developed for DPI use and shows differences and benefits in comparison to other mannitol qualities and a standard DPI lactose. Material and methods ®

Materials: Parteck M DPI and D-Mannitol were provided by Merck (Merck KGaA, Darmstadt, Germany). Comparable carrier particles were a processed mannitol quality for direct compression (Roquette Frères, Lestrem, France) and two DPI lactose qualities (Meggle Excipients and Technology, Wasserburg, Germany and DMV-Fonterra Excipients GmbH & Co. KG, Goch, the Netherlands) as gold standard. Micronised budesonide (mean particle size: 1.39 µm ± 0.03 µm; Farmabios,S.p.A., Cropello Cairoli, Italia) and micronised salbutamol sulphate (mean particle size: 1.61 µm ± 0.03 µm; Lusochimica S.p.A, Peschiera Borromoo, Italy) are used as model APIs in blends and NGI experiments. Scanning electron microscopy (SEM): By using SEM particle morphology and surface were visualised. Samples were investigated with a Zeiss Ultra 55 plus (Carl Zeiss NTS GmbH, Oberkochen, Germany) using the SE - 2 detector and a working voltage of 2 kV. Particle size distribution: Particle size was measured by laser light diffraction (HELOS, Sympatec GmbH, ClausthalZellerfeld, Germany, R5 lens). Powder was dispersed in an adjustable air jet by the RODOS system (Sympatec GmbH, Clausthal-Zellerfeld, Germany). The dispersing pressure was set to 3.0 bar ensuring a complete dispersion to individual particles without damage. To determine specifically fine carrier particles below 5 µm upon dispersion from ® the inhaler, carrier materials were dispersed by the Novolizer and size was measured with the INHALER system (Sympatec GmbH, Clausthal-Zellerfeld, Germany). The data was evaluated using the Windox 5.4.2.0 software (Sympatec GmbH, Clausthal-Zellerfeld, Germany). Reported data is the average of six measurements.

139


Drug Delivery to the Lungs 27, 2016 - Parteck® M DPI – a novel mannitol as carrier in dry powder inhalation formulations Specific surface area: According to the Ph.Eur. method 2.9.26 an eleven-point-measurement with a partial pressure p/p0 range between 0.05 and 0.30 was conducted. For this a NOVA 2200 BET system (Quantachrome Instruments, Byton Beach, USA) was used. Results were calculated with the multiple-point-BET-method and are given in m²/g. Preparation of interactive mixtures: By a double-weighing-method 1, 2 or 4 % budesonide and 1, 2 or 4 % SBS, ® ® respectively, were blended to Parteck M DPI. A Turbula blender (Type T2C, Willy A. Bachhofen AG Maschinenfabrik, Basel, Switzerland) at 42 rpm with blending times of 3 x 5 min was used for blending as validated in previous experiments. To eliminate agglomerates a sieving step was introduced before blending and after every five minutes. Ten samples (8-12 mg) of each powder mixture were taken for homogeneity testing and API content was analysed by reversed phase high performance liquid chromatography (RP-HPLC). Blends were judged to be homogeneous at a relative standard deviation (RSD) of less than 3 %. Impaction analysis: Impaction analysis was performed with the NGI (apparatus E, European Pharmacopoeia 8.6) ® ® utilising a reservoir inhaler (Novolizer ) and a capsule-based inhaler (Cyclohaler ), respectively. To avoid reentrainment of particles the cups were coated with a stage coating (Brij 35, Ethanol, Glycerol). Flow rates were adjusted to ensure 4 kPa pressure drop over the devices according to Ph. Eur. Budesonide and SBS content was identified by RP-HPLC. Data were evaluated with the Copley Inhaler Testing Data Analysis Software (Copley Scientific, Nottingham, United Kingdom). Fine particle fraction below 5 µm (of emitted) is calculated from the resulting aerodynamic particle size distribution. Reported data is average of three runs and measured at constant conditions (21 °C and 45 % relative humidity). Results and discussion ®

Physico-chemical characterisation of Parteck M DPI Particle size distribution of different materials is shown in Figure 1 and table 1. It can be seen that in contrast to commercial milled D-mannitol, mean particle size is larger which is advantageous for a material to be used as DPI ® [5] carrier, especially when used in an inhaler such as the Novolizer as shown in previous studies . There is also a comparison to two commercial lactoses for DPI, which are comparable in x50-value but have different span-values. 4

Parteck M DPI D-Mannitol DC Mannitol

1.5

3

1

2

0.5 0

Parteck M DPI Lactose A Lactose B

Probability density function q3

Probability density function q3

2

1

1

10 100 Particle size, µm

0

1000

0.1

1

10 100 Particle size, µm

1000

®

Figure 1 (left): Particle size distribution of Parteck M DPI, a commercial mannitol quality for direct compression (DC) and ® D-mannitol; 1 (right): Particle size distribution of Parteck M DPI and two commercial Lactose qualities for inhalation, n=6, error bars depict standard deviation. ®

Material

Parteck M DPI

D-Mannitol

DC-Mannitol

Lactose A

x50 (µm)

117 ± 5

66 ± 2

106 ± 2

128 ± 0

Span

2.03 ± 0.09

3.03 ± 0.04

1.71 ± 0.02

0.71 ± 0.00

Lactose B 118 ± 1 1.23 ± 0.03

Table 1: x50-values of the respective qualities and Span (Span = (x90-x10)/x50) ± standard deviation, n = 6 ®

In comparison to DPI lactoses and a processed mannitol for direct compression applications, Parteck M DPI shows a broader particle size distribution and especially a larger amount of large particles > 200 µm, but also a certain amount of intrinsic fines. The material does not show any increased affinity to adsorb water vapour up to 80 % rH, which is [6] ® advantageous in terms of storage stability . Further, Parteck M DPI is fully crystalline and is present in the most stable beta polymorph, which further underlines its physical stability (data not shown). Its low content of reducing sugars (below 0.02 % as granted by the specification) is advantageous for reaction-sensitive materials such as proteins.

140


Drug Delivery to the Lungs 27, 2016 – Nancy Rhein1 et al. Use as carrier material in interactive blends ® Due to the preparation method, Parteck M DPI particles exhibit an overall round shape with a structured surface (Figure 2A) comparable to DC Mannitol (2C). However, this his results in a uniquely high specific surface area compared to other materials (Figure 3 left), which allows balanced adhesion with small particulate API and results in homogeneous blends at modest blending conditions in a tumble blender. Blend homogeneity < 3 % RSD is reproducibly achieved with 3 x 5 min blending time regardless API (Fig. 3 right) and is comparable to blends with commercial DPI lactose as carrier. Blending conditions for other carrier materials ranged from 1 x 5 min (D-Mannitol) up to 5 x 5 min for lactose in combination with salbutamol sulphate to meet the criteria of RSD below 3 %. Further, the blends are stable and do not show any segregation or dosing problems when delivered from a reservoir device (data not shown). A

B

C

®

4

3%

3

RSD, %

Specific surface area, m²/g

Figure 2: SEM images of Parteck M DPI (A), D-Mannitol (B) and DC-Mannitol (C) (B: 250x magnification, A and C: 500x).

2 1

2% 1% 0%

0

Parteck® M DPI + Parteck® M DPI + 2 % Bud % SBS

2

®

Figure 3: BET surface area of Parteck M DPI, a commercial DPI lactose, a commercial mannitol quality for direct ® compression (DC mannitol) and D-mannitol (left) and homogeneity results of different blends with Parteck M DPI (right), n=3, error bars depict standard deviation. ®

Aerodynamic characterisation of interactive blends with Parteck M DPI as carrier and budesonide as hydrophobic, regularly shaped, model API or salbutamol sulphate as hydrophilic, needle-shaped, model API, respectively, show ® adequate fine particle fractions above 30 % when dispersed from the Cyclohaler as an example for capsule based ® system or the Novolizer as an example for reservoir devices (Fig .4). ®

Budesonide as the more hydrophobic material shows a clear trend towards smaller FPF when the Parteck M DPI ® ® blend is dispersed from the Cyclohaler in comparison to the Novolizer , whereas for salbutamol sulphate the trend is inverted. This is attributed to differences in the dispersing mechanism of the two devices with a preference for larger carrier materials of the Novolizer due to the classifier technology and differences in the carrier adherence between the two APIs. SBS is more difficult to distribute homogeneously (see RSD) with the carrier due to its needle-like shape. Possibly separate API agglomerates are formed instead which are not readily dispersed by the Novolizer classifier, but can get aerosolised as individual particles in the Cyclohaler. In this case, the smooth and regular surface of DPI lactose can be advantageous as it also allows needle-like particles to adhere closely. Hence, agglomerate formation is inhibited and individual particles get detached from the carrier in the Novolizer classifier. The two carrier materials also show differences in performance which can be due to differences in chemistry-specific API-carrier interactions. ® This is under further investigation. Parteck M DPI blend performance is reproducible and fine particle dose can directly be tuned by blend concentration as shown in Figure 5. Here, it can be seen, that an increase in blend drug content directly translates to a higher fine particle dose.

141


Drug Delivery to the Lungs 27, 2016 - Parteck® M DPI – a novel mannitol as carrier in dry powder inhalation formulations

Salbutamol sulphate

60

Lactose A

Lactose B

40 20 0

Novolizer

Fine particle fraction, %

Fine particle fraction, %

Budesonide Parteck® M DPI

60

Parteck® M DPI

Lactose A

Lactose B

40 20 0

Cyclohaler

Novolizer

®

Cyclohaler

Figure 4: Fine particle fraction of interactive blends with Parteck M DPI and budesonide (left) and salbutamol sulphate ® ® (right) from the Novolizer and the Cyclohaler , respectively (n=3, error bars are standard deviation). FPF of lactose-based interactive blends are given for comparison.

200 Fine particle dose, µg

®

Considering the particle size distribution of Parteck M DPI upon dry dispersion at 3 bar in comparison to commercial DPI lactose (Figure 1), it might be questionable whether the increased amount of intrinsic fines could lead to a noteworthy amount of mannitol ® being inhaled. Dispersion of pure carrier Parteck M DPI ® from the Novolizer however, showed a particle size fraction below 5 µm comparable to a commercial lactose carrier indicating that device dispersion does not result in pronounced detachment of intrinsic mannitol fines upon inhalation (data not shown).

1 % API

2 % API

4 % API

150 100 50 0

Budesonide

Salbutamol sulphate

Figure 5: Dependency of fine particle dose (FPD) on blend concentration from the Novolizer®, n=3, error bars depict standard deviation.

Conclusion ®

Particle engineered mannitol for dry powder inhaler formulations as available as “Parteck M DPI” offers a versatile alternative to lactose in carrier based formulations. The material features a special particle size distribution and unique surface structure employing a large surface area per gram. The material is especially suitable for regularly shaped micronized APIs and inhaler devices with high dispersion efficacy compared to the other carriers assessed in this study. Its large specific surface allows the preparation of homogeneous blends with low dose API without much effort. Blends result in balanced adhesion forces which allow constant dosing from reservoir devices without segregation and adequate aerodynamic characteristics. Acknowledgements The authors would like to thank Merck KGaA for funding this project. References 1

Rahimpour Y, Kouhsoltani M, Hamishehkar H: Alternative carriers in dry powder inhaler formulations. Drug Discovery Today 2014, 19(5):618-626.

2

Garnier S, Petit S, Mallet F, Petit M-N, Lemarchand D, Coste S, Lefebvre J, Coquerel G: Influence of ageing, grinding and preheating on the thermal behaviour of alpha-lactose monohydrate. International journal of pharmaceutics 2008, 361:131-140.

3

Pharmaxis: Australian Public Assessment Report for Mannitol. by Government Department of Health and Ageing; 2011.

4

FDA: Inactive Ingredient Search for Approved Drug Products - Mannitol. Internet data base, accessed 20.7. 2016.

5

Rhein N, Birk G, Scherließ R: Influence of different inhalers on fine particle fraction of mannitol carriers in dry powder inhaler formulations. In: World Meeting on Pharmaceutics, Biopharmaceutics and Pharmaceutical Technology. Glasgow, UK; 2016.

6

Rhein N, Birk G, Scherließ R: Characterisation of particle engineered Mannitol as alternative carriers in dry powder inhalation formulations. In: Drug Delivery to the Lungs 26: 2015; Edinburgh, UK; 2015.

142


Drug Delivery to the Lungs 27, 2016 - Valeria Carini et al. Dry powder intended for pulmonary delivery: Comparison between PGA-co-PDL and chitosan Nano Composite Microparticles 1,

2

3

3

Valeria Carini, Ayca Y. Pekoz , Gillian A. Hutcheon , Imran Y. Saleem , Adel A. Mohamed 1 2 3

3

School of Pharmacy, Unicam, Camerino University, Italy.

School of Pharmacy, Faculty of Pharmacy Istanbul University, Turkey

School of Pharmacy and Biomolecular Sciences, Liverpool John Moores University, Liverpool, L3 3AF, UK.

Summary Background: The use of polymer based nanoparticles (NPs), poly(glycerol adipate-co-ω-pentadecalactone) and chitosan for inhalation by spray drying process were evaluated for formulation parameters, mannitol sugar and leucine amino acid as excipient, concentration, and ratio of NPs to excipient.. The aim of this study was to design and formulate [PGA-co-PDL] and naturally occurring cationic chitosan nanoparticles within microparticle carriers (NCMPs) for pulmonary delivery. Methods : PGA-co-PDL, NPs were prepared via an oil in water (o/w) single emulsion solvent evaporation method using DOTAP (%w/v) to produce cationic NPs. Chitosan solution was prepared at concentration of 0.2% in 1% acetic acid buffer (pH:6). 5 formulations (F1-F5) were used to optimise different parameters such as sugar (mannitol) and amino acid (leucine) excipient, concentration and ratio of NPs to excipient, atomized air flow, aspirator capacity, feed rate, inlet/outlet temperature to formulate and achieve particles of desired yield, size and shape with low moisture content to target lung cancer cells. Results: Optimum PGA-co-PDL and chitosan NCMPs resulted in recovered nanoparticle size < 500 nm. Spray drying resulted in NCMPs over 65% yield. Optimum formulation theoretical dynamic diameter dae was 0.63±0.01 and 0.82±0.02 µm and formulations containing higher percentage mannitol showed lower moisture content 2.02±0.03 and 0.57±0.006 for PGA-co-PDL and chitosan NCMPs respectively. The tapped density of PGA-co-PDL and Chitosan 3 NCMPs were relatively similar (0.14±0.01 and 0.35±0.03 g cm- ). Conclusion: The results indicated optimum spray drying of NCMPs containing mixture of leucine and mannitol (25:75% w/w) with PGA-co-PDL and chitosan NPs, which can be applied for dry powder inhalation.

Background Polymer based NPs are considered attractive candidates for macromolecules as pulmonary delivery. They provide several advantages such as biocompatibility, sustained release, mucoadhesion and low toxicity. Furthermore, polymeric NPs can be combined with various drugs that lead to improved solubility of drugs and [1] achieve uniform distribution .A number of polymers have been studied for pulmonary macromolecules delivery including polyesters such as poly (glycerol adipate-co- ω-pentadecalactone) [PGA-co-PDL] as a pulmonary [2] carrier . Alfagihi et al., 2015 showed that PGA-co-PDL NPs could be incorporated into nanocomposite [3] microcarriers (NCMPs) using amino acids excipient with enhanced aerosol performance and high yield . Another study showed the the possibility of modification of PGA-co-PDL with cationic surfactant didodecyldimethylammonium bromide (DMAB) and surface adsorbed bovine serum albumin by double emulsion[4] spray drying with reasonable yield, good encapsulation efficacy and aerosolization performance .These polymeric NPs seem to be good alternative candidate for delivering small nucleic acids to the lung. It has been reported in a previous study, chitosan-based NPs were co spray-dried with lactose, mannitol, maltodextrin with or without leucine. Adding leucine and mannitol to the formulation produced a spherical shape of NCMPs with rough surface. Moreover, leucine improved the process yield due to its antiadherent properties and caused decreased [5] particle size . Aim To design, formulate and compare dry powders of PGA-co-PDL and chitosan NPs within microparticle carrier (NCMPs) for pulmonary delivery. Methods Production of NPs PGA-co-PDL, NPs were prepared via an oil in water (o/w) single emulsion solvent evaporation method using DOTAP (%w/v) to produce cationic NPs. Chitosan solution was prepared at concentration of 0.2% in 1% acetic acid buffer (PH:6). The TPP was dissolved in distilled water at concentration of 0.84 mg/ml, added dropwise to [6] 5ml polymer solution, and left stirring for 1 hour (250 rpm) . Both NPs were collected by centrifugation twice at 78,000 x g, 40 min and 4°C.

143


Drug Delivery to the Lungs 26, 2015 - Dry powder intended for pulmonary delivery: Comparison between PGA-coPDL and chitosan Nano Composite Microparticles Nanocomposite microparticles preparation 5 formulations (F1-F5) were used to optimise different parameters such as sugar (mannitol) and amino acid (leucine) excipient, concentration and ratio of NPs to excipient, atomized air flow, aspirator capacity, feed rate, inlet/outlet temperature to formulate and achieve particles of desired yield, size and shape with low moisture content. Both obtained polymeric NPs were incorporated into NCMPs using L-leucine and mannitol solution at polymer to carrier ratio of 1:1.5 (w/w) with different concentrations, L-leucine: mannitol F1 (100:0 %), F2 (75:25 %), F3 (50:50 %), F4 (25:75 %), F5 (:100 %). The prepared solutions were spray dried using a Büchi B-290 minispray drier (Büchi, B-290 Flawil, Switzerland). Characterisation The resultant NCMPs were characterised for particle size using a Zetasizer® NS (Malvern Instruments, UK), morphology (Scanning Electron Microscope), % yield was quantified as a percentage mass of expected total powder yield (n=3), water content (Thermogravimetric analysis, TGA Q50 UK), tapping density, theoretical aerodynamic diameter (d ae) and carr’s index. Results and discussion The identified optimum parameters (Table 1) applied for both PGA-co-PDL and chitosan NCMPs resulted in optimum formulation “F4” with particle size < 500 nm PGA-co-PDL NCMPs (409±10.05 nm), chitosan NCMPs (380.2±30.3 nm), (Table 2). Spray-drying resulted in NCMPs over 65% yield except chitosan NCMPs F5 which had 7.15±4.1. All formulations that incorporated higher percentage mannitol showed lower moisture content. Optimum parameters

Level

Unit

Feed Rate

0.5

ml/min

Atomized air flow

480

L/h

Aspirator capacity

70

%

Inlet

70

Outlet temperature

47

Table 1: Optimum Processing parameters used in the spray drying of F1-F5 Formulation (F)

Particle size (nm)

% Yield

Tapped density

Water content (%)

PGA-co-PDL F1

958.3±21

56±11.3

(-)

5.1±0.37

F2

1810.66±18

79.2±10.1

0.17±0.01

4.91±0.20

F3

3252.06±26

95.6±2.8

0.13±0.20

3.8±0.9

F4

409.7±10.05

86.05±15.01

0.14±0.01

2.02±0.03

F5

2174.6±13

84±2.2

0.20±0.05

3.77±0.11

F1

853.7±21

63.46±5.2

0.10±0.01

14.8±19.8

F2

1120±27

65.47±8.5

0.16±0.01

0.34±0.08

F3

1346±15

59.55±2.8

0.31±0.04

0.42±0.22

F4

380.2±30.3

69.68±3.10

0.35±0.03

0.57±0.006

F5

(-)

7.15±4.1

(-)

(-)

Chitosan

Table 2: PGA-co-PDL and chitosan particle size (nm), % Yield, tapped density and water content (n=3). F1 (PGA-co-PDL), F5 (Chitosan) (-) did not produce. The optimum formulation had geometric particle size less than 2µm. The tapped density of PGA-co-PDL NCMPs 3 (0.132±0.03 -0.20±0.01 g cm- ), were relatively similar chitosan NCMPs ( 0.10±0.01– 0.35±0.03) (Table 2). SEM analysis (Figure 1) indicated that NCMPs were spherical in shape with no particles fusion.

144


Drug Delivery to the Lungs 27, 2016 - Valeria Carini et al.

A

B

Figure 1: SEM images of F4 NCMPs optimum formulation prepared using chitosan (A) and PGA-co-PDL (B) in the same spray-drying conditions. Geometric particle size and tapped density was used to calculate the theoretical dynamic diameter (d ae). Optimum formulation (F4) dae was 0.63±0.01 and 0.82±0.02 µm for PGA-co-PDL and chitosan NCMPs respectively (Table 2) . Conclusion The results indicated that method has been optimised for the spray drying of NCMPs containing mixture of leucine and mannitol in PGA-co-PDL and chitosan. The nanoparticle size has been recovered after spray drying and geometric particle size suitable for targeting the respiratory bronchiole. Moreover, the optimum formulation had high yield, and low moisture content. Further studies will investigate aerosolisation performance, cellular toxicity therapeutic agents for treatment of respiratory diseases, uptake and macromolecules loading efficacy. References 1.

Patton, J. S. and P. R. Byron (2007). "Inhaling medicines: delivering drugs to the body through the lungs." Nature Reviews Drug Discovery 6(1): 67-74.

2.

Kumari, A., Yadav, S. K., & Yadav, S. C. (2010). Biodegradable polymeric nanoparticles based drug delivery systems. Colloids and Surfaces B: Biointerfaces, 75(1), 1-18. Alfagih, I., Kunda, N., Alanazi, F., Dennison, S. R., Somavarapu, S., Hutcheon, G. A., & Saleem, I. Y. (2015). Pulmonary Delivery of Proteins Using Nanocomposite Microcarriers. Journal of pharmaceutical sciences, 104(12), 4386-4398.. Kunda, N. K., Alfagih, I. M., Dennison, S. R., Somavarapu, S., Merchant, Z., Hutcheon, G. A., & Saleem, I. Y. (2015). Dry powder pulmonary delivery of cationic PGA-co-PDL nanoparticles with surface adsorbed model protein. International journal of pharmaceutics, 492(1), 213-222. Pourshahab, P. S., Gilani, K., Moazeni, E., Eslahi, H., Fazeli, M. R., & Jamalifar, H. (2011). Preparation and characterization of spray dried inhalable powders containing chitosan nanoparticles for pulmonary delivery of isoniazid. Journal of microencapsulation, 28(7), 605-613.

3. 4. 5.

6. Carrillo, C., Suñé, J. M., Pérez-Lozano, P., García-Montoya, E., Sarrate, R., Fàbregas, A., ... & Ticó, J. R. (2014). Chitosan nanoparticles as non-viral gene delivery systems: Determination of loading efficiency. Biomedicine & Pharmacotherapy, 68(6), 775-783.

145


Drug Delivery to the Lungs 27, 2016 - Mathias Mönckedieck et al. Detachment of various spray dried drugs from engineered mannitol carrier particles 1

2

2

3

4

Mathias Mönckedieck , Jens Kamplade , Peter Walzel , Hartwig Steckel , Nora Urbanetz & Regina 1 Scherließ 1

Department of Pharmaceutics and Biopharmaceutics, Kiel University, Grasweg 9a, 24118 Kiel, Germany 2 Technical University Dortmund, Emil-Figge Str. 68, 44227 Dortmund, Germany 3 Deva Holding AS, Halkalı Merkez Mah. Basın Ekspres Caddesi, Istanbul, Turkey 4 Daiichi Sankyo, Luitpoldstr. 1, 85276 Pfaffenhofen, Germany

Summary Most dry powder inhalers (DPIs) utilise the well-established system of coarse carriers (> 50 µm) with good flow properties and fine drug particles (< 5 µm) attached to the carrier surface to overcome the cohesiveness of micronised drug particles and to ensure adequate dosing. This project was investigating the influence of carrier size and shape on the dispersion of various drug particles during inhalation. Spray drying was used to prepare mannitol carriers with different particle shapes that concurrently ranged from 50 to 80 µm in size. Six batches of different characteristics were then blended with spray dried qualities of four model drugs (salbutamol sulphate, tiotropium bromide, budesonide, formoterol fumarate) that were different in hydrophilicity to examine effects of carrier characteristics and drug properties. All interactive powder blends were aerodynamically characterised by impaction analysis using the Next Generation Impactor to correlate Fine Particle Fraction (FPF) and particle properties. Carrier shape was detected as the main factor to influence the FPF while different drugs preferred differently shaped carriers to perform best. Salbutamol sulphate was preferably detached from spherical carriers, but got entrapped in indentions, while tiotropium bromide and budesonide preferred small indentions to hide from assumed press-on forces during blending as those increase particle-particle interactions. Contrary to other drugs tested, formoterol fumarate (FOR) exhibited reduced FPFs whenever agglomerates occurred upon blending. Introduction Drug delivery to the lungs by dry powder inhalation is a commonly used approach to locally treat respiratory diseases like asthma or chronic obstructive pulmonary disease (COPD) and is currently in focus of research for systemic drug delivery. Dry powder formulations are beneficial compared to nebulisers or pressurised metered dose inhalers in terms of stability issues that arise from the presence of a solvent or dispersing medium. Carrierbased systems consisting of coarse carbohydrate and fine drug particles have been applied in several marketed products to ensure adequate dosing of cohesive drug particles as those are required to range from 0.5 to 5.0 µm in aerodynamic diameter (so called fine particle fraction, FPF) to reach the deeper airways of the lungs. Lactose monohydrate as mostly used in those formulations reveals drawbacks that might affect storage stability of carrier [1] or drug particles, whereas the here used mannitol, a non-reducing sugar alcohol, is known for its inertness . Spray drying was applied for carrier preparation as this technique enables maximum control over drying parameters and therefore over product properties like particle size or the carrier morphology. Particle size [2] distributions were kept as narrow as possible by using a specially designed laminar rotary atomiser . This study investigated particle-particle interactions in interactive powder blends with mannitol carrier particles that were different in size and shape and various drug particles (salbutamol sulphate (SBS), tiotropium bromide (TIO), budesonide (BUD), formoterol fumarate (FOR)). Special focus was laid on the aerodynamic properties of those drugs in correlation to carrier shape as particle morphology was frequently described to influence drug dispersion [3] . The overall goal was to investigate the influences that arise from carrier and drug properties to in the best case draw conclusions for future experiments with other active pharmaceutical ingredients. Experimental Methods ®

Mannitol (Pearlitol 160C) as kindly provided by Roquette Frères (Lestrem, France) was used to prepare the carrier particles for inhalation. SBS (Selectchemie, Zurich, Switzerland) and TIO (Hangzhou Hyper Chemicals Ltd., Zhejiang, China) were chosen as a hydrophilic model APIs while BUD (Minakem SAS, Dunkerque, France) and FOR (Vamsi Labs Ltd., Maharashtra, India) served as hydrophobic counterparts. Spray drying of mannitol carriers was performed with a self-constructed spray tower at the TU Dortmund, Germany. Mannitol (15 % [w/w]) was atomised with a laminar rotary atomiser at different rotation speeds (8,000 to 14,000 rpm) and dried at different outlet temperatures (Tout = 70°C to 97 °C) to generate particles of different size and shape in the scope of a design of experiments. Droplets were generated at a constant feed rate of 10 L/h and particles collected in a container at the spray tower bottom. The product was kept in a dryer at 100 °C for 1 h to remove residual moisture. Resulting mannitol carriers were characterised with various techniques to finally choose six batches that were different in particle size and shape for the preparation of interactive powder blends. Those batches were named according to the resulting outlet temperature as marker for particle shape and with S/M/L to indicate the carrier size (e.g. M71(L) for 71 °C outlet temperature and large carrier with > 65 µm). Particle size distribution was measured with a Helos laser diffractometer (Sympatec GmbH, Germany) after dispersion by pressurised air of 1.2 bar (Rodos dispersing unit, Sympatec GmbH, Germany). Particle shape was examined by visual inspection of scanning electron microscope (SEM) images (Zeiss Ultra 55 Plus, Carl Zeiss NTS GmbH, Germany) and categorisation into categories from 1 (spherical) to 5 (indented).

146


Drug Delivery to the Lungs 27, 2016 - Detachment of various spray dried drugs from engineered mannitol carrier particles

Spray drying of all drug particles was performed with a Büchi Mini Spray Dryer B290 (Büchi Labortechnik AG, Switzerland) that was equipped with a 2 mm two-fluid nozzle and an Inert Loop system (Büchi B-295, Büchi Labortechnik AG, Switzerland) to remove organic solvents where applicable. Spray dried (SD) SBS was prepared from an aqueous solution (2.5 % [w/w]) at an outlet temperature (Tout) of 90 °C and with the spraying air set to 40 mm or 666 L/h. TIO was dried at the same conditions, but with a mass concentration of 2 % [w/w] and at an Tout of 83 °C. BUD was dissolved in methylene chloride (7 % [w/w]) and dried at Tout = 62 °C with the spraying gas set to 35 mm or 538 L/h in combination with the inert loop module and nitrogen (Linde AG, Germany) as spray gas. The inert loop was also applied for the generation of FOR particles, as those were gained from a methanolic solution (4.8 % [w/w]) at a Tout of 61 °C and a spraying gas of 33 mm or 498 L/h. Particle sizes were determined according to mannitol carrier particles, but with a dispersing pressure of 4.0 bar. ®

Drug (1 % [w/w]) and mannitol carriers (99 % [w/w]) were blended with a Turbula blender (Willy A. Bachofen AG Maschinenfabrik, Switzerland) for 3 x 15 min to gain interactive powder blends. All components were initially sieved with a 355 μm sieve to eliminate agglomerates. Components were weighed into stainless steel vessels using the double sandwich method. The homogeneity of all powder blends was tested from ten randomly drawn samples by drug quantification via high performance liquid chromatography (HPLC) targeting a relative standard deviation below 5 % and a minimum recovery of 90 % of the target drug content. Aerodynamic characterisation was performed with the Next Generation Pharmaceutical Impactor (NGI, Copley Scientific, United Kingdom). Flow rate was set to 78.2 L/min as impaction experiments were conducted with the ® Novolizer (Meda Pharma, Germany) as a commercial device. The reservoir was filled with 1.0 g of the respective powder blend. All cups of the NGI including the preseparator were coated with a stage coating (Ethanol, Brij 35, Glycerin) to prevent re-entrainment of particles. Impaction experiments were performed allowing 4.0 L of air to ® pass through the device. The Novolizer was primed by discharging the first twenty doses into an external tube. The following ten doses were released to the impactor. Drug particles were then dissolved in appropriate ® solvents. The contents all drugs were analysed by HPLC using a RP-18 column (LiChrospher 100, RP18 (5 μm), Merck KGaA, Germany). Data analysis was performed by using the Copley Inhaler Testing Data Analysis Software (Copley Scientific, United Kingdom). The FPF is shown as the emitted fraction of the appropriate emitted drug amount with an aerodynamic diameter of < 5 μm. All blends were analysed in triplicate. Results & Discussion The chosen carrier batches had particle sizes ranging from 51.5 µm to 76.8 µm and various particle shapes that were categorised to shape categories from 1.4 to 4.9 as summarised in Table 1. The shape was found to change based on the outlet temperature during the drying process, while particle size was mostly a factor of the rotation [4] speed of the rotary atomiser . The six chosen mannitol batches were blended with four drug batches ranging from 1.8 µm to 2.4 µm in size as listed in Table 1 to generate various interactive powder blends. Table 1 – particle characteristics of mannitol carrier particles and drug batches – mannitol batches were labelled according to outlet temperature (Tout) and size (S/M/L). Drug batches were not evaluated regarding their shape as all of them appeared mostly spherical.

Mannitol Carrier Particles

Outlet Temperature, Tout, °C Particle Size, d50, µm Particle Shape, Cat. 1-5

Drug Batches

M70(S)

M71(L)

M74(M)

M80(S)

M80(M)

M97(L)

SBS SD

TIO SD

BUD SD

FOR SD

70

71

74

80

80

97

90

82

62

61

51.5

68.1

57.5

53.7

60.8

76.8

2.4

2.2

1.8

2.1

1.7

1.4

2.5

4.3

3.7

4.9

spherical

Exemplarily, Figure 1 gives SEM images of FOR SD blended with mostly spherical carriers (Figure 1A) and deeply indented carriers (Figure 1B) as visualised after blending. The powder blends were then characterised with focus on their aerodynamic behaviour during impaction analysis.

147


Drug Delivery to the Lungs 27, 2016 - Mathias Mönckedieck et al.

Figure 1 – SEM images of A: M74(M) and FOR SD; B: M97(L) and FOR SD prior to impaction analysis

The FPF dealt as the main parameter to describe effects on drug dispersion during inhalation. Figure 2 displays the FPF plotted against carrier particle shape for powder blends with SBS SD, TIO SD, BUD SD and FOR SD. The FPF of indented carriers (high shape categories, right from dashed line in Figure 2) exhibited the same pattern for all drugs with M80(M) and M97(L) leading to reduced inhalable fractions compared to M80(S), which can be attributed to the indention depth of differently sized particles as small carriers comprise indentions of lower depth so that drugs cannot hide from shear forces during inhalation. It was further observed that spherical carriers (with low shape categories, left from dashed line in Figure 2) revealed improved dispersion in the case of SBS SD compared to indented carrier particles, while TIO SD and BUD SD exhibited the reverse trend with enhanced detachment from slightly indented carriers compared to spherical or deeply indented ones. FOR SD particles were detached best when large spherical (M71(L)) or deeply indented carriers (M97(L)) were used in the respective powder blends, while slightly indented ones exhibited reduced FPFs.

Figure 2 – Fine Particle Fraction (FPF, % ± standard deviation) for SBS SD and TIO SD, BUD SD and FOR SD plotted against the carrier particle shape (Category 1 (spherical) to 5 (indented)) of six different mannitol carrier batches. Labels depict the outlet temperature in °C during the spray drying process and suggest the size of the carriers (S/M/L) (n=3). Dashed lines separate the FPFs of spherical carriers and indented carriers.

Those observations can be attributed to the four blend qualities schematically depicted in Figure 3. The figure summarises how well different drugs got dispersed from different carrier qualities as it connects the results illustrated in Figure 2 with the evaluation of SEM images. The four different qualities assumed in Figure 3 were found for all drugs used in this project. (+) indicates that the respective carrier quality led to a good performance during impaction analysis for the respective drug, while (-) suggests lowered FPFs. It was observed that drugs were evenly spread over the surface of large spherical carrier particles, while drug bridges consisting of drug agglomerates were found for smaller carriers. It was assumed that cohesion and adhesion forces of the drugs overcome the weight forces of smaller carriers to enable those drug bridges (Figure 1A). Slightly indented carriers supported the generation of drug agglomerates in their cavities as observed in SEM images, while large indentions provided so much space for the entrapment of drugs that almost all drug particles were found within these large cavities.

148


Drug Delivery to the Lungs 27, 2016 - Detachment of various spray dried drugs from engineered mannitol carrier particles

The four cases are discussed with respect to the respective drug substance in the following section: SBS agglomerates in slight indentions or from drug bridges were dispersed easily due to low drug-to-drug interactions resulting in adequate FPFs, while the drug was tightly attached to the surface of deeper indentions and based on intense carrier-to-drug interactions. Shear forces were strong enough to overcome those interactions for particles or agglomerates not hidden in any indentions, but were too low for SBS in slight or deep indentions. The detachment of TIO SD and BUD SD might be governed by press-on forces during blending. The energy applied by the collisions of carrier particles during blend preparation was able to increase drug-carrier interactions and with this to lower the FPF for mostly spherical carrier particles. Improved dispersion from slightly indented carriers might occur as indentions shelter drug particles from press-on forces and drug-drug agglomerates were dispersed easily. Deep indentions, however, were found to be disadvantageous due to drug particle entrapment since shear forces during inhalation were not sufficient to overcome drug-to-carrier interactions. The dispersion of FOR SD from mannitol carriers was found to be best whenever build-up of drug agglomerates was prevented. This pertained mainly for large spherical carriers, where drug particles were evenly distributed over the whole surface or for deeply indented ones, where the volume of the indentions was large enough to provide space for evenly distributed single drug particles or small agglomerates. Presence of drug-drug agglomerates are detrimental for FOR dispersion. Particle-particle interactions between hydrophobic FOR and hydrophilic carrier surface appear to be weaker than for hydrophilic SBS and carrier resulting in facilitated dispersion.

Figure 3 – Distribution of drug particles on the surface of differently shaped carriers upon blending and dispersion of those drug particles during inhalation (carrier size: S = small / L = large). (+) and (-) indicate how well particles get dispersed from the carriers of different particle shape and size.

Conclusion The FPF of blends with spray dried mannitol carrier particles that were different in size and shape and various spray dried spherical drug particles was found to be based on the carrier properties particle shape and size, but also on the appearance of drug agglomerates as those might further influence dispersion from the carrier. It was observed that different drugs occur with different drug-to-carrier interactions, but also with divergent drug-to-drug interactions resulting in agglomerates of different strength and different aerodynamic performance. Acknowledgements The authors would like to thank Roquette Fréres for providing mannitol for the spray drying of mannitol carrier particles and the DFG for funding this project in the framework of SPP-1423. References 1

H. Steckel, N. Bolzen, Alternative sugars as potential carriers for dry powder inhalations, International Journal of Pharmaceutics 270 (2004), pp 297–306.

2

T. Schröder, P. Walzel, Design of Laminar Operating Rotary Atomizers under Consideration of the Detachment Geometry, Chemical Engineering and Technology 21 (1998) pp 349–354.

3

X. Zeng, G.P. Martin, C. Marriott, J. Pritchard, Lactose as a carrier in dry powder formulations: The influence of surface characteristics on drug delivery, Journal of pharmaceutical sciences 90 (2001) pp 1424–1434.

4

M. Mönckedieck, J. Kamplade, E.M. Littringer et. al, Spray drying tailored mannitol carrier particles for dry powder inhalation with differently shaped active pharmaceutical ingredients, in: U. Fritsching (Ed.), Process spray – functional particles produced in spray processes, Springer, 2016

149


Drug Delivery to the Lungs 27, 2016 - Zsófia Edit Pápay et al. Development and characterization of miconazole-loaded lipid nanoparticles against pulmonary mycoses 1

Zsófia Edit Pápay , Eszter Gulyás, Petra Füredi & István Antal Semmelweis University, Hőgyes Endre Street 7, Budapest, 1092, Hungary Summary The responsible pathogens for pulmonary mycoses can be commonly found in the environment, mainly Aspergillus and Candida fungal strains. The development of this severe condition is associated with chronic lung diseases and immunodeficiency. Azoles such as miconazole, is one of the first-line treatment with broad antifungal activity, however, it belongs to the II. group of Biopharmaceutical Classification System (BCS) with low water solubility. Therefore there is a need for development of an effective delivery tool for the local administration of this antifungal agent. Among colloidal carrier systems, solid lipid nanoparticles (SLN) have outstanding properties in terms of small particles size, high drug encapsulation capacity and tolerability. Therefore the aim of ® this study was to develop and characterize miconazole-loaded SLN against lung mycoses. Witepsol 35 was used as a lipid phase. For five high pressure homogenization cycles at 600 bar prove to be a suitable method for developing lipid nanoparticles with the optimal particle size (182.0 ± 1.9 nm), polydispersity index (0.297 ± 0.03) and as high encapsulation efficiency as 99%. The encapsulation of the miconazole was further verified by DSC thermograms and FTIR spectra. The antifungal study proved that miconazole maintained its effectiveness against A. flavus and C. glabra fungus strains, the clear zones of inhibition could be well observed. This study provides evidence that SLN are suitable carriers of the poorly water soluble miconazole and represents a promising delivery system against pulmonary mycoses with potential broad antifungal activity. Introduction Pulmonary mycoses are severe conditions, mainly associated with chronic lung diseases and immunodeficiency. The causing pathogens are usually Aspergillus and Candida species which can be commonly found in the environment. Those patients who dealing with asthma, COPD and diabetes are at higher risk and the diagnosis of [1] fungal infections of the lung is often delayed or missed . Azoles such as miconazole is one of the first-line [1, 2] . Miconazole has a broad spectrum of antifungal activity and it was proved to treatment besides amphotericin B [3] be effective against Aspergillus and Candida spp. as well . However, it has low water solubility therefore it [4] belongs to the BCS II. group (Biopharmaceutical Classification System) . Generally, oral or intravenous administration of antifungal agents are applied, however, there is a need for local administration in order to minimize the unpleasant systemic side effects. Solid lipid nanoparticles (SLN) are able to improve the water solubility and bioavailability of BCS II. drugs with outstanding properties in terms of particles size and high drug [5] encapsulation capacity . Moreover, prolonged drug release with high tolerability and low toxicity could be [6] achieved with the pulmonary administration of SLN . Therefore SLN is a good candidate to encapsulate miconazole and to increase the effectiveness. Furthermore, particles with smaller size than 260 nm are able to [7] escape from the phagocytosis of the macrophages . Therefore the aim of this study was to optimize the production of miconazole-loaded SLN with small particle size (below 200 nm) and high drug loading capacity against pulmonary mycoses. Materials and Methods Materials Miconazole was purchased from BrightGene Bio-Medical Technology Co. Ltd, (China). L-α-Phosphatidylcholine (PPC) and sterile discs (Whatman No.1) were purchased from Sigma-Aldrich Ltd., (Germany). Ethanol 96%, ® ® Witepsol 35 (W35), Tween 80 (Polysorbate 80) and stearic acid were obtained from Molar Chemical Ltd., (Hungary). All other reagents used in this study were of analitical grade and water was purified with Milli-Q (Merck ® Ltd., USA). Amicon Ultarcentrifugal filter devices with regenerated cellulose membrane (molecular weight cut off 30 kDa) were purchased from Millipore Ireland Ltd., (Ireland). Methods Preparation of lipid nanoparticles Preparation of miconazole loaded solid lipid nanoparticles (MCZ-SLN) was conducted by high pressure homogenization (HPH) method. Briefly, the miconazole (0.050, 0.100 or 0.150 g) was dissolved completely in ® Witepsol 35 (0.25, 0.50 or 0.75 g, ~45 °C) and the mixture of surface active agents (0.5 g ethanol, 0.0075 g ® Tween 80 and 0.03 g PPC). Later 10 ml of purified water (70 °C) was added under constant stirring (magnetic stirrer). These mixtures were prehomogenized for 3 minutes by Ultra-Turrax homogenizer (IKA® Works Inc., Germany) at 20,000 rpm and then further homogenized for 3, 4 or 5 cycles by an Avestin Emulsiflex B15 HPH (Avestin Europe GmbH, Germany) utilizing at 600 bar (26,106 psi) inlet pressure. The obtained MCZ-SLN nanoparticles were characterized in order to determine the effect of the number of homogenization cycles and amount of lipids on the particle size, polydispersity index (PDI) and entrapment efficiency (EE). The aim was to minimize the particle size, PDI and maximize the entrapment efficiency.

150


Drug Delivery to the Lungs 27, 2016 - Development and characterization of miconazole-loaded lipid nanoparticles against pulmonary mycoses Particle size measurement The mean particle size and the PDI of the samples were obtained by using Malvern Zetasizer Nano Zs (Malvern Instruments Ltd, UK). The equilibration time was 120 sec respectively, with PCS115 glass cuvette cell. All measurements were performed in triplicate (n=3) at 25 °C and presented as mean ± standard deviation (SD). Investigation of encapsulation efficiency (EE) The concentration of entrapped miconazole was determined indirectly by using ultrafiltration technique. The ® aliquot (0.5 ml) of undiluted samples were placed into an Amicon Ultarcentrifugal filter device and centrifuged for 10 minutes at 12000 g. The filtered aqueous phase contained the unencapsulated miconazol and its concentration was determined by HPLC (Agilent 1100, Agilent Technologies, USA) equipped with UV detector. Agilent C18 column (4.6 mm × 150 mm, 5 μm) was applied as a stationary phase and 10 µl of the samples were injected. The mobile phase consisted of 70% acetonitrile and 30% of 0.05 M sodium dihydrogen phosphate buffer and run isocratically at room temperature at a flow rate of 0.7 ml/min. Miconazole was detected by UV detector at 254 nm. The EE was calculated according to Eq.1. EE (%) = [(Total drug content- Amount of free drug)/ Total drug content] x 100

(1)

Differential scanning calorimetry (DSC) DSC analyses were performed on MCZ, empty SLN, physical mixture (miconazole and empty SLN) and MCZSLN. Accurately weighted samples (2 mg) were placed into aluminum pans and equilibrated to 25°C. The pans were then heated at a rate of 10°C/min in a range of 10–150 °C by using an Exstar 6000 DSC (Seiko Instruments Inc., Japan). An empty aluminum pan was used as reference. All measurements were performed in triplicates. Fourier transform infrared spectroscopy (FTIR) The FTIR spectra of MCZ, empty SLN, physical mixture (miconazole and empty SLN) and MCZ-SLN were -1 recorded by FTIR spectrometer (Alpha Bruker Corp., USA) in the range of 4000–400 cm . Antifungal study The antifungal activity of the miconazole-loaded lipid nanoparticles was investigated by paper disc diffusion method. Initially, a sterile agar medium was inoculated with the fungal suspensions of Candida glabrata and Aspergillus flavus strains (malt yeast agar media for C. glabrata, malt extract agar media for A. flavus). Then, sterile discs (5 mm in diameter) were impregnated with 2 µl SLN dispersions and placed into the Petri dishes. All the plates were incubated at 40°C up to 10 days due to development of the fungi. The zones of growth inhibition around the disc were measured after the specified period of incubation and empty SLN were used as negative control. The clear zone of inhibition was analysed in mm. Results Optimization of MCZ-SLN The aim of the preliminary experiments was to investigate the influence of HPH method on nanoparticle formation. It was found that 5 homogenization cycles resulted the lowest particle sizes and PDI values (Fig.1.). Therefore all of the miconazole loaded samples were homogenized 5 times. Table 1. summarizes the correlation between particle size, PDI and encapsulation efficiency. It can be seen that the particle size is decreased at higher lipid concentrations (0.50 and 0.75) but there was no significant difference between 0.5 g and 0.75 g W35 in terms of particle size. However, the encapsulation efficiency is extremely high (99%) even at the lowes lipid content (0.25). For the further characterization, the SLN 11 formulation was chosen due to the small size (182.0 ± 1.9 nm), low PDI value (0.297 ± 0.03) and high miconazole content (0.100 g, EE: 99.5%).

Figure 1 – The impact of homogenization cycles on particle size and polydispersity index (bars: particle size, line: PDI)

151


Drug Delivery to the Lungs 27, 2016 - Zsófia Edit Pápay et al. Table 1. The effect of lipid and miconazole on the particle size and PDI Sample name SLN 1 SLN 2 SLN 3 SLN 4 SLN 5 SLN 6 SLN 7 SLN 8 SLN 9 SLN 10 SLN 11 SLN 12

®

Witepsol 35 (g) 0.25 0.25 0.25 0.25 0.50 0.50 0.50 0.50 0.75 0.75 0.75 0.75

MCZ (g) 0.50 0.100 0.150 0.50 0.100 0.150 0.50 0.100 0.150

Particle size (nm) 139.9 ± 11.2 309.3 ± 28.5 266.2 ± 16.3 208.8 ± 8.6 178.9 ± 4.5 176.4 ± 2.3 178.1 ± 3.4 178.0 ± 2.7 164.9 ± 13.8 179.8 ± 4.2 182.0 ± 1.9 181.1 ± 3.8

Polydispersity index (PDI) 0.261 ± 0.03 0.482 ± 0.02 0.444 ± 0.06 0.382 ± 0.01 0.266 ± 0.03 0.303 ± 0.08 0.328 ± 0.02 0.313 ± 0.04 0.313 ± 0.02 0.318 ± 0.01 0.297 ± 0.03 0.367 ± 0.01

EE (%) 99.8 99.9 99.9 99.6 99.0 99.8 99.1 99.5 99.3

DSC analysis DSC measurement was performed to study the physical state of the samples. Thermograms of miconazole, empty SLNs, physical mixture and MCZ-SLN (SLN 11) are shown in Fig. 2. A sharp endothermic peak observed at 83°C is the melting point of miconazol and without impurities (Fig. 2. a). The recorded broad peak at 35°C is originated from the heterogenous W35 in the SLN formulation (Fig. 2. b). Thermogram of MCZ-SLN was similar to empty nanoparticles but only the thermal characteristics of W35 could be seen (Fig. 2. c). On the contrary, the endothermic peaks of both are presented in the physical mixture (Fig. 2. d).

Figure 2 and 3 – DSC thermograms and FTIR spectra of miconazole (a), empty lipid nanoparticles (b), physical mixture (c) and miconazole-loaded lipid nanoparticles (d) FTIR analysis FTIR analysis was conducted to measure the changes in chemical structure of miconazole and to detect the possible interactions between the excipients in the formulation. The spectrum of MCZ (Fig.3. a) exhibited -1 -1 characteristic peaks of imidazole C-N stretch at 3140 cm , aromatic C-H stretch at 3070 cm , C=C aromatic -1 -1 -1 vibration at 1566 cm and 1525 cm and C-C stretch at 1070 cm . In the spectrum of empty SLN (Fig. 3. b), the -1 -1 main characteristic peaks of W35 such as C-H stretching band at 2920 cm and 2850 cm of long fatty acid chain -1 -1 and C=O stretching band at 1727 cm and 1740 cm could be mainly observed. PPC also has the characteristic -1 -1 -1 C-H stretching bands of long fatty acid chain at 2920 cm and 2827 cm and C=O stretching band at 1734 cm . -1 -1 + -1 P=O stretching at 1237 cm , P-O-C stretching at 1058 cm and N (CH3)3 stretching at 974 cm could be identified. The spectrum of the physical mixture of empty SLN and miconazole (Fig. 3. c) showed the summation of the vibrational frequencies of the individual components. In the FTIR spectrum of the MCZ-SLN (SLN 11) (Fig. 3. d), the main characteristic peaks of W35 could be well observed and most of the banding vibrations of MCZ are decreased. Antifungal study Antifungal activity of MCZ-SLN (SLN 11) was proved by disk diffusion test. After 10 days clear zones of inhibition in case of MCZ-SLN were observed as shown in Fig. 4. On the contrary, empty SLN did not inhibited the growth of the A. flavus and C. glabra strains. In case of C. glabra, the diameter of inhibition zones was 25 mm and 10 mm for A. flavus.

152


Drug Delivery to the Lungs 27, 2016 - Development and characterization of miconazole-loaded lipid nanoparticles against pulmonary mycoses

Figure 4 – Antifungal activity of empty and miconazole-loaded lipid nanoparticles Discussion Smaller particle size of a colloidal carrier system could ensure low irritation and increased biocompatibility. HPH [5] method is commonly used technique to produce SLN . Generally, the more homogenization cycles applied, the [8] lower PDI could be achieved . In this study, 600 bar homogenization pressure was used for 3-5 homogenization cycles to avoid the excessive temperature increase of the samples. 5 cycles resulted the lowest particle size and PDI values, assuming that the population of nanoparticles tends to a monodisperse system (PDI~0.3). The effects of the concentration of lipid on particle size was investigated as well as the EE (%). To achieve small size and narrow distribution, the optimal proportion of lipid and drug should be used. However, the encapsulation efficiency is 99% even at the lowest lipid concentration applied (0.25 g). For the further characterization, the SLN 11 formulation was chosen due to the small particle size, low PDI value and optimal miconazole content (99.5 mg). Based on the available products in the market (tablet and gel), a single dose would be 100 mg. However, considering the pulmonary administration and the bioavailability improvement by SLNs, further investigations are required to determine the dosage. DSC measurement demonstrated that thermogram of MCZ-SLN was similar to the curve of empty nanoparticles due to the encapsulation of MCZ. This was further supported by the diagram of the physical mixture where the endothermic peak of MCZ appeared. FTIR analysis makes possible the quick and efficient identification of the compounds based on their functional groups and bond vibrations. The spectrum of empty SLN formulation showed the peaks of individual components, mainly W35 and PCC. This suggested that there was no recrystallization and interaction between the excipients during the HPH. The characteristic peaks of the drug could be seen in the physical mixture, as expected. In the spectrum of MCZ-SLN, some characteristic banding vibrations of MCZ could be observed which indicates the presence of drug without any chemical modification. Antifungal study proved that miconazole maintained its effectiveness. Empty SLN formulation was used as control. The lack of clear zone suggested that the excipients have no antifungal activity against these strains, however, the clear zone of inhibition against A. flavus and C. glabra indicated antimicrobial efficacy of MCZ-SLN formulations. The diameter of inhibition zones were different which can be attributed to the diverse fungus strains. Conclusion This study provides evidence that solid lipid nanoparticles are suitable carriers of the poorly water soluble miconazole. W35 is an applicable lipid to form nanoparticles with as high encapsulation efficiency as 99%. The developed MCZ-SLN represents a promising delivery system against pulmonary mycoses with potential broad antifungal activity. The authors intend to conduct further experiments to investigate the aerodynamic properties of the nanoparticles after spray drying. References 1. 2. 3. 4. 5. 6. 7. 8.

Bellmann R, Bellmann-Weiler R, Weiler S: Pulmonary mycoses. Memo 2008;1: pp15-19. Meersseman W, Lagrou K, Maertens J, Wijngaerden E V: Invasive aspergillosis in the intensive care unit. Clin Infecti Dis 2007;45: pp205-216. Schar G, Kayser F H, Dupont M C. Antimicrobial activity of econazole and miconazole in vitro and in experimental candidiasis and aspergillosis. Chemotherapy 1976;22: pp211-220. Tubic-Grozdanis M, Bolger M B, Langguth P: Application of gastrointestinal simulation for extensions for biowaivers of highly permeable compounds. AAPS J 2008;10: pp213-226. Mehnert W, Mäder K: Solid lipid nanoparticles: Production, characterization and applications. Adv Drug Deliv Rev 2001;47: pp165-196. Muller R H, Mader K, Gohla S: Solid lipid nanoparticles (SLN) for controlled drug delivery - a review of the state of the art. Eur J Pharm Biopharm 2000;50: pp161-177. Mansour H M, Rhee Y S, Wu X: Nanomedicine in pulmonary delivery. Int J Nanomedicine. 2009;4: pp299-319. Patravale V B, Ambarkhane A V. Study of solid lipid nanoparticles with respect to particle size distribution and drug loading. Die Pharmazie. 2003;58: pp392-395.

153


Drug Delivery to the Lungs 27, 2016 - Yingshan Qiu et al. Synthetic KL4 peptide as new carrier of siRNA therapeutics for pulmonary delivery Yingshan Qiu, Michael Y.T. Chow & Jenny K.W. Lam Department of Pharmacology & Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, 21 Sassoon Road, Pokfulam, Hong Kong

Summary Small interfering RNA (siRNA) has great potential for the treatment of various respiratory diseases through RNA interference (RNAi), but their clinical application is hindered by the lack of a safe and effective pulmonary delivery system. KL4 peptide is a synthetic amphipathic peptide that was previously developed to mimic the function of pulmonary surfactant protein B (SP-B). Its potential as siRNA carrier for pulmonary delivery was examined in this study. The cationic KL4 peptide was able to bind with siRNA to form complexes at 15:1 ratio (peptide to siRNA weight ratio) or above. It also mediated efficient gene silencing on lung epithelial cells, with 20:1 ratio as the optimal ratio for siRNA transfection. Furthermore, the KL4/siRNA complexes were not toxic at concentrations used for transfection in vitro. The study shows that KL4 peptide appears to be a promising candidate for siRNA delivery. Further investigation on animal study and work on aerosol formulation are required to develop KL4 peptide as siRNA carrier for clinical application. Introduction Small interfering RNA (siRNA) holds great promise as therapeutics to treat many diseases including respiratory diseases, by inhibiting the expression of the disease-causing gene(s) through a post-transcriptional gene 1 silencing mechanism, called RNA interference (RNAi) . One of the major barriers of siRNA therapeutics 2 development is the lack of a safe and effective delivery system suitable for clinical application . siRNA is a negatively charged, hydrophilic macromolecule, it is incapable of crossing the biological membrane unassisted. Moreover, it is extremely susceptible to enzymatic degradation. Therefore, a carrier is often required to facilitate the cellular uptake of nucleic acids as well as to protect the nucleic acids from nuclease degradation. For the treatment of respiratory diseases, the most direct way to deliver nucleic acid therapeutics is by pulmonary delivery, which is non-invasive and easily accepted by patients. It avoids the interaction with serum and the rapid nuclease degradation that occurs in the bloodstream. In addition, it can minimize systemic exposure and thereby 3 reducing systemic adverse effects . Cationic polymers and lipids are commonly used for transfection of nucleic acids. However, they are often associated with toxicity problem. A safe and efficient nucleic acid delivery system for pulmonary delivery remains highly sought after. KL4 peptide is a 21-residue amphipathic peptide containing repeating KLLLL sequences. This synthetic peptide was designed to mimic the overall ratio of cationic to hydrophobic amino acids in native surfactant protein B (SP4, 5 . KL4 peptide exhibits SP-B-like surface activity. It is one of the active components in Surfaxin, a FDA B) approved intratracheal suspension of pulmonary surfactant indicated for the prevention of respiratory distress syndrome in premature infants. In this study, we investigated the potential of KL4 peptide as siRNA carrier for pulmonary delivery. Due to its cationic nature, it is anticipated that the KL4 peptide can form complexes with siRNA and promote cellular uptake. The siRNA binding affinity of KL4 peptide was assessed by gel retardation assay. Its transfection efficiency and cytotoxicity were also evaluated on human lung epithelial cells. Experimental Methods Materials and cell culture – KL4 peptide (KLLLLKLLLLKLLLLKLLLLK) was purchased from ChinaPeptides (Shanghai, China) as 90% purity grade and used as provided. siRNAs (SilencerSelect GAPDH positive control siRNA and SilencerSelect negative control siRNA) were purchased from Ambion (Austin, TX, USA). GelRed™ nucleic acid stain was purchased from Biotium (Hayward, CA, USA). A549 cells (human lung adenocarcinoma epithelial cells) were obtained from ATCC (Manassas, VA, USA). The cells were maintained at 5% CO2, 37° C in DMEM supplemented in 10% fetal calf serum, 100 units/ml penicillin, 100 μg/ml streptomycin and 0.25 μg/ml amphotericin B. The cells were sub-cultured twice weekly. All other reagents were obtained from Sigma Aldrich (Saint Louis, MO, USA) and of analytical grade or better. Gel retardation assay – KL4/siRNA complexes were prepared at various ratios (from 2.5:1 to 30:1 of peptide to siRNA weight ratio) with 0.4 μg siRNA in 10 μl TAE buffer. The samples containing gel loading buffer were loaded into a 2% w/v agarose gel stained with GelRed™. Gel electrophoresis was run in TAE buffer at 100 V for 20 min and the gel was visualised under the UV illumination. siRNA transfection study – A549 cells were transfected with KL4/siRNA complexes containing 100 nM of GAPDH siRNA or negative control siRNA per well in 24-well plates. The complexes were prepared in Opti-MEM I reduced serum medium. After 4 h of incubation at 37 °C, the cells were washed with PBS and fresh DMEM supplemented with 10% FBS were added to the cells. After 72 h, the GAPDH expression was detected by Western blot. The cells were washed and lysed. The cell extracts containing 20 μg of protein were loaded into a 10% SDSpolyacrylamide gel and electrophoresis was run at 120 V for 90 min.

154


Drug Delivery to the Lungs 27, 2016 - Synthetic KL4 peptide as new carrier of siRNA therapeutics for pulmonary delivery After the proteins were resolved, they were transferred into a nitrocellulose membrane which was blocked in 5% non-fat dry milk for 1.5 h with shaking. The membrane was washed and incubated with primary antibody overnight at 4 °C on shaking. After rinsing, the membrane was incubated with horseradish peroxide conjugated secondary antibody for 2 h at room temperature. The bound secondary antibodies were detected with ECL™ Western blotting detection reagents. Densitometry was performed to analyse the protein expression on the Western blot. The experiments were conducted in three independent experiments. Cytotoxicity study – The cytotoxicity of the KL4 peptide was assessed by MTT (3-(4-5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide) assay. A549 cells were seeded in 24-well plates and the KL4/siRNA complexes were prepared as described above. After 4 h of incubation at 37 °C, the cells washed with PBS and fresh DMEM supplemented with 10% FBS were added to the cells. MTT assay was carried out at 4 h and 24 h posttransfection. MTT solution (0.8 mg/ml in PBS) was added into each well. The samples were incubated for 4 h at 37 °C. The MTT solution was removed and isopropanol was added to dissolve the insoluble formazan crystals for at least 15 min. The samples were examined by measuring the absorbance at 595 nm using UV/Vis spectrophotometer. The cell viability was calculated as percentage of the absorbance from cells treated with KL4/siRNA complexes against that obtained from cells treated with OptiMEM-reduced serum medium only. The experiments were conducted in three independent experiments.

Results KL4 peptide was evaluated for its binding affinity to siRNA by gel retardation assay. The disappearance of siRNA band in the gel image indicated that binding occurred between siRNA and KL4 peptide (Figure 1). The gel image shows that as the peptide to siRNA ratio increased, the siRNA band became fainter. siRNA was only partially bound to KL4 peptide at 5:1 and 10:1 ratios (peptide to siRNA weight ratio). At ratios 15:1 and above, the siRNA was completely bound to KL4 peptide as the siRNA band disappeared.

siRNA��2.5:1�����5:1�����10:1�����15:1�����20:1����30:1�

Figure 1 - Gel retardation assay of siRNA binding. KL4/siRNA complexes were prepared at various was ratios (2.5:1 to 30:1 peptide to siRNA weight ratio). Sample containing siRNA only was used as control.

To investigate the siRNA delivery efficiency of KL4 peptide on lung epithelial cells, A549 cells were transfected with KL4/siRNA complexes prepared at different ratios, using siRNA targeting glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and negative control. The cellular expression of GAPDH after transfection was evaluated by Western Blot and analysed by densitometry (Figure 2). At ratio 5:1 and 10:1, only moderate GAPDH knockdown of around 20% was observed. The most efficient gene silencing effect was observed at 20:1 ratio, with over 40% of GAPDH knockdown was detected. The result was comparable to Lipofectamine™2000, a lipidbased commercial transfection agent. As the ratios further increased, the gene silencing effect gradually subsided. The cytotoxicity of KL4 peptide was examined by MTT assay on A549 cells at 4 h and 24 h post-transfection, and the cell viability was calculated (Figure 3). There was no sign of cytotoxicity with KL4/siRNA complexes prepared at ratio up to 30:1. The cell viability was maintained over 90% for all the samples tested. Since the transfection efficiency of the complexes reduced at ratio above 20:1, the cytotoxicity of complexes formed at ratio above 30:1 was not investigated in this study.

155


Drug Delivery to the Lungs 27, 2016 - Yingshan Qiu et al.

A

50:1�

β-ac n� � GAPDH�

40:1�

30:1�

10:1�

20:1�

-������+� -������+� -������+�

-������+� -������+�

5:1�

-������+�

1.4

B Relative GAPDH expression

1.2 1.0 0.8 0.6 0.4 0.2

:1 50

:1 40

:1 30

:1 20

:1 10

5: 1

lip

of ec

ta m

in e

20 0

0

0.0

Figure 2 - GAPDH siRNA transfection on A549 cells analysed by Western blotting. (A) Cells were transfected with KL4 /siRNA complexes prepared at various ratios (5:1 to 50:1 peptide to siRNA weight ratio) using GAPDH siRNA (+) or negative control siRNA (-). Protein analysis was carried out at 72 h post-transfection. β-actin served as internal control for equal total protein loading. (B) The bands were analysed by densitometry with the density of GAPDH band normalised to that of the β-actin band of the corresponding sample. Lipofectamine™2000 (a lipid-based commercial transfection agent) was used as a control for comparison.

B 120

100

100

60 40

ol tr on C

tr C

on

30 :1

0 20 :1

0 10 :1

20

5: 1

20

30 :1

40

80

20 :1

60

10 :1

80

5: 1

cell viability %

120

ol

cell viability %

A

Figure 3 - Cytotoxicity of KL4 peptide was determined by MTT assay. A549 cells were transfected with KL4/siRNA complexes prepared at various ratios (5:1 to 30:1 peptide to siRNA weight ratio). The cell viability was evaluated at (A) 4 h and (B) 24 h post-transfection.

Discussion In this study, synthetic KL4 peptide was investigated as siRNA carrier for pulmonary delivery. KL4 peptide was originally designed to mimic SP-B to reduce alveolar surface tension. It is currently used clinically as one of the active components in pulmonary surfactant for intratracheal administration. Due to its cationic amphipathic nature, it is hypothesized that the KL4 peptide can bind with siRNA to form complexes through electrostatic interaction and promote cellular uptake of siRNA. First, the siRNA binding affinity of KL4 peptide was evaluated. It was found that complete binding with siRNA was achieved at 15:1 peptide to siRNA ratio.

156


Drug Delivery to the Lungs 27, 2016 - Synthetic KL4 peptide as new carrier of siRNA therapeutics for pulmonary delivery

The siRNA transfection efficiency was assessed on A549 cells, which are human lung epithelial cell line. The gene silencing effect was relatively low at ratios 10:1 or below. This was expected as the gel retardation assay revealed that binding with siRNA was not complete at these ratios. At high peptide to siRNA ratio (above 30:1), the gene silencing effect was also poor. This was probably due to the strong binding between KL4 peptide and siRNA at high ratios, leading to incomplete release of siRNA following cellular uptake. As a result, free siRNA was not readily available in the cytoplasm to initiate RNAi. The optimal ratio for siRNA transfection was found to be 20:1. To develop a delivery system for clinical application, the delivery agent must be non-toxic. KL4 peptide is clinically approved to be used for pulmonary administration. However, its current application is to lower the alveolar surface tension at the cell surface. It is not expected to be absorbed into the body. Hence, it is important to examine the cytotoxicity of the complexes formed between KL4 and siRNA. The MTT assay showed that KL4 peptide was not toxic after it bound and formed complexes with siRNA. Complexes with ratio above 30:1 was not tested as their transfection efficiency was not satisfactory.

Conclusions Overall, we demonstrated that KL4 peptide could mediate gene silencing effect of siRNA in lung epithelial cells. The peptide was not cytotoxic at concentrations used for transfection. Our result shows that KL4 peptide appears to be promising candidate as new siRNA carrier for pulmonary delivery. Further studies will be carried out to examine the siRNA transfection efficiency and safety profile of KL4 peptide in animal model following pulmonary administration. In addition, the formulation of KL4 peptide/ siRNA delivery system as powder aerosol will be investigated.

References 1.

Lam J K, Chow M Y, Zhang Y, Leung S W. siRNA Versus miRNA as Therapeutics for Gene Silencing, Mol Ther Nucl Acids. 2015; 4: e252.

2.

Ruigrok M J, Frijlink H W, Hinrichs W L. Pulmonary administration of small interfering RNA: The route to go? J Control Release, 2016; 235: pp14-23.

3.

Lam J K, Liang W, Chan H K. Pulmonary delivery of therapeutic siRNA, Adv Drug Deliver Rev 2012; 64: pp1-15.

4.

Cochrane C G, Revak S D. Pulmonary surfactant protein B (SP-B): structure-function relationships, Science, 1991; 254: pp 566-568.

5.

Mansour H M, Damodaran S, Zografi G. Characterization of the in situ structural and interfacial properties of the cationic hydrophobic heteropolypeptide, KL4, in lung surfactant bilayer and monolayer models at the air-water interface: implications for pulmonary surfactant delivery, Mol Pharm, 2008; 5: pp681-695.

157


Drug Delivery to the Lungs 27, 2016 – Flávia Musacchio Dextran Sulfate Microparticles Encapsulating Isoniazid and/or Rifabutin as Carriers for Pulmonary Tuberculosis Therapy Flávia Musacchio and Ana Grenha Center for Biomedical Research (CBMR), Faculty of Sciences and Technology, University of Algarve, Faro, 8005139, Portugal Center for Marine Sciences (CCMar), University of Algarve, Faro, 8005-139, Portugal Summary Tuberculosis (TB) is an active deadly pathology with high prevalence worldwide, occurring upon infection with Mycobacterium tuberculosis (MTB). Although conventional oral therapy is effective, it is associated with severe side-effects. The need to overcome these effects, which often lead to patient incompliance, demands alternative therapies, either by using new drugs or exploring new routes of administration of traditional drugs. The latter approach has been gaining popularity, as it could permit the reduction of used doses and the frequency of administration. Using the lung route would provide direct administration of antibiotics to the alveoli, where macrophages hosting the bacillus are located, which potentially comprises a successful approach. This study proposes the establishment of inhalable therapy, using dextran sulfate microparticles as carries of antitubercular drugs. This polysaccharide has structural residues/groups reported to be recognised in a preferential manner by alveolar macrophages, which increases the potential of the system in tuberculosis treatment. Microparticles were produced by spray-drying and characterized for morphology (SEM), Feret’s diameter, bulk and tap densities. The theoretical aerodynamic diameter (daer) was further calculated. The cytotoxic evaluation was performed in alveolar epithelial cells (A549) by the MTT assay. Microparticles were obtained with sizes varying within 1.1 and 1.9 μm, which potentiates phagocytosis. Theoretical daer between 1 and 1.4 μm suggests the adequacy of the carriers for the purpose of reaching the alveolar zone. The cytotoxic evaluation evidenced absence of toxicity. The generality of results provided encouraging indications to continue studying the potential of dextran sulfate microparticles as inhalable tuberculosis therapy. Introduction Tuberculosis is an infectious disease caused by Mycobacterium tuberculosis. Being transmitted by inhalation, it mostly affects the lungs, where the bacillus is hosted, multiplying inside alveolar macrophages. The mycobacterium can persist for a long period of time in the human body without symptoms, creating deposits from which the disease can actively develop during lifetime (1). The World Health Organization (WHO) has reported 9.6 million new cases in 2015 and it is known that HIV patients are more susceptible to develop the disease. Despite the existence of effective antitubercular therapy, this is still one of the most lethal diseases worldwide (2). Conventional treatment of TB involves oral antibiotic therapy with four drugs (isoniazid, rifampicin, ethambutol and pyrazinamide) for at least six months. However, along with a frequent patient noncompliance to the therapy, the appearance of multidrug-resistant TB demands alternative therapeutic approaches, either by proposing new drugs or by exploring other options (i.e. alternative delivery routes) for existing drugs. The current major goal is to develop a therapy that permits decreasing the severe side-effects, as well as the administration frequency and dose (3,4)]. In this context, direct delivery of antitubercular drugs to the alveoli entails a potential approach, because this is the location of macrophages hosting the bacteria, permitting a co-localisation of drugs and pathogens. This demands designing carries with suitable aerodynamic properties to reach the alveolar zone (aerodynamic diameter of 1-3 µm), which further enable macrophage capture, reported to be maximal for particles of 1-2 µm (5). This study proposes an alternative spray-dried inhalable formulation for antitubercular therapy, using the natural polymer dextran sulfate (DS) as microparticle matrix material and associating isoniazid (INH) and/or rifabutin (RFB), two first-line antitubercular drugs. The formulations with individual drugs may be used as add-on therapy, along with oral antibiotherapy (6), while the formulation co-encapsulating both drugs is expected to stand alone. Experimental Methods 1.

Preparation of Dextran Sulfate Microparticles by Spray-Drying

A solution of 2% (w/v) DS (Mw 8000 Da, Alfa Aesar, Germany) was prepared in purified water (room temperature), under magnetic stirring for 5 minutes. Whenever necessary, INH (Sigma-Aldrich, Germany) was dissolved separately in purified water (room temperature), while RFB (Chemos, Germany) was dissolved in absolute ethanol (VWR, Chemicals, France). After dissolution and immediately prior to spray-drying, the drugs were mixed with DS solutions at the following mass ratios: DS/INH = 10/1, DS/RFB = 10/0.2 and DS/INH/RFB = 10/1/0.2. These denominations were adopted to name microparticle formulations. When RFB is present in the spraying solution, the solvent is 80/20 (v/v) hydroalcoholic solution. Microparticles were produced using a laboratory scale spray-dryer equipped with a high performance cyclone (Büchi B-290 mini spray-dryer, Buchi Labortechnik, AG, Switzerland) at the following parameters: inlet temperature 115 ºC; aspirator setting 65%; feed speed 2.7 ± 0.1 mL/min; and spray flow rate 601 L/ h.

158


Drug Delivery to the Lungs 27, 2016 - Dextran Sulfate Microparticles Encapsulating Isoniazid and/or Rifabutin as Carriers for Pulmonary Tuberculosis Therapy 2.

Characterization of Microparticles

The morphological observation of microparticles was performed by field emission scanning electron microscopy (FESEM Ultra Plus, Zeiss, Germany). Dry powders were placed onto metal plates and 5 nm thick iridium film was sputter-coated (model Q150T S/E/ES, Quorum Technologies, Lewes, UK) on the samples before viewing. Microparticle size was estimated as the Feret´s diameter and was directly determined by optical microscopy (TR500, VWR Internation, Belgium) from the measurement of 300 microparticles (n = 3). Bulk and tap densities were determined using a tap density tester (Densipro 250410, Deyman, Spain) by measuring the volume of a known weight of powder before and after tapping, respectively (n = 3). The theoretical aerodynamic diameter (daer) was determined for each formulation using the Feret´s diameter (d) and tap density (ρtap) (7,8) using the equation below: 3

, where ρo is 1 g/cm and X is the shape factor. 3.

Preliminary determination of in vitro cytotoxicity

The determination of cytotoxic profile of DS-based microparticles was performed in A549 cells (human alveolar epithelial cell line) by the 3-(4 5-dimethylthiazol-2-yl)-2 5-diphenyltetrazolium bromide (MTT; Sigma-Aldrich) assay. Four microparticle formulations were tested: unloaded DS, DS/INH (10/1, w/w), DS/RFB (10/0.2, w/w) and DS/INH/RFB (10/1/0.2, w/w). The cells were exposed to microparticles at the concentrations of 0.1, 0.5 and 1.0 mg/mL, for 24 h. Cells were seeded in 96-well plates and allowed 24 h adherence, after which suspensions of the microparticles in cell culture medium were added to each well. After the exposure time, MTT solution (0.5 mg/mL in PBS, pH 7.4) was added to the plates and incubated for 2 h. After this, formazan crystals were dissolved with dimethyl sulfoxide (DMSO, VWR,Chemicals, France) and the absorbance measured by spectrophotometry (Infinite M200, Tecan, Austria) at 540 nm (background correction at 650 nm). A solution of 2% (w/v) sodium dodecyl sulfate (SDS, Panreac, Germany) and Dulbecco´s modified Eagle´s medium (DMEM, Lonza, Belgium) were used as positive and negative control of cell death, respectively. The formula to calculate cell viability was: Cell viability (%) = (A - S)/ (CM – S) x 100 Where A represents absorbance of tested formulation, while S and CM represent the measurement for SDS and DMEM, respectively. Results and Discussion Dextran sulfate easily dissolved in purified water, producing a translucent aqueous solution. The same aspect was obtained upon dissolution of INH, but the presence of RFB attributed the solution a red translucent colour. In line with this, RFB-containing microparticles were pink-coloured, contrasting to white microparticles of DS or DS/INH. All dry powders were obtained with high yields (73% – 83%), to which the use of a high performance cyclone that is included in the spray-dryer contributes strongly. The viewing of microparticles by SEM provided indications on morphology. As observed in Figure 1, unloaded DS microparticles have a tendency to exhibit a spherical shape, which becomes irregular with the incorporation of INH and RFB. DS/INH microparticles exhibited a spherical shape with little irregularities, while DS/RFB microparticles displayed morphology comparable to that of DS/INH/RFB microparticles (data not shown).

b

a

5µm

5µm

Figure 1. Microphotographs of DS-based microparticles viewed by scanning electron microscopy (SEM): a) unloaded DS microparticles and b) DS/INH/RFB microparticles. Several parameters considered to affect aerodynamic properties and dry powder performance were characterized in each microparticle formulation, including Feret´s diameter, physical morphology, bulk and tap density. The two properties most affecting deposition in the airways are the size and density of the particles (9,10). The Feret´s diameter of the produced microparticles varied between 1.1 and 1.9 µm (Table 1), which is indicated as adequate for macrophage uptake, in line with reported observations of favoured uptake for particles of 1-2 µm

159


Drug Delivery to the Lungs 27, 2016 – Flávia Musacchio 3

(5). Tap and bulk densities were around 0.2 – 0.6 g/cm , being observed that tap densities increase comparing with bulk densities, as expected from the proper method of determination. Theoretical aerodynamic diameters were calculated to range between 0.96 and 1.48 µm. The highest value corresponds to unloaded microparticles, while the lowest belongs to DS/RFB microparticles. As observed in SEM, the inclusion of RFB in microparticles provided irregular shapes, thus decreasing aerodynamic diameters, but the effect is somewhat balanced with the simultaneous inclusion of INH. Table 1. Yield of spray drying, Feret´s diameter, bulk and tap density and aerodynamic diameter of dextran sulfate microparticles (mean ± SD, n = 3) Yield (%)

Feret´s diameter (µm)

Bulk density 3 (g/cm )

Tap density 3 (g/cm )

Aerodynamic diameter (µm)

Unloaded

82.2 ± 4.2

1.91 ± 1.17

0.46 ± 0.01

0.61 ± 0.01

1.48 ± 0.02

#

DS/INH (10/1)

73.4 ± 2.8

1.69 ± 0.92

0.39 ± 0.01

0.56 ± 0.01

1.26 ± 0.01

#

DS/RFB (10/0.2)

82.8 ± 0.1

1.13 ± 0.77

0.24 ± 0.01

0.36 ± 0.02

0.96 ± 0.03*

DS/INH/RFB (10/1/0.2)

80.1 ± 4.3

1.57 ± 0.80

0.24 ± 0.01

0.41 ± 0.01

1.42 ± 0.02*

Microparticles

INH: Isoniazid; DS: Dextran Sulfate; RFB: Rifabutin; #: shape factor of 1; *: shape factor of 2

The results of cytotoxic evaluation in epithelial cell line, A-549, are depicted in Figure 2. The formulation was tested in concentrations ranging within 0.1 and 1.0 mg/mL, the latter being considered well above the real concentration that the pulmonary delivery of the formulations may induce (11). It is observed a general absence of toxicity towards alveolar epithelial cells, as all microparticle formulations registered cell viabilities above 76%, independently of the associated drug and tested concentration. The exposure to a solution of the polymer also resulted in similar observations. The assumption of absence of toxicity from this assay is based on cell viabilities above 70%, the threshold above which a cytotoxic effect is considered to occur according to ISO 10993-5 (12). As formulations are expected to reach the alveolar zone upon inhalation, testing cytotoxicity in an alveolar cell line was deemed adequate. However, studies in the target cells (macrophages) are also being performed. 140

Cell Viability (%)

120 100

DS Polymer

80

Unloaded MP

60

DS/INH MP

40

DS/RFB MP

20

DS/INH/RFB MP

0 0.1

0.5

1.0

Concentration (mg/mL) Figure 2. A549 cell viabilities after 24 h of exposure to DS-based microparticles. Results are expressed as mean ± SEM (n = 3, six experiments at each concentration).

Conclusion Microparticles based on DS and loaded with first-line antitubercular drugs, either alone or in combination, were efficiently produced by spray-drying. Th ch ct is p op ti s (F t’s i m t , nsity, theoretical aerodynamic diameter) gave positive indications regarding the potential of the formulations to reach the alveolar zone. Additionally, the exhibited size is deemed theoretically adequate for macrophage uptake. Nevertheless, addressing the aerosolization properties (emitted dose, fine particle fraction, aerodynamic diameter) of developed microparticles is expected in the near future. General absence of toxicity was observed in a metabolic assay performed in alveolar epithelial cells, but the need to perform other cytotoxicity studies and to use other relevant cell lines (i.e. macrophages) is identified. The determination of encapsulation efficacy and drug loading are currently being determined.

160


Drug Delivery to the Lungs 27, 2016 - Dextran Sulfate Microparticles Encapsulating Isoniazid and/or Rifabutin as Carriers for Pulmonary Tuberculosis Therapy Acknowledgements This study was supported by National funding from the Portuguese Foundation for Science and Technology through projects (PTDC/DTP-FTO/0094/2012, UID/BIM/04773/2013, UID/Multi/04326/2013. The PhD scholarship to Flavia Musacchio (SFRH/BD/88516/2012) is also acknowledged.

References 1.

Andersen P, Doherty TM, Pai M, Weldingh K. The prognosis of latent tuberculosis: can disease be predicted? Trends Mol Med. 2007;13(5):175–82.

2.

Global Tuberculosis Report 2015 [Internet]. Word Health Organization. 2015. p. 1–204. Available from: www.who.int/tb/data

3.

Sosnik A, Carcaboso AM, Glisoni RJ, Moretton MA, Chiappetta DA. New old challenges in tuberculosis: Potentially effective nanotechnologies in drug delivery. Adv Drug Deliv Rev [Internet]. 2010 Mar 18 [cited 2016 Oct 3];62(4–5):547–59. Available from: http://www.sciencedirect.com/science/article/pii/S0169409X09003640

4.

Smola M, Vandamme T, Sokolowski A. Nanocarriers as pulmonary drug delivery systems to treat and to diagnose respiratory and non respiratory diseases. Int J Nanomedicine. 2008;3(1):1–19.

5.

Ahsan F, Rivas IP, Khan M a., Torres Suárez AI. Targeting to macrophages: role of physicochemical properties of particulate carriers - liposomes and microspheres - on the phagocytosis by macrophages. J Control Release. 2002;79(1–3):29–40.

6.

Hoppentocht M, Hagedoorn P, Frijlink HW, De Boer AH. Developments and strategies for inhaled antibiotic drugs in tuberculosis therapy: A critical evaluation. Eur J Pharm Biopharm [Internet]. 2014;86(1):23–30. Available from: http://dx.doi.org/10.1016/j.ejpb.2013.10.019

7.

Anuradha Gupta, Garima Pant, Kalyan Mitra, Jitender Madan, Manish K. Chourasia, Amit Misra. Inhalable Particles Containing Rapamycin for Induction of Autophagy in Macrophages Infected withMycobacterium tuberculosis. Mol Pharm. 2014;11(4):1201–7.

8.

Palazzo F, Giovagnoli S, Schoubben A, Blasi P, Rossi C, Ricci M. Development of a spray-drying method for the formulation of respirable microparticles containing ofloxacin-palladium complex. Int J Pharm [Internet]. 2013;440(2):273–82. Available from: http://dx.doi.org/10.1016/j.ijpharm.2012.05.045

9.

Dockery D, Pope CAIII. Acute respiratory effects of particulate air pollution. Annu Rev Public Heal. 1994;15:107–32.

10.

Steckel H, Brandes HG. A novel spray-drying technique to produce low density particles for pulmonary delivery. Int J Pharm. 2004;278(1):187–95.

11.

Alves AD, Cavaco JS, Guerreiro F, Lourenço JP, Rosa da Costa AM, Grenha A. Inhalable Antitubercular Therapy Mediated by Locust Bean Gum Microparticles. Molecules [Internet]. 2016 Jan 28 [cited 2016 Aug 27];21(6):702. Available from: http://www.mdpi.com/1420-3049/21/6/702/htm

12.

International Organization for Standardization. Biological evaluation of medical devices - Part 5: Tests for in vitro cytotoxicity. Iso 10993–5. 2009. p. 1–52.

161


Drug Delivery to the Lungs 27, 2016 - LC Cunha1,2 et al. Inhalable chitosan microparticles as tools in tuberculosis therapy 1,2

3

LC Cunha , AM Rosa da Costa , A Grenha

1,2

1

CBMR – Centre for Biomedical Research Drug Delivery Laboratory, University of Algarve, Campus de Gambelas, Faro, 8005-139, Portugal 2 CCMAR – Centre of Marine Sciences, University of Algarve, Campus de Gambelas, Faro, 8005-139, Portugal 3 Algarve Chemistry Research Center (CIQA) and Department of Chemistry and Pharmacy, Faculty of Sciences and Technology, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugal

Summary: Lung tuberculosis (TB) represents approximately 80% of total cases and, therefore, the lung has been explored as an effective route for the delivery of drugs in the ambit of pulmonary TB. The pulmonary delivery of antitubercular drugs in a carrier system capable of reaching the alveoli, being recognized and phagocytosed by alveolar macrophages (mycobacterium hosts), would be a significant improvement to current oral drug regimens. Chitosan (CS) is a polysaccharide composed of N-acetylglucosamine and D-glucosamine residues, the former being recognised by macrophages and possibly potentiating phagocytosis. This work aimed at producing chitosan microparticles (CS MP) containing two first-line antitubercular drugs, isoniazid (INH) and rifabutin (RFB). A polymeric solution containing 10% (w/w) of INH and 5% (w/w) RFB (weight respective to CS), was spray-dried and the resulting microparticles evaluated as dry powder inhalation targeting alveolar macrophages. Spray-dried CS MP with theoretically adequate properties for deep lung delivery (aerodynamic diameter of 1.98 µm) were produced, efficiently associating isoniazid (INH) and rifabutin (RFB) – 73% and 97%, respectively – in combination. The effect of drug-loaded CS MP on the viability of two cell lines representative of the environment of relevance in pulmonary TB were assessed and absence of toxicity was observed in human alveolar epithelium (A549) and macrophage-differentiated (THP-1) cells. Human macrophage-differentiated THP-1 cells and rat alveolar macrophages NR8383 were exposed to fluorescein-labelled CS MP to assess the ability of CS MP to be taken up by alveolar macrophages. The analysis was performed by flow cytometry and CS MP evidenced strong ability to be captured by macrophages (percentage of phagocytosis >98%). Overall, the obtained data gave positive indications on the potential of the proposed system for an application as inhalable tuberculosis therapy. . Introduction: Although tuberculosis is a curable condition, it remains a major global health problem. The lung has been explored as an effective route for the delivery of drugs in the ambit of respiratory diseases, allowing direct targeting of the affected organ and possibly reducing systemic drug toxicity, which is a special advantage in diseases involving long-term treatments, such as TB. The approach in this case involves direct delivery of antibiotics to the infection site, thus possibly decreasing severe systemic side effects, such as hepatotoxicity and nephrotoxicity, and reducing the period of TB treatment, which are main reasons for patient incompliance. However, several limitations of pulmonary delivery have to be considered as well, mainly related with airway structure and specific defense mechanisms, such as the mucociliary clearance. Overcoming these limitations demands the design of aerodynamically suitable carriers that are capable of reaching the alveoli. Additionally, further benefit may be attained if the carriers can be recognised and phagocytosed by alveolar macrophages (mycobacterium hosts), an effect that can be mediated by a composition of the carriers favouring recognition by macrophage surface receptors (1,2). In this context, this work aimed at using CS to produce MP that efficiently associate a combination of the first-line antitubercular drugs INH and RFB, for an application in pulmonary tuberculosis therapy. Materials and Methods: CS MP were successfully produced by spray-drying (Buchi mini-spray dryer, B-290) a 2% (w/v) aqueous solution of CS (116 KDa; Sigma-Aldrich), with or without drugs – 10% (w/w) INH (3), and 2% (w/w) RFB (w/w respective to CS). Given the well-known potency of RFB as anti-TB agent (4) and after verifying the strong cell toxicity induced by RFB (data not shown), a concentration of 2% (w/w) was chosen for this drug. The formulation was denominated CS/INH/RFB = 10/1/0.2 (w/w). Briefly, the polymer was solubilized in a mixture of acetic acid/ethanol (10/1) at a concentration of 2% (w/v). Both drugs were separately ground in a porcelain mortar and solubilized in acetic acid (INH) or ethanol (RFB), prior to incorporation into CS dispersion previously prepared. In the case of unloaded CS MP, the polymer was solubilized in a mixture of acetic acid/ethanol (10/1) at a concentration of 2% (w/v) without drug association. The resulting dispersion was left under stirring for 1 h and then spray-dried. The spray-drying operating parameters were optimized as follows: inlet temperature 160 ± 1 ºC, aspirator 80%, feed flow 1.3 mL/min, and spray flow rate 473 L/h. MP were characterized regarding surface morphology, Feret’s diameter and real density. Aerodynamic diameter (Daer) was theoretically calculated based on 0.5 the Feret’s diameter and the real density (Daer = Feret’s diameter × (real density/f) ; where f represents the shape factor of MP, which in this case is 1. In order to determine the drug association efficiency, 30 mg of drugloaded CS MP were solubilized in 10 mL of HCl 0.1M, under magnetic stirring for about 20 min. After dissolution, aliquots (1 mL) were filtered (0.45 μm) and the drug content was quantified by UV-Vis spectrophotometry (Pharmaspec UV-1700, Shimadazu) at 268.5 nm (INH, aliquot diluted 1:10) and 500 nm (RFB). Calibration curves for each drug were previously established by solubilizing the drug at different concentrations in solutions of unloaded-CS MP produced in the same medium (HCl 0.1M).

162


Drug Delivery to the Lungs 27, 2016 - Inhalable chitosan microparticles as tools in tuberculosis therapy The cytotoxic evaluation of CS MP was performed by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Sigma-Aldrich) assay in A549 and macrophage differentiated THP-1 cells. CS, as well as unloaded and drug-loaded CS MP were tested at concentrations of 0.1, 0.5, and 1 mg/mL. Cell culture medium (CCM) and sodium dodecyl sulphate (SDS, Sigma-Aldrich) at a concentration of 2% (w/v) were tested as positive and negative controls of cell viability, respectively. CCM for A549 cells was Dulbecco’s modified Eagle’s (DMEM) medium and RPMI 1640 medium for THP-1 cells. Cells were seeded in 96 well plates and exposed to the samples for 3 or 24 h (37 ºC, humidified 5% CO2/95% atmospheric air). Macrophage ability to uptake unloadedCS MP was assessed by flow cytometry (FacScalibur cell analyser, BD Biosciences). Briefly, CS was labelled with fluorescein sodium salt activated by N-(3-dimethylaminopropyl)-N1-ethylcarbodiimide hydrochloride (EDAC), at pH 4. The stained solution was stirred for 12 h at room temperature and then dialyzed against water. Several washings with purified water removed nonspecific staining. The fluorescent chitosan was freeze-dried and solutions prepared with this polymer were then spray-dried under the same conditions described for unloaded-CS MP. NR8383 cells (rat alveolar macrophages) and macrophage-differentiated (by phorbol myristate acetate) THP1 cells were exposed to fluorescently-labelled CS MP, and after 2 h of incubation, cells were scraped and centrifuged (1500 rpm, 2 min) in 2 mL of PBS.3% FBS. The cycle of resuspension in PBS.3% FBS and centrifugation was repeated thrice. Cells were re-suspended in 1 mL of PBS.3% FBS, transferred to cytometry tubes (BD Biosciences) and the phagocytosis determined by flow cytometry. A total of 10,000 events were counted within a gated region and the data was presented as mean fluorescence (FL) intensity. The number of cells associated with fluorescence was considered the definition for uptake. Results and Discussion: The properties of dry powders play an important role in the development of inhalable formulations, as deep lung deposition depends mainly on characteristics like particle size, shape, and density (5). Morphological analysis by SEM showed that spray-dried CS MP are spherical, having a wrinkled surface that becomes smoother after drug incorporation (Figure 1). The observed morphologies are similar to those reported in other works involving spray-dried CS MP (6,7).

Figure 1 – Scanning electron microscopy microphotographs of unloaded (left) and drug-loaded (right) chitosan microparticles.

Spray-drying CS yielded up to 80% of CS MP with mean size ranging from 1.5 µm to 1.9 µm (Table 1). Although some microparticles presented diameters over 2 µm, the size of the majority lies between 1 and 2 µm. As expected, the incorporation of drugs had no effect on size, given the relatively low loading. Likewise, no significant 3 3 differences were observed in real densities (1.39 g/cm for unloaded MP and 1.37 g/cm for drug-loaded MP). Real density measurements are considered more accurate than bulk density, since the measured volume of particles excludes the interstitial space between the particles. The real density along with Feret’s diameter resulted in a theoretical aerodynamic diameter of 1.98 µm for drug-loaded MP, indicating theoretically suitable properties for deep lung delivery (8). Association efficiency and loading capacity of the developed CS MP were determined (Table 1). INH and RFB in combination were efficiently associated to CS MP, complying with the combined therapeutic regimen of TB, as recommended by WHO (9). INH was associated with an efficiency of 73%, resulting in a loading capacity around 7.5%. In turn, a higher association efficiency (97%) was observed for RFB, resulting in a loading capacity of 1.9%. Table 1 – Drug association efficiency, loading capacity, Feret’s diameter, real density and theoretical aerodynamic diameter of chitosan-based microparticles (mean ± SD, n = 3). PROPERTIES

Unloaded CS MP

Drug association efficiency (%)

-

Loading capacity (%)

-

Feret’s diameter (µm) Real density (g/cm3) Aerodynamic diameter (µm)

1.53 ± 0.86 1.39 ± 0.02 1.84 ± 0.12

163

Drug-loaded CS MP (CS/INH/RFB = 10/1/0.2; w/w) 73 ± 1 (INH) 97 ± 2 (RFB) 7.4 ± 0.1 (INH) 1.9 ± 0.1 (RFB) 1.86 ± 0.98 1.37 ± 0.04 1.98 ± 0.17


Drug Delivery to the Lungs 27, 2016 - LC Cunha1,2 et al. CS is one of the most studied natural polymers and its biocompatibility and biodegradability have been reported in the context of several delivery routes, including the pulmonary (5). To the best of our knowledge spray-dried CS MP loaded with a combination of INH and RFB are not described in the literature. We have tested the effect of CS/INH/RFB MP on the viability of two cell lines representative of the environment of relevance in pulmonary TB therapy, an alveolar epithelial line (A549 cells) and a line representing macrophages (macrophage-differentiated THP-1 cells). A solution of CS and a suspension of unloaded CS MP were tested as controls. As depicted in Figure 2, no cytotoxic effect was observed towards macrophage-like cells upon exposure to concentrations up to 1 mg/mL for 24 h. A similar response was obtained in A549 cells (Figure 3). The cell viability level of 70% (indicated with a dashed line) was considered the threshold beyond which a cytotoxic effect was assumed to occur, as designated by ISO 10993-5 (10).

Figure 2 – Macrophage-differentiated THP-1 cell viability after 3 h and 24 h of exposure to chitosan, unloaded (CS MP) and drug-loaded (CS/INH/RFB = 10/1/0.2, w/w) chitosan microparticles. Results are expressed as mean ± SEM (n = 3; six replicates per experiment at each concentration). Dashed line represents 70% cell viability.

Figure 3 – Alveolar epithelial line (A549 cells) cell viability after 3 h and 24 h of exposure to chitosan, unloaded (CS MP) and drug-loaded (CS/INH/RFB = 10/1/0.2, w/w) chitosan microparticles. Results are expressed as mean ± SEM (n = 3; six replicates per experiment at each concentration). Dashed line represents 70% cell viability.

Inhaled antitubercular therapies are desired to specifically target alveolar macrophages infected with TB bacilli (11). In order to assess the ability of CS MP to be taken up by alveolar macrophages, the polymer was labelled with fluorescein (fluorescein sodium salt was activated by N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide hydrochloride (EDAC), producing fluorescent-CS). This was then spray-dried and fluorescent unloaded-CS MP uptake was evaluated on human macrophage-differentiated THP-1 cells and rat alveolar macrophages NR8383. Microparticles were incubated with macrophages for 2 h, as it is reported that about 50-75% of deposited particles are phagocytosed within this period (12). The analysis was performed by flow cytometry and cells exhibiting fluorescence were assumed to have phagocytosed MP. Cells not exposed to fluorescence-labelled MP (control incubated with CCM) showed a certain degree of auto-fluorescence, while cells exposed to fluorescent-CS MP evidenced a stronger increase of the fluorescence signal.

164


Drug Delivery to the Lungs 27, 2016 - Inhalable chitosan microparticles as tools in tuberculosis therapy The percentage of CS MP taken up by macrophages was very high in both cases, 98.1 ± 1.8% for human macrophage-differentiated THP-1 cells and 99.9 ± 0.1 % for rat alveolar macrophages NR8383. Two 2 concentrations of CS MP were tested (80 and 240 µg/cm ) and a dose-dependent uptake was not observed. These results suggest a high affinity of macrophages for CS MP which can be explained by the molecular composition of the polymer. In spite of a natural ability of macrophages for the phagocytosis of particulate matter, a preference for certain particles/materials is observed, depending on the affinity for macrophage surface receptors. In fact, CS is described to interact with macrophage receptors by means of both the toll-like receptors (TLR-4) and the mannose receptor (13). However, further investigation on the ability for preferential macrophage capture remains to be done by a comparison with a polymer devoid of units/residues potentially recognized by macrophage receptors. Conclusion: Spray-dried inhalable powders consisting of chitosan microparticles loaded with a combination of isoniazid and rifabutin were efficiently produced for pulmonary administration. The delivery of these inhalable carriers using an adequate inhaler, could replace the conventional oral delivery of antitubercular drugs. The developed MP evidenced theoretically adequate aerodynamic properties for deep lung delivery. Drug-loaded MP had no effect on cell viability, as indicated by a metabolic assay performed on macrophage-like cells (THP-1) and rat alveolar macrophage cells (NR8383). Furthermore, CS MP evidenced strong ability to be captured by macrophages, upon 2 h incubation with macrophages. Local administration of these MP would possibly target alveolar macrophages hosting mycobacteria, allowing less frequent administration of lower drug doses and reducing major side effects at systemic level, thus contributing to therapeutic effectiveness. Acknowledgements: Funding from the Portuguese Foundation for Science and Technology (PTDC/DTPFTO/0094/2012, UID/Multi/04326/2013 and UID/BIM/04773/2013) is acknowledged. L.C Cunha, acknowledges PhD grant supported by CAPES – Brazil through the Brazilian Government scholarship programme Science without Borders. References 1.

Pandey R, Khuller GK. Antitubercular inhaled therapy: Opportunities, progress and challenges. J Antimicrob Chemother. 2005;55(4):430–5.

2.

António J. Almeida, Grenha A. Technosphere®: An inhalation system for pulmonary delivery of biopharmaceutical. In: Mucosal Delivery of Biopharmaceuticals. 2014. p. 483–98.

3.

Rojanarat W, Changsan N, Tawithong E, Pinsuwan S. Isoniazid Proliposome Powders for Inhalation — Preparation , Characterization and Cell Culture Studies. Int J Mol Sci. 2011;12:4414–34.

4.

Jo K, Ji W, Hong Y, Lee S, Kim WS, Kim DS, et al. The efficacy of rifabutin for rifabutin-susceptible, multidrug-resistant tuberculosis. Respir Med. 2013;107:292–7.

5.

Kundawala A, Patel V, Patel H, Choudhary D. Preparation, In vitro characterization, and in vivo pharmacokinetic evaluation of respirable porous microparticles containing rifampicin. Sci Pharm. 2014;82(3):665–81.

6.

Desai KGH, Park HJ. Encapsulation of vitamin C in tripolyphosphate cross-linked chitosan microspheres by spray drying. J Microencapsul [Internet]. 2005;22(2):179–92. Available from: http://www.ncbi.nlm.nih.gov/pubmed/16019903

7.

Corrigan DO, Healy AM, Corrigan OI. Preparation and release of salbutamol from chitosan and chitosan co-spray dried compacts and multiparticulates. Eur J Pharm Biopharm. 2006;62(3):295–305.

8.

Yang MY, Chan JGY, Chan HK. Pulmonary drug delivery by powder aerosols. J Control Release [Internet]. Elsevier B.V.; 2014;193:228–40. Available from: http://dx.doi.org/10.1016/j.jconrel.2014.04.055

9.

Inagawa H, Kohchi C, Soma G, Matsumoto M, Yamashita S, Terada H, et al. Pulmonary drug delivery systems for the treatment of tuberculosis. Drug Deliv Syst [Internet]. Elsevier B.V.; 2008;23(4):474–80. Available from: http://dx.doi.org/10.1016/j.ijpharm.2014.12.009

10.

ISO. International Organization for Standardization. Biological evaluation of medical devices Part 5: Tests for in vitro cytotoxicity. In: ISO 10993-5. Geneva,Switzerland: ISO; 2009. p. 42.

11.

Chan JGY, Chan H-KK, Prestidge C a., Denman J a., Young PM, Traini D. A novel dry powder inhalable formulation incorporating three first-line anti-tubercular antibiotics. Eur J Pharm Biopharm [Internet]. Elsevier B.V.; 2013 Feb [cited 2016 Jan 20];83(2):285–92. Available from: http://dx.doi.org/10.1016/j.ejpb.2012.08.007

12.

Geiser M. Update on macrophage clearance of inhaled micro- and nanoparticles. J Aerosol Med Pulm Drug Deliv [Internet]. 2010 Aug [cited 2015 Aug 11];23(4):207–17. Available from: http://www.ncbi.nlm.nih.gov/pubmed/20109124

13.

Rodrigues S, Grenha A. Activation of macrophages: Establishing a role for polysaccharides in drug delivery strategies envisaging antibacterial therapy. Curr Pharm Des [Internet]. 2015 Jan [cited 2016 Jan 5];21(33):4869–87. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26290207

165


Drug Delivery to the Lungs 27, 2016 – Hisham Al-Obaidi et el Preparation of hybrid silver/ciprofloxacin nanoparticles for pulmonary drug delivery 1 Hisham Al-Obaidi and Mridul Majumder 1 The School of Pharmacy, University of Reading, Reading RG6 6AD, UK 2 M2M Pharmaceuticals Ltd, Science & Technology Centre, Earley Gate, Reading RG6 6BZ, UK Summary Background. Ciprofloxacin (CFX) is a fluroquinolone antibiotic used as a first line treatment against infections caused by pseudomonas aeruginosa and streptococcus pneumonia, which are commonly, acquired by cystic fibrosis (CF) patients. However, no inhalation formulation is currently available for ciprofloxacin. A major problem with orally administered CP is the poor penetration through the thick mucus typically exists in CF. Silver has been shown to exert mucolytic activity hence a combination with CFX would be a potential approach to improve drug deposition. The aim of this study is to develop novel hybrid nanoparticles of CFX and silver coated with layers of silica. Coating with silica is essential to ensure uniform distribution of silver nanoaprticles as well as to prevent possible aggregation. The final particles are able to exert mucolytic activity as well as enhanced antibacterial activity. The matrix is formed of chitosan, which is a cationic polysaccharide able to further improve mucus adhesion. Methods. Silica coated silver nanoparticles were prepared using StÜber method based on hydrolysis of tetraethyl orthosilicate (TEOS) and analysed by photo correlation spectroscopy and transmission electron microscopy. The optimum ratio of chitosan and sodium triphosphate was used to encapsulate CFX. Particle deposition was assessed in vitro using twin stage impinge using Rotahaler device and based on peak flow of 50L/min. Results. Successful coating with silica was achieved using dimethylamine as a catalyst. Size measurements showed that the size of the silica coated silver nanoparticles ranged between 100-200nm. Entrapment efficiency % showed consistent results with approximate value of approximately 40% CFX. In vitro deposition results showed significant deposition in stage 2 using twin stage impinger (TSI) (~70%). Conclusions. Hybrid nanoparticles of silica coated-silver embedded in chitosan matrix and encapsulating ciprofloxacin were successfully prepared. The particles were significantly deposited into the second stage of TSI with some oropharyngeal deposition.

Introduction CF is a multisystem genetic disorder caused by a mutation in cystic fibrosis transmembrane conductance regulator (CFTR) gene located on chromosome 7. The functioning protein, an ion channel, produced by this gene regulates sodium and chloride levels in cells as well as the volume of the airway surface liquid (ASL). The mutated gene leads to a malfunctioning protein causing hypersecretion of thick and sticky mucus which is difficult to clear. Due to the impaired mucosal defences, recurrent infections caused by Psuedomonas aeruginosa leading to chronic pulmonary symptoms and deteriorating lung function in CF patients. Various antimicrobial agents are used to treat respiratory infections and the which is most commonly used is oral ciprofloxacin. Fig 1: Schematic showing processes involved in nanoparticles Ciprofloxacin (CFX; 1-cyclopropyl-6-fluoroformation 1,4-dihydro-4-oxo-7-(1-piperazinyl)-3-quinolone carboxylic acid) is a broad spectrum second generation fluroquinolone antibiotic used effectively against wide range of infections including P. aeruginosa and [1] Streptococcus pneumonia . The current delivery methods available for CFX are oral and intravenous infusion and no inhaled formulation is currently available. It is also worth mentioning that prolonged administration of current CFX doses/formulations could lead to severe gastrointestinal disturbances and arthropathy, therefore it is not recommended in young children. Pulmonary drug delivery in particular is seen as a non-invasive system for many different agents as the lungs provide a thin, yet extremely efficient absorptive mucosal membrane with a [2] good blood supply . It is however challenging to deliver the drug to a highly viscous environment typically found in CF patients. In order to overcome such limitations, the study of metal nanoparticles (NPs), functional core-shell colloidal NPs in particular, have gained extensive attention for their distinctive physicochemical properties showing potential for [3] mucosal delivery . It is a well-known fact that silver (Ag) ions have strong antimicrobial properties and is still [4] being used to treat wound infections . Recently, it has been shown that Ag has mucolytic properties thinning mucus and with limited toxicity (based on suggested concentrations) in mammalian cells hence it can be attractive [5] ingredient to be used antimicrobial preparations .

166


Drug Delivery to the Lungs 27, 2016- Preparation of hybrid silver/ciprofloxacin nanoparticles for pulmonary drug delivery However, silver NPs may aggregate so coating with a silica shell is essential before designing methods to deliver them. CFX is sparingly soluble in water, so when the drug is delivered via the airways there is an increased chance that the mucociliary mechanism will clear the drug, hence, the presence of silica can also serve to aid dissolution of CFX. Nanoparticles encapsulating CFX and silver will be promising solution to deliver the drug to the lungs to retain sufficient drug concentration above the minimum inhibitory concentration (MIC) entailing less frequent dosing and therefore to achieve a decline in fluctuation of drug concentration. Broad range of materials from natural to synthetic have been used with a particular interest in polysaccharides such as chitosan (CS), a [6] linear cationic polymer mainly degraded by lysozyme in the lungs . Its cationic nature dictates its mucoadhesive properties as a result of electrostatic forces between the cationic amino groups of chitosan and small anionic [7] molecules such as phosphates . This will provide a capacity allowing the entrapment of CFX, which will ensure effective delivery, and maintaining the drug at the target tissue. The aims of this study were to prepare hybrid nanoparticles of silica coated silver nanoparticles embedded into matrix of chitosan doped with ciprofloxacin. Secondly, to fully characterise the prepared nanoparticles and determine deposition efficiency using twin stage impinger, TSI. Experimental Materials Tetraethyl orthosilicate (TEOS; ≥99.0%, GC, Aldrich), silver (Ag; nanopowder, <100 nm particle size, contains PVP as dispersant, 99.5% trace metals basis, Aldrich), chitosan (CS; low molecular weight, Aldrich), sodium triphosphate (TPP; parum p.a., ≥98.0%, Sigma-Aldrich), ciprofloxacin (CFX; ≥98.0%, HPLC, Fluka), anhydrous lactose (Fluka), dimethylamine solution (DMA; 40 wt. % in H2O, Aldrich) and acetic acid (ReagentPlus®, ≥99%, Sigma-Aldrich) from Sigma-Aldrich, St Louis, USA. Ethanol absolute (EtOH) from VWR International, Leicestershire, UK. Methods Preparation of silica coated Ag and measurements of particle size using photo correlation spectroscopy (PCS) Various amounts (75 mL of EtOH, 20 mL distilled water (H2O) and 20 mg Ag) were mixed to followed by addition of 1 mL DMA combined with ultra sonication. Then to each of the flasks, 0.25, 0.5, 0.75 and 1 mL TEOS were added gradually whilst on vigorous stirring using mechanical stirrer. This created four solutions containing different amounts of TEOS whilst still having the same amount of Ag in each of them, the amount of Ag was also varied to allow comparison. The solutions were stirred for 2 hours, 2 mL, 1 mL and 0.5 mL of prepared solutions were drawn and were diluted by a factor of 10, 20 and 40 times respectively to be tested for particle sizing and their count rate by PCS (ZetaPlus, Brookhaven Instruments Corporation) at 0, 1, 3 and 5 hours post dilution undergoing 10 measurements. Determining the optimum ratio of CS:TPP Prior to optimising the NPs, the loading efficiency of CS nanoparticles with CFX had to be determined and as [8] indicated by a previous study . CS:TPP of variable ratios ranging from 0.6 to 12.0 were tested using 0.3% w/v CS in 0.2% v/v acetic acid, 0.8% w/v CFX in 0.2% v/v acetic acid and 2% w/v TPP in water (note that all solutions were freshly made before mixing). The controlled variable was the amount of CFX and the independent variables were the amounts of CS and TPP used. Determining drug encapsulation efficiency (% EE) of the CFX encapsulated silver NPs To the 16 solutions of silica coated Ag nanoparticle solutions prepared, 400 μL of acetic acid, 30 mL CS (=90 mg) and 15 mL CFX (=120 mg) were added followed by gradual addition of 9 mL TPP (=180 mg) with continuous stirring. The absorbance for each formula was measured spectroscopically at 271 nm (λmax of CFX). Transmission electron microscopy (TEM) Selected samples were run under TEM to determine the size of the nanoparticles using Philips Tecnai T20 microscope operating at 200 kV equipped with an EDS (energy dispersive spectrum) detector. In vitro equivalence testing using twin stage glass impinger (TSI) Selected samples were centrifuged (Allegra™ X-12 Centrifuge, Beckman Coulter) for 15 minutes at 13000 rpm and all the supernatant was removed. Each of the samples were left in a mini vacuum desiccator (0% RH) for 3 days to ensure samples were thoroughly dried. The mixtures were then individually milled with lactose (MixerMill MM 200, Retsch) for 5 minutes. The rationale was to generate uniform formulations so that impact observed could then be attributed to the presence of the hybrid nanoparticles not to the carrier lactose particles. Three gelatine capsules containing 40 mg of each of the powder were filled for in vitro testing using TSI. The best samples were selected based on entrapment efficiency and rate of dissolution. About 7 mL and 30 mL of 0.2% v/v acetic acid were introduced in stages 1 and 2 respectively. Particle deposition was assessed in vitro using twin stage impinge using Rotahaler device and based on peak flow of 50L/min. The capsule under test was placed in the DPI followed by attachment to the mouthpiece and the fractions of deposition were measured by collecting the powders of mouthpiece and upper chamber (stage 1) and lower chamber (stage 2). The concentration was measured at 271nm and the % CFX deposition in each stage was calculated accordingly.

Fig. 2: TEM images of silica coated silver nanoparticles. Both samples contain 2 mg Ag however, they contain (a) 5.82 μmol and (b) 92.44 μmol TEOS.

167


Drug Delivery to the Lungs 27, 2016 – Hisham Al-Obaidi et el Results and discussion Characterization of silica coated nanoparticles particles A factorial design of 32 formulations was selected based on the independent variables used: Ag at 2, 5 and 7.5 mg, and TEOS at 5.83, 34.75 and 92.44 μmol, giving a total of 9 preparations. Average particle size distribution for these selected formulations were determined which showed a particle size range 100-200nm The size of the NPs at time 0 hours increased with higher amounts of Ag and TEOS however, broader variation was seen with the particles containing lower amount of TEOS (5.82 µmol). It was interesting to see that the particles which contained 2 mg Ag and 92.44 μmol TEOS were of similar size to the ones that contained 7.5 mg Ag and 5.83 μmol TEOS (175.2 nm ± 8.1 and 167.7 nm ± 12.1 respectively). Dissolution rate of silica particles Dissolution rate was measured for particles using the count rate (intensity of light is proportional to the size of particles) obtained from PCS because it was seen as a controlling factor in the release of Ag and dissolution of CFX. It was seen that the majority of the particles completely dissolved when dispersed in water after 5 hours. It would have been safe to suggest that as the particle size increases the dissolution rate will be slower. However, by comparing the % dissolution (Fig. 3) with the particle sizes, the results are more distinct in that it was not necessarily the lower TEOS concentration that produced higher % dissolution. This is due to the fact that dissolution rate is influenced by particle aggregation whereby the surface area of the aggregated particles have substantially decreased. Fig. 3: Effect of varied TEOS amount on the % dissolution

of particles containing 7.5 mg of silver coated by silica post Entrapment efficiency % (EE%) of prepared chitosan dilution. nanoparticles CS:TPP ratios of 0.5, 1 and 2 were taken a step further by calculating the % Table 1: Average % EE of EE of CFX upon taking absorbance readings of supernatant for each sample ciprofloxacin in each formulation using UV spectroscopy at 271 nm (Table 1). By correlating the values to the Formu calibration curve for CFX at 271 nm, it was possible to calculate the amount la % EE ± sd of drug left in the supernatant and therefore the amount of encapsulated code drug. As shown in Table 1 (letter A-C indicate increased Ag content), EE% A1 41.9 ± 0.4 was consistent among prepared samples with samples containing 2mg A2 42.5 ± 1.5 showing highest %EE. Overall, the preparation method was optimised A3 40.6 ± 0.4 accordingly and best formulation was then selected to carryon in-vitro A4 21.9 ± 8.2 deposition experiments. B1 38.0 ± 4.5 B2 37.1 ± 1.1 Measurement of deposition efficiency using TSI B3 30.4 ± 0.6 As shown in Fig 4, significant fraction of the delivered particles was B4 29.0 ± 1.9 deposited in stage 2 (approximately 70%). These results suggest potential C1 41.6 ± 0.4 application of these particles to deliver the drug to the lungs versus C2 39.9 ± 0.9 compaction in oropharyngeal region or being swallowed. It is however C3 40.5 ± 0.5 acknowledged that there was some deposition of the particles in stage 1. C4 39.1 ± 8.2 Future work would be to optimise milling processing to ensure optimum aerodynamic properties.

Conclusions Using the modified Stöber method based on using DMA rather than ammonia as the catalyst was successfully used to form Ag core silica coated nanoparticles. Optimum ratio of chitosan was determined to form physically stable particles composed of chitosan as the matrix and contain silica coated silver nanoparticles. The matrix was doped with CFX to prepare particles with dual antibacterial activity. Many evidences support that having Ag as part of the nanoparticle enhances the antimicrobial activity which CFX also does, and that the use of the ionic gel formation of CS and TPP maximised EE%. Also the fact that the Fig.4: CFX % deposition at each stage of TSI ASL pH in CF patients is lower therefore if CFX can be formulated with enough drug encapsulated in the NPs then the lower pH will enhance the solubility of CFX allowing amplified antimicrobial activity. Successful deposition in TSI demonstrates potential application of these particles.

168


Drug Delivery to the Lungs 27, 2016- Preparation of hybrid silver/ciprofloxacin nanoparticles for pulmonary drug delivery References 1. Labiris, N. R.; Dolovich, M. B., Pulmonary drug delivery. Part I: physiological factors affecting therapeutic effectiveness of aerosolized medications. British journal of clinical pharmacology 2003, 56 (6), 588-99. 2. Calvo, P.; Remuñan-López, C.; Vila-Jato, J. L.; Alonso, M. J., Chitosan and Chitosan/Ethylene OxidePropylene Oxide Block Copolymer Nanoparticles as Novel Carriers for Proteins and Vaccines. Pharmaceut Res 1997, 14 (10), 1431-1436. 3. Baral, V. R.; Dewar, A. L.; Connett, G. J., Colloidal silver for lung disease in cystic fibrosis. Journal of the Royal Society of Medicine 2008, 101 Suppl 1, S51-2. 4. Rai, M.; Yadav, A.; Gade, A., Silver nanoparticles as a new generation of antimicrobials. Biotechnology advances 2009, 27 (1), 76-83. 5. Feng, Q. L.; Wu, J.; Chen, G. Q.; Cui, F. Z.; Kim, T. N.; Kim, J. O., A mechanistic study of the antibacterial effect of silver ions on Escherichia coli and Staphylococcus aureus. Journal of biomedical materials research 2000, 52 (4), 662-8. 6. Nordtveit, R. J.; Vårum, K. M.; Smidsrød, O., Degradation of partially N-acetylated chitosans with hen egg white and human lysozyme. Carbohydrate polymers 1996, 29 (2), 163-167. 7. Osman, R.; Kan, P. L.; Awad, G.; Mortada, N.; El-Shamy, A. E.; Alpar, O., Spray dried inhalable ciprofloxacin powder with improved aerosolisation and antimicrobial activity. Int J Pharm 2013, 449 (1-2), 44-58. 8. Rampino, A.; Borgogna, M.; Blasi, P.; Bellich, B.; Cesaro, A., Chitosan nanoparticles: preparation, size evolution and stability. Int J Pharm 2013, 455 (1-2), 219-28.

169


Drug Delivery to the Lungs 27, 2016 - Susana Rodrigues, et al. Chondroitin sulphate microparticles for tuberculosis treatment: a way to target macrophages Susana Rodrigues,

1, 2

3

Ana M Costa & Ana Grenha

1, 2

1

CBMR – Centre for Biomedical Research Drug Delivery Laboratory, University of Algarve, Campus de Gambelas, Faro, 8005-139, Portugal 2 CCMAR – Centre of Marine Sciences, University of Algarve, Campus de Gambelas, Faro, 8005-139, Portugal 3 Algarve Chemistry Research Center (CIQA) and Department of Chemistry and Pharmacy, Faculty of Sciences and Technology, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugal Summary Tuberculosis remains a leading cause of death; therapeutic failure being mainly due to non-compliance with prolonged treatments, often associated with severe side-effects. New therapeutic strategies are demanded and, considering that the lung is the primary site of infection, direct lung delivery of antibiotics is an interesting and, possibly, effective approach. Therapeutic success in this context depends on suitable carriers that reach the alveoli where Mycobacterium hosts (macrophages) reside, as well as on their ability to promote macrophage capture and intracellular accumulation of drugs. In this work, we propose inhalable chondroitin sulphate microparticles produced by spray-drying and tailored to suitable aerodynamic properties to reach the alveoli. Macrophage targeting will be driven by microparticle size, which is favoured for carriers of 1-2 µm, and composition based on chondroitin sulphate, a glycosaminoglycan comprised of alternating units of sulphated Nacetylgalactosamine and glucuronic acid residues, the former recognized by macrophage receptors. Spray-drying of chondroitin sulphate in combination with two first-line antitubercular drugs (isoniazid and rifabutin) was successful with a satisfactory production yield (> 70%). Microparticles have Feret’s diameter of 1.6 μm, potentiating macrophage uptake. Chondroitin sulphate as solution or microparticulate form and drug-loaded microparticles appear to not have a cytotoxic effect on alveolar epithelial cells. A more extended biocompatibility/safety assessment of this formulation needs to be performed. Taking into account the general trend of the results obtained so far, good indications are given that encourage the continuation of the studies in order to establish the potential of these microparticles as inhalable carriers in tuberculosis treatment. Introduction Tuberculosis (TB) is an infectious disease that is caused by Mycobacterium tuberculosis (Mtb), which is hosted by and multiplies inside alveolar macrophages, thus mostly affecting the lung. Mtb ability to block phagosomal maturation contributes to the long-term survival and persistence of the bacterium in host macrophages. TB remains the leading cause of preventable deaths, especially in developing countries, but is reappearing in Europe. Current therapy demands prolonged oral multi-drug antibiotherapy, frequently with strong side effects, thus decreasing patient compliance [1]. In this context, and considering lung tuberculosis as the major manifestation of the disease, the direct administration of antitubercular drugs to the alveoli, where macrophages reside, comprises a very promising approach, permitting the delivery of drugs to the primary site of infection, while bypassing first-pass metabolism. This strategy demands designing carriers that exhibit adequate aerodynamic properties to reach the alveolar zone (aerodynamic diameter of 1-5 µm), further enabling macrophage capture. The uptake of particles by alveolar macrophages was reported to be maximal for particles of 1-2 µm [2], but the innate ability of macrophages for unspecific uptake of particulate matter is considered a relevant advantage. Additionally, the clearance of particles within macrophages from the alveoli is slow (weeks/months) [3], ensuring high potential of action of phagocytosed particles. Importantly, particle phagocytosis may activate macrophages, providing additional bactericidal capacity [4]. Works reporting the use of carriers for macrophage targeting in tuberculosis therapy mostly apply synthetic polymers. To improve macrophage internalisation, some works report particle surface functionalisation with targeting ligands recognised by macrophage lectin receptors. Several natural polymers take an important advantage in this respect, having a composition based on units potentially undergoing direct recognition by macrophage receptors [5, 6]. Additionally, natural materials have higher probability of biocompatibility and biodegradability. Chondroitin sulphate (ChS) is a structural component of cartilage and belongs to the family of proteoglycans. These are macromolecules composed of a central core protein to which one or more glycosaminoglycan chains are covalently attached. Both the sulphate groups and Nacetylgalactosamine residues are potentially recognised by macrophage surface receptors [5-7]. Spray-drying is a mild and cost-effective technique to produce microparticles, which use in the production of dry powders for inhalation purposes is advantageous because of the possibility to tailor the aerodynamic properties and long-term stability comparing with liquid counterparts [8]. The application of carriers endowed with suitable properties to reach the alveolar zone upon inhalation and to promote alveolar macrophage uptake by surface recognition, would permit drug accumulation in the infection site, apart from possibly activating macrophages. This would improve therapeutic efficacy, reducing dosing frequency and increasing patient compliance. In this work, we propose spray-dried inhalable chondroitin sulphate microparticles, loaded with a combination of isoniazid (INH) and rifabutin (RFB) and tailored to suitable aerodynamic properties to reach the alveoli for an application in tuberculosis therapy. Macrophage targeting will be driven by microparticle size, which is favored for carriers of 1-2 μm [3], and composition based on chondroitin sulphate, constituted by sulphate groups and Nacetylgalactosamine residues recognized by macrophage receptors [4, 9].

170


Drug Delivery to the Lungs 27, 2016 - Chondroitin sulphate microparticles for tuberculosis treatment: a way to target macrophages Experimental methods Preparation of dry powders Aqueous solutions of ChS (Creative BioMart, USA) with or without associated drugs (INH and RFB - SigmaAldrich, Germany and Chemos GmbH, Germany, respectively) were prepared at 2% (w/v) ChS content. Due to hygroscopic properties of chondroitin, a hydroalcoholic solution (80/20, v/v) was used as solvent in order to increase the evaporation rate during spray drying process. Drugs were associated in order to obtain a final formulation of ChS/INH/RFB = 10/1/0.5 (w/w). Due to the hydrophobic character of RFB, this drug was pre® dissolved in the ethanol used for the mixture. Dry powders were obtained by spray-drying (laboratory scale Buchi Mini Spray Dryer, B-290, Switzerland), at the following conditions: airflow rate of 250 NL/h, solution feed rate of 1 mL/min, inlet temperature of 175 ± 2 ºC, aspirator set at 90%. Dry powders were collected and stored in a desiccator at room temperature until use. Characterization of microparticles The spray-drying yield was calculated by gravimetry, comparing the total amount of solids initially added for the preparation of solutions with the amount of resulting microspheres (n = 3). The morphological characterization of microspheres was performed by field emission scanning electron microscopy (FESEM; Ultra Plus, Zeiss, Germany). The particle size was estimated as the Feret’s diameter (distance between two tangents on opposite sides of the particles) and was directly determined by optical microscopy (Olympus IX51, Japan). The estimation was based on the mean diameter resulting from the evaluation of 300 particles (n = 3). Apparent tap density was obtained by measuring the volume of a known weight of powder in a 10 mL test-tube after mechanical tapping (Densitap, Deyman, Spain). After registration of the initial volume, the test-tube was submitted to tapping until constant volume was achieved (n = 3). Determination of drug content in microparticles was performed by incubating a known quantity of drug-loaded microparticles in HCl 0.1M, under magnetic stirring for 60 min. Samples were then filtered (0.45 µm) and the drugs quantified by HPLC (Agilent 1100 series, Germany). Evaluation of metabolic activity The developed microparticles and raw material (ChS) were assessed for cytotoxicity in A549 cells (lung alveolar cell line) by means of the MTT assay. Cells were exposed to the test materials for 3 h and 24 h at three different concentrations (0.1, 0.5 and 1.0 mg/mL). Cell culture medium (Lonza, Switzerland) and a solution of 2% (w/v) sodium dodecyl sulphate (SDS) were used as negative and positive controls of cell death, respectively. To initiate the assay, culture medium of cells at 24 h in culture in 96 well plates was replaced by 100 µL of fresh medium containing the test dispersions or controls at the referred concentrations. After 3 h or 24 h of cell exposure to microparticles, the test dispersions were removed and 30 µL of the MTT solution (0.5 mg/mL in PBS, pH 7.4) added to each well. After 2 h, any formazan crystals generated were solubilized with 50 µL of dimethylsulfoxide (DMSO). Upon complete solubilisation of crystals, the absorbance of each well was measured by spectrophotometry (Infinite M200, Tecan, Austria) at 540 nm and corrected for background absorbance at 650 nm. The relative cell viability (%) was calculated as follows: Viability (%) = (A – S)/(CM – S)× 100 where A is the absorbance obtained for each of the concentrations of the test substance, S is the absorbance obtained for 2% SDS and CM is the absorbance obtained for untreated cells (incubated with cell culture medium). The latter reading was assumed to correspond to 100% cell viability. Results/Discussion ChS microparticles were successfully produced by spray-drying, with yields of 73-77% (Table 1). A combination of two first-line antitubercular drugs was effective within the same carrier, complying with the recommendations from World Health Organisation regarding the establishment of combined tuberculosis therapy [10]. FESEM viewing revealed an irregular and convoluted morphology, as depicted in Figure 1. As shown in Table 1, 3 Feret’s diameters around 1.6 μm were obtained, along with a tap density of approximately 0.5 g/cm , for both drug-loaded and unloaded microparticles. The obtained microparticle size is considered adequate for macrophage uptake, reported to be favoured for particles of 1-2 μm [2]. In turn, the tap densities are similar to those reported for other polysaccharide-based formulations proposed for pulmonary delivery [11-13]. Density can be used along with the Feret´s diameter to calculate the theoretical aerodynamic diameter of microparticles [14]. However, an experimental estimation of the parameter is a more reliable assessment of the aerosolization properties of dry powders. This analysis is currently being performed at the University of Parma. If the results are not adequate for deep lung delivery, as required for the proposed application, microparticle properties will be optimized. Both drugs were efficiently associated to the microparticles, although with different efficiencies. INH revealed 73% association efficiency, contrasting with 59% for RFB. The lower association of the latter is probably due to the hydrophobic character of the drug, which is probably lost during the spray drying process.

171


Drug Delivery to the Lungs 27, 2016 - Susana Rodrigues, et al. Table 1 - Microparticle characterization (mean ± SD, n = 3).

3

Microparticles

PY (%)

Feret´s diameter (µm)

Tap density (g/cm )

EE (%)

ChS

73.3 ± 4.4

1.67 ± 1.05

0.50 ± 0.01

n. a.

ChS/INH/RFB

77.2 ± 2.2

1.63 ± 1.13

0.52 ± 0.02

INH: 73.4 ± 7.3 RFB: 59.3 ± 6.8

ChS: chondroitin sulphate; EE: Encapsulation efficiency; INH: isoniazid; n.a.: not applicable; PY: production yield; RFB: rifabutin

Figure 1 - Representative microphotograph of chondroitin sulphate microparticles obtained by field emission scanning electron microscopy (FESEM).

Cell viability (% of control)

The use of ChS in lung delivery was only proposed occasionally and, therefore, it is important to determine the cytotoxic profile of this material. The metabolic assay MTT demonstrated that A549 cell viability did not decrease beyond 70% independently of the formulation, the tested concentrations and duration of the assay, as is depicted in Figure 2. This value is referred at ISO 10993-5 as the level beyond which a cytotoxic effect is considered to exist [15]. It is considered that the concentration of 1 mg/mL is possibly much higher than that needed for a therapeutic effect [16]. Notwithstanding the good indication of these results, it is deemed very important to extend the type of assays that are performed to encompass other aspects of cell toxicity (membrane integrity, for instance) and also to use a cell line representative of macrophages, which are also relevant cells in the environment related with tuberculosis infection. The latter is currently being addressed.

120 100 80 60 40 20 0 0.1 ChS polymer

0.5

1.0

Concentration (mg/mL) ChS microparticles Drug loaded ChS microparticles

Figure 2 - Cell viability of A549 cells upon 24 h exposure to ChS polymer and ChS-based microparticles, as obtained from the MTT assay. Values are represented as % of control (cells incubated with cell culture medium) (mean ± SEM, n = 3). Dashed line represents 70% cell viability.

172


Drug Delivery to the Lungs 27, 2016 - Chondroitin sulphate microparticles for tuberculosis treatment: a way to target macrophages Conclusions Chondroitin sulphate microparticles associating a combination of antitubercular drugs were successfully obtained by spray-drying. The microparticles are suggested to have adequate properties (size, density) to reach the alveolar zone upon inhalation, although the experimental estimation of aerodynamic properties is still in course. Both the polymer and the microparticles appear to not have a cytotoxic effect on alveolar epithelial cells. This assessment will be extended to macrophages in the near future. The results obtained so far are encouraging regarding the progression of studies in order to demonstrate the adequacy of the system as antitubercular drug carrier for inhalable therapy. Acknowledgements This work was supported by National Portuguese funding through FCT - Fundação para a Ciência e a Tecnologia, through the projects PTDC/DTP-FTO/0094/2012, UID/Multi/04326/2013 and UID/BIM/04773/2013. The PhD scholarship to Susana Rodrigues (SFRH/BD/52426/2013) is also acknowledged. References 1. 2. 3. 4. 5. 6.

7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

Ranjita, S., A. Loaye, and M. Khalil, Present Status of Nanoparticle Research for Treatment of Tuberculosis. Journal of Pharmacy and Pharmaceutical Sciences, 2011. 14(1): p. 100-116. Ahsan, F., et al., Targeting to macrophages: role of physicochemical properties of particulate carriers— liposomes and microspheres—on the phagocytosis by macrophages. Journal of Controlled Release, 2002. 79(1–3): p. 29-40. Geiser, M., Update on macrophage clearance of inhaled micro- and nanoparticles. Journal of Aerosol Medicine and Pulmonary Drug Delivery, 2010. 23(4): p. 207-217. Rodrigues, S. and A. Grenha, Activation of macrophages: Establishing a role for polysaccharides in drug delivery strategies envisaging antibacterial therapy. Current pharmaceutical design, 2015. 21(33): p. 4869-4887. Rachmilewitz, J. and M.L. Tykocinski, Differential effects of chondroitin sulfates A and B on monocyte and B-cell activation: evidence for B-cell activation via a CD44-dependent pathway. Blood, 1998. 92(1): p. 223-229. Leteux, C., et al., The Cysteine-Rich Domain of the Macrophage Mannose Receptor Is a Multispecific Lectin That Recognizes Chondroitin Sulfates a and B and Sulfated Oligosaccharides of Blood Group Lewis(a) and Lewis(x) Types in Addition to the Sulfated N-Glycans of Lutropin. The Journal of Experimental Medicine, 2000. 191(7): p. 1117-1126. Chong, A.S. and C.R. Parish, Cell surface receptors for sulphated polysaccharides: a potential marker for macrophage subsets. Immunology, 1986. 58(2): p. 277-284. Atkins, P., Dry powder inhalers: an overview. Respiratory Care, 2005. 50(10): p. 1304-1312. Higashi, N., et al., The Macrophage C-type Lectin Specific for Galactose/N-Acetylgalactosamine Is an Endocytic Receptor Expressed on Monocyte-derived Immature Dendritic Cells. Journal of Biological Chemistry, 2002. 277(23): p. 20686-20693. Atif, M., et al., Duration of treatment in pulmonary tuberculosis: are international guidelines on the management of tuberculosis missing something? Public Health, 2015. 129(6): p. 777-782. Rodrigues, S., et al., Hybrid nanosystems based on natural polymers as protein carriers for respiratory delivery: Stability and toxicological evaluation. Carbohydrate Polymers, 2015. 123(0): p. 369-380. Koc, B. and F. Kaymak-Ertekin, The Effect of Spray Drying Processing Conditions on Physical Properties of Spray Dried Maltodextrin. 9th Baltic Conference on Food Science and Technology - Food for Consumer Well-Being: Foodbalt 2014, 2014: p. 243-247. Yang, X.-F., et al., The influence of amino acids on aztreonam spray-dried powders for inhalation. Asian Journal of Pharmaceutical Sciences, 2015. 10(6): p. 541-548. Hinds, W.C., Aerosol technology : properties, behavior, and measurement of airborne particles1999, New York: Wiley. ISO, Biological evaluation of medical devices Part 5: Tests for in vitro cytotoxicity, in 10993-5, I.O.f. Standardization, Editor 2009. Alves, A.D., et al., Inhalable Antitubercular Therapy Mediated by Locust Bean Gum Microparticles. Molecules, 2016. 21(6).

173


Drug Delivery to the Lungs 27, 2016 – Abigail Martin et al.

Optimisation of phorbol myristate acetate (PMA) mediated differentiation of U937 human lung monocytes to alveolar like macrophages A Martin, D Murnane, DYS Chau, MB Brown, V Hutter Department of Pharmacy, School of Life and Medical Sciences, University of Hertfordshire, AL10 9AB. Summary Respiratory disease remains an increasing global health burden, however few new medicines for asthma and chronic obstructive pulmonary disease (COPD) have reached the market in the past decade. Whilst there has been a considerable amount of research investment into the development of new inhaled medicines, many fail due to safety or efficacy. One of the main reasons for this is the observation of a foamy alveolar macrophage responses in rats in pre-clinical studies which limits the dose of these compounds in subsequent studies despite not knowing if these observations are toxic in humans. The ultimate aim of this work is to develop an accurate, immune responsive, human alveolar macrophage based co-culture model for the assessment of inhaled medicines. Accordingly, U937, a human lung-derived monocyte cell line, was differentiated to a macrophage phenotype in the presence of 5 nM and 100 nM phorbol myristate acetate (PMA) over 24-96 h. The activation, differentiation and functionality characteristics of native and treated cells were assessed using a combination of cell viability and morphological analyses, microarray technology, flow cytometry and phagocytic ability. The concentration of PMA was found to have a more significant impact than incubation time on the expression of macrophage cell surface markers, morphology and viability. This work suggests that the differentiation protocol for U937 cells is important to consider as it may impact cell response. The U937 cell line offers good potential to explore macrophage responses to inhaled pharmaceutics in a robust in vitro immunocompetent co-culture model with airway epithelial cells. Introduction Despite considerable research investment into the development of new medicines for the treatment of respiratory 1,2 disease, only a very limited number have reached the market in the past decade . A key factor for the attrition of potential candidate inhaled drug compounds is the observation of an immune response in the airways of rats in 2 pre-clinical studies, characterised by the appearance of foamy alveolar macrophages . There is increasing evidence that inhaled medicines can be taken up by alveolar macrophages resident within the lung giving them a 1-3 foamy, vacuolated appearance . However, the mechanism for induction of this ‘foamy’ alveolar macrophage 1,2 response and its significance in relation to lung pathophysiology are currently poorly understood . Despite not knowing if the foamy macrophage effects observed in rats are a safety concern for humans, many new drug 2 candidates are halted due to safety or lack of efficacy in subsequent studies . Table 1: Reported phorbol myristate acetate (PMA) mediated differentiation protocols for U937 cells Cell density

Concentration

Incubation

Rest period

Surface marker expression

Compared control

Reference

3x105 cells/ml

160 nM

72 h

0h

Y

(2)

1x10 cells/ml 5×105 cells/ml 2x104 cells/well n/a 5×105 cells/ml

40 nM 81 nM 5 nM 16 nM 1-10 nM

48 h 48 h 72 h 72 h 48 h

0h 0h 72 h repeated* 72 h 0h

CD11a/b/c, 29, 54, 56, 59, 36, 44, 61, 41b CD36 n/a CD11a/b. 14, 18 CD14, 206 CD14,13, 4, 71

Y N Y Y Y

(3) (4) (5) (6) (7)

4x104- 7.5x105 cells/ml 2×105 cells/ml

162 nM 20 nM+

24, 48 h 0-24 h

0h 0h

n/a CD11b, 36

Y Y

(8) (9)

4×10 cells/ml 5 5×10 cells/ml 6 1x10 cells/ml

32 nM 150 nM 4 nM

CD11a/b/c, 15, 33, 14 n/a CD11b

Y N Y

(10) (11,12) (13)

n/a 6 1x10 cells/ml 5×105 cells/ml 5 5×10 cells/ml

16 nM ++ 160 nM 16 nM 20 nM

48-72 h 0h 96 h 0h 96 h (media change 24 h after 48 h) 12 h 72 h** 48 h 48 h 24 h 0h 12-72 h 0h

CD14 CD11b n/a CD11a/b, 14

Y Y n/a Y

(14) (15) (16) (17)

4x105cells/ml

6

5

32 nM

8h

36-46 h***

n/a

N

(18)

5×10 cells/ml 5 6 2x10 -10 cells/ml 5×105 cells/ml

16 nM 81 nM 16 nM

72 h 48 h 12, 24, 48,72 h

24 h Not disclosed**** 0h

n/a n/a n/a

Y N Y

(19) (20) (21)

n/a 1x106 cells/ml

100 nM 1.6 nM+++

48 h 72 h

0h 0h

CD14 n/a

N Y

(22) (23)

100 nM

72 h

0h

n/a

N

(24)

5

6

1x10 cells/ml

*fresh complete culture medium every 72 h up to 2-3 weeks, **fresh complete culture medium every 24 h up to72 h, ***included 3 h incubation with LPS, ****did not disclose length of time for rest period. Concentrations were corrected to nM, +Concentration range: +0200 nM, ++80-162 nM for 24 and 48 h, +++0.16 nM- 24.3 nM. Rest period: the amount of time cells were incubated in fresh culture medium following treatment. Compared to control: Differentiated U937 cells were either compared (Y- yes) or not compared (N-no) to their untreated phenotype.

174


Drug Delivery to the Lungs 27, 2016 – Abigail Martin et al.

In vitro cell culture models provide an ideal platform to assess cell response to exposure of compounds in a high throughput manner and to investigate the mechanisms involved which generate the foamy macrophage phenotype. It is first necessary to identify an appropriate, well characterised cell line in order to develop a robust in vitro human model. To date, no human alveolar macrophage cell line is readily available. As such, current models rely on lung and blood derived monocyte cell lines, which are differentiated using various stimuli to generate a mature monocyte/macrophage-like cell line that can be exposed to aerosols or particulate matter of 3 interest . However, the methodologies available to differentiate and validate these cell lines are poorly defined, especially for U937 cells whereby differentiation protocols vary significantly from use of different stimuli, incubation and resting times (Table 1). Furthermore, the characterisation of differentiated U937 cells differs in comparison with other monocyte/macrophage cell lines, suggesting a possible need to characterise individual cell 25 lines to their specific profiles . The aim of this research was to investigate if varying the differentiation protocol for U937 had an impact on the resulting cell characteristics. Optimising a cell line representative of human alveolar macrophages will provide the basis for a robust in vitro immunocompetent co-culture model for the safety assessment of new candidate inhaled medicines. Materials and Methods Cell culture: The U937 human monocyte cell line (LCG Standards, Teddington, Middlesex, UK) was cultured in RPMI supplemented with 10 %v/v FBS, 100 IU/ml penicillin-100 µg/ml streptomycin solution, 2 mM L-glutamine and incubated in a humidified atmosphere at 37 °C with 5 %v/v CO2. Cell number was maintained between 1 x 5 6 10 to 2 x 10 cells/ml and experiments performed with cells between passage 2 and 15 from purchase. For 6 differentiation to a macrophage phenotype, cells were seeded at a density of 5 x 10 cells/mL in complete cell culture medium supplemented with PMA (5 or 100 nM) and incubated for either 24, 48, 72 or 96 h followed by a 24 h resting period in complete cell culture medium. Characterisation of differentiated U937s: Cells were characterised for cell differentiation (CD) surface markers using array analysis (Sciomics, Heidelberg, Germany) and flow cytometry techniques (CD11a, CD11b, CD14, CD36 and CD206). Phagocytosis was assessed using Cayman's phagocytosis assay kit (Bar Hill, Cambridge, UK) containing 0.1 µm latex beads coated with FITC-labelled IgG. Viability of cells was assessed every 24 h for 18 days following exposure to PMA using Guava® ViaCount® Reagent from Millipore Limited (Watford, Hertfordshire, UK). Cell morphology was assessed by haematoxylin and eosin stain on sterilised milinex films from Agar Scientific (Stanstead, Essex, UK). Results and Discussion Protein microarray results showed that PMA treated U937 cells were rich in surface markers specific to 26 macrophages and alveolar macrophages . CD11a, CD11b, CD14, CD36 and CD206 were selected for characterising PMA-mediated differentiation of U937 cells based on the protein microarray results and literature. Cell surface marker expressions for CD11a, CD11b and CD36 were more abundant in the PMA differentiated U937 cells in comparison with undifferentiated cells. CD11b was significantly more highly expressed (p<0.05) in cells incubated with 100 nM PMA in comparison with 5 nM and untreated control. Expression levels of CD14 and CD206 (data not shown) were lower than for the other markers tested and were minimally impacted by the presence of PMA. (Figure 1). No significant (p>0.05) difference in expression level were observed for the cell surface markers investigated with the length of PMA exposure time. 8,27,28

. No In accordance with the literature, U937 cells differentiated with PMA display phagocytic characteristics significant difference (p>0.05) was observed between PMA treated and PMA naïve cells at each time point tested (Figure 2). However, a significant decrease (p<0.05) in phagocytic activity was observed for all treatments with increasing time in culture. Prolonged viability and growth of PMA naïve and treated U937 cells was observed for 18 days following PMA treatment (Figure 3). No significant difference (p<0.05) was observed for the health and proliferation of PMA naïve cells and cells treated with 5 nM PMA for 24 h or 96 h. In contrast, a significant reduction (p<0.05) in the number of cells present was observed in the 100 nM PMA treatment group in comparison with cells exposed to 5 nM PMA at both 24 h and 96 h exposure times. These results indicate that U937 cells exposed to 100 nM PMA stop proliferating and continue to be viable for at least 18 days after differentiation. Whilst there are conflicting reports in the literature regarding U937 proliferation post differentiation, the low proliferation rate generated in this 29,30 . Morphological differences study may be more representative of resident alveolar macrophages in situ including increased pseudopodia were observed in cells treated with 100 nM PMA in comparison with exposure to 5 nM concentrations (data not shown). This supports the findings that the concentration of PMA used for the differentiation of the U937 cell line to a macrophage-like phenotype is an important factor in determining the characteristics of the cell.

175


Drug Delivery to the Lungs 27, 2016 – Abigail Martin et al.

Figure 1: Expression profile of CD markers in U937 cells incubated with PMA. U937 cells were seeded 5 x 105 cells per 100 µl on a 96 well plate and incubated with 0 nM, 5 nM or 100 nM PMA for (A) 24 h, (B) 48 h, (C) 72 h and (D) 96 h followed by a 24 h rest period in fresh culture medium. Expression of CD11a, CD11b, CD14 and CD36 was measured by a flow cytometry using a direct staining method alongside relevant isotype controls. Data is presented as the mean +/- standard deviation of three experiments performed in triplicate. * indicates p<0.05.

Figure 2: Phagocytic behaviour in U937 cells treated with PMA. U937 cells were seeded 5 x 105 cells per 100 µl on a 96 well plate and incubated with 0 nM, 5 nM or 100 nM PMA for (A) 24 h, (B) 48 h, (C) 72 h and (D) 96 h followed by a 24 h rest period in fresh culture medium. Phagocytosis was assessed by incubating cells with 0.01 %v/v FITC-tagged 1.0 µm beads for 24 h in fresh culture media analysed using flow cytometry. Data is presented as the mean +/- standard deviation of three experiments performed in triplicate.

Figure 3: Viability of U937 cells after differentiation with PMA. Cells were seeded at 5x105 cells per 100 µl on a 96 well plate and incubated with no PMA (control), 5 nM or 100 nM PMA for either (A) 24 h or (B) 96 h followed by a 24 h rest period in fresh culture medium. Cell viability was assessed using Guava ViaCount assay where day 1 was assigned 24 h after media change. Data is presented as the mean +/- standard deviation of three experiments performed in triplicate.

176


Drug Delivery to the Lungs 27, 2016 – Abigail Martin et al.

Conclusion We have shown that modifications in the differentiation protocol for U937 cells result in cells with altered levels of cell differentiation markers, proliferation characteristics and morphology. Our observations suggest that the concentration of PMA that U937 cells are exposed to has a more significant impact on cell characteristics than the length of incubation. This highlights the need for an optimised and standardised protocol for PMAdifferentiation of U937 cells to allow for comparable research between laboratories. Our results suggest that exposure to 100 nM PMA for 48-72 h may provide the optimal conditions for the generation of an alveolar macrophage-like phenotype for U937 cells based on morphology, cell proliferation after differentiation, phagocytic activity and cell surface marker expression data presented. Future work will characterise U937 responses as part of robust in vitro immunocompetent co-culture model for the safety assessment of new candidate inhaled medicines. References 1.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

Lewis DJ, Williams TC, Beck SL. Foamy macrophage responses in the rat lung following exposure to inhaled pharmaceuticals: a simple, pragmatic approach for inhaled drug development. J Appl Toxicol. 2014;34(4):319–31. Prieto J, Eklund A, Patarroyo M. Regulated expression of integrins and other adhesion molecules during differentiation of monocytes into macrophages. Cell Immunol. 1994 Jun;156(1):191–211. Alessio M, De Monte L, Scirea A, Gruarin P, Tandon NN, Sitia R. Synthesis, processing, and intracellular transport of CD36 during monocytic differentiation. J Biol Chem. 1996 Jan 19;271(3):1770–5. Ruijters EJB, Haenen GRMM, Weseler AR, Bast A. The cocoa flavanol (−)-epicatechin protects the cortisol response. Pharmacol Res. 2014 Jan;79:28–33. Hass R, Lonnemann G, Mannel D, Topley N, Hartmann A, Köhler L, et al. Regulation of TNF-α, IL-1 and IL-6 synthesis in differentiating human monoblastoid leukemic U937 cells. Leuk Res. 1991 Jan;15(5):327–39. Minafra L, Di Cara G, Albanese NN, Cancemi P. Proteomic differentiation pattern in the U937 cell line. Leuk Res. 2011 Feb;35(2):226– 36. Hewison M, Brennan a, Singh-Ranger R, Walters JC, Katz DR, O’Riordan JL. The comparative role of 1,25-dihydroxycholecalciferol and phorbol esters in the differentiation of the U937 cell line. Immunology. 1992;77:304–11. Minta J, Pambrun L. In vitro induction of cytologic and functional differentiation of the immature human monocytelike cell line U-937 with phorbol myristate acetate. Am J Pathol. 1985;(119):111–26. Yamamoto T, Sakaguchi N, Hachiya M, Nakayama F, Yamakawa M, Akashi M. Role of catalase in monocytic differentiation of U937 cells by TPA: hydrogen peroxide as a second messenger. Leukemia. Nature Publishing Group; 2009 Apr;23(4):761–9. Garc a A, Serrano A, Abril E, Jimenez P, Real LM, Cantón J, et al. Differential effect on U937 cell differentiation by targeting transcriptional factors implicated in tissue- or stage-specific induced integrin expression. Exp Hematol. 1999 Feb;27(2):353–64. Ralph P, Williams N, Moore MAS, Litcofsky PB. Induction of antibody-dependent and nonspecific tumor killing in human monocytic leukemia cells by nonlymphocyte factors and phorbol ester. Cell Immunol. 1982 Aug;71(2):215–23. Stoppelli MP, Cortit A, Soffientinit A, Blasit F, Assoian RK. Differentiation-enhanced binding of the amino-terminal fragment of human urokinase plasminogen activator to a specific receptor on U937 monocytes. Biochemistry. 1985;82(August):4939–43. Shepherd MC, Baillie GS, Stirling DI, Houslay MD. Remodelling of the PDE4 cAMP phosphodiesterase isoform profile upon monocytemacrophage differentiation of human U937 cells. Br J Pharmacol. 2004 May;142(2):339–51. Verhoeckx KCM, Bijlsma S, de Groene EM, Witkamp RF, van der Greef J, Rodenburg RJT. A combination of proteomics, principal component analysis and transcriptomics is a powerful tool for the identification of biomarkers for macrophage maturation in the U937 cell line. Proteomics. 2004 Apr;4(4):1014–28. Sintiprungrat K, Singhto N, Sinchaikul S, Chen S-T, Thongboonkerd V. Alterations in cellular proteome and secretome upon differentiation from monocyte to macrophage by treatment with phorbol myristate acetate: insights into biological processes. J Proteomics. Elsevier B.V.; 2010 Jan 3;73(3):602–18. Sordet O, Bettaieb A, Bruey JM, Eymin B, Droin N, Ivarsson M, et al. Selective inhibition of apoptosis by TPA-induced differentiation of U937 leukemic cells. Cell Death Differ. 1999;6(4):351–61. Sordet O, Hermine O, Vainchenker W, Garrido C, Solary E, Dubrez-daloz L. Specific involvement of caspases in the differentiation of monocytes into macrophages. Hematopoiesis. 2002;100(13):4446–53. Sajjadi FG, Takabayashi K, Foster AC, Domingo RC, Firestein GS. Inhibition of TNF-a Expression by Adenosine. J Immunol. 1996;(156):3435–42. Tenney DJ, Morahan PS. Effects of differentiation of human macrophage-like u937 cells on intrinsic resistance to herpes simplex virus type 1. Am J Pathol. 1987;139(9):3076–83. Vogel CF a, Garcia J, Wu D, Mitchell DC, Zhang Y, Kado NY, et al. Activation of inflammatory responses in human U937 macrophages by particulate matter collected from dairy farms: an in vitro expression analysis of pro-inflammatory markers. Environ Health. BioMed Central Ltd; 2012 Jan;11(1):17. Twomey B, McCallum S, Isenberg D, Latchman DS. Elevation of heat shock protein synthesis and hsp gene transcription during monocyte to macrophage differentiation of U937 cells. Vol. 93, Clinical and Experimental Immunology. Wiley-Blackwell; 1993. p. 178. Kuroda A, Sugiyama E, Taki H, Mino T, Kobayashi M. Interleukin-4 Inhibits the Gene Expression and Biosynthesis of Cytosolic Phospholipase A2 in Lipopolysaccharide Stimulated U937 Macrophage Cell Line and Freshly Prepared Adherent Rheumatoid Synovial Cells. Biochem Biophys Res Commun. 1997 Jan 3;230(1):40–3. Fukunaga M, Tsuruda K. Actinobacillus actinomycetemcomitans induces lethal effects on the macrophage-like human cell line U937. Oral Microbiol Immunol. 2001;16(5):284–9. Matheson LA, Fairbank NJ, Maksym GN, Paul Santerre J, Labow RS. Characterization of the Flexcell Uniflex cyclic strain culture system with U937 macrophage-like cells. Biomaterials. 2006 Jan;27(2):226–33. Harris P, Ralph P, Division S. Human Leukemic Models of Myelomonocytic Development : A Review of the HL-60 and U937 Cell Lines. J Leukoc Biol. 1985;422:407–22. Martin A, Hutter V, Murnane D, Chau D, Brown M. In vitro assessment of alveolar macrophage characteristics for the application of inhaled pharmaceuticals. In 2015. p. Drug delivery to the lungs conference. Daigneault M, Preston JA, Marriott HM, Whyte MKB, Dockrell DH. The identification of markers of macrophage differentiation in PMAstimulated THP-1 cells and monocyte-derived macrophages. PLoS One. Public Library of Science; 2010 Jan 13;5(1):e8668. Taniguchi K, Hikiji H, Okinaga T, Hashidate-Yoshida T, Shindou H, Ariyoshi W, et al. Essential Role of Lysophosphatidylcholine Acyltransferase 3 in the Induction of Macrophage Polarization in PMA-Treated U937 Cells. J Cell Biochem. 2015 Dec;116(12):2840–8. Davies LC, Jenkins SJ, Allen JE, Taylor PR. Tissue-resident macrophages. Nat Immunol. 2013 Oct;14(10):986–95.

177


Drug Delivery to the Lungs 27, 2016 – Mikael Brülls. Development of novel efficient rodent nose-only inhalation exposure systems Mikael Brülls 1

1

AstraZeneca R&D, AstraZeneca R&D Gothenburg, SE-431 83 Mölndal, Sweden

Summary It was decided within AstraZeneca R&D to develop novel efficient passive nose-only inhalation exposure systems to enable inhalation administration to rodents in early preclinical development phase. It was decided from the start, when a first novel inhalation system was developed, that it would be a nose-only inhalation exposure system to be used with commercially available restraining tubes. The most important design feature of the novel inhalation exposure system was to develop a closed exposure chamber instead of a flow-pass system in order to improve the utilization of the drug compound. The limited amount of test material available in this early development phase together with practical limitations regarding time and resources for formulation development precluded the possibility to use dry powder aerosols and nebulization was therefore selected as the aerosol generation technique for the novel inhalation exposure system. It was decided to use a vibrating mesh nebulizer with the first novel exposure system. The in-vitro investigations made with a first real prototype system was successful and it was therefore decided to manufacture two fully equipped exposure units for in-vivo use. The first novel inhalation system did unfortunately not meet the expected improvement in utilization of drug compound in in-vivo use. Two other new novel inhalation systems were therefore developed at AstraZeneca R&D. The two new systems have been shown to be significantly more efficient than the first system and also proven to be able to replace intratracheal instillation in the early development phase at AstraZeneca R&D. Introduction It was decided within AstraZeneca R&D to develop novel efficient passive nose-only inhalation exposure systems to enable inhalation administration to rodents in early preclinical development phase when new compounds are continuously synthesized and tested to rank compounds in order to identify a lead structure or optimize a lead compound. A direct administration method, intratracheal instillation, was used instead because no commercially available passive inhalation exposure system existed that could deliver appropriate lung doses with the limited amount of test material available in this early preclinical phase at AstraZeneca R&D. Concerns regarding intratracheal instillation There are a number of concerns regarding intratracheal instillation. It is invasive delivery of a large amount of vehicle as well as it is applying a dose rate substantially greater than that which would have occurred during passive inhalation and it poses the risk of overwhelming lung defences and causing effects that are not relevant [1] . Other problems are that the intratracheal intubation may cause local irritation and that an unknown amount of [2] the drug may be coughed up or swallowed . Hatch et al showed that mice retained 70% of the instilled [3] radiolabeled albumin in the lungs while the head and carcass (mainly the stomach) contained the other 30% . These figures are in agreement with a similar study in rats performed at AstraZeneca R&D. Perhaps the most consistently reported disparity between passive inhalation and intratracheal instillation relates to the distribution of the deposited dose. Passive inhalation results in a homogenous distribution throughout the lungs, whereas intratracheal instillation generally results in less homogeneity of the dose distribution in the alveolar region and [1] can result in focally high doses of material . Zecchi et al showed in an imaging study that the deposited dose was more concentrated around central airways in intratracheal instillation in comparison with passive inhalation, in [4] which the dose was more uniformly distributed among all the lung sections, reaching also parenchymal regions . A study performed at AstraZeneca where rats were administered a blue dye either via passive inhalation, where the aerosol was generated via nebulisation or via intratracheal instillation also showed that the passive inhalation generated a uniform lung deposition but this was not the case for the intratracheal instillation, see figure 1. The lung was uniformly coloured light blue by the dye after passive inhalation whereas the dye was deposited centrally and patchy after intratracheal instillation. A

B

C

D

Figure 1 - Four pictures of lungs from rats administered with blue dye via passive inhalation and intratracheal instillation. A = control (no blue dye administered), B = passive inhalation, C = intratracheal instillation, D = dissected central airways from an intratracheal instilled lung

178


Drug Delivery to the Lungs 27, 2016 - Development of novel efficient rodent nose-only inhalation exposure systems A study to compare the effect of budesonide on sephadex induced lung edema in rats after intratracheal instillation compared with passive inhalation was performed at AstraZeneca R&D. The results showed that the potency of budesonide decreased approximately a tenfold when administered via intracheal instillation in comparison with passive inhalation. Design of the first novel exposure chamber It was decided from the start, when a first novel rodent inhalation system was developed, that it would be a noseonly inhalation exposure system to be used with commercially available restraining tubes, see figure 2. A wholebody inhalation exposure system was considered inappropriate because the contamination of the fur when using such a system leads to a very high extent of oral drug administration.

Figure 2 - A rat in a commercially available restraining tube for nose-only inhalation exposure.

Commercially available nose-only rodent inhalation exposure systems are based on the principle of continuously generating aerosol on a carrier air supply and the flow of aerosol is equally divided so that a fraction of the total flow passes in front of the nose of each animal. Each animal will continuously be exposed to freshly generated aerosol. This secures that there is constant quality of the inhaled aerosol but the drug compound will not be utilized efficiently in the system because a major part of the drug will pass through the system without being inhaled. The most important design feature of the novel inhalation exposure system was therefore to develop a closed exposure chamber instead of a flow-pass system in order to improve the utilization of the drug compound. Other design features that were decided for the first novel inhalation system was that ten restraining tubes would be connected to the closed exposure chamber, that the restraining tubes would be positioned at the same level from the floor of the chamber and that a impeller be positioned in the bottom of the chamber. The impeller would decrease the settling velocity of the aerosol by creating a lifting airflow and thereby increase the residence time of the aerosol in the chamber. The geometrical design of the exposure chamber was first evaluated using computational fluid dynamic simulations on a few different theoretical models, see figure 3.

Figure 3 - Three different theoretical models and a plot of the hollow walls of the closed exposure chamber that were evaluated using computational fluid dynamic simulations.

One of the theoretical models was selected and a first real prototype system was manufactured for in-vitro investigations. It was necessary to solve the problem of keeping an acceptable atmosphere in the closed chamber, i.e. keeping acceptable levels of oxygen and carbon dioxide. One way to solve this issue would be to use pure oxygen in the closed exposure chamber initially and combine this with a short enough administration time. This solution was used successfully by Wu et al when a radioactive aerosol generation and inhalation system was developed and [5] used for an imaging study . Another method was however selected for the design of the first closed exposure chamber system developed by AstraZeneca and that was to use soda lime as a carbon dioxide absorbent that would preclude a build-up of high carbon dioxide levels. Ten rats results in an influx of approximately 60 ml of carbon dioxide per minute into the closed exposure chamber. It was investigated how much absorbent that was needed to avoid build-up of high levels of carbon dioxide and where in the chamber the adsorbent should be placed. It was concluded that 0.5 to 1 kg of soda lime should be sufficient for administration times up to 30 minutes and it should be available both in the walls as well as on the floor of the exposure chamber.

179


Drug Delivery to the Lungs 27, 2016 – Mikael Brßlls. The system was aimed to be used for an administration time of up to ten minutes per group of dosed animals. A ratio of 2:1 of absorbent in the walls compared with the floor was found to be appropriate. The walls were made hollow with a net on the inner wall in order to hold the soda lime in place, see figure 6. Vibrating mesh nebulizer used as aerosol generator in the first novel rodent inhalation exposure system The limited amount of test material available in this early preclinical phase together with practical limitations regarding time and resources for formulation development precluded the possibility to use dry powder aerosols and nebulization was therefore selected as the aerosol generation technique for the novel exposure system. The fact that it was a closed chamber system precluded the use of a jet nebulizer because it delivers a volume flow of aerosol, i.e. droplets in a flow of air. It was therefore necessary to use either a vibrating mesh nebulizer or an ultrasonic nebulizer. It was decided to use a vibrating mesh nebulizer, the e-Motion from PARI GmbH. The output rate from this nebulizer was one of the highest from any commercially available nebulizer and it was desired to minimize the exposure time and thus obtain a high aerosol concentration in the chamber as fast as possible. The residual volume in this nebulizer was minimal and this was beneficial from a utilization of drug compound perspective. The vibrating mesh nebulizer generates the aerosol through the pumping action of a perforated plate that vibrates at a high frequency. The plate contains more than a thousand precision-formed holes, surrounded by a vibrating piezoelectric element, which makes the plate vibrate at a high frequency. During each vibration the plate is displaced about one micrometer and it will act as a micro pump, drawing liquid through the holes to form [6] micron sized droplets . A Computational Fluid Dynamic (CFD) investigation was made in order to get a better understanding of the aerosol generation from the vibrating mesh nebulizer into the closed chamber. In order to be able to compare the CFD simulated aerosol generation with a real aerosol generation a vibrating mesh nebulizer was connected to the top of a transparent holding chamber with on open outlet in the bottom and this system was studied both theoretically and visually. Due to the complex nature of the aerosol generation from the vibrating mesh nebulizer the focus of the study was to get a basic understanding of the physics of the system. There are a lot of changes in the flow characteristics over very small time scales when the aerosol is generated and it was therefore difficult to study the spray with the naked eye and the aerosol was therefore studied with the aid of a high speed camera. A rounded plume and tumultuous swirls could be observed. It was concluded that the droplets from the nebulizer entrained the air, which induced a flow that propelled the droplets beyond the calculated individual stopping distances i.e. there was a two-way momentum coupling between the droplets and the air. The entrainment of the initially stagnant air was so extensive that turbulence was generated in the air and a turbulence model was needed to model the system. When modelling the droplets sprayed into the chamber, calculations were made to roughly estimate the amount of evaporation that could occur to saturate the air in the chamber. It could be concluded that a relatively small amount of water was needed to saturate the air and only a small amount of the water sprayed from the nebulizer would therefore evaporate. To simplify the calculations evaporation was not taken into consideration in the simulations. Droplets that hit the walls of the holding chamber were assumed to stick to the walls and wet them. The CFD calculations were evaluated by comparing the simulated results with the visual observations using the high speed camera and it was concluded that there was a good qualitative agreement between the calculated results and the visual observations. The extent of turbulence was estimated, see figure 4. The turbulence was calculated to be generated mostly at the inlet where the spray of droplets was most dense and that it created swirls in the air in the chamber.

A

B

Figure 4 - A. Contour plot of the turbulent kinetic energy in the holding chamber. B. Illustration of swirls in the air velocity vectors

180


Drug Delivery to the Lungs 27, 2016 - Development of novel efficient rodent nose-only inhalation exposure systems

In-vivo evaluation of the first rodent inhalation exposure system The in-vitro investigations made with the first real exposure system was successful and it was therefore decided to manufacture two fully equipped units for in-vivo use, see figure 5. An in-vivo test of a fully equipped unit was performed using Wistar rats. Mometazone furoate 10 mg/ml nanosuspension was nebulized continuously for three minutes into the chamber. The rats continued to inhale and exhale from the chamber for an additional two minutes after the nebulizer had been stopped, i.e. the total administration time was five minutes per group of animals. The rats were then dismounted and sacrificed immediately, by an intraperitoneal injection of pentobarbital sodium. The lungs were dissected and homogenized. The substance concentration was determined by LC-MS-MS. The amount of Mometazone suspension nebulized, was determined by weighing the nebulizer before and after each test-run. The result was that 0.4 to 1.1% of the nebulized amount of drug deposited in the lungs of the rats. This was significantly lower than expected from the in-vitro investigations of the system.

Figure 5 - A first novel functional fully equipped nose-only aerosol exposure system for in-vivo use

Further development – redesign of the novel rodent inhalation exposure system The first novel rodent inhalation exposure system did not meet the expected improvement in utilization of drug compound in in-vivo use and was therefore not a suitable replacement for intratracheal instillation in the early development phase. The exposure system was therefore thoroughly redesigned and two new novel closed systems with significantly different designs were developed. An individual rodent system was developed and initially used to administer a radioactive tracer in a study of mucociliary clearance using SPECT imaging. A four-rodent system was also developed and it was initially used for pharmacokinetic studies. Other uses of these two systems will be tested and evaluated. The most important design differences was that the new systems were significant smaller, lacked an impeller to increase the aerosol residence time and also lacked carbon dioxide absorbent that precluded a build-up of high carbon dioxide levels. A vibrating mesh nebulizer, Aeroneb Lab from Aerogen, was used initially with both of the two new systems, but also a jet nebulizer, Cirrus2 from Intersurgical Ltd., has been used successfully with both systems and the use of an ultrasonic nebulizer is currently being evaluated for use. The two new systems have been thoroughly tested and evaluated in both in-vitro and in-vivo studies and they have been shown to be significantly more efficient than the first exposure system and also proven to be able to replace intratracheal instillation in early development phase at AstraZeneca R&D. References 1

Driscoll K E, Costa D L, Hatch G, Henders R, Oberdorster G, Salem H, Schlesiger R B: Intratracheal Instillation as an Exposure Technique for the Evaluation of Respiratory Tract Toxicity : Uses and Limitations, Toxicol Sci 2000; 55; pp 24-35.

2

Cryan S. Sivadas N. Garcia-Contreras L: In vivo animal models for drug delivery across the lung mucosal barrier, Adv Drug Deliv Rev 2007; 59; pp 1133-1151.

3

Hatch G E, Slade R, Boykin E, Hu P C, Miller F J, Gardener D E: Correlation of effects of inhaled versus intratracheally injected metals on susceptibility to respiratory infection in mice, Am Rev Respir Dis 1981; 124: pp 167-173.

4

Zecchi R, Trevisani M, Pittelli M, Pedretti P, Manni M E, Pieraccini G, Pioselli B, Amadei F, Monetia G, Catinellac S: Impact of drug administration route on drug delivery and distribution into the lung: an imaging mass spectrometry approach, Eur. J. Mass Spectrom 2013; 19; pp 475–482.

5

Wu Y, Kotzer C J, Makrogiannis S, Logan G A, Haley H, Barnette M S, Sarkar S K: A Noninvasive [99mTc]DTPA SPECT/CT Imaging Methodology as a Measure of Lung Permeability in a Guinea Pig Model of COPD, Mol Imaging Biol 2011; 13: pp 923-929.

6

Lee S H., Nano spray drying: A novel method for preparing protein nanoparticles for protein therapy, Int J Pharm, 2011; 17: pp192-200.

181


Drug Delivery to the Lungs 27, 2016 - Nicoletta Zallocco et al. Chitosan nanocarrier systems for delivery of pneumococcal vaccine via nebulization 1,4

2

2

2

1

Nicoletta Zallocco , Eliane N. Miyaji , Viviane M. Goncalves , Douglas B. Figueiredo , Imran Y. Saleem , 3 Ronan MacLoughlin 1

School of Pharmacy & Biomolecular Sciences, Liverpool John Moores University, Liverpool, UK. 2 Centro de Biotecnologia, Instituto Butantan, Sao Paulo, Brazil. 3 Aerogen Ltd. Galway Business Park, Dangan, Galway, Ireland. 4 School of Pharmacy, University of Camerino, Camerino, Italy.

Summary Background: Streptococcus pneumoniae enters the human body via the lungs and is the main cause of pneumonia. The current vaccine is administered via the parenteral route and has limited protection in the lungs and towards the different strains of S. pneumoniae. Pulmonary vaccination is an attractive alternative targeting dendritic cells (DCs) in the lungs initiating an immune response. To optimise targeting and uptake by DCs, chitosan (CHT) and chitosan hydrochloride (CHCl) nanoparticles (NPs) were loaded with pneumococcal surface protein A (PspA) and delivered via nebulization for pulmonary delivery. Methods: CHT/PspA and CHCl/PspA NPs were prepared by ionic gelation method. The formulations were then collected by centrifugation (35,000rpm, 10min, 25°C) to remove unbound PspA and re-suspended in 1mL of 0.9% NaCl for nebulization. NPs were characterised in terms of particle size, surface charge, polydispersity index (PDI), drug loading after centrifugation, post nebulization dose, particle morphology, in vitro release and cell viability (DCs) after 24h. Results: Results indicated the size of CHT/PspA NPs (236.36±5.57nm, +10.78±0.47mV, PDI 0.366±0.045), CHCl/PspA NPs (372.633±7.02nm, +31.8±1.16mV, PDI 0.216±0.025) were suitable for targeting DCs. PspA loading after centrifugation and nebulization (CHT/PspA NPs: 9.92µg/mg,9.69 µg/mg), and (CHCl/PspA NPs: 1.62µg/mg, 1.29 µg/mg) indicated little loss. The release studies revealed continuous release of protein after 24h (CHT/PspA NPs: 26%, CHCl/PspA NPS: 99%). The NPs appear to be well tolerated by DCs. Conclusion: The results indicated chitosan NPs could be a promising candidate for pulmonary delivery of pneumococcal vaccine. Introduction The increase in mortality and morbidity, mainly correlated with pulmonary diseases, such as pneumonia, has [1] gained significant attention to produce a non-invasive approach to enhance immunogenity . The main portal of entry of Streptococcus pneumoniae into the body is the respiratory tract. Among the non-invasive routes of delivery, the pulmonary route can overcome some challenges such as invasiveness, low stability and integrity of the antigen. The current vaccine is administered via the parenteral route and has limited protection towards the different strains of S. pneumoniae and in the lungs. A new formulation is required to achieve higher efficacy and [2] level of protection against mucosal diseases . In order to generate a stronger immune response, particulate antigens have been preferred to soluble antigen, and research has focused on using nanoparticles (NPs) as a delivery carrier and potential adjuvant. In fact, NPs of appropriate size (150–500 nm) and charge (−40 mV - +35 mV) are able to enter the lymphatics and travel to [3] DCs within the lymph nodes, while larger particles (> 500 nm) are not so efficient and selective . Surface charge of NPs are known to play an important role in determining the cellular uptake, and cationic NPs compared to [1] anionic or neutral, have better interactions with the negatively charged cell membrane . Nebulization of aqueous NPs suspensions seems to be an appropriate delivery mechanism to achieve deposition of NPs in the lungs to [4] target DCs . Chitosan has attracted particular interest as a biodegradable material for mucosal delivery systems, especially for [5] its important capacity to enhance drug permeability and absorption at mucosal site . Despite all its positive biological properties, it has a major drawback; it is soluble in acidic conditions, which may affect the stability and integrity of some antigens. As an alternative, chitosan can be functionalised for the synthesis of chitosan hydrochloride (CHCl), a water soluble derivative that retains its mucohadesive and uptake properties, and can be considered as a promising candidate for vaccine delivery. Aim The aim of this study was to prepare and compare different chitosan NPs (chitosan in acetic acid (CHT) and chitosan hydrochloride (CHCl)) complexed pneumococcal surface protein A (PspA) in terms of nebulization for targeting lung DCs.

182


Drug Delivery to the Lungs 26, 2015 - Chitosan nanocarrier systems for delivery of pneumococcal vaccine via nebulization Methods: Nanoparticle preparation Chitosan NPs were prepared by ionic gelation method. Briefly, 10 mg of CHT (Sigma, UK) were dissolved in 5 ml of 1% acetic acid solution, while 10 mg of CHCl (Heppe Medical Chitosan GmbH, Halle, Saale) were dissolved in 5 ml of water. The pH was adjusted to 6 using NaOH 0.5M. Subsequently, various amount of TPP aqueous solution with the ratio CHT:TPP 3:1, with PspA (0.1 mg/ml) incorporated in TPP were added drop-by-drop to the above solutions under magnetic stirring (250 RPM) at room temperature for 1 hour. The colloidal suspension was centrifuged at 35,000 rpm for 10 min at 25 °C and washed twice with distilled water to remove unbound PspA. Nanoparticle characterization Particle size, Zeta-Potential and Polydispersity index (PDI): the NPs were characterised using a laser particle size analyser (ZetasizerNano ZS, Malvern Instruments Ltd., UK). 1 ml of NPs suspensions was loaded into a cuvette and the measurements were recorded at 25 °C (mean ± SD, n=3). Drug loading (DL): PspA loaded in the NPs was determined by measuring the amount of protein remaining in the supernatant after centrifugation, using a QuanicPro bicinchoninic acid (BCA) protein assay kit (micro BCA assay, Sigma-Aldrich) by UV spectroscopy at 562 nm (mean ± SD, n=3). Empty CHT and CHCl NPs were used as a control. Post nebulization dose: Loaded NPs after centrifugation were resuspended in 1 ml of 0.9% NaCl and nebulized for 5 minutes using Aerogen Pro Lab Nebulizer (Aerogen, Galway, Ireland), a vibrating mesh nebulizer with multiple apertures to generate fine-particles and low velocity aerosol. The post nebulization dose was calculated to determine the compatibility of the nebulizer with the chitosan NPs/PspA formulation. Particle morphology: 50 µl of loaded NPs were mounted on gold coated aluminium stubs (EmiTech K 550X Gold Sputter Coater, 25mA for 3 min) and visualised by scanning electron microscopy (SEM). In vitro Release studies: Loaded NPs after centrifugation were dispersed in 2 ml of PBS, pH 7.4. The samples TM were incubated at 37 °C and left rotating at 20 RPM on a HulaMixer Sample Mixer (Life Technologies, Invitrogen, UK). The samples were then centrifuged after 30 min, 1, 2, 4 and 24 hours at 17,000 x g for 10 min and 1 ml of the supernatant was removed and replaced with fresh medium. The supernatant was analysed with micro BCA assay. [2]

Cell Viability study: The toxicity of the NPs were investigated in DCs (ATCC, JAWS II) using MTT assay . The treatment was carried out for 4 hours at 37 °C, followed by the addition of MTT for 2 hours. 100 µl PspA NPs dispersions in complete medium were added at concentration (0.001–1 mg/ml) and 10% dimethyl sulfoxide (DMSO) as a positive control. The absorbance was measured at 570 nm using a plate reader (Molecular Devices, SpectraMAX 190) and the percentage of viable cells was calculated as the absorbance between NPs complexed with PspA and untreated control DCs. Results and Discussion The ionic gelation method produced particles in the nanometer size range with positive zeta potential values, which are associated with the free amine groups on the chitosan following complexation with TPP (Table 1). Loaded NPs were larger in size than blank NPs, indicating encapsulation of PspA. Moreover, loaded NPs revealed a lower charge than blank NPs. This was due to the presence of the protein, which has a negative charge when the pH (6) is above its pI (4.9). Table1 - The Particle size, PDI and zeta potential for CHT and CHCl non-loaded (blank) and loaded nanoparticles before nebulization (mean ± SD, n=3). Particle Size (nm)

PDI

Z-Potential (mV)

CHT NPs

206.70 ± 3.75

0.370 ± 0.02

+11.9 ± 0.48

CHT/PspA NPs

236.36 ± 5.57

0.366 ± 0.05

+10.78 ± 0.47

CHCl NPs

347.00 ± 3.24

0.189 ± 0.014

+33.20 ± 1.13

CHCl/PspA NPs

372.63 ± 7.02

0.216 ± 0.025

+31.80 ± 1.16

.

183


Drug Delivery to the Lungs 27, 2016 - Nicoletta Zallocco et al. The micro BCA assay after centrifugation and nebulization confirmed the high drug loading (DL) and aerosolised dose in each formulation of CHT and CHCl, as shown in Table 2 Table 2 - Drug loading (DL) of PspA NPs after centrifugation and aerosolised dose after nebulization Concentration of protein (μg/ml)

DL after centrifugation (µg/10 mg NPs)

Post nebulization dose (µg/10 mg NPs)

CHT/PspA NPs

100

99.20

96.90

CHCl/PspA NPs

20

16.24

12.98

A

B

Figure 1 - SEM pictures of CHCl non-loaded NPs (A) loaded (B) NPs after nebulization. Scale bar 30 µm. Owing to their small size and low inertia, NPs are commonly exhaled after inhalation resulting in low doses in the lungs. To overcome this problem, the NPs have been nebulized in order to achieve the appropriate characteristics of size suitable for the pulmonary delivery. In fact, it is known that particles in the size range of 1 to 5 µm are required to reach the respirable airways. SEM analysis (Figure 1) of non-loaded (A) and loaded (B) nanoparticles after nebulization revealed spherical NPs with rough surface morphology. The particle sizes observed by SEM were pproximately 1 µm, indicating the NPs were aggregated post nebulization. 120 CHT

CHCl

% cumulative release

100 80 60 40 20 0 0

5

10 Time (h) 15

20

25

Figure 2 - In vitro release of PspA from CHT (A) and CHCl (B) NPs in phosphate buffer saline (pH 7.4) over 24 hours at 37 °C.

184


Drug Delivery to the Lungs 26, 2015 - Chitosan nanocarrier systems for delivery of pneumococcal vaccine via nebulization CHT and CHCl NPs showed significant difference in release of PspA. The cumulative percentage of PspA released over time from CHT and CHCl NPs (Figure 2) indicated an initial burst release of 8% (CHT) and 20% (CHCl), followed by continuous release of 26% (CHT) and 99% (CHCl) after 24 hours. The much higher burst release of PspA from CHCl NPs rather than the CHT NPs, may be due to the higher charge of the polymer and the water-soluble properties which could result in faster dissolution and release of PspA. 120

CHT

CHCl

Cell Viability (%)

100 80 60 40 20 0 0,001

0,01

0,025

1

Concentration (mg/ml)

10%DMSO

Figure 3 - Dendritic cells (DCs) viability measured by MTT assay after 24 hours exposure to NPs. The cell viability of DCs incubated with CHT/PspA and CHCl/PspA NPs for 4 hours was evaluated by the MTT assay. The NPs (Figure 3) revealed reduced cell viability with an increasing NPs concentration. The NPs showed a 97% cell viability for CHT and 77% for CHCl at 0.001 mg/ml concentration that reduced to 63% (CHT) and 52% (CHCl) at 1 mg/ml concentration. The lower cellular viability of CHCl may be due to the higher positive zeta potential that has been attributed to the interactions with the plasma membrane. Conclusion: The results obtained indicated high PspA loading in the CHT and CHCl formulations, with almost no loss post nebulization. The high delivery efficiency of the vibrating mesh nebulizer has the advantage of reducing the vaccine dose and to improve the cost-effectiveness compared to the conventional ultrasonic nebulizers. In addition, our results indicated CHT and CHCl have a low in vitro cell toxicity towards DCs which is required for pulmonary drug delivery carriers. Further studies would focus on the in vivo immune studies. References: 1. Kunda, N.K., et al., Dry powder pulmonary delivery of cationic PGA-co-PDL nanoparticles with surface adsorbed model protein. Int J Pharm, 2015. 492(1-2): p. 213-22. 2. Kunda, N.K., et al., Pulmonary dry powder vaccine of pneumococcal antigen loaded nanoparticles. Int J Pharm, 2015. 495(2): p. 903-12. 3. Heidi M Mansour , Y.-S.r., Xiao wu Nanomedicine in pulmonary delivery. International Journal of Nanomedicine, 2009. 4: p. 299-319. 4. Ruge, C.A., et al., Disintegration of nano-embedded microparticles after deposition on mucus: A mechanistic study. Colloids Surf B Biointerfaces, 2016. 139: p. 219-27. 5. Kunda, N.K., et al., Nanocarriers targeting dendritic cells for pulmonary vaccine delivery. Pharm Res, 2013. 30(2): p. 325-41.

185


1

Drug Delivery to the Lungs 27, 2016 - Mohammad AM Momin et al. Enrichment of the surface of spray-dried powder particles with a hydrophobic material to improve their aerosolization 1

1

2

3

Mohammad AM Momin , Ian G Tucker , Colin Doyle , John Denman , & Shyamal C Das

1

1

2

New Zealand's National School of Pharmacy, University of Otago, Dunedin 9054, Dunedin, New Zealand 3 The University of Auckland, 20 Symonds Street, Auckland, New Zealand, Future Industries Institute, University of South Australia, Mawson Lakes, SA 5095, Australia

Summary High dose delivery of drugs to the lung using a dry powder inhaler is an emerging approach to combat multi-drug resistant local infections. To achieve high dose delivery, high aerosolization of the powders is important. Hygroscopic drugs are a particular challenge since moisture uptake and resultant agglomeration reduces aerosolization. We hypothesize that, under appropriate conditions, a solution of hygroscopic drug and a hydrophobic compound can be spray-dried to produce particles with surfaces rich in the hydrophobic compound and this surface enrichment improves the aerosolization of these spray-dried powders. A hygroscopic drug, kanamycin was combined with a hydrophobic drug, rifampicin (at a ratio of 3:2 w/w), and spray-dried using a Buchi Mini Spray-dryer under optimized conditions. X-ray photoelectron spectroscopy (XPS) and time-of-flight secondary ion mass spectrometry (ToF-SIMS) were used to study the surface composition. The in vitro aerosolization efficiency was investigated using a next generation impactor (NGI). The powders were also investigated for physicochemical properties. All the spray-dried powders were within the inhalable range (1.1-5.9 µm). XPS and ToF-SIMS showed the surface of the powder was enriched with rifampicin (approximately 98% rifampicin). The fine particle fraction (FPF) significantly increased from 31.3 ± 1.8% (kanamycin-only) to 76.6 ± 3.8% (kanamycin/rifampicin combination). All the spray-dried powders were amorphous in nature. Surface enrichment of kanamycin by hydrophobic rifampicin improves aerosolization. Improved aerosolization may be helpful to deliver high dose to the lung and combat multi-drug resistant local infections. Introduction [1]

Drug-resistance is a major challenge in the treatment of pulmonary infections . Sub-therapeutic levels of drugs [2] in the infected lung is the principal cause of the development of drug-resistance , so high dose delivery to the target site is essential. Dry powder inhalation (DPI) is a useful delivery route and has been used for many years for low dose drugs; but high dose delivery using DPI depends on the production of highly aerosolizable powders [3-5] . The hygroscopic nature of a drug is a risk for good aerosolization efficiency due to with appropriate properties [6] the tendency of moisture uptake and resultant agglomeration . This work aims to develop an approach to improve the aerosolization of hygroscopic drugs by hydrophobic enrichment of the surface of particles which are made by spray-drying. Kanamycin sulfate was used as a hygroscopic drug and rifampicin as a hydrophobic compound. Kanamycin alone is used for drug-resistant tuberculosis and the kanamycin/rifampicin combination is synergistic against Mycobacterium avium-intracellulare [7] complex . In this study, the combination powder of kanamycin sulfate and rifampicin was prepared at a mass ratio of 3:2 based upon the currently recommended dosage regimens of both drugs. Experimental methods Materials Kanamycin sulphate and rifampicin were purchased from Hangzhou Dayangchem Co., Ltd., Zhejiang, China. Acetonitrile and ethanol (High performance liquid chromatography, HPLC, grade) were purchased from Merck, Germany. Sodium dihydrogen phosphate, phenylisocyanate, triethylamine, orthophosphoric acid (analytical reagent grade) and silicone oil (viscosity 10 cSt) were purchased from Sigma–Aldrich, St. Louis, USA. Size 3 hard gelatin capsules were kindly donated by Capsugel Co., Ltd., Tokyo, Japan. Fresh Milli-Q water was collected and filtered through 0.45 μm membrane filter before use. Preparation of powders Powder particles were produced using a Buchi B-290 Mini Spray-Dryer (Buchi Labortechnik AG, Flawil, Switzerland) with a high performance cyclone in a closed-mode. Feed solutions (kanamycin with or without rifampicin) were prepared in a co-solvent system of ethanol and water (70:30, v/v). Rifampicin was first dissolved in ethanol; kanamycin sulfate was dissolved in water. The two solutions were then mixed (at rifampicin and kanamycin ratio of 40 and 60% w/w) and sonicated for 5 min to give a total concentration of 0.67% w/v. The formulations were spray-dried using the aspiration 100%, drying gas flow rate 670 L/h (55 mm height), inlet temperature 170 ºC, outlet temperature 89–92 ºC, pump feeding rate 2 mL/min, and the nozzle diameter 0.7 mm. The spray-dried powders from the sample collector were transferred into screw-capped glass vials and stored at room temperature and used within 3 days.

186


Drug Delivery to the Lungs 27, 2016 - Enrichment of the surface of spray-dried powder particles with a hydrophobic material to improve their aerosolization In vitro aerosolization The in vitro aerosolization performance of the spray-dried powders was determined using a next generation impactor (NGI) equipped with Copley HCP5 vacuum pump and Copley TPK 2000 critical flow controller (Copley Scientific Ltd., Nottingham, UK). Approximately 20 mg of powder samples were dispersed and collected from different stages (1 to 7 and micro-orifice collector, MOC) and analysed using a validated high-performance liquid chromatography (HPLC) method. The aerosolization efficiency was represented as fine particle fraction (FPF), defined as a quotient of drug deposited on stages 2 to MOC and drug emitted from the inhaler device (emitted dose, ED). The Copley Inhaler Testing Data Analysis Software (CITDAS 3.10) was used to calculate the mass median aerodynamic diameter (MMAD) and geometric standard deviation (GSD). Surface composition Surface compositions (to a depth of 5-10 nm) of the single-drug and combination powders were evaluated using X-ray photoelectron spectroscopy (XPS), (AXIS Ultra DLD Spectrometer, Kratos Analytical Ltd., Manchester, UK). The elemental distribution on the surface (top 1-2 nm) was determined using time-of-flight secondary ion mass spectrometry (ToF-SIMS), (PHI TRIFT V nanoTOF instrument, Physical Electronics Inc., Chanhassen, MN, USA). Physicochemical properties The powders were characterized for particle size, morphology and crystallinity by laser diffraction, scanning electron microscopy and X-ray diffractometry, respectively. Statistical analysis All data were expressed as mean ± standard deviation. Statistical analyses were performed by one-way analysis of variance (ANOVA) with Student-Newman-Keuls test (compare all pairs) as a post hoc test at (P < 0.05) using Instat Graphpad Prism software (version 4.00; GraphPad Software, San Diego CA). Results All the spray-dried powder particles were within the inhalable size range (1.1 to 5.9 µm) with the maximum process yield of 75.9 ± 1.7% for kanamycin/rifampicin combination powder. The surface composition measured by XPS showed the surface of the kanamycin/rifampicin combination powder was composed of approximately 98% rifampicin (Table 1). The enrichment of hydrophobic rifampicin was also evident in the ToF-SIMS distribution study of kanamycin/rifampicin combination powder (Figure 1). All the spray-dried powders had high emitted doses of above 85.0% (Table 2). The combination powder particles showed higher aerosolization efficiency (%FPF: 76.6 ± 3.8) than kanamycin-only powder (%FPF: 31.3 ± 1.8) (Table 2). The surface morphology of kanamycin-only powders was smooth and spherical with some elongated particles (Figure 2a) whereas flake-shaped particles were present in the combination powder (Figure 2c). The X-ray diffraction results showed that all the spray-dried powders were amorphous (Figure 3). Table 1- Elemental composition of spray-dried powders determined by X-ray photoelectron spectroscopy (XPS) (expressed as percentage relative atomic concentrations) (SD means spray-dried, K, R and KR represent kanamycin, rifampicin and kanamycin/rifampicin combination, respectively: Last two rows are molar percentages of kanamycin and rifampicin based on elemental composition).

Elements C N O S Kanamycin (%) Rifampicin (%)

SD-K 56.6 9.3 31.9 2.2 100 0

SD-R 74.5 6.0 19.4 0 100

SD-KR 72.6 6.2 21.2 2.0 98.0

Table 2- In vitro aerosolization performances of spray-dried powders (mean ± standard deviation, n=3, SD means spray-dried; K, R and KR mean kanamycin, rifampicin and kanamycin/rifampicin combination; K-KR means kanamycin in the KR combination and R-KR means rifampicin in the KR combination; ED means emitted dose, FPF means fine particle fraction, MMAD and GSD mean mass median aerodynamic diameter and geometric standard deviation, respectively.

Powders SD-K SD-R K-KR R-KR

ED (%) 91.0±0.9 85.8±1.2 88.5±1.4 89.5±1.9

FPF (%) 31.3±1.8 82.2±3.1 76.6±3.8 83.4±0.9

187

MMAD (µm) 4.2±0.7 2.1±0.2 1.9±0.2 1.4±0.2

GSD (µm) 3.0±1.1 2.8±1.0 2.7±0.5 2.7±0.3


1

Drug Delivery to the Lungs 27, 2016 - Mohammad AM Momin et al.

Figure 1- Distribution of kanamycin (red) and rifampicin (green) on the surface of kanamycin/rifampicin combination powder determined by ToF-SIMS (scale bar represents 10 Âľm).

Figure 2- Representative scanning electron micrographs of spray-dried (a) kanamycin-only; (b) rifampicin-only; (c) kanamycin/rifampicin combination.

Figure 3- X-ray diffractograms of spray-dried powders (SD means spray-dried; K, R and KR mean kanamycin, rifampicin and kanamycin/rifampicin combination).

188


Drug Delivery to the Lungs 27, 2016 - Enrichment of the surface of spray-dried powder particles with a hydrophobic material to improve their aerosolization Discussion The XPS study, which measures the elemental composition of powder surfaces to a depth of 5-10 nm, gave elemental compositions which were consistent with the theoretical values of the pure kanamycin sulfate and rifampicin (theoretical values not shown). The percentage of rifampicin in the upper 5-10 nm layer of the kanamycin/rifampicin powder was calculated based on the elemental composition of the XPS data. These data indicate that the surface layer is approximately 98% rifampicin. This finding was supported by the ToF-SIMS study. The surface (top 1-2 nm layer) distribution in the 50 x 50 μm field (Figure 1) is dominated by rifampicin (green) although there is some evidence of kanamycin (red dots) also present. The presence of some kanamycin signals indicates that the coverage of rifampicin on the combination powder surface is not be complete, in agreement with the XPS results which indicates about 2% kanamycin sulphate on the surface. Thus it is concluded that the co-solvent system of water and ethanol used for spray-drying of kanamycin-rifampicin combination formulation produced a surface-modified powder, enriched with hydrophobic rifampicin. This is probably due to the higher solubility of rifampicin in ethanol and migration towards droplet surfaces during the [8] rapid drying process . Further increase of rifampicin ratio in the combination formulation may produce almost completely rifampicin enriched particle surfaces. The in vitro aerosolization efficiency of kanamycin-rifampicin combination powder was significantly (P < 0.05) higher than kanamycin-only powder (Table 2). If we encapsulate 100 mg of the combination powder (ED 88.5% and FPF 76.6%), 67.8 mg will reach the deep lungs. The improved aerosolization of the combination powder may be due to the enrichment of the surface of the powder particles with hydrophobic rifampicin, as the surface [9] modification with hydrophobic materials should reduce particle interactions and improve powder dispersibility . However, the improved aerosolization may be due to the flake-shaped morphology of the combination powder [10] . Further research is required to distinguish these particles (Figure 2) which reduces particle contact area potential mechanisms. The MMAD (1.4 and 1.9 µm) and GSD (2.7 µm) values in the combination powder indicate [11] that the formulation is appropriate for deep lung delivery . All the spray-dried powders were amorphous in nature and their behaviour on storage should be investigated. The average recoveries of kanamycin calculated from drug contents were more than 95% which is within the [12] pharmacopoeial range for inhalation powders . Conclusion Spray-dried powder particles with hydrophobic surface enrichment were successfully produced using a co-solvent system of ethanol and water. Hydrophobic surface enrichment and/or change in particle morphology significantly improved the aerosolization efficiency of hygroscopic kanamycin possibly by reducing interparticulate interactions. Improved aerosolization may be helpful to deliver high doses of kanamycin/rifampicin to the lung to treat multidrug resistant local infections. Further research is required to investigate the effect of other ratios of kanamycin and rifampicin on the surface enrichment and the applicability of the current approach to other drugs used for inhalation. References Edwards B., Jenkins C., 2007. The Australian Lung Foundation’s Case Statement: respiratory infectious disease burden in Australia. Australian Lung Found. p9.

1

2

Millard, J., Ugarte-Gil, C., Moore, D.A., 2015. Multidrug resistant tuberculosis. BMJ 350, h882.

3

Atkins, P., 2005. Dry powder inhalers: an overview, Respir. Care pp. 1304-1312.

4

Claus, S., Weiler, C., Schiewe, J., Friess, W., 2014. How can we bring high drug doses to the lung? Eur. J. Pharm. Biopharm. 86, 1-6.

5

Das, S., Tucker, I., Stewart, P., 2015. Inhaled dry powder formulations for treating tuberculosis. Curr. Drug Del. 12, 26-39.

6

Hoppentocht, M., Akkerman, O.W., Hagedoorn, P., Frijlink, H.W., de Boer, A.H., 2015. The Cyclops for pulmonary delivery of aminoglycosides; a new member of the Twincer family. Eur. J. Pharm. Biopharm. 90, 8-15.

7

Zimmer, B.L., DeYoung, D.R., Roberts, G.D., 1982. In vitro synergistic activity of ethambutol, isoniazid, kanamycin, rifampin, and streptomycin against Mycobacterium avium-intracellulare complex. Antimicrob. agents chem. 22, 148-150.

8

Kim, E.H.J., Dong Chen, X., Pearce, D., 2003. On the mechanisms of surface formation and the surface compositions of industrial milk powders. Drying Technol. 21, 265-278.

9

Pfeffer, R., Dave, R.N., Wei, D. and Ramlakhan, M., 2001. Synthesis of engineered particulates with tailored properties using dry particle coating. Powder Technol. 117, 40-67.

10

Chew, N.Y., Chan, H.K., 2001. Use of solid corrugated particles to enhance powder aerosol performance. Pharm. Res. 18, 1570-1577.

11

Glover, W., Chan, H.K., Eberl, S., Daviskas, E., Verschuer, J., 2008. Effect of particle size of dry powder mannitol on the lung deposition in healthy volunteers. Int. J. Pharm. 349, 314-322.

12

BP, C., 2015. British Pharmacopoeia. Stationery Office, London England.

189


Drug Delivery to the Lungs 27, 2016 - Nashwa Osman et al. Comparative Nanotoxicology of Novel Polymeric Nanocarriers with Different Surface Charge against Human Lung Epithelial Cells Nashwa Osman, Darren Sexton, Gillian Hutcheon, and Imran Saleem School of Pharmacy & Biomolecular Sciences, Liverpool John Moores University, Liverpool, L3 3AF, UK. Summary Background: Biodegradable polymers are fuelling the development of drug delivery systems due to their biocompatibility, biodegradability, and ease fabrication. Poly (lactic-co-glycolic acid) (PLGA) is one of such biodegradable polymer and is Food and Drug Administration (FDA) approved. Poly glycerol adipate-co-ωpentadecalactone (PGA-co-PDL) is such a new polymer that has been developed and characterized in house laboratory to improve upon the physicochemical properties of PLGA. It was successfully formulated into nanocarriers. The study aims to evaluate the in-vitro cytotoxicity profile of PGA-co-PDL Nanoparticles (NPs) carriers in comparison to PLGA NPs for pulmonary drug delivery. Methods: NPs were formulated from PGA-co-PDL and PLGA by single emulsion using different (anionic and cationic) emulsifiers. The NPs were characterized for size and charge. In-vitro cytotoxicity was evaluated by Alamar Blue (AB), and Reactive Oxygen Species (ROS) assays using Calu-3 cells. Results: PGAco-PDL NPs showed good toxicity profile in comparison to PLGA NPs. The anionic NPs were more compatible to the cationic NPs. Less ROS production was detected with the anionic NPs. Conclusion: PGA-co-PDL polymer was successfully formulated into NPs with a suitable size range for pulmonary drug delivery. The results showed a good toxicity profile of PGA-co-PDL NPs in comparison with the PLGA NPs. The anionic particles showed better compatibility confirming future suitability for pulmonary drug delivery. Introduction Pharmaceutical nanocarriers are designed to achieve better drug compatibility and reduce its toxicity through much enhanced cellular uptake, targeted drug delivery to the diseased cells, and more controlled-release delivery. This is especially important for poor soluble, less bioavailable, and hydrophobic drugs. Biodegradable polyesters are commonly used for pharmaceutical and therapeutic drug delivery purposes due to many advantages they offered and the most commonly used and FDA-approved for a variety of drug delivery applications is PLGA. Their main drawbacks are the acidic and long bulk hydrolytic degradation. That affects the pH-sensitive drugs and the local acidity at the site of drug action that promotes the inflammatory response. PGA-co-PDL has been investigated as an alternative to PLGA for drug and macromolecule delivery [1, 2]. It has successfully been formulated into NPs for delivery of macromolecules, i.e. bovine serum albumin, pneumococcal protein showing promising results for treating lung disease and vaccine delivery [3, 4]. The use of NPs to improve macromolecules and drug delivery is accomplished by controlling the size and surface properties, the surface charge. The surface charge plays an important role in enhancing the cellular uptake and the efficiency of delivery. Positive-charged NPs are usually used to achieve the highest NPs-cellular interaction and uptake. This presents a challenge due to the cytotoxicity of the positive charge [5]. The study aims to evaluate the in-vitro cytotoxicity profile of PGA-co-PDL NP carriers with two distinct surface charge in comparison to PLGA NPs for pulmonary drug delivery. Experimental methods Acid terminated PLGA (50:50) with a molecular weight (MW) of 7000 - 17000 KDa was purchased (Sigma Aldrich). PGA-co-PDL was synthesized via enzyme catalysed co-polymerization of three monomers; glycerine, vinyl adipate and ω-pentadecalactone in 1:1:1 molar ratio as previously reported by Thompson et al [4]. This co-polyester was characterized using 1H-NMR spectroscopy (Bruker AVANCE 300 MHz, Inverse probe with B-ACS 60, Auto sampler with gradient shimming). The initial MW of both polymers was determined using Gel Permeation Chromatography (GPC) (Viscoteck TDA Model 300 operating OmniSEC4 software) calibrated with polystyrene standards [1, 4]. NPs were prepared from both polymers by the single-emulsion solvent-evaporation method using polyvinyl alcohol; PVA as an anionic emulsifier (-VE charge) and 1,2-dioleoyl-3-trimethylammonium-propane (chloride salt); DOTAP as a cationic emulsifier (+VE charge). 200 mg of each polymer were dissolved in 2 ml of dichloromethane (DCM) then added drop wise to 5 ml 10% PVA under probe sonication (26 µm for 2 minutes). The positive charged NPs were prepared using 10% DOTAP to polymer mass. The organic phase was added drop wise to 20 mL of 0.75% PVA under magnetic stirring 600 rpm for 10 minutes then adjusted to 500 rpm for 3 hours to allow the evaporation of the DCM. NPs were centrifuged twice at 76,000 xg, and 4 ºC for 40 min using Beckman Coulter Optima XPN-80 ultracentrifuge. Triplicates of PLGA and PGA-co-PDL NPs were characterized for size and surface charge (ζ Potential) using Zetasizer™ Nano ZS. NPs optimization experiments were performed to achieve a NPs target size of approximately 200 nm using design of experiments (Taguchi). In-vitro cytotoxicity was evaluated by Alamar Blue (AB) and Reactive Oxygen Species (ROS) assays [6]. The NPs were re-suspended with serum-free cell culture media (Eagle’s Minimum Essential Medium (EMEM) (ATCC® 30-2003™)) prior to cell culture assays. Calu-3 cells were seeded in a density of 40 x103 cells per well in 96 well plates for 48 hours using the complete media (containing10% Fetal Bovine Serum).

190


Drug Delivery to the Lungs 27, 2016 - Comparative Nanotoxicology of Novel Polymeric Nanocarriers with Different Surface Charge against Human Lung Epithelial Cells For AB assay, media was removed prior to the treatment with a 200 µL of a serial concentration of NPs (0.125- 2 mg/ml) in triplicate and incubated for 24 hours. Following incubation, the media was removed and cells were washed with PBS prior to addition of a 100 µL of 10% AB solution and further incubated for 3 hours. ROS was detected by 2,7-dichlorofluoresin diacetate reagent (DCFH-DA). Media was removed from the 96 well plate and the cells were washed with PBS prior to addition of DCFH-DA reagent (100 µL of 100 µM solution) and incubated at 37 °C in the dark for 40 minutes. The DCFH-DA reagent was then removed and the cells washed with PBS then followed by NPs treatment for 24 hours. Fluorometric evaluation of AB (560EX nm/590EM nm) and ROS (485EX nm/530 EM nm) was performed using the 96 well plate reader. The results were expressed as a percentage of the untreated control and compared to the positive control (10% DMSO for AB, 200 µM H2O2 for ROS assay). Results and Discussion Polymer characterization: The synthesized co-polyester PGA-co-PDL was a white powder with MW of 14.73 KDa as determined by the GPC. The co-polymer integration pattern was confirmed by 1HNMR spectra, (δH CDCl3, 300 MHz): 1.34 (s, 22 H, H-g), 1.65 (m, 8 H, H-e, e′, h), 2.32 (m, 6 H, H-d, d′, i), 4.05 (q)-4.18 (m) (6 H, H-a, b, c, f), 5.2 (s, H, H-j). The MW of PLGA was 17.57 KDa. NP characterization: Size and charge of the formulated NPs are shown in table1. Table 1.The physicochemical properties of NPs prepared from PGA-co-PDL and PLGA polymers using two different emulsifiers (Mean ±SD, n=3). NP type

Average (nm)

PLGA –VE NP

diameter

PDI

Zeta potential in mV

148.8 ± 0.8

0.21

-10.7 ± 2.1

PLGA +VE NP

189.6 ± 6.0

0.32

+12.5 ± 1.0

PGA-co-PDL –VE NP

195.6 ± 4.0

0.12

-13.6 ± 1.5

PGA-co-PDL +VE NP

242.2 ± 2.0

0.20

+13.0 ± 2.0

In-vitro cytotoxicity of PGA-co-PDL NPs: AB viability assay: Control cells had a percent viability of 100±12.33. PGA-co-PDL –VE NPs showed nearly equal compatibility to the PLGA –VE NPs within the same dose range. The PGA-co-PDL +VE NPs showed lower viability compared to the –VE NPs and more cytotoxicity to the +VE PLGA NPs at higher concentrations (Fig 1). At low concentration, there was no statistically significant difference between the –VE types of NPs and the +VE types NPs. There is statistically significant difference between the –Ve and the +VE NPs of PGA-co-PDL. At a high concentration, there was no statistically significant different between the –VE types/or between the +VE types of each polymer, but was significantly difference between the +VE types of NPs (reduced viability at high concentration with the +VE PGAco-PDL NP compared to PLGA NPs).

191


Drug Delivery to the Lungs 27, 2016 - Nashwa Nashwa Osman et al.

Alamar Blue Assay

120

PGA-co-PDL -VE NPs

% Viability

100

PLGA -VE NPs

80

PGA-co-PDL +VE NPs

60

PLGA +VE NPs

40 20 0

control

0.125

0.25

0.5

1

2

10% DMSO

Concentration mg/ml Fig 1. The percent of the viability after 24hrs exposure to both -VE and +VE NPs of both polymers by AB assay (Mean ± SD, n=3). ROS assay: Control cells had a percent ROS production of 100±15.2. PGA-co-PDL NPs showed similar percentage of ROS production to the untreated control cells at low concentrations but gradually started to be decreased with high concentration. Both –VE NPs of the two polymers showed similar effect that had less ROS production compared to the +VE NPs. The scavenging effect observed with the anionic NPs with the increase of the concentration is either real antioxidant effect that needs further verification with another oxidative stress quantification assays or a false decrease in the ROS production at higher concentrations attributed to reduced cellular uptake of the aggregated NPs. The + VE PGA-co-PDL NPs showed increase ROS production with the increase in concentration that was more pronounced than the +VE PLGA NPs , suggesting that their underlying cytotoxic mechanism can be attributed to ROS production (Fig 2). There was statistically significant difference between the –VE and the +VE types PGA-co-PDL NPs and to PLGA NPs.

450

PGA-co-PDL -VE NPs

ROS Assay

400

PLGA -VE NPs

% ROS Production

350

PGA-co-PDL +VE NPs

300

PLGA +VE NPs

250 200 150 100 50 0

control

0.125

0.25

0.5

1

2

200 um H2O2

Concentration mg/ml Fig 2. The percentage of ROS production after 24hrs exposure to both -VE and +VE NPs of both polymers by ROS assay (Mean ± SD, n=3). Conclusion: PGA-co-PDL polymer was successfully formulated into NPs with a suitable size range for pulmonary delivery. The results showed a good toxicity profile of PGA-co-PDL NPs in comparison with the PLGA NPs. The –VE NPs were more compatible than the +VE NPs and had less oxidative stress potential. Further studies will be addressed to uncover their molecular mechanisms of toxicity, cellular uptake and transport, internalization, and metabolism in pulmonary cell lines and co-culture models under both air and liquid interfaced conditions. In addition, the NPs limitations for aerosol delivery will be determined in terms of mucous transport, inflammatory induction, and genotoxicity.

192


Drug Delivery to the Lungs 27, 2016 - Comparative Nanotoxicology of Novel Polymeric Nanocarriers with Different Surface Charge against Human Lung Epithelial Cells References 1 Kunda N, Alfagih I, Dennison SR, Tawfeek HM, Somavarapu S, Hutcheon GA & Saleem IY; Bovine Serum Albumin Adsorbed PGA-co-PDL Nanocarriers for Vaccine Delivery via Dry Powder Inhalation. Pharm Res, 32:1341-1353 (2015). 2 Tawfeek H. M., Khidr S.H., Samy E.M, Ahmed S.M, Murphy M, Mohammed A, Shabir A, Hutcheon GA & Saleem IY: Poly(Glycerol Adipate-co-ω-Pentadecalactone) Spray-Dried Microparticles as Sustained Release Carriers for Pulmonary Delivery. Pharm Res, 28(9), 2086-2097 (2011). 3 Alfagih I, Kunda N, Alanazi F, Dennison SR, Somavarapu S, Hutcheon GA, Saleem IY. Pulmonary Delivery of Proteins Using Nanocomposite Microcarriers. Pharm Sci, 104:4386–4398, (2015). 4 Thompson CJ, Hansford D, Higgins S, Hutcheon, G.A., Rostron, C., Munday, D.L. Enzymatic synthesis and evaluation of new novel o-pentadecalactone polymers for the production of biodegradable microspheres. J Microencapsulation, 23:213–26 (2006). 5 Eleonore Fröhlich. The role of surface charge in cellular uptake and cytotoxicity of medical nanoparticles. Int J Nanomedicine, 7: 5577–5591(2012). 6 Handbook of Toxicological Characterization, p22-25, p46-50, Ver.1.0 (2012).

193


Drug Delivery to the Lungs 27, 2016 - Shyamal C. Das et al. The influence of lung surfactant, 1,2-dipalmitoyl-sn-glycero-3-phosphatidylcholine (DPPC) on the aerosolization of two anti-tubercular drugs, pyrazinamide and moxifloxacin Shyamal C. Das1, Bhamini Rangnekar & Basanth Babu Eedara

New Zealand's National School of Pharmacy, University of Otago, Dunedin 9054, New Zealand Summary Inhaled therapy is potentially more efficient than current oral and parenteral anti-tubercular treatments due to its ability to deliver a higher drug concentration to the lungs. For treating TB, a high dose of the drug (many milligrams) needs to be delivered to the lungs requiring to develop highly aerosolizable powders. This study investigated the influence of lung surfactant, DPPC, on the aerosolization of a first line hydrophilic anti-TB drug, pyrazinamide and a second line relatively hydrophobic drug, moxifloxacin HCl in the presence or absence of Lleucine. Using spray drying, individual powders of supplied pyrazinamide (PSD) and moxifloxacin HCl (MSD) alone and with 10% L-leucine (PL, ML) and 10% DPPC (PLD, MLD) were produced. The powders were characterized for physicochemical properties and aerosolization behaviour. The particle size of all powders except PSD was < 5 µm. The emitted doses of all the spray dried powders were very high (75-89%). The PSD showed poor aerosolization behaviour (FPF of 18.7 ± 3.4%) due to the presence of large (>10 µm) crystalline particles. However, the addition of L-leucine (PL) produced spherical hollow particles and improved aerosolization (FPF 53.0 ± 3.2%). The addition of DPPC and L-leucine to pyrazinamide (PLD) further improved aerosolization (FPF 74.5 ± 5.3%). However, the aerosolization of MSD although increased by the addition of L-leucine (FPF from FPF 55.6 ± 3.3% to 74.1 ± 1.3%), it was not further increased when both DPPC and L-leucine were added. In conclusion, DPPC in the presence of L-leucine significantly increased aerosolization of pyrazinamide, but not of moxifloxacin. Introduction Despite the availability of anti-TB drugs for almost a century, TB remains a major health problem with around 2 billion people currently carrying the causative microorganism of TB, Mycobacterium tuberculosis; and in 2012, 1.4 [1] million people died and 8.7 million new cases were reported worldwide . Pulmonary delivery of powders containing anti-TB drugs is potentially more efficient than current oral and parenteral anti-tubercular treatments due to its ability to deliver a higher drug concentration to the primary site of infection i.e. the lungs. In treating TB, it is important to deliver higher doses of the drug (many milligrams as opposed to <500 µg for asthma) to the [2] lower respiratory tract in the form of an aerosol of particles 1-5 µm in size . However, these tiny particles are, in general, forming agglomerates due to high cohesiveness which are difficult to de-agglomerate resulting in low fine particle fractions. To ensure high dose delivery, highly aerosolizable inhalable powders which are carrier-free or contain minimum amount of excipients (‘high drug-load’) are essential. One of the strategies to improve aerosolization of ‘high drug-load’ powders is by coating the particle surface with a hydrophobic material. We reported the achievement of high aerosolization by the dry coating of inhalable drug [3] particles using magnesium stearate . However, the surface coating of particles can also be achieved by spray drying through appropriate selection of spray drying conditions and excipients. Lung surfactants (LS) can be used for such coating since they have a hydrophobic chain. LS, being endogenous, can be rapidly metabolized and [4] eliminated from the lungs after inhalation . Dry powder formulations composed of phospholipids either alone or in combination with other carriers have demonstrated improved aerosolization behaviour with maximum fine [4-5] . We have recently reported achievement of >70% fine particle fraction using particle fraction (FPF) ~50% [6] phospholipids ; however, the excipient concentration was limiting high drug-load. Thus, this study was designed to investigate the influence of lung surfactant on the aerosolization of ‘high drugload’ inhalable powders of two anti-TB drugs, pyrazinamide and moxifloxacin HCl produced by spray drying. [7] Pyrazinamide is a first line anti-TB drug and the only drug which is effective in latent TB . Moxifloxacin HCl is a second line anti-TB drug belonging to the fluoroquinolone class, and if taken concomitantly with a first line anti-TB [8] drug, has the potential to shorten the TB treatment period . These two drugs were chosen since pyrazinamide is hydrophilic, moxifloxacin HCl is relatively less hydrophilic. Pyrazinamide, and moxifloxacin HCl will be spray dried with 10% of 1,2-dipalmitoyl-sn-glycero-3-phosphatidylcholine (DPPC) and/or L-leucine to achieve the high drug load. DPPC was used as this is the most abundant component in the lung surfactant. L-leucine was used since it is one of the widely used amino acids in the spray dried inhalation powders to improve aerosolization. The concentrations of excipients were kept low (≤ 20%) to ensure high drug loading. This results will hopefully facilitate the development of powder formulations for high dose delivery, and an effective anti-TB inhalable powder.

194


Drug Delivery to the Lungs 27, 2016 - The influence of lung surfactant, 1,2-dipalmitoyl-sn-glycero-3phosphatidylcholine (DPPC) on the aerosolization of two anti-tubercular drugs, pyrazinamide and moxifloxacin Materials and methods Materials Pyrazinamide and moxifloxacin HCl were purchased from Amsal Chem. Pvt. Ltd. (Gujarat, India) and Hubei Yuancheng Saichuang Technology Co., Ltd. (Wuhan, China), respectively. L-leucine and DPPC were obtained from Hangzhou Dayangchem Co., Ltd. (Hangzhou, China) and Lipoid (Ludwigshafen, Germany), respectively. All other reagents and chemicals utilized in this study were of high-performance liquid chromatography (HPLC) grade and purchased from Merck (Darmstadt, Germany). Preparation of the spray dried powders: Spray dried powder formulations of pyrazinamide and moxifloxacin HCl alone or with 10% of L-leucine and 10% DPPC were produced using a laboratory scale spray dryer (BUCHI B-290 Mini Spray Dryer, BUCHI Labortechnik AG, Switzerland), with a stainless steel standard 2-fluid nozzle with a 0.7 mm (internal diameter) nozzle tip and 1.5 mm cap. Feed solutions (0.5% w/v) were prepared in ethanol-water mixture (70:30% v/v) and were spray dried at a pump flow rate of 2 mL/min, air flow rate of 670 L/h, aspiration rate of 50%, the inlet temperature of 70 °C and the outlet temperature of 41 ± 1 °C. All the obtained powders were collected and stored in a desiccator at room temperature. Surface morphology and physicochemical characterization: Morphological features of the powder particles were visualized using JEOL 6700F FE-SEM (SEM, JEOL Ltd., Tokyo, Japan) at an acceleration voltage of 5 kV. The geometric mean diameter (dg; n > 300 particle count; mean ± SD) of the spray dried powders was measured from the SEM images (×2000 magnification) of the respective powder samples using ImageJ 1.48 software (National Institutes of Health, Maryland, USA). Solid state nature of the supplied drugs, excipients and spray dried powders, and the physicochemical interaction between the components of the spray dried powders was characterized using X-ray powder diffraction (XRPD) analysis, differential scanning calorimetry (DSC) study, hot stage microscopy (HSM) study, thermogravimetric analysis (TGA) and attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopy. In vitro aerodynamic performance test: ®

The aerodynamic behaviour of the spray dried powders from the Aerolizer device (Novartis, Surrey, UK) was assessed using the Next Generation Impactor (NGI; Copley Scientific, Nottingham, UK). NGI set up comprises of a stainless steel induction port attached with a mouthpiece adaptor, eight removable stainless steel collection cups located at seven plus a micro orifice collector (MOC) stages, a high capacity vacuum pump (Model HCP5), and a critical flow controller (TPK 2000). Before each run, the NGI collection cups were coated with silicone oil to minimize the particle entrainment after deposition. The air flow rate (Q) was tested and tuned at 100 L/min using ® the Copley’s flow meter (Model DFM 2000). Aerolizer device with a powder sample (~ 20 mg) filled hard gelatine PEG capsule (size 3; colourless; Qualicaps, Osaka, Japan) was activated for 2.4 s into the NGI assembly at an ® air flow rate of 100 L/min. The Aerolizer device with empty capsule shell, mouth piece, induction port and collection cups were rinsed with specific volumes of water: methanol (80:20% v/v) mixture and transferred into the volumetric flasks, shaken, filtered, and assayed using HPLC. Statistical analysis ® All data were processed and their mean ± standard deviation values determined using Microsoft Excel spreadsheets. Statistical analysis of data was performed by one way analysis of variance (ANOVA) and StudentNewman-Keuls post-hoc testing using GraphPad Prism 5 software (GraphPad Software, CA, USA) with p ≤ 0.05 as the minimum level of significance. Results All the spray dried powders (Fig. 1B-F) except PSD (Fig. 1A) produced fairly spherical shaped porous particles. While pyrazinamide before spray drying was non-spherical (Fig 1A), the addition of L-leucine in pyrazinamide resulted in spray dried particles of corrugated surface (1B and C). Although MSD particles were found to agglomerate, the agglomeration was less evidenced in L-leucine and DPPC containing powders. The particle size of (geometric diameter) of all spray dried powders was < 5 μm (PL 2.5 ±1.0 μm, PLD 3.8 ± 1.4 μm, MSD 1.9 ± 0.8 μm, ML 2.4 ± 1.0 μm and MLD 2.2 ± 1.0 μm) except P SD (> 10 μm). The X-ray powder diffraction analysis of the supplied pyrazinamide (Fig. 2A) and moxifloxacin HCl (Fig. 2B) showed intense diffraction peaks which represent the crystalline nature of both the drugs in their supplied form. XRPD diffractograms of the pyrazinamide spray dried powder formulations (Fig. 2E-G) showed the presence of pyrazinamide in crystalline form with altered peak positions indicating polymorphic transformation compared to its supplied form (-polymorphic form).

195


Drug Delivery to the Lungs 27, 2016 - Shyamal C. Das et al. The diffraction pattern of the pyrazinamide in the PSD without any excipients (Fig. 2E) matched with the δpolymorphic form. However, the spray dried powder formulations PL and PLD composed of L-leucine and DPPC produced a γ-polymorphic form of the pyrazinamide. In contrast, moxifloxacin spray dried powders (Fig. 2H-J) representing the amorphous form of the moxifloxacin HCl after the spray drying process as characterized by a hollow diffraction pattern without any diffraction peaks at their respective positions compared to the supplied crystalline form. Similarly, DSC, TGA, HSM results exactly match with the findings of the XRPD and confirms the presence of pyrazinamide in the crystalline form with altered polymorphic form and moxifloxacin in the amorphous form after spray drying process. ATR-FTIR studies reveal no chemical incompatibility between the combined drugs and excipients. All the spray dried powders showed high emitted doses of >80% except MSD (75.7 ± 1.3%) as determined by NGI. The spray dried pyrazinamide (PSD) showed very poor aerosolization behaviour with a minimal depositions on the deeper stages of the NGI and showed an FPF of 18.7 ± 3.4% only. Of the particles emitted, 39.9 ± 7% were deposited in the inhalation port and mouthpiece. The addition of L-leucine improved FPF to 53.0 ± 3.2%. The addition of DPPC along with L-leucine to the pyrazinamide further improved the FPF to 74.5 ± 5.3%. Figure 3 shows the % drug deposited on stages 1-7 of the Next Generation Impactor™ (NGI™) for the developed spray dried powder formulations. Similarly, spray dried moxifloxacin HCl composed of L-leucine showed the significantly (p<0.05) improved FPF (74.7 ± 4.7%) compared to the spray dried moxifloxacin without excipients (FPF 55.6 ± 3.3%). As opposed to pyrazinamide, the addition of DPPC along with L-leucine has not shown any significant (p<0.05) change in the FPF (74.1 ± 1.3%).

Figure 2. X-ray powder diffractograms of the supplied pyrazinamide (A), moxifloxacin HCl (B), L-leucine (C), DPPC (D), and spray dried formulations: pyrazinamide (E), pyrazinamide with 10% L-leucine (F), pyrazinamide with 10%Lleucine and 10% DPPC PLD (G), moxifloxacin HCl (H), moxifloxacin with 10% L-leucine (I) and moxifloxacin with 10% L-leucine and 10% DPPC (J).

Figure 1. Representative scanning electron microscopic images of spray dried pyrazinamide (A), pyrazinamide with 10% L-leucine (B), pyrazinamide with 10%L-Leucine and 10% DPPC PLD (C), moxifloxacin HCl (D), moxifloxacin with 10% L-leucine (E) and moxifloxacin with 10% Lleucine and 10% DPPC (F). Scale bar represents a length of 10 m in Fig. 1A and 1 m in all others.

Figure 3. Aerosol dispersion performance as % deposited on stage 1-7 of the Next Generation Impactor™ (NGI™) for spray dried powder formulations: spray dried pyrazinamide (PSD), pyrazinamide with 10% L-leucine

196


Drug Delivery to the Lungs 27, 2016 - The influence of lung surfactant, 1,2-dipalmitoyl-sn-glycero-3phosphatidylcholine (DPPC) on the aerosolization of two anti-tubercular drugs, pyrazinamide and moxifloxacin (PL), pyrazinamide with 10%L-leucine and 10% DPPC PLD (PLD), moxifloxacin HCl (MSD), moxifloxacin with 10% L-leucine (ML) and moxifloxacin with 10% L-leucine and 10% DPPC (MLD). Discussion Although the particle size distributions of the two powders (PL and PLD) were similar, the addition of DPPC improved the aerosolization of pyrazinamide which could be due to the surface active nature of DPPC and migration of DPPC to the droplet surface during the spray drying process and settling on the droplet surface with [5] the hydrophobic tails positioning outward . These hydrophobic chains of DPPC might have decreased the cohesive forces between the spray dried particles improved aerosolization. However, determination of the surface composition of the spray dried particles and measuring particle interactions are necessary to confirm these. The very poor aerosolization behaviour of the spray dried pyrazinamide (P SD) was due to the presence of larger size crystalline particles which were deposited in the inhalation port and mouthpiece resulting in less drugs available for the following stages of NGI. The significant improvement in the FPF of pyrazinamide due to the addition of Lleucine could be due to the production of porous particles of inhalable range. On the other hand, the aerosolization of moxifloxacin HCl (MSD) improved significantly when L-leucine was added (ML) by the formation of porous particles and reducing agglomeration. Since moxifloxacin HCl itself is relatively hydrophobic, the addition of L-leucine might have converted the surface hydrophobic to its peak level leaving the little scope of further change. Thus, the aerosolization was not further increased when DPPC was added with L-leucine in moxifloxacin HCl (MLD). Conclusion The lung surfactant, DPPC can improve aerosolization of a relatively hydrophilic anti-TB drug, pyrazinamide in presence of L-leucine. However, the aerosolization of relatively hydrophobic moxifloxacin HCl can be increased with L-leucine, the addition of DPPC with L-leucine cannot increase aerosolization. Although further studies are required, it is postulated that improved aerosolization could be due to the migration of DPPC on to the surface of the liquid droplet during the spray drying process. References 1.

Das S, Tucker I, Stewart P: Inhaled dry powder formulations for treating tuberculosis, Current Drug Deliv 2015; 12: pp26-39.

2.

Claus S, Weiler C, Schiewe J, Friess W: How can we bring high drug doses to the lung? Eur J Pharm Biopharm 2014; 86: pp 1-6.

3.

Shi J N, Das S C, Morton D A V, Stewart P J: The kinetics of de-agglomeration of magnesium stearate coated salbutamol sulphate powders: dry coating did not change the rate but increased the extent of deagglomeration, KONA 2015; 32: pp 131-142.

4.

Bosquillon C, Lombry C, Préat V, Vanbever R: Influence of formulation excipients and physical characteristics of inhalation dry powders on their aerosolization performance, J Control Release 2001; 70: pp 329–339.

5.

Cuvelier B, Eloy P, Loira-Pastoriza C, Ucakar B, Sanogo, A A, Dupont-Gillain C, Vanbever R: Minimal amounts of dipalmitoylphosphatidylcholine improve aerosol performance of spray-dried temocillin powders for inhalation, Int J Pharm 2015; 495: pp 981-990

6.

Eedara B B, Tucker I G, Das S C: Phospholipid-based pyrazinamide spray-dried inhalable powders for treating tuberculosis Int J Pharm 2016; 506: pp 174–183.

7.

Mitchison D A, Fourie P B: The near future: improving the activity of rifamycins and pyrazinamide, Tuberculosis 2010; 90: pp 177-81.

8.

Chan J G, Tyne A S, Pang A, Chan H K, Young P M, Britton W J, Duke C C, Traini D: A rifapentinecontaining inhaled triple antibiotic formulation for rapid treatment of tubercular infection, Pharm Res 2014; 31: pp 1239-53.

197


Drug Delivery to the Lungs 27,2016, Rob Bischof Preclinical sheep models for pulmonary disease and drug delivery Rob Bischof The Ritchie Centre, Hudson Institute of Medical Research, Clayton VIC 3168, AUSTRALIA

Summary This work outlines the development of the Allergenix house dust mite (HDM) sheep model of acute and chronic allergic asthma, a result of ongoing research investigations based in Melbourne, Australia. As a research model, sheep are ideally suited for respiratory diseases to bridge the gap between rodent and humans. A suite of new preclinical in vitro and in vivo assays have been developed with the purpose of delivering enhanced mechanistic data, better informing efficacy study design. Central to this, our proprietary sheep model of acute and chronic allergic asthma uses the relevant human allergen house dust mite, and represents many features of human asthmatic disease. Further, the similar features of ovine and human airways makes sheep an enhanced model for appropriate preclinical investigations of respiratory drug delivery and efficacy, with strong translational outcomes. Our models provide an improved pathway for preclinical validation of new drug targets or evaluation of emerging drug delivery technologies. Introduction The similarities in anatomical and physiological features of ovine and human airways makes sheep an ideal model for appropriate preclinical investigations of respiratory drug delivery and efficacy. Our research group has a strong focus on large animal (sheep) models of asthma (based on the relevant human allergen house dust mite, HDM) and other respiratory disease (COPD, pulmonary fibrosis), pulmonary inflammation, and drug delivery systems. Our skills and facilities have enabled development of novel assays that have delivered important mechanistic [1-3] . Such in vivo efficacy studies can be shown to offer an improved data, better informing efficacy study design pathway for preclinical validation of new drug targets and evaluation of emerging drug delivery technologies, with strong translational outcomes. Methods For the sheep asthma model, animals are immunised with solubilised HDM (with alum as adjuvant), and atopic sheep selected from animals that show elevated levels of HDM-specific IgE. Sheep are then given repeated HDM [1, 2] . airway challenges to induce features of acute and chronic disease Results In contrast to smaller animal models (Table 1), sheep exhibit similar breathing patterns, lung characteristics and pathophysiology to humans, evident with respect to:Airway'characterisGcs'and'pulmonary'physiology' lung size, anatomy and development; bronchial circulation and airway innervations; characteristics of mast cells and epithelial mucus production. The adaptability and suitability allow a broad range of techniques to be incorporated into studies in this sheep model, particularly with respect to target identification and endpoint analysis. This level of flexibility accommodates a number of study designs, including different drug/compound administration routes (oral; intrapulmonary (whole lung or targeted to individual lung lobes/segments); intranasal; injections (iv, im, id, sc, ip); in-life monitoring (body weight, temperature, blood pressure, heart rate, O2/CO2 bloodgas levels). In sheep we are able to conduct comparable measures of lung mechanics, performed in un-sedated (awake) animals, and simultaneous measurement of physiological and immunological parameters in blood and airways.

'

Parameter

''''

' body'mass ''''''' ' nose'and/or'''''''''''' mouth'breathers' ' branching'system' of'tracheaUbronchial' airways ' ' ' Gdal'volume'(mL)' ' respiratory'rate ''''''' (breaths/minute)'

''''''Human

''''''''Sheep

'''

'Rat'

''''''''80'kg

''''''''''45'kg

'''''''''''''0.3'kg'

''''mouth/nose

'''''''mouth/nose ''''''''''''''nose'

'' ''''dichotomous''''''''''dichotomous'''''''''''''monopodial' '''''400'U'616

''''''''180'U'405''''''''''''''''''0.87'U'2.08'

'''''''12'U'20

''''''''''15'U'40

'

'''85'

Table 1. Airway physiology and pulmonary characteristics in ConfidenGal' human versus sheep versus rat

198


Drug Delivery to the Lungs 27, 2016 - Preclinical sheep models for pulmonary disease and drug delivery Measures and biomarker analyses available in sheep include: arterial/venous cannulation; endobronchial brushings and tissue samples, bronchoalveolar lavage (BAL) sampling; histology (H&E, Alcian Blue, PAS, Masson’s Trichrome, etc); immunohistochemistry and immunostaining/immunochemistry (surface, intracellular, tissue structure); electron microscopy; flow cytometry (surface, intracellular, multicolour); ELISA detection of cytokines and biomarkers of inflammation/disease; microarray and other molecular analyses. In the context of allergic asthma, the HDM sheep asthma model displays a host of significant features (Figure 1) including: high serum IgE and allergic inflammation in the lungs after allergen challenge; activated inflammatory cells (eosinophils, mast cells, macrophages, dendritic cells, T cells), Th2 and other inflammatory cytokines, mucus Sheep'model'of'allergic'asthma'' hypersecretion, airway wall remodelling and compromised lung function.

Sheep'model'of'allergic'asthma'' unchallenged unchallenged (saline control)

(saline control)

BAL BAL

Acute ))) )))))Acute)))) ''''''''''''''''

)))))Acute)))) ))) ''''''''''''''''

Chronic )))))))))))))))))Chronic)

Airways ''''''''''''Airways'

)))))))))))))))))Chronic) ''''''''''''Airways'

*!*! inflammaGon'

inflammaGon'

airway'wall'remodeling'

airway'wall'remodeling'

challenged

(HDM) challenged (HDM)

Meeusen'EN'et'al'Drug'Discovery'Today:'Disease'Models'2009'

Meeusen'EN'et'al'Drug'Discovery'Today:'Disease'Models'2009' • induced'by'sensiGsaGon'and'exposure'to'house'dust'mite'(HDM)'allergen' Figure 1.• Features of the Allergenix sheep asthma model; sensitisation followed by airway allergen challenge displays'many'pathophysiological'features'of'human'asthma' produces (acute and chronic), as well as airway wall remodelling in the chronic condition • induced'by'sensiGsaGon'and'exposure'to'house'dust'mite'(HDM)'allergen' • inflammation invesGgate'pathways'of'disease'pathogenesis'and'assess'efficacy'of'asthma'therapies'

• displays'many'pathophysiological'features'of'human'asthma' ' Our model is based on the relevant human allergen HDM, and offers an improvement to the • invesGgate'pathways'of'disease'pathogenesis'and'assess'efficacy'of'asthma'therapies'

Ascaris sheep ConfidenGal' asthma model, allowing for a more standardised sensitisation protocol and significant features of allergic asthma ' are not replicated in the Ascaris model (Table 2). The HDM sheep asthma model has also been shown to be that responsive to human drug/antagonist treatments.

Sheep�model�of�allergic�asthma:��HDM�(Allergenix)�vs�Ascaris�

ConfidenGal'

Human�

HDM�sheep�model� Ascaris�sheep�model�

Human-relevant�allergen/IgE�responses� Allergen�induced�airway�constric=on� Early/late�airway�responses� Airway�hyperresponsiveness� Mucus�hypersecre=on� Airway�inflamma=on� Airway�eosinophils� Airway�mast�cells� Th2�driven�cytokines/mechanisms� Chronic�decline�in�lung�func=on� Chronic�=ssue�changes�-�remodelling�

Table 2. Sheep model of allergic asthma: HDM (Allergenix) versus Ascaris Confiden' al�

Conclusions

A broad range of techniques are incorporated into studies undertaken in the sheep model including: pilot studies to assess compound activity in the sheep model; controlled experimental studies of disease processes involved in asthma, COPD and pulmonary fibrosis; testing of different routes of delivery/application, including nebulisation and dry-powder delivery, and other parenteral routes; whole and targeted ‘segmental’ lung delivery; evaluation of novel airway delivery technologies; detailed immunological analyses and pharmacokinetics/pharmacodynamics; immune analyses at the cellular and molecular level, and cytokine and inflammatory marker expression; capacity for repeated in-life blood and BAL sampling, as well as tissue sampling of lung segments through endobronchial brushings/biopsies; post-mortem analyses; real-time measurement of airway mechanics/lung function before,

199


Drug Delivery to the Lungs 27,2016, Rob Bischof during and after
experimental treatments, in un-sedated animals; surgical techniques including blood and lymphatic cannulation. There are significant and unique features of this sheep model of acute and chronic allergic asthma and its differentiation from other sheep models in using the relevant human allergen house dust mite. References 1.

Bischof RJ, Snibson K & Meeusen ENT (2003). Induction of allergic inflammation in the lungs of sensitized sheep after local challenge with house dust mite. Clinical and Experimental Allergy 33: 367375.

2.

Snibson KJ, Bischof, RJ, Slocombe, RF & Meeusen EN (2005). Airway remodeling and inflammation in sheep lungs after chronic airway challenge with house dust mite. Clinical and Experimental Allergy 35: 146-152.

3.

Liravi B, Piedrafita D & Bischof RJ (2015). Dynamics of IL-4 and IL-13 expression in the airways of sheep following allergen challenge. BMC Pulmonary Medicine 15:101.

200


Drug Delivery to the Lungs 27, 2016 – Athina Skemperi, et el A Study on the Drug Deposition Mechanisms of Surface-treated pMDI Canisters 1

1

2

Athina Skemperi , David Worrall Gary Critchlow & Phil Jinks 1

3

Chemistry Department, Loughborough University, Loughborough, Leicestershire, UK Materials Department, Loughborough University, Loughborough, Leicestershire, UK 3 3M Drug Delivery Systems, Loughborough, Leicestershire, UK

2

Summary The aim of this project was to explore drug deposition mechanisms on the aluminium canisters employed with Pressurised Metered Dose Inhalers (pMDIs). The investigation explored the influence of various surface treatments, applied to the canisters, to drug deposition phenomena. Physicochemical characterisation of the canisters was performed after the application of different surface treatments to explore potential links between: surface topography, surface chemical composition, total surface energy, and drug caking appearing on the canister walls. The coating treatments which were tested were; Fluorinated Ethylene Propylene (FEP) lacquer, vapour deposited Parylene and a 3M-proprietary, fluid-applied fluorosilane. Anodisation was also explored both with and without additional fluorosilane treatment. A number of surface physicochemical characterisation techniques were employed, namely; Scanning Electron Microscopy (SEM), X-Ray Photoelectron Spectroscopy (XPS) and Contact Angle (CA) analysis. The data from these analyses were correlated with those from a Drug Deposition test employing drug quantification by UV spectrophotometry. The results obtained indicated a direct correlation of drug deposition on canister walls to the total surface free energy. The lowest total surface free energy values and lowest deposition values were seen when the 3M-proprietary fluorosilane coating was applied as the final treatment. In this case a surface free energy value of 15.73 mN/m and a percentage drug deposition of 7% compared to the calibration can, were achieved. Introduction Aerosol Inhalation therapy is the most widely used treatment for the common diseases of the lungs namely: Chronic Obstructive Pulmonary Disease (COPD), and asthma. The key advantages of inhalation therapy arise from the direct targeting of the drug to the site of action. This results in more rapid onset of therapeutic effect, fewer systemic side effects and lower doses, giving an improved safety profile in comparison with for example, oral therapy [1, 2]. PMDIs are the most popular aerosol devices used for inhalation therapy. However, one of the most significant drawbacks of pMDI systems is that drug deposition commonly appears on the hardware of the device, from suspension-based formulations [2, 4]. Such drug deposition (caking), may result in ineffective drug therapy due to low or inconsistent dosing [5, 6]. There are several published studies on drug deposition phenomena in pMDI devices, most of which focus on stabilisation of the formulation by the use of either excipients, surfactants or particle engineering techniques [7, 8]. Further studies focus on the effect of both moisture and temperature storage conditions to the drug caking appearing on the device [9, 10]. Other investigations explore the effect of hardware design to the performance of pMDIs [11, 12]. However, there are limited studies concerning surface treatments applied to the hardware of the inhaler i.e. cleaning methods and coating procedures [13, 14]. The aim of this study was to explore drug deposition behaviour on pMDI canisters employing a range of surface treatments and to relate this to surface topography and surface chemistry. Experimental Section 1 Physicochemical characterisation of the aluminium canisters, In order to observe the effect of different surface treatments on surface morphology, surface elemental composition and total surface energy of the samples, three different techniques were applied to the samples: Scanning Electron Microscopy (SEM), X-Ray Photoelectron Spectroscopy (XPS), and Contact Angle (CA) analysis. The results obtained are presented below.

201


Drug Delivery to the Lungs 27, 2016 – A Study on the Drug Deposition Mechanisms of Surface-treated pMDI Canisters Scanning Electron Microscopy (SEM) The SEM technique was used to acquire high resolution information on sample surface topography. The results are presented in Fig 1.

Figure 1a: Surface of the Aluminium Can after solvent cleaning process x1000 Magnification

Figure 1b: Surface of the Aluminium Can after Anodisation Surface treatment x1000 Magnification

Figure 1c: Surface of the Aluminium Can after Parylene coating Surface treatment x1000 Magnification

Figure 1d: Surface of the Aluminium Can after FEP coating Surface treatment x1000 Magnification

Figure 1: Surface topography of aluminium canisters after various surface treatments were applied In Figure 1, it is seen that the anodisation treatment produced a significantly smoother surface topography (Figure1b) compared to the original surface (Figure 1a). In contrast the Parylene coating treatment masked only the fine topography of the sample (Figure 1c). Finally, complete masking of the aluminium surface irregularities was observed with the FEP coating treatment (Figure 1d). X-Ray photoelectron Spectroscopy Analysis The aim of the XPS analysis was to determine the surface chemical composition of the aluminium canisters before and after coating application. The results are shown in Table 1. Table 1: Surface chemical composition atomic%, excluding H and He, of the aluminium canisters after various coating treatments applied Sample

Al

C

Cl

F

Mg

O

S

Si

N

Ni

Novec 72DE cleaned can

26.46

26.80

0.00

0.00

5.30

41.40

0.00

0.00

0.00

0.00

FEP coated can

0.00

35.30

0.00

64.7

0.00

0.00

0.00

0.00

0.00

0.00

Parylene coated can

0.00

83.80

12.68

0.31

0.00

2.28

0.00

0.92

0.00

0.00

Anodised can

22.57

33.00

0.00

0.00

0.00

38.50

3.10

0.00

0.86

2.00

Anodised + 3M fluorosilane can

13.71

31.73

0.00

26.89

0.00

22.87

1.83

0.00

0.67

2.30

3M flurosilane can

6.55

29.22

0.00

48.62

0.00

15.61

0.00

0.00

0.00

0.00

202


Drug Delivery to the Lungs 27, 2016 – Athina Skemperi, et el In Table 1 the surface elemental composition of the cans, after the application of various treatments is shown. These data indicate that the coatings have created specific surface chemistry. The absence of detectable aluminium at the surface for both the FEP and the Parylene coatings suggests that both coating types completely mask the surface. The high percentages of fluorine for both the FEP coating and for the 3M fluorosilane coating applied directly onto the aluminium canister suggest that effective non-stick performance might be expected for these systems. The high percentages of both carbon and chlorine are both characteristic of the Parylene coating and the high percentages of aluminium along with oxygen were seen with the anodised canister surface. Contact angle analysis The aim of the Contact Angle technique was to quantify the surface hydrophobicity and the total surface energy of the samples. The results are shown n Table 2. Table 2 Contact Angle and total Surface Energy of the aluminium canisters after surface treatment application CA DIM (°)

Total Surface Energy (mN/m)

Dispersive component (mN/m)

Polar component (mN/m)

87.90

56.80

30.05

27.46

2.59

FEP coated alum can

101.70

77.40

18.59

17.27

1.32

Parylene coated alum can

84.10

39.70

39.17

37.48

1.69

Anodised alum can

80.60

41.30

38.44

35.36

3.08

Anodised+3M fluorosilane alum can

106.50

86.00

15.87

14.53

1.34

3M fluorosilane coated alum can

112.60

84.20

15.73

15.41

0.32

Sample

CA Water(°)

Novec 72DE cleaned alum can

The sessile drop technique was used for the acquisition of the contact angle data. The Surface Energy calculation employed was that of Owens-Wendt and Kaelble (OWK). A two-liquid method was employed using Deionised Water (a polar liquid) and Diiodomethane (DIM) (a non-polar liquid). This method allows the surface energy to be expressed in terms of polar and dispersion components. The DIM is used in order to detect surface energy due to dispersive interactions and water is used to detect surface energy due to polar interactions [15, 16]. The results indicated a range of values of the total surface energy and of the polar and dispersive elements of the surface energy across the samples. The lowest total surface energy was seen with the use of the 3M fluorosilane final treatment, while the highest total surface energy was observed with the Parylene coated sample. Notable is the significant decrease of the total surface energy observed with the anodised can, following application of the 3M fluorosilane as a final treatment. Section 2 Application of drug deposition testing to the treated aluminium canisters.

Aluminium canisters Figure 2 Deposition performance of aluminium canisters with different surface treatments

203

3M fluorosilane

3M fluorosilane

3M fluorosilane

Anodised 3M fluorosilane Anodised 3M fluorosilane

Anodised

Parylene

Parylene

Parylene

FEP

FEP

FEP

Solvent washed

Solvent washed

Solvent washed

Calibration

Calibration

100 90 80 70 60 50 40 30 20 10 0

Calibration

% Drug Deposition

The canisters were subjected to a previously published rapid screening deposition test based on the controlled deposition of micronised salbutamol sulphate particles on the internal surface of canisters, followed by rinsing steps, followed by the assay of residual surface salbutamol sulphate deposition [14]. The results are shown in Figure 2.


Drug Delivery to the Lungs 27, 2016 – A Study on the Drug Deposition Mechanisms of Surface-treated pMDI Canisters In Figure 2, the deposition performance of the aluminium canisters, after different surface treatments, is shown. Two to three replicates were used in each case. The canisters labelled “Calibration” provide a test reference point defining 100% deposition, as the salbutamol sulphate suspension has been applied to these canisters but without subsequent rinsing. The salbutamol sulphate suspension was applied to the remaining test canisters with the addition of two separate fluid rinses, employing decafluoropentane (5ml) with five 180 degree inversion shakes of 1 second cycle duration for each rinse. From the results obtained it is observed that the highest drug deposition occurred with the Parylene coated sample, while the lowest drug deposition occurred with the 3M fluorosilane coated sample. Notable is the decrease in drug deposition for the solvent-washed and the anodised sample, when followed by the 3M fluorosilane as a final treatment. Discussion Viewing all experimental data as a whole, it can be seen that, in general terms, FEP and fluorosilane coatings give rise to high measured surface fluorine levels, which in turn correlate to low deposition performance. Although FEP provides the smoothest coating, the key driver for deposition appears to be surface energy, since the fluorosilane coatings do not significantly alter the topography of the can surface, yet perform slightly better than the FEP coating in terms of measured deposition. The Parylene coated canister shows both higher surface energy and higher deposition than the uncoated canister yet is topographically smoother then the uncoated canister, which further supports the argument that surface energy is the key driver to reducing deposition. Conclusions Drug deposition on pMDI canisters was shown to vary considerably across a range of samples with different coatings applied. Deposition was found to correlate very strongly with total surface energy rather than surface topography. Hence coatings depositing high surface fluorine and providing low surface energy, irrespective of topography, gave the best performance. Further work will focus on developing a deeper understanding of the variables in coating application, deposition test methodology and fluorosilane chemistry on measured deposition performance. References [1]

J. L. Rau: The inhalation of Drugs: Advantages and Problems Respiratory care, 2005; 50:3: pp367-382. J. C. Virchow, G. K. Crompton, R. D. Negro, S. Pedersen, A. Magnan, J. Seidenberg, P. J. Barnes: Importance of inhaler devices in the management of airway disease, Respiratory Medicine, 2008; 102: pp10-19. [3] D. E. Geller: Comparing Clinical Features of the Nebulizer, Metered-Dose Inhaler, and Dry Powder Inhaler, Respiratory care, 2010; 50:10: pp1313-1322. [4] C. Harder, E. Lesniewska, C. Laroche: Study of ageing of dry powder inhaler and metered dose inhaler by atomic force microscopy, Powder Technology, 2011; 208: pp252-259. [5] M. B. Dolovich, R. Dhand: Aerosol drug delivery: developments in device design and clinical use, Lancet, 2011; 377: pp10321045. [6] H. D. Smyth: The Influence of formulation variables on the performance of alternative propellant-driven metered dose inhalers, Advanced Drug Delivery Reviews, 2003; 55: pp807-828. [7] L. Wu, M. Al-Haydari, S. R. P. da Rocha: Novel propellant-driven inhalation formulations: Engineering polar drug particles with surface-trapped hydrofluoroalkane-philes, European Journal of Pharmaceutical Sciences, 2008; 33: pp146-158. [8] D. D’Sa, H. K. Chan, W. Chrzanowski: Predicting physical stability in pressurized metered dose inhalers via dwell and instantaneous force colloidal probe microscopy, European Journal of Pharmaceutics and Biopharmaceutics, 2014; 88: pp129135. [9] R. O. Williams III, C. Hu: Moisture Uptake and its influence on Pressurized Metered-Dose Inhalers, Pharmaceutical Development and Technology, 2000; 5: 2: pp153-162. [10] F. M. Shemirani, S. Hoe, D. Lewis, T. Church, R. Vehring, W. H. Finlay: In Vitro Investigation of the Effect of Ambient Humidity on Regional Delivered Dose with Solution and Suspension MDIs, Journal of Aerosol Medicine and Pulmonary Drug Delivery, 2012; 26:0: pp1-8. [11] J. Berry, S. Heimbecher, J. L. Hart, J. Sequeira: Influence of the Metering Chamber Volume and Actuator Design on the Aerodynamic Particle Size of a Metered Dose Inhaler, Drug Development and Industrial Pharmacy, 2003; 29:8: pp. 865-876 [12] D. W. Grimble, S. Theodossiades, H. Rahnejat, M. Wilby: Thin film tribology of pharmaceutical elastomeric seals, Applied Mathematical Modelling, 2013; 37: pp406-419. [13] D. Traini, P. M. Young, P. Rogueda, R. Price: The Use of AFM and Surface Energy Measurements to Investigate DrugCanister Material Interactions in a Model Pressurized Metered Dose Inhaler Formulation, Aerosol Science and Technology, 2006; 40: pp227-236. [14] P. A. Jinks: The development of a new rapid screening test for evaluating the non-stick performance of MDI canisters, 2010; Conference proceedings of Drug Delivery to the Lungs 21. [15] D. P. Subedi: Contact Angle Measurements for the Surface Characterisation of Solids, The Hivalayan Physics, 2011; 2: pp14. [16] O. Santos, T. Nylander, R. Rosmaninho, G. Rizzo, S. Yiantsios, N. Andritsos, A. Karabelas, H. M. Steinhagen, L. Melo, L. B. Petermann, C. Gabet, A. Braem, C. Tragardh, M. Paulsson: Modified stainless steel surfaces targeted to reduce fouling-surface characterization, Journal of Food Engineering, 2004; 64: pp63-79. [2]

204


Drug Delivery to the Lungs 27, 2016 – Donovan B. Yeates et al. Generation of Respirable Particles from Surfactant Suspensions and Viscous Solutions at High Dose Rates Donovan B. Yeates & Xin Heng KAER Biotherapeutics, 926 S. Andreasen Drive, Escondido, California 92029, USA Summary To enable the delivery of high concentrations of aerosolized therapeutic agents at high payloads from viscous solutions or sols, aerosols were generated from surfactant and polyvinylpyrrolidone, PVP with SUPRAER®. These fluids were atomized and the water evaporated using warm dilution air and infrared radiation. The resultant dry aerosol was concentrated using a virtual impactor. The masses of aerosol collected on filters at the output were determined gravimetrically. Particle size distributions were measured with a Marple-Miller cascade impactor. Viscosities were measured with a capillary rheometer. Surface tensions were measured with a contact angle analyzer. Aerosols between 2.6 and 4 μm MMAD were generated from 10% PVP solutions and 9.33% surfactant suspensions (34 cP) at aerosolization rates between 0.5 and 3 ml/minute with a modest increase in particle size with aerosolization rate that was less evident when surfactant was aerosolized. Particle size increased with viscosity over the range of 4 to 39 cSt. When aerosolizing 10% 8 kDa PVP solutions, the output increased in proportion to the aerosolization rate with 193 mg/min being delivered at 64% efficiency. Outputs of 0.7 and 1.2 g were collected in 3.3 and 6.7 min, respectively. When 9.33% surfactant suspensions were aerosolized, up to163 mg/min was delivered at 59% efficiency. The surface tension of surfactant was not changed following aerosolization and resuspension. SUPRAER increases the viscosity range over which respirable aerosols can be generated at up to 3 mg/s opening up possibilities for new treatments for respiratory ailments with surfactant and large molecules. Introduction Common to all patients with acute respiratory distress syndrome ARDS is the increased surface tension of surfactant and its continued degradation. The mortality rate remains between 40 and 60%. It is estimated that the treatment of acute respiratory distress syndrome with aerosolized surfactant requires 300-1000 mg to be delivered to the lungs. High concentrations of surfactant have high viscosities which prevent its rapid aerosolization as well as the effective treatment with the therapeutically needed high masses of surfactant. Pulmonary fibrosis, neonatal RDS, pneumonia, COPD, atelectasis, cystic fibrosis and severe asthma are all candidates for aerosol surfactant therapies. In addition, surfactant has been proposed as a carrier to augment the delivery of co-delivered drugs. Biologics comprise some 50% of the product development pipeline but most of these agents under development have been proposed to be delivered intravenously. We reported data on [1] in which liquid aerosols are generated from polyvinylpirrodidone, optimized aerosol delivery with SUPRAER® PVP, dried, concentrated and delivered as respirable particles at high dose rates. We now extend this data to include the aerosolization and delivery of surfactant and viscous solutions of PVP. IR Reflector

Evaporation Chamber

Virtual Impactor

Fluid Input Flow Distributor

Exit Cone

Output

IR Control

Exhaust Port With a Filter Flowmeter

Nozzle and Counter Flow Pressure Gauge

Temperature Controller

Figure 1 - The laboratory version of SUPRAER showing the aerosol processing system atop the console.

205


Drug Delivery to the Lungs 27, 2016 – Generation of Respirable Particles from Surfactant Suspensions and Viscous Solutions at High Dose Rates Experimental Methods The laboratory model of SUPRAER is shown in FIG 1. A syringe pump (not shown) is used to feed an aqueous solution/suspension to the aerosolizing nozzle that is inserted into a flow distributor. The proprietary nozzle aerosolizes 100% of the fluid to form a liquid aerosol with a narrow size distribution (σg <2). This aerosol plume is arrested with a co-axial counter-flow air and the fluid evaporated from the particles using a combination of warm compressed air, and dilution air together with infrared radiation whose wavelength is optimized for the absorption band of water. The resultant dry particle aerosol is concentrated using a virtual impactor with radially aligned acceleration and deceleration nozzles. The particles gain momentum as they pass through the acceleration nozzles. They cross a small gap and loose momentum as they pass through the deceleration nozzles to provide a low velocity inhalable aerosol. Most of the air (70-90%) exits from the gap between these nozzles into a plenum and is exhausted through two output ports with filters. Thus the output aerosol is comprised of a higher concentration of particles in a much smaller volume of air. The resultant aerosol flows to the output port at 3 cm of water pressure where it is delivered on demand. We explored the effects of nozzle diameter and parameter settings in SUPRAER, nozzles with different exit orifice diameters (KB-N-xx), aerosolization rate, and solution/suspension viscosity on the mass median aerodynamic diameter (MMAD) of aerosols. Aqueous solutions (weight (g)/volume (ml)) of the polymeric excipient, polyvinylpyrrolidone, PVP (Acros Organics, USA) with different molecular weights were used as a protein or biologic surrogate. Surfactant suspensions (Molecular Express Inc, US) 9.33% were used. Particle size distributions were measured by using a Marple-Miller cascade impactor at output flow rate of 30 l/min. 4% Allura Red AC (Sigma-Aldrich, USA) was added to the PVP solutions and surfactant suspensions to facilitate the spectrographic measurements of the mass distributions. The MMAD and geometry standard deviation (GSD) of aerosols were determined from the cumulative log-probability plot. To evaluate the relationship between viscosity and particle size, the viscosities of PVP solutions with nominal molecular weights 8, 29, 40 and 58 kDa and surfactant suspension were measured with a capillary rheometer and expressed in cSt. The outputs (delivered dose) by SUPRAER for PVP and surfactant were measured at output flow rate of 44 l/min. To evaluate the effect of aerosolization process on the surface tension of the surfactant, the surface tensions prior to and following aerosolization by SUPRAER-CA were measured with Contact Angle Analyzer (FTA-200) using the pendant drop shape method. Results Evaluation of nozzles for high dose rate, high payload aerosol delivery When 5% 8 kDa and 20% 58 kDa aqueous solutions of PVP, were aerosolized using atomizing nozzles with a wide range of aerosol exit orifice diameters, the MMAD of aerosols markedly decreased with increasing aerosol exit orifice diameter (FIG 2a).

KB-N-300, CA: 60 psi

KB-N-700, CA: 60 psi

Figure 2 - (a) The MMAD of aerosols corresponding to different nozzle exit orifice diameters. Aerosols were generated from 5% 8 kDa and 20% 58 kDa PVP solutions at an aerosolization rate of 1 ml/min at compressed air pressure (CA) of 30 and 60 psi, respectively. (b) Photograph of an aqueous aerosol generated at 3 ml/min using nozzles KB-N-300 and KB-N-700. The aerosol exits through the center of the aerosol exit orifice surrounded by a sheath of air essentially devoid of particles. On inspection of photographs of these nozzles (FIG 2b), it can be seen that the aerosol was created external to the exit orifice of the small diameter nozzles (KB-N-300) but when nozzles 400 μm and larger were used (KB-N700) there was a marked decrease in break-up length with the aerosol being created within an aerosolizing space

206


Drug Delivery to the Lungs 27, 2016 – Donovan B. Yeates et al. between the fluid orifice and the aerosol exit orifice. This aerosol formation mechanism, together with the increased airflows leads to the generation of smaller primary liquid particles. Effect of Fluid Aerosolization Rate FIG 3 shows the modest increase in particles size for PVP at aerosolization rates between 0.5 and 3 ml/min. When surfactant is aerosolized, the particle size appears almost independent of fluid flow rate in the range of 1 to 3 ml/min.

Figure 3 - The MMAD of aerosols corresponding to different fluid aerosolization rates. Nozzle KB-N-500 was used to aerosolize 10% 8 kDa PVP solutions and 9.33% surfactant suspensions with a compressed air pressure (CA) of 40 psi.

Figure 4 - The MMAD of aerosols corresponding to different viscosities of the PVP solution. Nozzle KB-N-700 was used to aerosolize 10% and 20% PVP solution with the fluid flow rate of 1 ml/min at compressed air pressure (CA) of 40 psi.

Effect of Fluid Viscosity As the viscosity of surfactant suspensions increases rapidly with increasing surfactant concentration, we evaluated the relationship between viscosity and particle size. As predicted, from the Ohnesorge number, there was a modest increase in particle size with increasing viscosities between 4 and 39 cSt (FIG 4). Output Efficiency and Total Dose Delivered It can be seen in Table 1 that the output of 10% PVP increases to 192 mg/min at 3 ml/min. When the surfactant suspension was aerosolized the output efficiency (59%-63%) was marginally lower than with PVP (64%). Table 1 - PVP solution and surfactant suspension dose rate and output efficiency for nozzle KB-N-500 at compressed air pressure of 40 psi.

10% 8 kDa PVP

9.33% Surfactant

Fluid Flow Rate (ml/min)

Efficiency (Dose Rate)

1

64% (64 mg/min)

2

64% (128 mg/min)

3

64% (192 mg/min)

1

63% (59 mg/min)

2

63% (118 mg/min)

3

59% (165 mg/min)

Using nozzle KB-N-500 at compressed air pressure of 40 psi to aerosolize 10 ml ant 20 ml 10% 8 kDa PVP solution at an aerosolization rate of 3 ml/min, 0.7 and 1.2 g were collected at the output in 3.3 and 6.7 min, respectively. Surfactant Surface Tension Prior To and Following Aerosolization.

207


Surfactant 3

59% (165 mg/min)

Drug the Lungsat27, 2016 – Generation of Respirable from Surfactant UsingDelivery nozzle to KB-N-500 compressed air pressure of 40 psiParticles to aerosolize 10 ml antSuspensions 20 ml 10% and 8 kDa PVP Viscous at High Doserate Rates solution Solutions at an aerosolization of 3 ml/min, 0.7 and 1.2 g were collected at the output in 3.3 and 6.7 min, respectively. [3] Consistent with previous preliminary data , showing that biologics retained their physicochemical and functional Drug Delivery topost the Lungs 27,Prior 2016To – Generation of Respirable Particles from andtension Surfactant Surface Tension and Following Aerosolization. characteristics aerosolization, the aerosolized surfactant was seen to Surfactant have the Suspensions same surface Viscous Solutions at High Dose Rates following aerosolization (FIG 5). [3]

Consistent with previous preliminary data , showing that biologics retained their physicochemical and functional characteristics post aerosolization, the aerosolized surfactant was seen to have the same surface tension following aerosolization (FIG 5).

Figure 5 - The surface tensions of the surfactant prior to and following aerosolization by SUPRAER-CA through using Nozzle KB-N-500 to aerosolize 9.33% surfactant suspension with the aerosolization rate of 1 ml/min at compressed air pressure (CA) of 40 psi. Discussion Figure 5 -and TheConclusions surface tensions of the surfactant prior to and following aerosolization by SUPRAER-CA through using Nozzle KB-N-500 to aerosolize 9.33% surfactant suspension with the aerosolization rate of 1 Weml/min demonstrate that fine air particles aerosols than 4 μm MMAD can be readily generated by nozzles with large at compressed pressure (CA) ofless 40 psi. orifices from high concentrations of surfactant suspensions and large molecules solutions with SUPRAER and Discussion and Conclusions delivered at dose rates of 2.5 to 3 mg/s with total doses delivered of 700-1200 mg. That the aerosol exits the nozzle through a sheath of gas devoid of particles makes these nozzles essentially uncloggable and suitable for We demonstrate fine particles aerosols 4 aerosols μm MMAD can readily generated by nozzles with each large high payload drugthat delivery. These data implyless thatthan such can bebe delivered at these rates throughout orifices from high concentrations of surfactant suspensions and large molecules solutions with SUPRAER and and every inhalation throughout the treatment period. delivered at dose rates of 2.5 to 3 mg/s with total doses delivered of 700-1200 mg. That the aerosol exits the [2] through sheath of gas devoidofofsolutions particlesaerosolized makes these essentially uncloggable suitable for . SUPRAER Itnozzle is notable thatathe highest viscosity bynozzles mesh-type nebulizers is <4 cSt and high payload drug delivery. These data imply that such aerosols can be delivered at these rates throughout markedly extends the range of large molecule and viscous solutions from which fine particle aerosols caneach be and every inhalation readily generated andthroughout delivered the to attreatment least 40 period. cSt. The maximum aerosol generation rate for jet type nebulizers that produce particles of less than 3 μm in diameter is <0.3 ml/min. These droplets generally only[2]contain up to SUPRAER It isofnotable that theSUPRAER highest viscosity of solutions aerosolized mesh-type nebulizers <4 cStthis .rate 5% active agent. produces solid particles of pure by agent of similar size at 10is times with 20 markedly extends the range of large molecule and viscous solutions from which fine particle aerosols can be times the dose per particle. readily generated and delivered to at least 40 cSt. The maximum aerosol generation rate for jet type nebulizers that laboratory produce particles than 3 used μm inindiameter is <0.3 ml/min. These droplets generally contain up to The version of of less SUPRAER these experiments was designed to determine its only performance and 5% of active agent. produces particles of pure agent similar setting. size at 10 timesreduction this rate with 20 specifications prior toSUPRAER its conversion into asolid device designed for use in aofclinical Further in size times the dose perwould particle. and simplification facilitate domiciliary use. The aerosol can be generated continuously or the fluid flow controlled such that aerosol is only delivered on demand to the patient during an inhalation. In this configuration Theaerosol laboratory of SUPRAER in these experiments to determine its performance and the canversion be delivered at up toused 3 mg/s throughout each was and designed every inhalation throughout the treatment specifications prior conversion into use in aforclinical reduction in size period. Thus, at 0.7tol/sitsinhalation rate, upatodevice 15 mgdesigned could befor delivered each 5setting. secondFurther inhalation. SUPRAER and simplification would tofacilitate use. The can be generated or the delivery. fluid flow could also be configured provide domiciliary a respiratory assist foraerosol both patient activated and continuously continuous aerosol controlled such that aerosol is only delivered on demand to the patient during an inhalation. In this configuration the aerosol be deliveredanticancer at up to 3 mg/s and throughout each and every inhalation throughout the treatment High doses ofcan anti-infectives, agents biotherapeutics delivered intravenous administration not only period. Thus, at patient 0.7 l/s inhalation to own 15 mg could be delivered for each 5 second inhalation. SUPRAER has considerable resistance,rate, but up their attendant complications. Intravenous administration is likely to could also betoconfigured provide a respiratory for both patient and continuous require a 10 100 timestohigher dose than if it assist were administered byactivated aerosol inhalation. The aerosol ability todelivery. deliver these agents directly to the lungs markedly reduces the total dose of the agent, the cost of therapy, as well as High doses of anti-infectives, anticancer agents and biotherapeutics delivered intravenous administration not only systemic toxicities. has considerable patient resistance, but their own attendant complications. Intravenous administration is likely to require a 10 to 100 times higher dose than if it were administered by aerosol inhalation. The ability to deliver References these agents directly to the lungs markedly reduces the total dose of the agent, the cost of therapy, as well as systemic Yeates D toxicities. B, Heng X: Targeting the optimum particle size and output for aerosol drug delivery using SUPRAER, Respiratory Drug Delivery 2016; pp395-399. References Chan J G Y, Traini D, Chan H -K, Young P M, Kwok, P C L: Delivery of high solubility popyols by vibrating mesh Yeates D B, Heng X: mucociliary Targeting the optimum J. particle sizeMed andand output for aerosol drug delivery using nebulizers to enhance clearance, Aerosol Respiratory Drug Delivery 2012; 25 SUPRAER, (5) pp297Respiratory Drug Delivery 2016; pp395-399. 305. Chan J D G Y, Chan -K, Young PofM,fine Kwok, P C L:aerosol Deliverywith of high solubility popyols by vibrating mesh Yeates B: Traini High D, dose rateH generation particles SUPRAER™ compared to two mesh nebulizers Respiratory to enhance mucociliary clearance, J. Aerosol Med and Respiratory Drug Delivery 2012; 25 (5) pp297nebulizers, Drug Delivery 2013; pp313-316. 305. Yeates D B: High dose rate generation of fine particles aerosol with SUPRAER™ compared to two mesh nebulizers, Respiratory Drug Delivery 2013; pp313-316.

208


Drug Delivery to the Lungs 27, 2016 - A. A. Desta 1, et al. Nebuliser Issues To Be Improved For The Pulmonary Administration Of Nano Encapsulations 1

2

1

A. A. Desta , K. C. Carter , T. Gourlay &A. B. Mullen

2

1

2

Biomedical Engineering, 106 Rottenrow East, Glasgow, G4 0NW, United Kingdom Strathclyde Institute of Pharmacy and Biomedical Sciences, 161 Cathedral Street, Glasgow, G4 0RE, United Kingdom

Summary Nano based formulations which encapsulate drugs, nano-encapsulations (NE), can be used to provide economical and effective delivery of existing and novel drugs. When administered via the pulmonary route they can enhance therapeutic outcomes whilst reducing adverse side effects. They can be used for local or systemic delivery, broadening the types of conditions that can be treated by inhalation. It is believed that atomising NEs developed as a simple colloidal suspension fluid rather than a dry powder format may offer clinical advantages. However, existing devices have a number of limitations with respect to delivery of these systems. This paper will highlight the aspects of current nebuliser devices that negatively impact NE viability and delivery efficiency. These drawbacks are being addressed in our ongoing work to develop a novel nebuliser. In addition, new data is presented surveying nebulisers in current clinical use based on anonymised adverse incident reports obtained from the UK’s Medicines & Healthcare product Regulatory Agency (MHRA). Plotting the findings in a priority matrix revealed that Increased NE reservoir concentration and device reliability are the most important variables, whereas issues such as ease of use and user error were the least important issues to the delivery and viability of NE. It is concluded that NE integration for pulmonary administration needs device improvements to uphold the economic characteristics of NE. Introduction Liposomes and niosomes are NE that entrap medicine in a vesicle systems. They have proved a suitable delivery vehicle for drugs by enhancing cellular drug uptake and reducing the required drug dose for effective treatment. NE preparation methods mostly result in a colloidal fluid system. For pulmonary administration however, nebulisers have been shown to have a detrimental effect on NE causing aggregation, bursting and leakage of the [1] encapsulations when atomising . Various studies report that post-nebulisation NE drug leakage is approximately [2] 40% resulting in significant delivery of un-encapsulated drug and wastage of lipids or surfactants . As a result, transforming NE into a dry powder through freeze drying has become a post-production necessity. This addresses some of the instability issues when formulated as a suspension such as leakage, aggregation and sedimentation. Also, as a dry powder, NE can simply be inhaled through dry powder inhalers (DPI) instead of [3] nebulisers which only deliver liquid based drugs .However, to prepare NE for freeze drying, further excipients such as lyoprotectants are also required to ensure steric stabilisation. These however increase production costs [4] and result in additional toxicity concerns for pulmonary delivery . The potential of NE inhalation therapy with nebulisers without resorting to freeze dried preparations is explored. Key existing nebuliser issues are also prioritized when developing a nebuliser for better NE administration. Methodology To effectively qualify the current needs for improvement, reported issues with nebulisers in research and clinics are explored. In order to highlight the key nebulisation issues associated with detrimental effect to NE, previous research articles are reviewed involving nebulisers and NE drug delivery; reporting non-formulation factors only. Thus avoiding issues in regards to NE suspension stability which has little to do with nebuliser device design. MHRA incident report data is used to understand how the wider nebuliser therapy issues in clinics affect NE delivery. The anonymised nebuliser device incident report data was collected from 2011 to 2014 and categorized 72 incidents based on type of incident and subdivided by potential injury level. Key nebuliser issues are displayed on a priority matrix of NE viability against the importance of delivering NE.NE viability refers to the intactness of the NE particles and NE delivery efficiency refers to their transport efficiency from the reservoir through the atomiser in to aerosol micro droplets and in the lung. The position of the issues in the priority matrix was based on the author’s opinion formed through understanding the severity and impact on worst case scenarios.

209


Drug Delivery to the Lungs 27, 2016 - Nebuliser Issues To Be Improved For The Pulmonary Administration Of Nano Encapsulations

Results

Figure 1: MHRA's nebuliser incident report between 2011 and 2014. MHRA adverse incident reports with manufacturer, device model and location anonymised. Comparing reported device faults and total incident count, n = 72.

Figure 2: Priority index of nebuliser issues and their importance for the efficient NE viability and delivery.NE viability refers to the intactness and post nebulisation entrapment efficiency whereas NE delivery efficiency refers to the concentration of NE within the aerosol droplets and delivery of the aerosols into the lung. Upper right quadrant contains the highest priority and lower left quadrant is the lowest priority issues for both NE viability and NE delivery.

210


Drug Delivery to the Lungs 27, 2016 - A. A. Desta 1, et al. Discussion Incident report findings are displayed as a bar chart in Fig. 1. Although 72 incidents in 4 years don’t provide statistical significance, it does present the issues deemed adverse enough to be reported to the MHRA. It is found that majority of the reported incidents had serious potential injury. Temperature and mechanical faults were found to be the highest number of reported issues. Flow related faults have been reported with two potential injuries of death and one reported actual injury of death. Flow faults are particular to air jet nebulisers where the driving gas can be either oxygen or air depending on the patient’s clinical condition. Previous incidents have [13, 14] . resulted from mistaking air and oxygen supply resulting in hypoxemia or hyperoxia The gathered issues from both research and the MHRA incident reports were then displayed on a priority matrix in Fig.2. The importance levels are assigned based on their understood effect towards NE rather than statistical occurrence as it helps present the key priorities novel solutions should consider The priority matrix shows that increased NE reservoir concentration is the highest priority issue for both the NE viability and delivery efficiency. This was previously noted in studies comparing aerosol output vs drug output where the emitted aerosol droplets had less NE concentration than the original solution. This results in higher NE concentration within the reservoir, leading to aggregation that causes leakage of encapsulated drugs and affects [7] the size polydispersity of the NE . Reservoir temperature is another issue that is significantly detrimental to the NE viability but less so with the NE delivery efficiency. Its detrimental effect of weakening the NE walls results in susceptibility to mechanical agitation [5, 6] . MHRA incident reports associate temperature with serious potential injury and from the atomisation process one actual serious injury. Although it is important to highlight that the incident report data associates 55% of [8] reported temperature issues with “nebuliser compressor”, which is only present in air jet nebuliser types . These nebuliser types use a motor driven air compressor to provide the air jet and are normally fitted with a thermal switch which renders the device inoperable until it cools down. This affects NE delivery efficiency more than NE viability. Mechanical issues are believed to arise from air jet nebuliser types. This is the only nebuliser that makes use of mechanical actuation to provide air jets for atomisation. The mechanical issues arise from the diaphragm air compressor which constitutes a motor, piston and valves. It is also believed that flow issue arises mostly from air jet nebulisers. The fault again is associated with the air compressor unit where the air flow circuit can be restricted or the air compression system composed of motor, piston or valve is not providing the necessary force to provide the advised flow rate emitted by the air compressor. Electronics faults are related by the internal circuitry in relation to arcing and potential of causing fire. These issues found in the MHRA incident reports were grouped into performance reliability. It is believed that issues such as mechanical, flow and electronics faults did not occur as a fundamental design flaw but rather as in use wear and tear of the device. Performance deviations can affect both the viability and delivery efficiency of NE. One key performance index of nebulisers and an NE delivery factor is the aerosol droplet size. Control of aerosol droplet size determines the deposition profile of the drug ranging from the nebuliser passages through to the [11] alveoli and exhaled drug . Majority of the mentioned factors are underpinned by reliability of the device where functionality of the device is required over the course of the therapy and deviation of performance over time can affect both NE viability and [9, 10] . delivery efficiency In addition to reliability ease of use would be an essential feature in relation to treatment adherence. This would affect drug loading process, adherence to the therapy schedule and correct device operation. Therefore, incorrect operation will negatively impact both NE viability and delivery efficiency. Although ease of use can be countered with effective training there are signs in the industry where nebuliser design is incorporating autonomous features like breath sensors which only activate the atomiser during the first 80% of inhalation breath [10] cycle to reduce loss of aerosol during exhalation .

211


Drug Delivery to the Lungs 27, 2016 - Nebuliser Issues To Be Improved For The Pulmonary Administration Of Nano Encapsulations Conclusion This study helps to highlight key device issues which if improved will positively impact the viability and delivery efficiency of NE through nebulisers. Increased NE reservoir concentration and performance reliability are key issues in both NE viability and delivery efficiency. Reservoir temperature plays an important role in NE viability and aerosol droplet size as an important determinant of drug deposition. Not associating the issues with nebuliser types helps bring focus to possible issues to minimize in the design of future nebulisers. The only exception was the temperature issue associated with Air jet nebuliser type. As it was important to distinguish the device temperature issue presented in the MHRA report from the reservoir temperature issue presented in research. Although this paper focuses on NE inhalation therapy, improvement in reported issues is not limited to NE and can bring benefits to sensitive medication such as inhaled protein therapy. The purpose of this study was to help guide decisions on which nebuliser issues to focus on when adapting future nebulisers for NE delivery. It is believed efforts for NE therapy integration should not be reserved to formulation activity only. But should incorporate device improvements to effectively adapt NE for inhalation therapy. Since NE are delivery vehicles, any effort in upholding their benefits will result as an economic advantage for drug delivery to the lungs.

Reference [1]

Elhissi A, Hidayat K, Phoenix D A et al. Air-jet and vibrating-mesh nebulization of niosomes generated using a particulate-based proniosome technology. Int J Pharm.2013;1:193-199

[2]

Lehofer B, Bloder F, Jain PP, Marsh LM, Leitinger G, Olschewski H, Leber R, Olschewski A, Prassl R: Impact of atomization technique on the stability and transport efficiency of nebulized liposomes harboring different surface characteristics. Eur J Pharm Biopharm 2014;3:1076-1085.

[3]

Pattini B S, Chupi V V&Torchilin V P. New Developments in liposomal drug delivery.Chemical Review.( Accessed 11 July 2015)doi:10.1021/acs.chemrev.5b00046

[4]

Uchegbu IF, Vyas SP: Non-ionic surfactant based vesicles (niosomes) in drug delivery. Int J Pharm: 1998; 1:33-70.

[5]

Hertel S, Pohl T, Friess W, Winter G: That’s cool!–Nebulization of thermolabile proteins with a cooled vibrating mesh nebulizer. Eur J Pharm Biopharm. 2014; 2: 357-365.

[6]

Bridges PA, Taylor KM: Nebulisers for the generation of liposomal aerosols. Int J Pharm 1998; 1:117125

[7]

Elhissi AM, Faizi M, Naji WF, Gill HS, Taylor KM: Physical stability and aerosol properties of liposomes delivered using an air-jet nebulizer and a novel micropump device with large mesh apertures. Int J Pharm 2007; 1: 62-70.

[8]

Received Incident Reports on Nebulisers 2011- 2014. MHRA.2015

[9]

Fink JB, Rubin BK:Problems with inhaler use: a call for improved clinician and patient education. Resp Care 2005; 10: 1360-1375.

[10]

Denyer J &Dyche T:The Adaptive Aerosol Delivery (AAD) Technology: Past, Present and Future. J Aerosol Med Pulm Drug Deliv 2010; 1: 165-176

[11]

O'Callaghan C, Barry PW: The science of nebulised drug delivery. Thorax 1997; 2:S31.

[12]

Tena AF, Clarà PC: Deposition of inhaled particles in the lungs. Arch Bronconeumol 2012;7: 240-246.

[13]

Selecting ‘oxygen’ or ‘medical air’ to give nebulisers | Signal. NHS patient safety (Accessed 25 Jul 2016);http://www.nrls.npsa.nhs.uk/resources/?EntryId45=134717

[14]

Trust-wide combined oxygen and nebuliser therapy alert. East North Hertfordsh. NHS Trrust Press Releases (Accessed 25 Jul 2016);http://www.enherts-tr.nhs.uk/blog/813/press-releases/trust-statementtrust-wide-combined-oxygen-and-nebuliser-therapy-alert/

212


Drug Delivery to the Lungs 27, 2016 – Cuong Tran et al. Opt2Fill™Dispersible Tablet – A Novel Method for the Manufacture of pMDIs Cuong Tran, Chen Zheng, Simon Warren & Glyn Taylor i2c Pharmaceutical Services, Cardiff Medicentre, Cardiff, CF14 4UJ, UK. Summary Manufacture of suspension-type pressurised metered dose inhalers (pMDI) requires judicious in-process homogenization and agitation to ensure homogeneity of suspended drug(s) and accurate and reproducible canister filling. In this study, the aerosol characteristics of pMDIs manufactured by the addition of propellant dispersible tablets (Opt2Fill™ tablet) to pMDI canisters prior to the addition of HFA134a/ethanol blend was assessed. The tablets contained a combination of salbutamol sulphate (SS) and beclometasone dipropionate (BDP), a dispersant (menthol) and lactose. Each tablet weighed approximately 90 mg and contained sufficient drug for 200 metered doses each containing 120 µg of SS and 50 µg of BDP. The aerosol particle size distributions (PSD) of the Opt2Fill™ products were compared with marketed formulations containing either SS or BDP, i.e. Airomir® and QVAR® 50. The PSD characteristics clearly demonstrated effective dispersion of the Opt2Fill™ tablets with fine particle fraction (FPF) properties similar to the comparators. The PSD of Opt2Fill™ products showed acceptable pMDI performance from a solid dosage form consisting of two drugs of differing physicochemical properties, i.e. one insoluble, one soluble. It is proposed that this technology can overcome some of the challenges commonly encountered during conventional single stage and two-stage pMDI filling processes. Dispensing the solid dosage form into the canister simplifies in-process checks and eliminates the requirements for homogenisation / recirculation of suspension systems and the need for large pressure vessels. Cleaning processes between products can also be greatly simplified. Batch sizes can be readily varied and scaled ensuring process continuity from pilot batches to large commercial batches. Introduction Suspension-type pMDI formulations require judicious in-process homogenisation and agitation controls during manufacture to ensure reproducible canister content uniformity. Both single-stage and two-stage pMDI filling processes have other associated challenges in addition to homogeneity issues. These may include drug losses by adsorption onto filling lines and pressure vessels, sensitivity to product flow and pressure variations, and suspension concentration due to propellant losses into pressure vessel headspace. In addition, stability issues may arise during two-stage filling processes as a consequence of partial drug solubility in excipients such as ethanol. Many of these problems, i.e. drug losses and suspension homogeneity also apply to cold filling processes with additional issues related to possible condensation / moisture inclusion. The objective of this study was to evaluate the novel manufacturing method of adding the active pharmaceutical ingredients API(s) to the [1] aerosol container in the form of a propellant dispersible tablet (Opt2Fill™ tablet), prior to addition of propellant , thereby circumnavigating the challenges described above. Opt2Fill™ tablet formulations contained API and optionally a dispersant e.g. menthol and an inert carrier / bulking agent e.g. lactose. The propellant dispersible tablets were formulated to produce a combination product containing a micronised suspended API (SS) and a non-micronised soluble API (BDP). Aerosol characterisation tests were conducted to compare the novel formulation with the performance of marketed reference products. Previous studies have shown that Opt2Fill™ [2] powders can be used to effectively formulate suspension pMDI products . Methods Micronized SS (Jayco Chemical Industries Ltd, India) (d50 1.94 µm) and unmicronized BDP (Farmabios SpA, Italy) and were blended together by low shear mixing (Turbula® Mixer, Wiley A, Bachofen AG, Switzerland) with commercial inhalation grade -lactose monohydrate (Respitose SV010, DFE-Pharma, Germany) (d50 112 µm) previously coated with 1% (w/w) menthol (Sigma-Aldrich, UK). The final powder blend contained approximately 12% (w/w) SS:Lactose and 5% (w/w), BDP:Lactose. The powder blends were used to produce SS/BDP Opt2Fill™ tablets. Tablets were produced on a single punch tablet machine (RIVA Minipress MII, Argentina) using a 6 mm round flat bevelled edge punch operating with a compression force not exceeding 60 kN to produce tablets weighing approximately 90 mg for a formulation calculated to produce metered doses of 120 and 50 µg for SS and BDP respectively. In addition, uncompressed SS/BDP blended with 1% (w/w) menthol coated lactose powders (Opt2Fill™ powder) were produced. Single Opt2Fill™ tablets or appropriate amounts of Opt2Fill™ powder were dispensed into either plain aluminium aerosol canisters (Presspart Manufacturing Ltd, UK) or polyethylene terephthalate (PET) vials (19 mL, Createchnic, Germany). Appropriate volumes of pre-mixed 5% (w/w) absolute ethanol in propellant HFA 134a were cold transferred to the containers, which were crimp sealed with Kemp KHFA 50 µL metering valves (VARI SpA, Italy). Containers were shaken and briefly sonicated, in order to ensure full dissolution of the BDP. Canisters were then stored for a minimum of two days quarantine at ambient temperature / humidity prior to testing. Using actuators with 0.25 mm orifice diameters aerosol PSD, was determined via the next generation impactor (NGI) at a flow rate of 30 L/min. The dose characteristics and PSD of the Opt2Fill™ tablet and Opt2Fill™ powder formulations were compared with two marketed reference products i.e. a suspension SS pMDI Airomir® (Teva, UK Ltd.) and a solution BDP pMDI QVAR® 50 (Teva, UK Ltd.). Analyses of SS and BDP were conducted by validated high-performance liquid chromatography (HPLC).

213


Drug Delivery to the Lungs 27, 2016 – Opt2Fill™Dispersible Tablet – A Novel Method for the Manufacture of pMDIs Briefly, following deposition from the pMDI, SS/BDP were quantitatively recovered from NGI components in 80% methanol:water. The actuator, induction port (IP) and collection cups were recovered with 25, 50 and 10 mL respectively. HPLC analysis was performed using a Waters Sphersiorb S5ODS2, column (4.5 x 250 mm, 5 µm), at a flow rate of 1.0 ml/min (75% methanol v/v), 100 µl injection volume, UV detection at 276 nm. Results The measured content uniformity of the Opt2Fill™ tablet and Opt2Fill™ powder formulations were consistent, with RSD values below 2 % (data not shown). Figure 1 shows an example of an Opt2Fill™ tablet prior to 5% ethanol/HFA 134 addition and following agitation and brief sonication.

Figure 1. Photographs of a SS/BDP Opt2Fill™ tablet prior to 5% ethanol/HFA134a addition and following dispersion.

The key aerosol characterisation results are presented in Table 1. i.e. fine particle fraction (FPF) (% < 5 µm), mass median aerodynamic diameter (MMAD) and geometric standard deviation (GSD). The data showed broad similarities in aerosol characteristics between the non-optimised experimental formulation and the appropriate marketed comparator products. Table 1. Aerosol particle size distribution of SS and BDP from Opt2Fill™ Powder and Tablet formulations compared to Airomir® and QVAR 50®. Opt2Fill™ Powder

Opt2Fill™ Tablet

Airomir®

SS FPF (% < 5 µm)

43.02

48.47

48.80

SS MMAD (µm)

3.23

3.49

2.00

SS GSD

1.77

1.80

1.57

BDP FPF (% < 5 µm)

70.70

68.90

72.96

BDP MMAD (µm)

1.11

1.50

0.72

BDP GSD

3.95

2.95

1.96

QVAR 50®

The aerosol particle size distribution from NGI analysis of the formulations for SS and BDP are shown in Figures 2 and 3 respectively. The distribution profiles were broadly similar for each test formulation and the appropriate reference product. Inspection of the SS profile shows there was a trend for reduced IP deposition for the test formulations compared to Airomir®, this trend was reversed on Stage 1, consequently overall the total combined fractions on IP and Stage 1, were similar. On Stages 3 to Filter the test formulations deposition tended to be shifted to the left, relative to Airomir®, indicating a larger particle size distribution within the ‘respirable range’. Inspection of the BDP profile shows there was similar IP deposition for the test formulations compared to QVAR®. Very little BDP was recovered from Stages 1 – Stage 4 for QVAR® with the majority of the drug recovered from Stage 5 – Filter. For the test formulations BDP was evenly distributed over Stages 3 to Filter.

214


Active Recovery (%)

Drug Delivery to the Lungs 27, 2016 – Cuong Tran et al.

Opt2Fill Powder

50 45 40 35 30 25 20 15 10 5 0

Opt2Fill Tablet Airomir

NGI Stages

Active Recovery (%)

Figure 2. Aerosol particle size distribution of SS from Opt2Fill SS/BDP tablet and powder pMDI formulations compared to Airomir® using the NGI at 30 L/min.

Opt2Fill Powder

50 45 40 35 30 25 20 15 10 5 0

Opt2Fill Tablet QVAR

NGI Stages Figure 3. Aerosol particle size distribution of BDP from Opt2Fill SS/BDP tablet and powder pMDI formulations compared to QVAR® 50 determined using the NGI at 30 L/min.

Discussion and Conclusion Previous studies have shown that Opt2Fill™ powders i.e. API(s) blended with inert carriers (without disintegrant) [2] can be used to effectively formulate suspension pMDI products . In this study we have demonstrated that Opt2Fill™ tablet technology can be used to prepare a pMDI combination formulation containing both a suspended and a soluble drug. The performance of this non-optimised formulation compared favourably with appropriate reference products.Differences in PSD for the suspended API (i.e. SS) could have resulted from disparities in the primary particle size of the micronised SS in Airomir® and Opt2Fill™. In the case of the solution API (i.e. BDP) it is possible that differences in the ethanol content and also hardware differences e.g. valve materials and actuator orifice diameter may have contributed to the contrasting PSD profiles. It is also likely that some propellant droplets contained both particulate SS and solubilised BDP and thus co-deposition of BDP and SS may have occurred thereby increasing the particle size characteristics of the solubilised BDP.

215


Drug Delivery to the Lungs 27, 2016 – Opt2Fill™Dispersible Tablet – A Novel Method for the Manufacture of pMDIs [3]

It was previously reported that an Opt2Fill™ tablet formulation containing menthol coated lactose and salmeterol xinafoate/fluticasone propionate was efficiently dispersed in HFA 134a. The resultant pMDI formulation demonstrated aerosol particle size distribution results similar to those of the marketed comparator, Seretide™. Experiments to date indicate that effective pMDI formulations may be manufactured by the novel approach of dispensing APIs in the form of propellant dispersible tablets. This technology offers manufacturing advantages compared to established single-stage and two-stage filling methods. It is anticipated that upon scale up the technology will provide a flexible approach to pMDI manufacturing using essentially standard equipment. Batches of bulk tablets complying with release specifications will be dispensed into canisters followed by valve crimping and check weighing. The final step i.e. propellant filling will only require standard pressure filling and check weighing equipment with no necessity for homogenization/mixing, pressure lines, or pressure vessels. Propellant filling may be conducted immediately after valve crimping or alternatively following a predetermined delay and could if desired be performed at a separate facility. Cleaning processes between products will be greatly simplified. Batch sizes can be readily varied and scaled ensuring process continuity from pilot batches to large commercial batches. References 1

Taylor G, Warren S, Tran C: Pressurised metered dose inhalers and method of manufacture. WO 2015/121653 A1 Patent Application.

2

Tran CH, Zheng C, Warren S, Taylor G: Investigations of tiotropium pMDI suspension formulations. In Drug Delivery to the Lungs 24. The Aerosol Society. Bristol, UK: 2013: 122-125.

3

Taylor G, Tran CH, Zheng C and Warren S: Application of an Opt2Fill dispersible tablet to the production of a novel salmeterol/fluticasone pMDI. In proceedings Respiratory Drug Delivery 2016, volume 2, 2016: 381 -384.

Acknowledgements Generous gifts were provided by: VARI SpA – metering valves; Mexichem – HFA134a; Presspart – canisters; DFE-Pharma – lactose monohydrate.

216


Drug Delivery to the Lungs 27, 2016 - Philip Chi Lip Kwok et al. Droplet Sizes of Electronic Cigarette Aerosols 1

2

Philip Chi Lip Kwok , Philippe Rogueda & Lu Hou

1

1

Department of Pharmacology and Pharmacy, Li Ka Shing Faculty of Medicine, The University of Hong Kong, 21 Sassoon Road, Pokfulam, Hong Kong SAR 2 Aedestra Limited, 11/F, Rykadan Capital Tower, 135-137 Hoi Bun Road, Kwun Tong, Kowloon, Hong Kong SAR Abstract Electronic cigarettes (ECs) produce aerosols by heating e-liquids that subsequently condense into droplets upon cooling. The e-liquid vehicle commonly consists of a mixture of common solvents including water, ethanol, glycerol, and propylene glycol. However, the effects of the solvents on EC aerosol characteristics are unknown. This study measured the droplet size distribution of five e-liquids (four self-made solvent-only mixtures and one commercial e-liquid) by laser diffraction. The reason for using this sizing technique was to gather data that can be compared with those obtained from pharmaceutical inhalers in the future. Interestingly, the droplet size distribution was monomodal and comparable between all five e-liquids, despite their different compositions. The volumetric droplet diameters ranged from 1-10 μm, with high variability around the peak sizes. The variability could be due to the poor control of the heating of the e-liquid at the coil. Glycerol and propylene glycol were essential for aerosol generation. On the other hand, pure water and pure ethanol could not produce aerosols. Their presence in the mixtures did not affect the droplet size either. The micron-sized droplets measured in this study were much larger than the nano-sized droplets reported in other studies. This could be due to a difference in the testing setup, experimental procedure, and/or the EC model. Further investigation using aerodynamic impaction is needed to confirm the laser diffraction data. Introduction Electronic cigarettes (ECs), or e-cigarettes, are devices that produce aerosols through heating and vaporising [1] solutions, called e-liquids, that may or may not contain nicotine . Most of them are currently marketed for recreational use. Some have even been approved as smoking cessation aids. A user inhales through the mouthpiece; power is supplied to the heating element to heat the e-liquid; the e-liquid is soaked from the reservoir [1] into a wick that is in contact with the heating coil . The heating can be triggered by an internal airflow sensor that detects the inhalation airflow or by an external manual switch. The heated e-liquid will evaporate and travel through the mouthpiece. In general there is a single air path over the heater with cooling in the space post-heater. In some designs, ambient air is channelled into the mouthpiece during inhalation through small holes on the sides [1] of the ECs that maintain the airflow and pressure drop . Upon cooling, the e-liquid vapour condenses into droplets that are inhaled. Since ECs are not considered as pharmaceutical products and they are not clear-cut tobacco products (since eliquids may or may not contain nicotine) their regulation has been difficult. Even if regulations are in place, they are of a lower standard than those for pharmaceutical products. However, some countries (Brazil, Singapore, [2] Canada, the Seychelles, and Uruguay) have banned all EC products to avoid potential concerns to the public . On the other hand, others (e.g. Hong Kong SAR, New Zealand, and Switzerland) ban only nicotine-containing eliquids but allow unregulated sale of nicotine-free ones. The revised European Union Tobacco Products Directive and the Tobacco and Related Products Regulations of the United Kingdom have similar stipulations. According to these, ECs are regulated as tobacco products for liquids ≤ 20 mg/mL nicotine and reusable or disposable devices with refillable containers ≤ 10 mL or cartridges ≤ 2 [3,4] . There are also requirements on the labelling and mL. Nicotine-free e-liquids are not subjected to this directive information supplied to consumers with the EC products. In May 2016, the United States Food and Drug Administration extended tobacco regulation to cover ECs because they are now deemed to be tobacco products. Thus the manufacture, labelling, packaging, and sale of EC products are now controlled under the Family Smoking Prevention and Tobacco Control Act. For example, they cannot be sold to consumers younger than 18 years old and free samples cannot be given out. A health warning statement regarding the adverse effects of nicotine must be displayed on the packaging. An alternative statement [5] (‘This product is made from tobacco’) is required even if the product is self-certified to be nicotine-free . The major difference between ECs and conventional pharmaceutical inhaler products is that thermal vaporization ® is required to form the aerosol. In this respect, it is similar to the Staccato delivery system, in which a drug pre[6] coated onto a solid substrate surface is rapidly heated upon inhalation by the patient . The vaporized drug condenses into an aerosol due to cooling while travelling through the mouthpiece. However, ECs also differ from ® the Staccato in that they contain a range of Generally Recognised As Safe (GRAS) but not fully tested excipients ® while the latter is excipient-free. Furthermore, ECs’ formulations are liquids whereas that of the Staccato is a ® solid. Nicotine has been successfully delivered from a Staccato system, although this product has not been [7] commercialised .

217


Drug Delivery to the Lungs 26, 2016 - Droplet Sizes of Electronic Cigarette Aerosols

Since nicotine is a liquid under ambient conditions, it cannot form a physically stable film on the substrate. Therefore, it was chemically combined with thermally reversible zinc halides (ZnBr2 and ZnCl2) to form molecular complexes that can be loaded as solid coatings. Highly pure aerosols containing 99% nicotine could be generated with an emitted dose of 116.8 μg, of which 57% was droplets with a volumetric mean diameter of 800 nm and [7] 43% was vapor . The small droplet size indicated that they can deposit in the deep lungs. It should be noted that ® the Staccato requires much higher airflow rates than ECs to emit an aerosol (15-80 L/min vs 1-3 L/min). If we compare their aerosol performance in the light of this, then ECs are much more efficient because it requires much lower inhalation effort by the patient in actuating the dose. This property is beneficial in pulmonary drug delivery because inhalation effort can be adversely affected by the disease state, such as that in asthma. Two EC-like products have been approved by the Medicines and Healthcare products Regulatory Agency (MHRA) for nicotine replacement therapy, namely Voke Inhaler and e-Voke Electronic Inhaler. The Voke Inhaler is a breath-actuated MDI in the shape of an EC containing nicotine. It uses the propellant, HFA-134a to provide [8] the energy for aerosol generation . Unlike ECs, no heating is involved in the Voke Inhaler. Besides the [8] propellant, it includes propylene glycol, ethanol, saccharin, and levomenthol as excipients . This product was ® [8] deemed to be therapeutically equivalent to the reference product, Nicorette Inhalator . Unlike the Voke Inhaler, the e-Voke Electronic Inhaler is actually an EC in its design and operation principle. Thus it is the first EC that has been approved as a pharmaceutical product. It uses a cartridge system containing a simple formulation, with [9] glycerin and water for injections as the vehicle . The small number of excipients in the e-Voke formulation is probably for reducing the amount of likely thermal degradants. This product has also been demonstrated to be ® [9] therapeutically comparable to Nicorette Inhalator . The effects of the e-liquid formulation (e.g. solvent and excipient compositions etc) and device operation factors (e.g. heating temperature and duration, airflow rate etc) on the aerosol properties have not been extensively studied. Thus it is worth to explore these aspects. E-liquids are solutions that are comprised of a solvent vehicle [10] and solutes. The solvent base can be a mixture of glycerol, propylene glycol, ethanol, and water . The solutes include nicotine and food flavourings. Several hundred flavours are available on the market, ranging from conventional (e.g. fruits, chocolate, menthol) to exotic ones (e.g. bubble gum, butterscotch, gourmet cinnamon, pie crust, root beer). E-liquids may or may not contain nicotine, just as they may or may not contain all four solvents listed above. Since regulation requirements for e-liquids are relatively loose, the labelling of the identity and quantity of their ingredients are often unclear or even erroneous. The relationship between droplet size distribution and the formulation composition has not been studied for ECs. Thus the aim of this study was to measure the inter-liquid effect on the size distributions of droplets produced from a commercial EC using laser diffraction, which is a technique for testing inhaled products, with a view to compare ECs with pharmaceutical inhalers in the future. Experimental methods and materials Aerosols produced from five solutions (see below) were sized by laser diffraction. All concentrations are expressed as % w/w. 

100% glycerol

100% propylene glycol

Mixture A (14.3% water, 0.1% ethanol, 71.3% glycerol, 14.3% propylene glycol)

Mixture B (25% each for water, ethanol, glycerol, propylene glycol)

Commercial nicotine-free e-liquid (grapefruit flavoured Check-Point Premium E-Liquid; Check-Point, Hong Kong SAR) o

The contents of this e-liquid are unclear. The label on the packaging listed ‘food grade vegetable glycerine, propylene glycol, food grade glycerol, distilled water, natural flavours, artificial flavours’ but no amount or proportion of the ingredients were disclosed.

The investigators did not add nicotine or flavourings to the above liquids. Mixture A was intended to feature equal amounts of the four solvents commonly used in e-liquids, whereas Mixture B was a random combination to demonstrate that the droplet size distribution was independent of the solvent composition (see Results). Each solution was tested with one V12 reservoir-type EC (Kimree, Shenzhen, China). The volume loaded into the EC was 0.5 mL. Aerosols were sampled by a vacuum pump at 2 L/min for 3 s across the measuring zone in a HELOS KR laser diffractometer (Sympatec, Clausthal-Zellerfeld, Germany). There was no dilution air through the laser diffractometer. The flow rate (2 L/min) and duration (3 s) for the sampling are typical for a puff inhaled by an EC [11,12] . Data were collected with the WINDOX 5.8.0.0 software (Sympatec, Clausthal-Zellerfeld, Germany). Six user runs were conducted per liquid.

218


Drug Delivery to the Lungs 27, 2016 - Philip Chi Lip Kwok et al.

Results The volumetric droplet size distributions of the aerosols generated with the five testing liquids are shown in Figure 1. Overall, the distributions were monomodal and were comparable between the different liquids. They were particularly variable near the peak sizes. Figure 2 shows the 10, 50, and 90% undersize diameter parameters (i.e. D10, D50, and D90, respectively). D10 and D50 were comparable with relatively low variability across the liquids, while D90 was highly variable. There were minimal nano-droplets. 50

16

Glycerol

D10 D50

Propylene glycol

40

Mixture A

Commercial e-liquid

20

Diameter (μm)

% Volume

Mixture B

30

D90

12

8

4

10

0

0 0.1

1

Diameter (μm)

10

Glycerol

100

Propylene glycol

Mixture A

Mixture B

Commercial eliquid

Figure 2. Diameter parameters for aerosols produced from Kimree V12 EC. Data presented as mean ± standard deviation (n = 6).

Figure 1. Droplet size distribution of aerosols produced from Kimree V12 EC. Data presented as mean ± standard deviation (n = 6).

Discussion The Kimree V12 EC produced droplets of an inhalable size, with volumetric median diameters between 3-4 µm. It is interesting that similar size distributions were obtained regardless of the liquid used. This suggests that the aerosol generating mechanism through heat vaporisation and subsequent condensation did not depend on the vehicle composition. At least that was the case for the four solvents tested (water, ethanol, glycerol, propylene glycol). Pure water and pure ethanol were tested in preliminary experiments but no aerosol could be produced. In fact, they leaked out of the small holes on the sides of the EC that were intended for the entrance of dilution air. This was probably due to the lower viscosity of water and ethanol compared to glycerol and propylene glycol (Table 1). Their lower boiling points would also render them more difficult to condense into droplets upon cooling. Therefore, glycerol, propylene glycol, and mixtures containing them are more able to form aerosols using the EC. The droplet size appeared to be driven by propylene glycol or glycerol regardless of the exact composition of the liquid vehicle (i.e. 100% glycerol or propylene glycol produced the same droplet size distribution as their diluted forms). The high variation in the larger droplet sizes may be due to the low quality of the devices. Since most ECs are not regulated as pharmaceutical products, their production and quality control standards are lower than those for medical inhalers. The high variability in the size distributions, especially around the peak sizes, could be due to the poor control in the heating of the e-liquid at the coil. The heating temperature and duration in the EC may be variable. This could cause variable temperatures in the resultant vapour, hence affecting the condensation rate and droplet size. Dynamic viscosity (mPa. s) Boiling point (°C)

Water 0.89 at 25°C

Ethanol 1.22 at 20°C

100

78.5

Glycerol 111.0 at 20°C (for 83% w/w aqueous glycerol solution) 290

Propylene glycol 58.1 at 20°C

188

Table 1. Dynamic viscosity and boiling point of the solvents [14]

It has been reported in the literature that the diameter of the droplets/particles generated from two unnamed commercial ECs were in the nanometer range (250-450 nm), which were comparable to those produced from [13] tobacco cigarettes . The small droplets implied that they should deposit in the deep lungs.The aerosols in that study were sampled and measured by a differential mobility spectrometer coupled to a smoking cycle [13] simulator . On the other hand, the volumetric diameters in the micron range were measured by laser diffraction in this study. While droplets < 5 μm would enter the lungs, the larger ones may deposit in the oropharynx and be absorbed buccally or swallowed.

219


Drug Delivery to the Lungs 26, 2016 - Droplet Sizes of Electronic Cigarette Aerosols The difference in the measured droplet size range between the two studies could be due to differences in the testing setup, experimental procedure, and/or the EC model. Laser diffraction and differential mobility analysis have different mechanistic principles. The former technique measures the scattering pattern of a laser beam imparted to the aerosol sample. It is an optical technique that detects particles from about 100 nm to the millimeter range and can be affected by variations in the refractive index. On the other hand, the latter technique imparts electrostatic charges to the particles and then measures the mobility of the charged particles in an electric field. It can measure particles down to about 2 nm and is not affected by the refractive index. Thus differential mobility analysis is more sensitive for sizing nanoparticles. The Sympatec HELOS KR laser diffractometer employed in the current study can measure particle diameters from 0.1-8750 μm. Thus if the droplets produced from the Kimree V12 EC were a few hundred nanometres in diameter, they should still be detectable. More data using another sizing technique are required for confirmation of the laser diffraction measurements. Droplet sizing using an aerodynamic impaction would be conducted in the future to provide more information. In essence, ECs share a common purpose with conventional pharmaceutical inhalation devices: they produce aerosols intended to be orally inhaled into the respiratory tract. However, ECs are much cheaper and more popular than pharmaceutical inhalers, as evident from their widespread recreational use. The fact that droplets of inhalable and comparable sizes were produced in this study using different solvent vehicles shows a potential advantage in this EC system. Drugs of different solubility may be accommodated by using different co-solvent mixtures, while the droplet size distribution would not be affected. This provides more formulation freedom and a wider design space. Therefore, it is worth to examine ECs further from the viewpoint of inhalation drug development to explore what can be applied to respiratory drug delivery in the future. Conclusion The Kimree V12 EC produced respirable droplets from a range of solutions containing various amounts of glycerol and propylene glycol. There was relatively high variation in the sizes of the large droplets but overall the distributions were comparable between the different liquids. EC technology may be investigated further to gain insights that may be useful for the development of novel inhaled products in the future. References 1. Brandon TH, Goniewicz ML, Hanna NH, Hatsukami DK, Herbst RS, Hobin JA, Ostroff JS, Shields PG, Toll BA, Tyne CA, Viswanath K, Warren GW. 2015. Electronic nicotine delivery systems: a policy statement from the American Association for Cancer Research and the American Society of Clinical Oncology. Clin Cancer Res 21(3):514-525. 2. Grana R, Benowitz N, Glantz SA. 2013. Background Paper on E-cigarettes (Electronic Nicotine Delivery Systems). 2013. San Francisco, US: Center for Tobacco Control Research and Education, University of California. 3. European Parliament and European Council of the European Union. 2014. Directive of the European Parliament and of the Council on the Approximation of the Laws. ed.: European Parliament and European Council of the European Union. 4. 2016. The Tobacco and Related Products Regulations 2016. Norwich, UK: The Statutory Office. 5. 2016. Deeming Tobacco Products To Be Subject to the Federal Food, Drug, and Cosmetic Act, as Amended by the Family Smoking Prevention and Tobacco Control Act; Restrictions on the Sale and Distribution of Tobacco Products and Required Warning Statements for Tobacco Products. US: Food and Drug Administration. 6. Dinh KV, Myers DJ, Noymer PD, Cassella JV. 2010. In Vitro Aerosol Deposition in the Oropharyngeal Region for Staccato® Loxapine Journal of Aerosol Medicine and Pulmonary Drug Delivery 23:253-260. 7. Simis K, Lei M, Lu AT, Sharma KC, Hale RL, Timmons R, Cassella J. 2008. Nicotine aerosol generation from thermally reversible zinc halide complexes using the Staccato® system. Drug Development and Industrial Pharmacy 34(9):936-942. 8. 2014. UKPAR Voke/Nicotine 0.45mg Inhaler. ed., London, UK: Medicines & Healthcare Products Regulatory Agency. 9. 2015. UKPAR e-Voke 10 & 15mg Electronic Inhaler. ed., London, UK: Medicines & Healthcare Products Regulatory Agency. 10. Etter J-F. 2012. The Electronic Cigarette: an Alternative to Tobacco? ed.: CreateSpace Independent Publishing Platform. 11. De Jesus S, Hsin A, Faulkner G, Prapavessis H. 2015. A systematic review and analysis of data reduction techniques for the CReSS smoking topography device. Journal of Smoking Cessation 10:12-28. 12. Lopez AA, Hiler MM, Soule EK, Ramôa CP, Karaoghlanian NV, Lipato T, Breland AB, Shihadeh AL, Eissenberg T. 2016. Effects of electronic cigarette liquid nicotine concentration on plasma nicotine and puff topography in tobacco cigarette smokers: A preliminary report. Nicotine & Tobacco Research 18:720-723. 13. Ingebrethsen BJ, Cole SK, Alderman SL. 2012. Electronic cigarette aerosol particle size distribution measurements. Inhalation toxicology 24(14):976-984. 14. Rowe RC, Sheskey PJ, Quinn ME editors. 2009. Handbook of Pharmaceutical Excipients. ed., London, UK: Pharmaceutical Press.

220


Drug Delivery to the Lungs 27, 2016 – Nazli Nezami et al. A Study of Factors Affecting Nucleation and Bubble Growth in Pressurised Metered Dose Inhalers Nazli Nezami & Hendrik Versteeg Loughborough University, Wolfson School of Mechanical, Electrical and Manufacturing Engineering, Ashby Road, Loughborough, LE11 3TU, UK Summary Various hypotheses have been introduced to explain disintegration of the continuous liquid phase into individual droplets leading to spray formation in pressurised metered dose inhalers (pMDIs). In a practicable system, the liquid formulation to be discharged from the pressurised container needs to be nucleated to ensure spray generation. Nucleation can be described as the generation of a nucleus of the vapour phase within the bulk liquid. As a stable nucleus is formed, it grows significantly and then detaches from its nucleation site to move upwards in the liquid phase. In our research, the effects of various parameters on the nucleation of HFA227 was systematically analysed with the aim of gaining a better understanding of bubble formation and the nucleation process in HFA propellants , including the surface geometrical properties, actuator orifice size and the mass flow rate through the orifice. Other important factors influencing the nucleation process that were considered comprised the viscosity and surface tension of the formulation, thermodynamic state variables including temperature, pressure and degree of superheat. The results highlighted the effect of surface imperfections on the rate of nucleation and bubble growth. A comparison of different actuator orifice sizes was made and a significant change in the shape and motion of the bubbles was observed. An intense nucleation was also observed at higher mass flow rate of HFA227 through the valve. It is anticipated that recognising the factors affecting nucleation and bubble growth of HFA227 may lead to potential routes of influencing the medical aerosol generation mechanism inside the pMDI and control the fine particle size distribution. Introduction The function of the pMDIs is highly dependent on the coordination of patient inhalation and device actuation. Correct synchronizing of the breathing and firing the inhaler leads to a better drug deposition in the lungs. However, even with good inhalation technique, only 10-20% of the emitted dose reaches the biological target with [1] most of the dose being trapped in the mouth and oropharynx . High oropharyngeal deposition, which is due to the large particle size distribution and spray high velocity, may lead to the systemic and localized side effects. To some extent, recent developments in HFA pMDIs have addressed poor lung deposition and high oropharyngeal deposition. Despite the significant importance of pMDIs in treating respiratory diseases, there is a poor understanding of the inhaler internal flow characteristics and spray formation mechanism. The present research is a study of aerosol formation in pressurised metered dose inhalers. Its major aim is to improve understanding of the thermo-fluid dynamic processes leading to droplet generation to predict the characteristics of the aerosol cloud and eventually control the fine particle size distribution. The process of disintegration of the liquid formulation within the inhaler into the discrete droplets leading to spray formation is termed atomization. There are many hypotheses developed to explain atomization process. [2] According to Wiener disintegration of the continuous liquid phase into the individual droplets happens due to [3] flash evaporation. Finlay also suggested that as there are numerous vapour bubbles and cavities trapped on the internal surface of the pMDI components, flash evaporation or cavitation are the main mechanisms [4] dominating atomization in pMDIs. According to Polanco et al flash evaporation of pressurised liquid formulation is critically reliant on the nucleation and existence of nucleation sites. Nucleation can be described as the onset of a phase transition through which the system is temporarily brought into a thermodynamic unstable condition from its initial sub-cooled state and becomes superheated. In the superheated liquid state, a nucleus can overcome the nucleation energy barrier and is free to grow. In a pMDI, the superheated condition is achieved by an isothermal pressure drop and results in the appearance of vapour bubbles in the low-pressure regions. These vapour bubbles grow rapidly to a comparatively large size. The faster pressure drop results in higher degree of superheat for liquid formulation, and the more intense vaporisation. The nucleation process is categorised into two types: homogeneous nucleation and heterogeneous nucleation. Superheat is crucial for the generation of vapour phase to overcome the cohesive strength of the continuous liquid phase and tear it up. Homogeneous nucleation occurs entirely within the bulk liquid without the presence of vapour nuclei and normally requires high degrees of superheat. The heterogeneous nucleation initiates from pre-existing vapour nuclei in the bulk liquid or when the superheated liquid is in contact with a gas phase or a solid phase (e.g. Container walls, valve components or suspended drug particles). Relatively low degrees of superheat are sufficient for heterogeneous nucleation, so the [5] probability of heterogeneous nucleation is much higher than that of homogeneous nucleation . Newly formed vapour bubbles in a pMDI actuator grow to form small nuclei, which detach from the bulk liquid as the flashing jet is moving downstream from the spray orifice. Among the important parameters governing the liquid disintegration process are the number and size of the bubbles at the spray orifice. Bubble growth rate has been extensively researched in the past studies and it has been concluded that the growth process is generally controlled by the liquid inertia, [6] surface tension, and the pressure difference between the ambient and inside of the bubble . This paper reports

221


Drug Delivery to the Lungs 27, 2016 – A Study of Factors Affecting Nucleation and Bubble Growth in Pressurised Metered Dose Inhalers the findings of a visualisation study to investigate nucleation and bubble growth in HFA placebo formulations used in pMDI applications. Experimental Method The major aim of this research is the microscale analysis of the nucleation process and bubble growth taking place within a pressurised container filled with mixtures of liquefied pharmaceutical propellants. To fulfil this goal a test rig, which is shown in Figure 1, has been designed and manufactured. The experimental procedure involved imaging the bubble formation and growth within the propellant mixture along with measuring the changes in pressure and temperature of the mixture during the nucleation and bubble growth processes. Continuous valve

Thermocouples feedthrough

Gas phase and liquid phase thermocouples

Pressure sensor Figure 1 - Experimental setup

As can be seen in the figure above, the test rig is made from aluminium with two large circular openings that receive the windows; the windows are made from transparent PET (polyethylene terephthalate) to allow clear imaging. The test apparatus with a volume of about 44 mL is relatively large in size compared to a typical metering chamber (25 to 100 ÂľL in volume), but this choice facilitates a fundamental study of vapour nucleation. Small openings are made in the aluminium enclosure for receiving pressure sensors, thermocouples, and the continuous valve. The filling and depressurisation of the test rig was performed using the continuous valve. Discharge of the pressurised formulation was accomplished by two methods: (i) through continuous flow valve for small mass flow rates and (ii) by opening one of the plugs to achieve larger flow rates. A high speed video camera connected to a computer was used to record the nucleation and growth of vapour bubbles while propellant was discharging into the atmosphere (Figure 2). The captured videos then were analysed by means of a Matlab program and parameters such as the bubble diameter, bubble growth rate along with the changes in the bubble shapes and dynamics were calculated. In addition, a second computer equipped with data logging software was utilised in order to monitor the changes in the temperature and pressure during each actuation.

High speed video camera Test rig

Light source Figure 2 - Experimental layout

Results Initial Observations

The videos captured during the initial experiments showed bubbles originating from the bottom and the sidewall where the pMDI valve and the pressure tapping were located, as shown in Figure 3. It is assumed that defects and crevices on the container internal surface led to air being trapped at these surface imperfections during the filling process, and resulted in bubble nucleation originating from pre-existing gas cavities and vapour nuclei trapped at these locations. When the bulk liquid is in contact with a solid phase or a gas phase the interfacial energy is lower, which reduces the [7] minimum energy required for the initiation of nucleation, the so-called nucleation free energy barrier . Videos taken when propellant discharged through the pressure-tapping plug showed intense nucleation. This was due to the fact that a larger propellant flow rate discharged from the container, the pressure difference between

222


Drug Delivery to the Lungs 27, 2016 – Nazli Nezami et al. [8]

the container and the ambient further more rapidly . This resulted in an increase in the degree of superheat; hence, the conditions were more favourable for bubble nucleation and consequently the number of nucleation sites increased, resulting in intense nucleation. A

B

Figure 3 – (A) Initial observation of nucleation and (B) Nucleation during rig exhaustion Bubble Shape Observations

Experiments using actuators with different orifice sizes showed an interesting change in the shape and motion of the bubbles. Videos captured during experiments with smaller actuator orifice showed bubbles moving in a vertical line. In this case, as can be seen in Figure 4(A), bubbles remained relatively spherical. However, bubbles produced while the propellant was discharging through the larger actuator orifice, deviated from spherical as they were travelling in a zigzag manner in the liquid phase (Figure 4(B)). B

A

Figure 4 – (A) Nucleation using smaller actuator orifice and (B) Nucleation using larger actuator orifice Bubble growth rate and rise velocity

The changes in the diameter of a specific bubble, the clearest bubble in the captured videos, along with the fluctuations in its rise velocity, were calculated. A sequence of images was analysed by following a bubble while it was rising in the liquid phase whilst it was visible in the viewing window until it coalesced with the liquid-vapour interface and eventually entered the vapour phase. These results were plotted against the time in Figure 5. They showed an increase of approximately 0.21mm for the bubble diameter while it was moving upwards in the liquid phase. As can be seen in the figure, the bubble diameter grew gradually up until 3.3 seconds and then it started to oscillate. The same trend can be seen in terms of the bubble rise velocity. It was relatively constant until 3.3 seconds, and then it quickly dropped and started to oscillate. The reason was that after 3.3 seconds the vapour bubble reached the liquid-vapour interface and before entering the vapour phase it rebounded back into the liquid phase a few times. As the vapour bubble approached the liquid-vapour interface the following forces were acting on the bubble determining its size and trajectory: (i) the downward force due to the surface tension in the thin layer of liquid between the bubble and the interface, (ii) the bubble buoyancy force, (iii) inertia of the accelerating liquid (including added mass) and (iv) the bubble internal pressure. The combination of these forces resulted in oscillations in the bubble diameter and bubble rise velocity; until the vapour bubble overcame the pressure build [9] up made by liquid layer, and coalesced with the interface .

223


Drug Delivery to the Lungs 27, 2016 – A Study of Factors Affecting Nucleation and Bubble Growth in Pressurised Metered Dose Inhalers

Figure 5 - Bubble growth rate and rise velocity

Discussion and Conclusions Bubble nucleation and growth in pharmaceutical propellants is dominated by device macro- and micro-geometry (i.e., metering chamber volume, expansion chamber geometry, orifice dimensions and geometrical properties of the internal surface), physicochemical properties of the fluid (viz. nature of the liquid formulation, liquid surface tension, liquid density, gas density and liquid viscosity), and operating parameters, such as the propellant flow rate and degree of superheat at the beginning of the process. These parameters generally govern the bubble size at the discharge orifice by controlling the mode of bubble formation, bubble frequency, and also bubble [10] detachment . This detailed study will reveal the most important factors controlling the modes of vapour phase generation. In this work a procedure for the study of these factors affecting nucleation and bubble growth was developed and demonstrated. The results suggested that vapour nucleated from pre-existing gas cavities due to surface imperfections. It was shown that how different mass flow rates through actuators with different orifice sizes affect bubble shapes and dynamics. In particular, an intense nucleation at high mass flow rates was observed. Furthermore, the changes in the diameter and displacement of bubbles while they were travelling up through the liquid were studied. Once bubbles reached the liquid-gas interface, they were found to bounce back into the liquid phase a few times before they burst through the interface. As a result, an oscillation was observed in terms of bubble growth rate and bubble rise velocity. The results showed that nucleation process and bubble growth can be affected by changes in the surface geometrical properties and mass flow rate through the valve. Increasing the number of heterogeneous nucleation sites inside the experimental setup gives rise to a significant increase in the rate of nucleation and bubble growth. Further work will also seek to investigate the role played by surface roughness in nucleation as well as the effects of scale and volume/surface area ratio. These results may ultimately lead to potential routes of influencing spray generation mechanism inside the pMDI with potential contributions to the design and development of next-generation inhaler devices meeting the patient’s requirements more effectively. References 1 2

3 4

SP Newman: Principles of metered-dose inhaler design, Respir. Care J 2005; 50 (9): pp 1177-1190. MV Wiener: How to formulate aerosols to obtain the desired spray pattern, Soc. Cos. Chem. J 1958; 9: pp 289-297. W Finlay: The Mechanics of inhaled pharmaceutical aerosol: an introduction, Academic Press, 2001. Polanco, Geanette, AE Holdø, and G Munday: General review of flashing jet studies, J Hazard Mater 2010; 173(1): pp 2-18.

5

D Kawano, H Ishii, H Suzuki, Y Goto, M Odaka, and J Senda: Numerical study on flash-boiling spray of multicomponent fuel, Heat Tran Asian Res. 2006; 35(5): pp 369-385. 6

MS Plesset and SA Zwick: The growth of vapour bubbles in superheated liquids, J. Appl. Phys. 1954; 25: pp 493-500.

7

SF Jones, GM Evans, and KP Galvin: Bubble nucleation from gas cavities_ a review , Adv. Colloid Interface Sci. 1999; 80(1): pp 27-50. 8

AR Clark: Metered atomization for respiratory drug delivery, Doctoral dissertation, Loughborough, UK, 1991.

9

R Manica, E Klaseboer, and DYC Chan: The impact and bounce of air bubbles at a flat fluid interface, Soft Matter. J 2016; 12: pp 3271-3282. 10

AA Kulkarni and JB Joshi: Bubble formation and bubble rise velocity in gas-liquid systems: a review, Ind. Eng. Chem. Res. 2005; 44(16): pp 5873-5931.

224


Drug Delivery to the Lungs 27, 2016 - Nurcin Ugur et al. Leucine Coated Inhalable Insulin and Thymopentin Peptide Powders Produced by Aerosol Flow Reactor Method 1

1

2

2

1

Nurcin Ugur , Ville Vartiainen , Luis M. Bimbo , Jouni Hirvonen , Esko I. Kauppinen and Janne Raula

1

1

2

Department of Applied Physics, Aalto University School of Science, Espoo, 00076, Finland Division of Pharmaceutical Chemistry and Technology, Faculty of Pharmacy, University of Helsinki, 00014, Finland

Summary Pulmonary route of drug delivery has been the focus of intense research in order to improve the efficacy of therapeutic peptides both for local and systemic treatment. However, formulation of such inhalable powders is challenging since the peptide may be in inactive conformation due to the thermodynamic effects of particle preparation process. In this research, inhalable insulin and thymopentin peptide powders were prepared by the aerosol flow reactor method from precursor solutions of peptide, trehalose, sodium citrate and leucine for the protection of the labile peptides. Scanning electron microscopy images of inhalable peptide powders showed intact and separated particles with spherical and partly-buckled shapes. Particle size distribution of insulin particles was 1.13 m and thymopentin particles was 0.63 m. Aerosolization of inhalable powders were measured with an inhalation testing device showed that emitted dose (ED) of the insulin powders were 1.9 mg/dose and fine particle fractions (FPF) of the insulin powders were 55-59%. Thymopentin powders exhibited 1.5-1.7 mg/dose of ED and 27-36% of FPF. Additionally, inhalable insulin powders were analysed by circular dichroism after the particle preparation process and results indicated similar -helical conformation which assures the conformational stability of insulin after the aerosol process. Furthermore, high performance liquid chromatography analysis of both thymopentin and insulin powders indicated a similar corresponding solute peak as the pristine samples at the same retention time. Protective encapsulation of peptides was shown to be successful and temperature pulse did not affect the peptide stability. In the near future, cellular interaction, cytocompatibility and drug permeation properties of the peptide powder particles will be investigated. Introduction The discovery of new therapeutic peptides and the improvement of conventional therapeutic treatments demand new strategies for an efficient drug delivery. Innovative biopharmaceutical molecules are proposed to treat the rising number of diseases such as for cancer treatments with monoclonal antibodies [1], vaccines [2] and insulin for the treatment of diabetes [3]. However, the efficient delivery of the therapeutic peptides has remained a challenge in current pharmaceutical research due their large molecular size, hydrophilicity and chemical and enzymatic lability [4]. Peptides are rapidly degraded by enzymes or by low pH conditions in gastrointestinal tract and as injected the peptides face rapid hepatic degradation and low patient compliance. These challenges stimulated much of the research towards the pulmonary delivery of the peptides as a promising non-invasive alternative route for the delivery of labile peptides. An ultimate demand in the particle engineering of peptides is that the peptides are still active in their resulting inhalable formulation. Therefore, we have developed an aerosol-based method [5] where peptide particles are gently dried followed by rapid consecutive decrease and increase in the saturation condition of a coating material L-leucine. This renders particles flowable and dispersible [6] and also provides protective encapsulation [7] for peptides while the activity of peptides in sustained. Materials and Methods Human insulin in powder form was purchased from Benjamin Pharmaceutical Chemical Co. Ltd., China. Insulin solution was prepared by lowering the pH to 2.8 with 0.1 M acetic acid solution. After dissolution of the insulin, the pH was adjusted to 7.5 with 1 M NaOH solution. D-Trehalose (Sigma Aldrich, USA), sodium citrate tribasic dihydrate (Sigma Aldrich, Japan) and L-Leucine (Alfa Aesar, Germany) were weighed and dissolved in the insulin solution. Drug-to-excipient ratio was 1:9. Final solution was used in the aerosol process. Resulting insulin particles is referred as Ins-Tr-L particles in this paper. Peptide-free excipient (trehalose, leucine and sodium citrate) particles are referred as Tr-L particles. Thymopentin (TP5) in powder form was ordered from Huajin Pharma, China. Precursor thymopentin (2.5 g/l) solution contained trehalose (12.9 g/l), L-leucine (6.4 g/l) and sodium citrate (3.2 g/l) dissolved in ion-exchanged water. Final solution was used in the aerosol process. The resulting thymopentin particles is referred as TP5-Tr-L particles in this paper. Droplets were generated with an ultrasonic nebulizer and transferred with nitrogen gas into a laminar flow reactor -4 that was set to 50C. Liquid volume consumption was 710 with a gas flow rate 20 l/min. Thereafter the temperature was pulsed up to 230 C for 0.25 s of residence time. At the downstream of the reactor, the dry particles were cooled and diluted in a porous tube. Insulin microparticles were collected and stored at 20C at 0% relative humidity for further analysis.

225


Drug Delivery to the Lungs 27, 2016 - Leucine Coated Inhalable Insulin and Thymopentin Peptide Powders Produced by Aerosol Flow Reactor Method

Particle size distributions in the gas phase were determined with an electrical low-pressure impactor, ELPI (Dekati LTD., Finland). Morphology of the Ins-Tr-L particles, TP5-Tr-L particles and Tr-L particles were imaged with a scanning electron microscope (Zeiss Sigma VP) at 1.5 kV. High Performance Liquid Chromatography (HPLC) elution profiles of pristine insulin, pristine thymopentin, Ins-Tr-L particles, TP5-Tr-L particles and Tr-L particles were analysed using Agilent-Hewlett Packard 1050 Series HPLC System with UV detector at ambient temperature. Chromatography was performed with a SunFire RP-C18 column (4.6x150 mm, 5µm). Sample concentrations of 3 to 5 mg/ml were run with the injection volume of 40 l and the peptide was detected by UV at wavelength 220 nm. The mobile phase consisted of acetonitrile and 0.1% trifluoroacetic acid. A linear gradient from 0 to 60% acetonitrile was applied in 30 min at 1 ml/min. Circular dichroism (CD) data were collected using a JASCO J-720 CD spectropolarimeter. Pristine insulin and insulin aerosol samples were diluted in 0.1 M acetic acid solution to get final concentration of protein 0.1 mg/ml. CD spectra were collected in the wavelength range of 190-260 nm using 1mm path length cuvette. Temperature controller was used and measurements were carried out at 20C. Each final spectrum was averaged from three consecutive scans. Spectra of the acetic acid solution and the excipients were recorded and subtracted from InsTr-L particle sample spectrum. The resulting spectra were smoothed by the Savitzky–Golay function using a convolution width of 20 points. Fine powder aerosolization of powders was studied using an inhalation testing device developed in-house. The detailed operating principles has been discussed before [6][8]. Briefly, Easyhaler reservoir was filled with 1 g of the peptide powder and inhalation was performed 10 times. Inspiration flow rates of 40 L/min and 55 L/min were adjusted by pressure drops to 2 kPa and 4 kPa, respectively. Particles were collected on the stages of a Bernertype low pressure impactor (BLPI) wherefrom mass distributions were measured gravimetrically. Mass median aerodynamic diameters (MMAD), geometric standard deviation (GSD), of each deposited powders were calculated and fine particle fractions (FPF) were calculated according to the emitted dose (ED). Moisture sensitivity analysis of peptide particles (Ins-Tr-L and TP5-Tr-L particles) and excipients particles (Tr-L particles) were conducted in desiccators, where 44%, 65%, and 75% relative humidity (RH) were achieved with o saturated solutions of the salts NaNO2, K2CO3, and NaCl at 25 C, respectively. Solid state properties of the samples were then characterized with scanning electron microscopy (SEM), differential thermal analysis (DTA) and x-ray diffraction (XRD) at 2nd and 7th days. XRD and DTA results are still underway. Results and Discussion The SEM images (Figure 1) of inhalable powders show intact and separated particles with spherical and partlyfolded shape. The incorporation of insulin or thymopentin did not affect the surface morphology of the particles since no significant morphological differences are observed between Figure 1A, 1B and 1C. Number average size of Tr-L (only excipients: trehalose, sodium citrate and leucine) particles was 0.93 m, Ins-Tr-L particles was 1.13 m and TP5-Tr-L particles was 0.63 m in gas phase. A

B

C

Figure 1 - SEM images of (A) Tr-L particles (excipients particles) (B) Ins-Tr-L particles (insulin particles) (C) TP5-Tr-L particles (thymopentin particles).

Powder aerosolization of excipient powders and peptide powders were compared in Table 1. Excipient (Tr-L) powders emitted dose (ED) was higher than insulin particles (Ins-Tr-L) and thymopentin (TP5-Tr-L) particles. On the other hand, fine particle fraction (FPF) of insulin (Ins-Tr-L) powders was highest compared to excipient (Tr-L) particles and thymopentin (TP5-Tr-L) particles. The pristine peptides and peptide powders were further characterized and compared by reversed-phase high performance liquid chromatography. Both pristine insulin and insulin (Ins-Tr-L) particles indicated a similar corresponding solute peak and elution profile with retention time at 24.9 min and similarly, retention time of both pristine thymopentin and thymopentin (TP5-Tr-L) particles were at 13.9 min.

226


Drug Delivery to the Lungs 27, 2016 - Nurcin Ugur et al.

Table 1 flows.

SAMPLE

Aerosolization results of the L-leucine coated powders from Easyhaler at two pressures and inhalation

ED (mg/dose)

CVED

FPF ( 5m,%)

GSD

MMAD (m)

2 kPa

4 kPa

2 kPa

4 kPa

2 kPa

4 kPa

2 kPa

4 kPa

2 kPa

4 kPa

Ins-Tr-L

1.9

1.9

0.3

0.4

55

59

1.6

1.3

1.9

1.7

TP5-Tr-L

1.6

1.7

0.6

0.5

36

27

1.2

1.1

2.1

2.0

Tr-L

2.3

2.6

0.2

0.1

46

53

2.0

2.0

1.8

1.9

Circular dichroism spectra (Figure 2) of pristine insulin (green line) and insulin (Ins-Tr-L) particles (red line) were almost the same indicating similar -helical conformation which assures the conformational stability of insulin after the aerosol process. Therefore, it can be stated that the new way of engineering the leucine coating can be successfully applied for the preparation of inhalable peptide particles. Stability of insulin particles were further studied with particles that were stored at room temperature (RT) at 0% relative humidity for 87 days. The CD spectra of stored particles (blue line) did not exhibit irregular shaped peptide structure and exhibited similar spectra as the pristine insulin, indicating an -helical conformation. Therefore, the insulin particles are stable at room temperature at 0% relative humidity.

Figure 2 – Circular dichroism spectra of pristine insulin (grey line), insulin particles (orange line) and insulin particles stored at room temperature at 0% relative humidity for 87 days.

Stability of the inhalation powders were further investigated under humid conditions and morphology analysis of each particle is shown in Figure 3. In general, all the particles under humid conditions did not strongly coagulated as clumps and remained as individual particles. However, formation of necks with adjacent particles was observed on insulin particles (Ins-Tr-L) at 75% RH on Day 7. The aerosolization and agglomerate formation of particles under humid conditions will be further analysed by inhalation test device. Moreover, solid state characteristics of particles under humid conditions will be investigated by XRD analysis and DTA analysis.

227


Drug Delivery to the Lungs 27, 2016 - Leucine Coated Inhalable Insulin and Thymopentin Peptide Powders Produced by Aerosol Flow Reactor Method

Figure 3 – The SEM images of the particles under humid conditions (44%, 65%, 75% RH) at room temperature. The samples were collected and imaged on Day 2 and Day 7 after conditioned.

Conclusions We have demonstrated the production of conformationally stable insulin and thymopentin inhalable dry powders. A rapid change in temperature needed for the leucine coating did not affect the structure and the conformation of the peptides. Protective encapsulation of peptides was shown to be successful and temperature pulse did not affect the peptide stability. We will further analyse the solid state properties of the peptide powders with other methods such as XRD, DTA and FTIR spectroscopy to confirm the peptide stability. Furthermore, particle permeation through cell monolayer experiments will be studied in the near future for analysing the efficacy of peptide particles. Acknowledgements We thank Academy of Finland (project no. 289080) for financial support. This work made use of Aalto University Nanomicroscopy Center (Aalto-NMC) facilities. Maryna Chukhlieb is acknowledged for providing the CD device. Maxim Antopolsky acknowledged for providing HPLC device. We also thank Teicos Pharma for collaboration. References 1. 2. 3. 4. 5. 6. 7.

8.

Louis M. et al. Monoclonal antibodies: versatile platforms for cancer immunotherapy. Nat Rev Immunol 10 (2010) 317327 Vanneman, M. and Dranoff, G. Combining immunotherapy and targeted therapies in cancer treatment Nat Rev Cancer 12 (2012) 237-251 Hovorka, R., et al. Manual closed-loop insulin delivery in children and adolescents with type 1 diabetes: a phase 2 randomised crossover trial. Lancet 375 (2010) 743–751. Almeida A. J. and Grenda, A. Technosphere®: An inhalation system for pulmonary delivery of biopharmaceuticals in Mucosal Delivery of Biopharmaceuticals Eds. Eds. das Neves, J. and Sarmento B. 2014, pp 483-498 Raula, J., Kuivanen, A., Lähde, A., Kauppinen, E.I., 2008. Gas-phase synthesis of L-leucine-coated micrometer-sized salbutamol sulphate and sodium chloride particles, Powder Technol., 187, 289–297 Raula, J., Lähde, A., Kauppinen, E.I., 2009. Aerosolization behavior of carrier-free L-leucine coated salbutamol sulphate powders, Int. J. Pharm., 365, 18–25 Raula, J., Thielmann, F., Kansikas, J., Hietala, S., Annala, M., Seppälä, J., Lähde, A., Kauppinen, E.I., 2008. Investigations on the humidity-induced transformations of salbutamol sulphate particles coated with L-leucine, Pharm. Res., 25, 2250-2261 Kauppinen, E., Kurkela, J., Brown, D., Jokiniemi, J., Mattila, T., 2002. Method and apparatus for studying aerosol sources, WO 02/059574.

228


Drug Delivery to the Lungs 27, 2016 – Lei Mao et al. Application of Effective Data Analysis (EDA) and Abbreviated Impaction Method (AIM) in the Metered Dose Inhaler (MDI) Product Screening Lei Mao, Alexandra Hughes & Carole Evans Catalent Pharma Solutions, 160N Pharma Drive, Morrisville, North Carolina, 27560, USA Summary Effective Data Analysis (EDA) and Abbreviated Impaction Method (AIM) concepts have been recommended and advocated by the IPAS-RS Cascade Impaction Working Group as a quality control (QC) tool for inhalation product testing due to its efficiency and effectiveness in distinguishing product performance. The purpose of this study is to evaluate the feasibility of applying EDA and AIM in the early phase of product development. We retrospectively analyzed the aerodynamic particle size distribution (APSD) results generated using the next generation pharmaceutical impactor (NGI) on a group of suspension metered dose inhalers (MDIs) consisting of different formulation/container closure variants before and after three month storage at 40°C/75%RH using the EDA approach. We assessed the correlation between the mass median aerodynamic diameter (MMAD) and large particle mass (LPM) to small particle mass (SPM) ratio, fine particle dose (FPD) and impactor sized mass (ISM) and their changes after storage. A good correlation was observed between the MMAD and LPM/SPM at an effective cut off diameter (ECD) of 2.30 µm which is close to MMAD range of 2.5-3.3 µm for this group of product variants. A good correlation was also observed between FPD and ISM as well as their change after storage. It was concluded from the data analysis that LPM/SPM ratio and ISM can replace MMAD and FPD as an indication for the product performance evaluation therefore an AIM with stages of a properly selected impactor size and ECD can be used for initial product screening. Introduction Effective Data Analysis (EDA) and Abbreviated Impaction Method (AIM) concepts have been recommended and advocated by the IPAS-RS Cascade Impaction Working Group as a quality control (QC) tool for inhalation product [1][2] . testing although it is not yet included in the regulatory guidance and pharmacopeia EDA is a matrix approach of using the impactor sized mass (ISM) and large particle mass (LPM) to small particle mas (SPM) ratio to present aerodynamic particle size distribution and enables detection of the outliers which is not readily achievable by fine particle dose (FPD) and mass median aerodynamic diameter (MMAD) alone. In our previous evaluation, the EDA concept was demonstrated appropriate in correlating the MMAD and large particle mass (LPM) to small particle mass (SPM) ratio when an appropriate effective cut off diameter (ECD) was selected using the impaction data generated during solution metered dosed inhaler (MDI) and dry powder inhaler [3][4] . (DPI) evaluation The purpose of this presentation is to demonstrate suitability of EDA approach and AIM in the early phase MDI product screening. MDI performance is determined by the formulation, container closures and device (actuator). In the early MDI product development phase, we would like to screen as many combinations as possible and quickly rule out those candidates with little potentials. Therefore, a high efficient rather than high resolution screening tool is more desirable. For example, the AIM with stages of an appropriate impactor size and ECD could expedite testing and serve as a better option compared to the full stage impaction methods if its suitability can be justified. To demonstrate suitability of EDA and AIM approach, we took one of our MDI development programs as an example and retrospectively analysed the APSD results generated from the NGI to evaluate the correlation between the ISM and FPD, LPM/SPM ratio and MMAD as well as the correlation between the changes of these parameters during feasibility stability. If the correlation can be demonstrated, then the ISM and LPM/SPM results generated from AIM can be predictive and replace FPD and MMAD for the product screening purpose. Experimental Methods For preparing the MDI canisters, the suspension formulations were compounded in a 10L vessel and filled into Presspart cans crimped with either Bespak or Aptar valves using laboratory scale equipment to give a matrix of ten variants. The filled canisters were then quarantined for fourteen days in the testing laboratory prior to performance testing. The canisters were also stored in a valve down orientation at a 40°C/75%RH condition. APSD profile of each variant delivered from the same actuator was measured using a Next Generation Pharmaceutical Impactor (NGI) at 30L/min. The impactor cups were recovered and analyzed using HPLC to determine drug deposition on each stage and filter.

229


Drug Delivery to the Lungs 27, 2016 - Application of Effective Data Analysis (EDA) and Abbreviated Impaction Method (AIM) in the Metered Dose Inhaler (MDI) Product Screening Results Tables 1 presents the stage deposition of the APSD by NGI results (n=2) generated before and after the canisters were stored valve down at 40°C/75%RH. For each variant, two canisters were tested. MMAD, FPD, ISM, LPM/SPM at an ECD of 3.99 µm, 2.30 µm and 1.36 µm were calculated per the notes for Table 1. Table 1

V-C

NGI stage deposition measured from MDI variants before and after storage at 40°C/75%RH and calculated fine particle dose (FPD, less than 5 µm), MMAD, ISM and LPM/SPM at different effective cut off diameters NGI stage deposition (µg) 3 4 5 6

Ind

1

2

1-1 1-2 2-1 2-2 3-1 3-2 4-1 4-2 5-1 5-2 6-1 6-2 7-1 7-2 8-1 8-2 9-1 9-2 10-1 10-2

23.9 31.7 21.2 22.2 23.5 17.0 11.8 13.2 9.9 10.4 9.7 11.7 12.0 11.4 12.9 13.4 13.6 12.3 10.4 10.7

0.62 0.60 0.60 0.59 0.58 0.47 0.38 0.38 0.34 0.32 0.32 0.41 0.38 0.41 0.32 0.37 0.29 0.27 0.21 0.25

1.69 2.05 1.34 1.23 1.63 1.29 0.92 0.83 0.61 0.84 0.65 0.55 0.78 0.92 0.88 0.98 0.95 0.62 0.34 0.40

4.46 6.29 3.68 3.31 4.48 3.00 2.59 2.20 1.69 2.17 2.43 2.20 2.33 2.66 2.42 2.69 2.50 1.88 1.03 1.14

8.17 11.3 6.95 6.61 8.45 5.78 5.04 4.83 3.57 4.52 5.49 4.84 4.74 5.69 5.01 5.86 4.98 4.26 5.56 5.95

1-1 1-2 2-1 2-2 3-1 3-2 4-1 4-2 5-1 5-2 6-1 6-2 7-1 7-2 8-1 8-2 9-1 9-2 10-1 10-2

26.3 25.5 25.0 20.5 24.9 23.9 15.8 17.6 20.3 12.7 15.5 10.8 18.6 14.8 15.3 14.8 11.8 12.4 11.2 10.4

0.46 0.50 0.60 0.73 0.56 0.37 0.41 0.15 0.52 0.41 0.38 0.36 0.28 0.25 0.20 0.27 0.31 0.47 0.17 0.24

2.03 1.38 1.76 1.51 1.69 1.51 0.71 0.61 0.52 0.87 0.59 0.55 0.49 0.84 0.58 0.79 0.91 0.93 0.38 0.56

5.45 4.18 3.82 3.75 3.81 3.84 1.74 1.67 0.97 2.00 2.03 1.92 1.05 2.05 1.41 1.83 2.21 2.06 1.02 1.43

8.78 7.35 6.42 6.08 6.58 6.67 3.43 3.48 2.11 3.80 4.51 4.51 2.16 3.82 3.00 3.50 4.39 3.95 5.18 6.26

7

MO Filter C Before storage 4.98 1.72 0.55 0.37 0.12 6.46 1.97 0.77 0.31 0.14 4.11 1.37 0.50 0.25 0.09 3.96 1.57 3.37 0.35 0.10 5.06 1.62 0.53 0.27 0.07 3.64 1.20 0.55 0.26 0.08 3.09 1.19 0.46 0.23 0.08 3.41 1.18 0.42 0.20 0.07 2.58 0.89 0.31 0.16 0.08 2.98 0.99 0.38 0.17 0.08 3.95 1.47 0.43 0.67 0.07 3.37 1.30 0.50 0.20 0.06 3.12 1.25 0.35 0.14 0.06 3.61 1.22 0.53 0.24 0.06 3.30 1.27 0.45 0.19 0.06 4.01 1.27 0.39 0.21 0.07 3.19 0.59 0.11 0.05 0.05 2.88 0.54 0.10 0.05 0.08 4.56 0.52 0.09 0.01 0.04 4.68 0.53 0.08 0.03 0.05 After three month storage at 40°C/75%RH 4.91 1.51 0.71 0.31 0.13 4.18 1.33 0.53 0.26 0.08 3.02 0.90 0.30 0.13 0.03 2.96 0.95 0.30 0.15 0.05 3.87 1.26 0.51 0.24 0.05 3.66 1.07 0.32 0.15 0.04 2.12 0.72 0.27 0.15 0.02 2.38 0.77 0.34 0.18 0.04 1.62 0.71 0.26 0.11 0.02 2.45 0.85 0.31 0.14 0.03 3.41 1.38 0.53 0.25 0.04 3.24 1.62 0.66 0.30 0.05 1.61 0.52 0.34 0.13 0.03 2.31 0.71 0.30 0.10 0.02 2.20 0.74 0.28 0.11 0.02 2.21 0.72 0.27 0.13 0.03 3.17 0.51 0.13 0.07 0.02 2.59 0.56 0.15 0.08 0.02 3.82 0.47 0.07 0.02 0.01 4.28 0.54 0.10 0.03 0.01

MM AD

FPD (µg)

ISM (µg)

LPM/SPM at 3.99 2.30 1.36 (µm) (µm) (µm)

2.93 2.97 2.95 2.52 2.94 2.89 2.84 2.72 2.71 2.78 2.56 2.64 2.81 2.81 2.75 2.78 2.98 2.86 2.50 2.56

18.3 21.8 14.3 18.5 20.2 13.1 11.1 11.0 8.6 10.2 11.7 10.7 9.8 10.0 10.2 12.2 11.0 9.6 9.5 9.9

22.1 29.3 18.3 20.5 22.1 15.8 13.6 13.2 9.9 12.1 15.2 13.0 12.8 14.9 13.6 15.5 12.4 10.4 12.1 12.9

0.39 0.40 0.38 0.28 0.38 0.37 0.35 0.30 0.30 0.33 0.25 0.27 0.32 0.32 0.32 0.31 0.38 0.32 0.13 0.14

1.85 2.03 1.89 1.19 1.93 1.76 1.69 1.49 1.46 1.64 1.30 1.40 1.59 1.64 1.58 1.60 2.11 1.85 1.33 1.40

7.00 8.19 7.28 2.80 7.86 6.56 5.94 6.01 5.88 6.49 4.73 5.34 6.06 6.28 5.90 6.98 14.4 12.5 17.4 17.8

3.07 3.00 3.28 3.22 3.04 3.08 2.91 2.76 2.77 2.92 2.59 2.52 2.81 2.99 2.74 2.94 2.94 3.03 2.59 2.67

19.5 16.2 12.8 12.7 14.7 14.1 7.7 8.2 5.3 8.7 11.3 11.5 5.3 8.4 7.1 7.8 9.5 8.6 10.1 12.0

23.8 19.3 16.4 15.8 18.0 17.3 9.2 9.5 6.3 10.5 12.7 12.9 6.3 10.2 8.3 9.5 11.4 10.3 11.0 13.2

0.46 0.40 0.52 0.50 0.44 0.45 0.37 0.32 0.31 0.38 0.26 0.24 0.32 0.40 0.31 0.38 0.38 0.41 0.15 0.18

2.15 2.02 2.74 2.57 2.04 2.29 1.79 1.55 1.32 1.76 1.27 1.19 1.41 1.95 1.49 1.82 1.93 2.04 1.50 1.66

7.96 7.77 11.0 9.86 7.74 9.92 6.90 6.12 4.75 6.86 4.79 3.89 5.21 7.98 6.25 7.24 14.6 11.8 18.3 18.4

V-C: Variant No-Can No; Ind: NGI induction port Cut of diameter was 11.7 µm, 6.4 µm, 3.99 µm, 2.30 µm, 1.36 µm, 0.83 µm, 0.54 µm for NGI stages 1 to 7, respectively; MOC: Micro orifice collector, MMAD: unit in µm, FPD: Fine particle dose less than 5 µm calculated using the Copley CITDAS software (v2.0) LPM/SPM at 3.99µm: Sum of mass deposited from cup 2 to cup 3/Sum of mass deposited from cup 4 to filter LPM/SPM at 2.30µm: Sum of mass deposited from cup 2 to cup 4/Sum of mass deposited from cup 5 to filter LPM/SPM at 1.36µm: Sum of mass deposited from cup 2 to cup 5/Sum of mass deposited from cup 6 to filter MMAD: Calculated using the Copley CITDAS software (v2.0) ISM: Sum of mass deposition from stage 2 to filter

230


Drug Delivery to the Lungs 27, 2016 – Lei Mao et al. Discussion Correlation between MMAD and LPM/SPM ratio ISM, LPM, SPM and LPM/SPM at different ECDs (1.36 μm, 2.30 µm, and 3.99 µm) were calculated using the equations noted under Table 1. Impactor size for this study is 11.7 µm (cup one). MMAD and FPD less than 5 µm were calculated using Copley CITDAS software (version 2.0, Copley Scientific Ltd., Nottingham, UK). The MMAD measured from ten variants was in the ranges of 2.52 μm to 2.98 μm and 2.52 μm to 3.28 μm before and after three month storage, which was close to the ECD value of 2.3 µm.

20.00

20.00

15.00

15.00

10.00 R² = 0.007

5.00 R² = 0.9219 R² = 0.7862

0.00 2.30

2.80

ISM and SPM3.99µm)(µg)

ISM and SPM3.99µm)(µg)

The LPM/SPM ratios calculated at different ECDs (1.36 μm, 2.30 µm, and 3.99 µm) were plotted against the [3][4] , a good MMAD (Figure 1a before storage and Figure 1b after storage). Similar to our previous findings correlation was observed between the MMAD and LPM/SPM ratio when an ECD close to the MMAD, i.e. 2.30 µm 2 was selected. The correlation coefficient between LPM/SPM and MMAD (R ) was 0.9219 and 0.8877 at an ECD of 2.30 µm whereas the correlation coefficient was 0.007, 0.004 at an ECD of 1.36 µm and 0.7862 and 0.9030 at an ECD of 3.99 µm before and after storage at 40˚C/75%RH. This suggests that an appropriate ECD of 2.30 µm should be selected if AIM is to be used in the initial screening of this product.

10.00

R² = 0.004

5.00

R² = 0.8877 R² = 0.903

0.00

3.30

2.30

2.80

FPD less than 5 µm (µg) LPM/SPM3.99µm

LPM/SPM2.30µm

3.30

FPD less than 5 µm (µg) LPM/SPM3.99µm

LPM/SPM1.36µm

LPM/SPM2.30µm

LPM/SPM1.36µm

a b Figure 1 - Correlation between MMAD and LPM/SPM calculated from different ECDs before (a) and after (b) storage

Correlation between ISM and FPD Figure 2 presents the correlation between the FPD and ISM before (a) and after (b) storage. A good correlation 2 between two parameters was observed (R = 0.9245 before storage and 0.9853 after storage). 35.0

30.0

30.0

R² = 0.9853

20.0

20.0

ISM (µg)

ISM (µg)

25.0

R² = 0.9245

25.0

15.0 10.0

15.0 10.0 5.0

5.0 0.0

0.0 0.0

10.0

20.0

30.0

0.0

FPD less than 5 µm (µg)

10.0

20.0

30.0

FPD less than 5 µm (µg)

a Figure 2 - Correlation between FPD and ISM before (a) and after (b) storage

b

Correlation between changes in the ratio between MMAD and LPM/SPM, PDF and ISM ratio before and after storage Figure 3 presents the correlation between changes in MMAD and LPM/SPM2.30µm (a), FPD and ISM (b) after stability storage. 2

A good correlation was observed between changes in MMAD and LPM/SPM2.30µm (R = 0.9285) and changes in 2 FPD and ISM (R = 0.7413).

231


Drug Delivery to the Lungs 27, 2016 - Application of Effective Data Analysis (EDA) and Abbreviated Impaction Method (AIM) in the Metered Dose Inhaler (MDI) Product Screening

5.00

0.00

Change in ISM (µg)

-10.00

-5.00

0.00 -5.00

-10.00

-15.00

5.00

Change in LPM/SPM at 2.30 µm (µm)

5.00

R² = 0.7413

3.75

2.50 R² = 0.9285

1.25 0.00

-0.50

-20.00

0.00

0.50

1.00

-1.25 -2.50

Change in FPD < 5µm (µg)

Change in MMAD (µm)

Figure 3 Correlation between changes in MMAD and LPM/SPM 2.30µm (a), FPD and ISM (b) after stability storage.

Conclusions APSD by NGI results generated from suspension MDIs were retrospectively analyzed using the EDA approach to exemplify its application in the initial product development. A good correlation was observed between the MMAD and LPM/SPM at ECD2.30µm. A good correlation was also observed between the FPD and ISM. There was a good correlation between the changes in the MMAD and LPM/SPM2.30µm as well as the changes in FPD and ISM after stability storage, It was concluded from this example that AIM can be used for the initial product screening when stages of an appropriately selected impactor size and ECD are selected (cup one at 11.7 µm and cup four at 2.30 µm in this example). The ISM and LPM/SPM values can be indicative therefore to replace FPD and MMAD for the product performance evaluation and initial screening. 1

IPAC-RS_Newsletter_February_2016 online http://ipacrs.org/assets/uploads/outputs/IPACRS_Newsletter_February_2016.pdf

2

Tougas TP, Mitchell JP, Lyapustina S (eds) Good Cascade Impactor Practices, AIM and EDA for Orally Inhaled Products, Springer Science & Business Media, 2013

3

Mao L, Ponder D, Hughes A, White J, Joshi K, Glaab V, Mitchell JP, Lyapustina S: Efficient Data Analysis Case Study I: Pressurized Metered Dose Inhalers Containing a Solution Formulation, Respiratory Drug Delivery 2012; 2:pp 441-446, 2012

4

Joshi K, Carlson D, Finney C, Fairbrother T, Mao L, Mitchell JP, Lyapustina S: Efficient Data Analysis Case Study II: Dry Powder Inhaler Formulation, Respiratory Drug Delivery 2012; 2:pp447-452, 2012

232


Drug Delivery to the Lungs 27, 2016 – Bridie Dutton. Development of an inhaled ion-paired salbutamol formulation 1

1

2

2

1

Bridie Dutton , Arcadia Woods , Robyn Sadler , Nathalie Fa , Ben Forbes & Stuart Jones 1

1

Institute of Pharmaceutical Science, King’s College London, 150 Stamford Street, London, SE1 9NH, UK 2

GlaxoSmithKline, Park Road, Ware, Hertfordshire, SG12 0DP, UK

Summary Ion-pairing is a promising technique to control drug delivery that has been shown to be effective for both transdermal and oral medicines. This strategy could also be very interesting for inhaled formulations as it has the potential to control the delivery of a drug without changing the structure of the molecule. Previously, salbutamol has been shown to bind the negatively charged counter ions sodium sulfate, gluconate and octanoate with varying degrees of strength. However, to administer these ion-pairs to the lung an inhalable formulation must be developed. Spray drying is an attractive method to engineer particles for an inhalable formulation as it allows the control of particle size, which influences lung deposition. In this study spray dried ion-pair formulations were developed using lactose, as the bulking agent, PVP, as a physical stabilizer, and l-leucine, to improve powder dispersion. In vitro deposition studies with the spray dried formulations using the Next Generation Impactor demonstrated that the gluconate and octanoate ion-pairs improved the dispersibility of the salbutamol microparticles. The addition of PVP to the microparticles generally increased their size except for the octanoate formulation, which produced a smaller particle and better dispersion. The addition of l-leucine generally improved the dispersibility of the powders, but it didn’t affect the particle size, hence resulted in a smaller MMAD in comparison to the counter ion/PVP formulations. The results of this study show that it is possible to make an ionpair formulation that is suitable for inhalation. Introduction Oppositely charged ions held together through non covalent interactions have the potential to create a controlled release system for inhaled medicines because the ion-paired complex can display different physicochemical properties compared to the parent drug. Previous studies have showed that salbutamol binds to sulfate, gluconate and octanoate with differing strengths (pK = 1.57, 2.27, and 2.56 respectively) to form biocompatabile complexes [1] . However, the formulation used in this previous work was a simple lactose based microparticle, which is unlikely to covey the stability or aerosolizability suitable for a pharmaceutical product. Spray drying is a method of particle engineering that is commonly used to generate inhalable particles, mainly because it allows precise control of the [2] particle size of the powder , but it often creates amorphous particles, which require additional excipients to [3, 4] . physically stabilize the powder to prevent agglomeration The aim of this study was to engineer inhalable particles containing salbutamol ion-pairs in a chemically and physically stable state that could be deposited in the lung. Due to the possibility of amorphous lactose causing the powder to have poor results the inclusion of polyvinylpyyrolidone (PVP) and l-leucine in to the formulation was also studied. PVP has been proven to increase the stability of amorphous lactose over time due to its ability to [4] inhibit crystallization, therefore it was chosen as the excipient to physically stabilize the microparticles . In [5] addition, l-leucine was added to the particles to improve the dispersibility of the powder . The aersolisation and [5] stability of the powders was studied . Methods Formulations containing 1 % w/w salbutamol and 99% lactose were spray dried using a Buchi B-191 spray dryer o with the following settings: aspirator: 80%, feed pump: 10%, inlet temperature: 180 C, air flow: 800L/h. The solid content of the feed solution was 3 g/ 100 mL of deionized water. The counter ions were added to the feed solution in a 20:1 counter ion to drug ratio. The PVP (10k) was added as 0.5 % of the total solid content and l-leucine was added as 5% of the total solid content. The spray dried powder was collected and sized using a Helos/Rodos laser diffraction particle-size analyzer. Approximately 100 mg of powder was used to obtain each size distribution. The R1 lens was used (size range 0.1 μm – 35 μm), the dispersion pressure was set at 6.00 bar and the vacuum was set at 30 - 32 mbar. The powders were also sized and surface morphology examined using scanning electron microscopy (SEM). A Zeiss Supra 25 SEM was used. The sample was deposited with a brush on a carbon disc stuck to an aluminium stub. The stub was then coated with platinum for 60 seconds. Physical stability of the powders was assessed using differential scanning calorimetry (DSC). Experiments were performed on a TA Q2000 modulated DSC. A sample of 2 – 5 mg was used in a non-hermetically sealed o o aluminium pan. The spray dried powder components were heated from 25 to 180 C at a heating rate of 10 Cmin 1 o before the temperature was brought back to 25 C. The instrument was calibrated using indium prior to the experiments. DSC thermograms were analysed using TA Instruments Universal Analysis 2000 software. In-vitro drug deposition was studied using a next generation impactor (NGI). The powders were fired into the system using a Pheonix monodose dry powder inhaler. A total of three assessments were carried out for each

233


Drug Delivery to the Lungs 27, 2016 – Development of an inhaled ion-paired salbutamol formulation spray dried formulation and 5 capsules were used per assessment. Capsules were prepared by manually loading 21 12.5 (± 0.5) mg of spray dried powder in to a size 3 hard gelatin capsule. The NGI was set up according to USP . The NGI collection cups were coated with 19.25 % PEG200 in Acetone, 4 mL for stage 1, and 2 mL for all other stages. The formulations were tested at 100 L/min for 2.5 seconds. After testing the throat, all stages and the external filter paper were washed with 80 % (v/v) water-methanol mixture. HPLC was used to quantify the amount of salbutamol deposited at each stage. Data were represented as a mass median aerodynamic diameter (MMAD), geometric standard deviation (GSD), emitted dose (ED), and fine particle fraction (FPF). The ED of the formulation was calculated as the mass deposited in the throat and stages 1-8 of the NGI as a percentage of the mean content of salbutamol found in the content uniformity analysis. The FPF was determined as the percentage of the ED that has a particle size less than 5 μm and it was calculated from a plot of the cumulative mass deposited vs. effective cut off diameter. MMAD and GSD were calculated by plotting the inverse normal of the cumulative % under the stated aerodynamic diameter versus the log effective cut off diameter at 100 L/min. The MMAD was taken as the value at which the trend line for this data intersects with 50 % cumulative % value. GSD was calculated as the 84th percentile/16th percentile from this data. Results and Discussion The effect of adding counter ions The spray drying process produced a white powder with a good yield for all the test formulations. The median particle size was generally around 3 μm with each counter ion apart from sodium gluconate, which produced a significantly smaller particle size of around 2.5 μm (p <0.001). DSC results showed a glass transition (Tg) for the o base and sulfate formulations at around 120 C, however none was seen for the gluconate or octanoate. SEM pictures of all formulations showed smooth circular particles, which suggests that the particles were amorphous. The NGI results showed that addition of the sulfate to the salbutamol in the microparticles had no effect on the aerosolisation of salbutamol from the inhaler device. The addition of the gluconate counterion resulted in a significantly smaller MMAD (p = 0.024) and higher FPF value compared to base and sulfate, presumably due to its smaller particle size. The octanoate formulation displayed a similar MMAD value (p = 0.032) to that of the gluconate, despite its larger particle size, suggesting that the addition of the counter ion improved the dispersion of the powder. The effect of adding PVP The addition of PVP in to the spray drying mixtures resulted in a larger particles being generated for all the counter ion drug mixtures apart from the octanoate formulation, which was significantly smaller than the counter ion formulation. SEM pictures show that the PVP had an effect on the surface morphology of the particles; it generated “dimples” which suggested that the polymer had migrated to the particle surface during the spray drying process. The DSC results showed no glass transition for any of the formulations, which was presumably a consequence of the complex mixture containing four components resulting in signal overlap. NGI results show that due to the larger size of particle the general trend is that the MMAD for each formulation is higher and the FPF lower. The ED remains the same. This trend is not followed by the octanoate formulation which exhibits a much smaller MMAD value and higher FPF; however it also has a smaller ED. This could be due to the fact that the sodium octanoate has its own effect on the surface properties of the particles, and creates a more aerosolisable particle. The effect of adding l-leucine Spray drying with l-leucine created a much more free flowing powder for all apart from the octanoate formulation. The particle size increased for each formulation, however the dispersibility also increased. This resulted in a smaller MMAD value for the sulfate counter ion compared to the equivalent formulations without l-leucine. There was also no significant difference in the MMAD value for the gluconate/leucine formulation compared to the PVP even though the particle size was increased. The addition of l-leucine to the octanoate powder, however did not have the same effect. The resulting MMAD value was significantly higher than either of the 2 previous formulations, however the ED was greatly increased. There were also no Tg seen for any of these formulations.

234


Drug Delivery to the Lungs 27, 2016 – Bridie Dutton.

Table 1 – Particle size, in vitro deposition results and SEM pictures (Scale = 1 μm) for all formulations (± standard deviation, n=3)

Formulation

Counter ion

x50 (μm)

ED (%)

MMAD (μm)

GSD (μm)

FPF (%)

1% Salbutamol

None

3.68 (± 0.26)

74.24 (± 2.14)

6.30 (± 0.71)

2.31 (± 0.04)

25.34 (± 4.08)

99% Lactose

Sulfate

3.61 (± 0.28)

74.28 (± 1.35)

6.24 (± 0.36)

2.20 (± 0.09)

25.14 (± 1.42)

Gluconate*

2.57 (± 0.35)

76.57 (± 1.71)

4.85 (± 0.35)

2.09 (± 0.01)

33.52 (± 3.24)

Octanoate*

3.28 (± 0.06)

79.00 (± 3.06)

4.85 (± 0.73)

2.39 (± 0.18)

27.83 (± 2.84)

1% Salbutamol

Sulfate

3.92 (± 0.07)

68.78 (± 2.41)

5.50 (± 0.11)

2.24 (± 0.03)

22.87 (± 3.41)

98.5% Lactose

Gluconate

3.36 (± 0.13)

73.91 (± 3.79)

5.15 (± 0.62)

2.30 (± 0.16)

25.74 (± 3.44)

0.5 % PVP

Octanoate**

3.06 (± 0.12)

64.10 (± 3.54)

4.41 (± 0.21)

2.30 (± 0.11)

36.23 (± 1.45)

1% Salbutamol

Sulfate

4.21 (± 0.05)

79.38 (± 1.12)

5.19 (± 0.15)

2.51 (± 0.03)

34.93 (± 3.36)

93.5% Lactose

Gluconate

3.94 (± 0.06)

87.58 (± 2.74)

5.30 (± 0.26)

2.44 (± 0.22)

29.96 (± 5.08)

0.5 % PVP

Octanoate

3.20 (± 0.04)

84.42 (± 2.09)

5.09 (± 0.38)

2.56 (± 0.04)

33.39 (± 1.21)

Counter ion

Counter ion

5 % l-leucine Counter ion

235


Drug Delivery to the Lungs 27, 2016 – Development of an inhaled ion-paired salbutamol formulation

A)

B)

C)

Figure 1 – Example SEM images of the A) counter ion, B) counter ion/PVP and C) counter ion/PVP/leucine formulations (scale = 1 μm)

Conclusion In conclusion, addition of the gluconate and octanoate counter ions in to the formulation had a beneficial effect, resulting in a smaller MMAD and larger fine particle fractions. The addition of PVP resulted in a general trend of the powder having larger particles and therefore a larger MMAD value for each formulation except the octanoate. The addition of PVP to the octanoate formulation instead resulted in a smaller particle size and also the lowest MMAD value of the study. Leucine caused a slight increase in particle size for each formulation. Only the sulfate formulation appeared to benefit from the leucine in terms of improved dispersion. The results of this study show that it is possible to create a spray dried ion-pair formulation, but that it is important to understand how the chosen counter ion and excipients interact as this can vary the aerosolisability of the powders. References 1

Dutton, B. L., Sadler, R., Forbes. B., Jones, S. A. “Ion-pairs – a novel formulation strategy to alter drug disposition in the lungs” Drug Delivery to the Lungs, Edinburgh, Scotland, 2015.

2

Masters, K. (1991) Spray Drying Handbook. Harlow: Longman Scientific & Technical ; New York : Wiley

3

Naini, V., Byron, P. R., Philips, E. M. (1998) Physicochemical stability of crystalline sugars and their spray dried forms:Dependence upon relative humidity and suitability for use in powder inhalers. Drug Development and Industrial Pharmacy. 24. 895-909

4

Berggren, J., Alderborn, G. (2004) Long-term stabilization potential of poly(vinylpyrrolidine) for amorphous lactose in spray-dried composites. European Journal of Pharmaceutical Sciences, 21(2–3), 209-215 5

Mangal, Sharad, Meiser, Felix, Tan, Geoffrey, Gengenbach, Thomas, Denman, John, Rowles, Matthew R., . . . Morton, David A. V. (2015). Relationship between surface concentration of l-leucine and bulk powder properties in spray dried formulations. European Journal of Pharmaceutics and Biopharmaceutics, 94, 160-169

6

Samiei, N., Mangas-Sanjuan, V., Gonzalez-Alvarez, I., Foroutan, M., Shafaati, A., Zarghi, A., Bermejo, M. (2013) Ion-pair strategy for enabling amifostine oral absorption: Rat in situ and in vivo experiments. European Journal of Pharmaceutical Sciences. 49. 499-504

7

Xi, H., Wang, Z., Chen, Y., Li, W., Sun, L., Fang, L. (2012) The relationship between hydrogen-bonded ion-pair stability and transdermal penetration of lornoxicam with organic amines. European Journal of Pharmaceutical Sciences. 47. 325-330 8

Song, W., Cun, D., Xi, H., Fang, L. (2012) The control of skin-permeating rate of Bisoprolol by ion-pair strategy for long-acting transdermal patches. AAPS PharmSciTech. 13. 811-815

236


Drug Delivery to the Lungs 27, 2016- Segolene Sarrailh, et al. The continued development of a cyclic olefin copolymer elastomer sealing system for metered dose inhalers Segolene Sarrailh, Chris Baron, Gerallt Williams Aptar Pharma, Le Vaudreuil, 27100, France Summary This work reports further proof around the potential value of cyclic olefin copolymer elastomer (COCe) as an innovative sealing material for metered dose inhalers. As previously reported cyclic olefin copolymers are a new class of polymeric materials with property profiles which can be varied over a wide range during polymerization. These new materials exhibit a unique combination of properties which can be customized by varying the chemical structure of the copolymer. Interesting performance benefits for metered dose inhalers (MDIs) applications could be considered as low water absorption, excellent water vapour barrier properties, rigidity, strength and compression set, excellent biocompatibility, good resistance to acids and alkalis and their ability to be moulded into specific component parts. This work details further proof around the performance of COCe as an innovative sealing material, during this study key performance attributes were investigated such as leak rates and moisture ingress, and new investigations have been made into formulation compatibility and the characterisation of any potential extractables from the COCe material. Other manufacturing related parameters such as crimping and recycling have previously been investigated and reported. Introduction The sealing of a metered dose valve to a canister is a key part of the overall integrity of the MDI system and this work reports on the continued development of a novel COCe elastomer sealing system. As previously reported cyclic olefin [1] copolymers are a new class of polymeric materials with property profiles which can be varied over a wide range during polymerization. These new materials exhibit a unique combination of properties which can be customized by varying the chemical structure of the copolymer. Interesting performance benefits for MDI applications could be considered as - low water absorption, excellent water vapour barrier properties, rigidity, strength and compression set, excellent biocompatibility, good resistance to acids and alkalis and their ability to be moulded into specific component parts. COCs have already found successful industrial applications in the parenteral drug arena and their unique [2] combination of properties, detailed above, make them ideal solutions for applications such as prefilled syringes . This work reports further proof around the potential value of COCe as an innovative sealing material, during this study key performance attributes were investigated such as leak rates and moisture ingress, and new investigations have been made into formulation compatibility and the characterisation of any potential extractables from the COCe material. Other manufacturing related parameters such as crimping and recycling have previously been investigated (1) and reported . Experimental methods Moisture ingress levels were assessed by manufacturing MDIs containing COCe type sealing gaskets coupled with various gathering rings and filled with hydro fluoro alkane (HFA) without ethanol. Samples were stored at 40°C/75%RH, inverted position, over 12 months and their moisture contents were determined using a validated Karl 3 Fisher method and were compared to EPDM (ethylene propylene diene monomer) and nitrile sealing gaskets. Leak rates were assessed by weight loss on MDIs containing COCe type sealing gaskets and filled with HFA without ethanol stored at 40°C/75%RH inverted position, and their annual leak rates calculated and compared to EPDM and nitrile sealing gaskets. Crimping parameters were optimized in order to get the best fit for each of the material evaluated. Compatibility was evaluated for both suspension and solution types formulations upon accelerated ageing, considering that the main compatibility risks for such formulations are respectively drug adhesion on materials and drug loss. Drug adhesion was investigated using a suspension of micronized salbutamol sulphate with HFA134a, which was manufactured in-house and pressure-filled through Aptar DF30 metering valves, stored for 6 months at 40°C/75%RH, inverted position. Adhesion of powder to canister, neck gasket & gathering ring and the rest of the metering valve were determined using a validated HPLC (high performance liquid chromatography) method and compared to EPDM sealing gaskets.

237


The continued development of a cyclic olefin copolymer elastomer sealing system for metered dose inhalers

Two different model solution formulations were also employed, chosen to evaluate the compatibility of COCe with both weak organic and strong mineral acids: a mixture of Ipratropium bromide, ethanol, water, citric acid and HFA134a and a mixture of formoterol fumarate, ethanol, hydrochloric acid, isopropyl myristate and HFA134a. Both were manufactured in house, by first preparing a concentrated solution (ethanol, active pharmaceutical ingredient and excipients) which was introduced into canisters, and then adding the propellant through the crimped valve. Although the second formulation shows rapid and high loss of active ingredients when stored at accelerated conditions, it serves as an investigative method only to assess material compatibility over a shorter time period. An Aptar DF30 metering valve was used and the remaining percentage of the active ingredient was determined after storage for 6 months at 40°C/75%RH, inverted positon, using validated HPLC methods and compared to EPDM sealing gaskets. Extractable levels were assessed for EPDM and COCe raw materials components following reflux extraction with ethyl acetate for each material type and quantified using validated HPLC (high performance liquid chromatography, UV detection, reverse phase C18 column, eluting solvents water/ethyl acetate/acetonitrile) and GCFID (gas chromatography flame ionisation detection, dimethylpolysiloxane capillary column, helium carrier gas) analytical methods.

Results The results of moisture ingress and leak rate testing for the various MDIs assessed are given in Figure 1 and 2 respectively. Leak rates (mg, n=10)

Moisture ingress up to T12M (n=15) T1M T6M T12M

800

EPDM & PA

Nitrile & PA

COCe & PA

637.1 545.6 716.6 776.8

200.0 305.2

595.7 706.8 734.4 749.6

0

138.9 253.0 416.1 442.6 477.6 602.9

200

248.0 318.6

400

340.8 284.4 471.5 502.0 662.3 629.8

600

500

Annual static leakage (mg)

T0 T4M T9M

1000

Water ingress (ppm)

600

400 300 200

100 0

COCe & PE

EPDM & PA

Sealing gasket / gathering ring combination

Nitrile & PA

COCe & PA

COCe & PE

Sealing gasket / gathering ring combination

Figure 1 (left) - Moisture level results of various configurations of MDIs (n=5) with different sealing gaskets containing HFA 134a without ethanol tested using Karl Fisher. Error bars represent +/- one standard deviation. Figure 2 (right) - Leak rate testing of various configurations of MDIs containing HFA with different sealing gaskets. Error bars represent one standard deviation.

238


Drug Delivery to the Lungs 27, 2016- Segolene Sarrailh, et al.

The results of COCe compatibility with suspension and solution type formulations are detailed in Figure 3 and 4 respectively. Compatibility of COCe with MDI mineral acid solution type formulation (n=5)

Compatibility of COCe with suspension type MDI formulation (n=5) 8000

can

gasket & ring

120

rest of valve

7000

100

T0

T1M @40°C/75%RH

T3M @40°C/75%RH

T6M @40°C/75%RH

80

98

1000

59

61

83

40

62

2000

60

55

3000

85

4000

98

5000

% recovery FF

Salbutamol sulphate adhesion (µg)

6000

20

0

T0 (+3j)

T1M @ 40/75

T3M @ 40/75

T6M @ 40/75

T0 (+3j)

T1M @ 40/75

COCe & PA

T3M @ 40/75

T6M @ 40/75

EPDM & PA

0

POM/PBT - EPDM & PE

Gasket and gathering ring type

POM/PBT - COCE & PE Valve, sealing gasket & gathering ring

Figure 3 (left) – Adhesion of salbutamol sulphate to various MDI components following storage for up to 6 months at 40°C/75%RH. Figure 4 (right) – Recovery of Formoterol Fumarate following storage for up to 6 months at 40°C/75%RH.

The results of extractables testing for the various materials assessed are given in Figure 5. Calculations are based on elastomers only (internal and neck gaskets), considering them fitted into a final product (DF30 valve).

Extractables, mg/valve

0,5

DF30 valve & EPDM

DF30 valve & COCe

0,4 0,3 0,2

0,1 0,0 Semi-volatiles

Fatty acids

Antioxidants

Alkanes

Total

Figure 5 - Extractable levels per MDI valve containing different gasket sealing materials.

239


The continued development of a cyclic olefin copolymer elastomer sealing system for metered dose inhalers

Discussion The results of the moisture testing, see Figure 1, reveal that COCe overall has good moisture protection for the contents of the MDI during storage for up to 12 months under accelerated conditions. When coupled with a polyamide (PA) ring (which is a material well known for its capacity to absorb water), it shows the best results, while when coupled with a polyethylene (PE) ring, results are equivalent to those obtained for nitrile gasket and PA ring. COCe is [2] known to have a low water vapour transmission rate and for aerosol products that may be sensitive to moisture content, COCe could be an interesting choice for the sealing material of the MDI system. The results for leak rates, see Figure 2, show that the COCe material is superior to both EPDM and Nitrile when stored at accelerated stability test conditions. Results of drug adhesion, see Figure 3, show that there is no more drug deposition on a valve fitted with COCe neck gasket, compared to EPDM neck gasket. When evaluated with a model formulation containing mineral acid, the MDI valves fitted with COCe material show equivalent chemical compatibility to valves fitted with EPDM sealing gaskets, see Figure 4. Extractables levels measured from MDI valves containing COCe, see Figure 5, were found to be significantly less compared to those found in valves with EPDM sealing materials. The sealing gaskets in an MDI account for the majority of the elastomeric content of the MDI, and if extractables can be minimized from this source, it will significantly reduce the overall extractables burden of the MDI container. The main extractables components from COCe were noted to be semi-volatiles, fatty acids, antioxidants and alkanes.

Conclusion This work provides further evidence for the potential value of COCe as an innovative sealing material, it can offer several advantages in terms of the overall performance of the MDI system, specifically with regard to leak rates, moisture resistance and formulation compatibility thus eventually to the overall performance of such MDI products incorporating this kind of valve to canister sealing technology. COCe leak rates and extractables levels were noted as measurably better in comparison to traditionally used materials such as Nitrile and EPDM.

References 1

Sarrailh et al, The development of a novel metered dose inhaler cyclic olefin copolymer elastomer sealing system, Drug Delivery to the Lungs 26, 2015, p366

2

Cyclic olefin copolymers, www.topas.com/

3

Williams G, Tcherevatchenkoff A, Moisture Transport into CFC-Free MDIs, Respiratory Drug Delivery VI. Volume 1, 1998: 471-474.

240


Drug Delivery to the Lungs 27, 2016 – Scott Courtney et al. Investigation Into The Effects Of Valve Pressure On Total Delivered Dose (TDD) From Vhcs 1 1 Scott Courtney , PhD; Benjamin Pratt , B.App.Sci.

1

Medical Developments International, Research & Development Dept., 4 Caribbean Drive, Scoresby, Melbourne, AUSTRALIA

Summary Total delivered dose and valve pressure differential tests have been conducted and compared for several valved holding chambers (VHCs). The results of total delivered dose were compared to the corresponding valve pressure results to identify whether or not there is a correlation between valve pressure and the total delivered dose (TDD). Both total delivered dose and valve pressure differential (VPD) measurements of several valve holding chambers have been conducted to test the hypothesis that the pressures generated by valves have an impact on the VHCs delivered dose. Results show that there is variance in the pressure differentials created by different valve designs, with the cross-valve producing the lowest pressure differentials on inhalation at tidal breathing simulations for both adult and paediatric rates (0.447 and 0.167 cm/H2O) followed by the disc valve (0.661 and 0.291 cm/H 2O) then the single slit valves (0.702 – 0.727 and 0.284 – 0.413 cm/H2O). VHCs exhibiting a lower tidal pressure differential deliver a higher dose (82.41 – 74.69 ug, adult; 71.75 – 69.60 ug, paediatric) when compared to VHCs with higher pressure generation (61.28 – 65.90 ug, adult; 60.10 – 66.45 ug, paediatric). In most cases the cross valve and disc valve generated less pressure than the 2 single slit valves; which correlates with the TDD results where the cross valve and disc valve gave a higher dose than the single slit valves in both adult and paediatric breathing rates. Introduction Valved holding chambers are designed to hold the active ingredients in the chamber, ready for inhalation. This mechanism ensures that it is less critical to time actuation with inhalation, making VHCs beneficial for the very young or [1] old, or an individual suffering an asthmatic episode . VHCs have the added advantage of allowing multiple inhalations without removing the device from the patient’s mouth. There are a wide variety of VHCs on the market today, many with different valve designs. There is little known about the importance of VHC valve design and the impacts of superior or inferior valve design on the TDD. It is hypothesised that increased valve resistance may prevent the valve from fully opening, which may impact how drug inside the chamber deposits onto the surface of the chamber, resulting in a lower delivered dose. This would be particularly relevant in paediatric cases, where the patient is less capable of producing enough force to open higher resistance valves. Other factors including high back pressure during exhalation though the VHC may discourage the patient from keeping the device in their mouth, which in turn may increase the delay time between successive breaths. In this study, we compare 4 available VHCs to identify whether or not the differing valve designs have an effect on the pressures generated by normal breathing through the device, as well as identifying any correlation between valve pressure differentials and VHC TDD. Experimental Methods TDD and VPD testing was conducted on four unique VHCs as outlined in Table 1. Table 1: List of VHC’s tested, VHC abbreviation, and sample distribution/testing matrix VHC

Abbrev.

Static Vol. (ml)

No. of samples

VPD testing (ALL)

TDD Adult

TDD Paed.

Anti-static Compact Space Chamber Plus

AS-CSCP

160

9

3

3

3

AeroChamber Plus Flow-Vu Anti-static VHC

ACFV+

149

9

3

3

3

OptiChamber Diamond Anti-static Chamber

OCD-AS

140

9

3

3

3

VORTEX Non Electrostatic Holding Chamber

VOR-NE

194

9

3

3

3

241


Drug Delivery to the Lungs 27, 2016 - Investigation Into The Effects Of Valve Pressure On Total Delivered Dose (TDD)

From Vhcs

Salbutamol sulfate was primed according to manufacturer’s recommendations before use in testing. All VHCs were used either straight from the original packaging without priming (AS-CSCP, ACFV+ & OCD-AS) or washed before use (VOR-NE) as per manufacturer’s instructions. VPD testing was conducted by simulating a consistent tidal breathing rate which was maintained using a waveform generator. As outlined in the Canadian standards for testing the performance of VHCs, measurements of performance made under static flow conditions may not fully describe device behaviour [2] under the continuous varying flow and pressure changes that are associated with the respiratory cycle . The waveform generator was calibrated to take 14 x 500mL breaths per minute when testing adult lung volumes and 18 x 200mL breaths per minute when testing paediatric lung volumes. The simulation continued for 5 minutes per VHC. The pressure differential was captured and tabulated, then the average of the triplicate VHCs was graphed. The same 4 brands of VHC were then tested for TDD at both paediatric and adult tidal breathing rates using salbutamol sulfate. TDD tests were run in triplicate using 1 pMDI which was actuated 3 times into each VHC. Each VHC was connected directly to the Dosage Unit Sampling Apparatus (DUSA) which in turn was connected to the waveform generator. Once the waveform generator was set to the appropriate parameter (adult or paediatric), the pMDI was shaken well and prepared for actuation by inserting the pMDI into the device to be tested. The actuation was timed so that the medication was injected at the beginning of inhalation (to mimic 0s delay testing). Once the first actuation had taken place, 2 full breaths were allowed to pass before repeating for a second actuation at the beginning of the third [3] inhalation . This process was completed a total of 3 times for each VHC at each breathing rate. The VHCs and DUSA filters were washed/extracted using 50:50 MeOH:H2O and analysed by HPLC. Results Experimental TDD results, displayed in units of µg of salbutamol as sulfate, for each VHC tested at both adult and paediatric breathing rates are summarised in Table 2 below. Graphs 1 – 8 display the difference in average pressures produced by the VHCs when monitored over 5 minutes at both tidal breathing rates and static flow rates. Table 2: Experimental results for Total Delivered Dose VHC

Total Delivered Dose Adult breathing rate (µg)

Total Delivered Dose Paediatric breathing rate (µg)

AS-CSCP

82.41

69.60

ACFV+

74.69

71.75

OCD-AS

65.90

60.10

VOR-NE

61.28

66.45

Table 3 summarises the data represented in graphs 1 to 8. The initial 4 columns show the overall averages of inhalation and exhalation pressure over the entire 5 minute periods of tidal breathing simulation. The final 2 columns display the average pressures produced after 5 minutes of constant static flowrates of 30 and 15L/min. When testing using adult flow rates, the VHCs returned TDD results in the following order from largest to smallest AS-CSCP > ACFV+ > OCD-AS > VOR-NE. The correlating average tidal inhalation pressure from lowest to highest was as follows AS-CSCP > ACFV+ > OCD-AS > VOR-NE. The order of performance when testing the pressures produced at a static 30L/min flow rate was ACFV+ > AS-CSCP > VOR-NE > OCD-AS. When testing using paediatric flow rates, the VHCs returned TDD results in the following order from largest to smallest ACFV+ > AS-CSCP > VOR-NE > OCD-AS. The correlating average tidal inhalation pressure from lowest to highest was as follows AS-CSCP > VOR-NE > ACFV+ > OCD-AS. The order of performance when testing the pressures produced at a static 15L/min flow rate was AS-CSCP > ACFV+ > VOR-NE > OCD-AS.

242


Drug Delivery to the Lungs 27, 2016 – Scott Courtney et al.

0 -0.5 -1

V O R - N E 3 0 L / m i n V a l ve P r e s s u r e a t T i d a l a n d St a t i c f l o w r a t e s

7

0.5

PRESSURE (CM H2O) ←INHALATION : EXHALATION→

PRESSURE (CM H2O) ←INHALATION : EXHALATION→

A S - C SC P 3 0 L / m in V a lve P r e s s u r e a t T id a l a n d St a t ic f lo w r a t e s

Zero Pressure Line Linear (Static Flow Pressure) 10 per. Mov. Avg. (Tidal Flow Pressure)

5 3 1 -1 Zero Pressure Line Linear (Static Flow Pressure) 10 per. Mov. Avg. (Tidal Flow Pressure)

-3

Graph 1: AS-CSCP adult static vs tidal flow rates

Graph 2: VOR-NE adult static vs tidal flow rates O C D - A S 3 0 L / m i n V a l ve P r e s s u r e a t T i d a l a n d St a t i c f l o w r a t e s

AC FV + 3 0 L / m in V a lve P r e s s u r e a t T id a l a n d St a t ic f lo w r a t e s PRESSURE (CM H2O) ←INHALATION : EXHALATION→

PRESSURE (CM H2O) ←INHALATION : EXHALATION→

4 3 2 1 0 -1

Zero Pressure Line Linear (Static Flow Pressure) 10 per. Mov. Avg. (Tidal Flow Pressure)

-2

Graph 3: ACFV+ adult static vs tidal flow rates

-2 Zero Pressure Line Linear (Static Flow Pressure) 10 per. Mov. Avg. (Tidal Flow Pressure)

V O R - N E 1 5 L / m i n V a l ve P r e s s u r e s a t T i d a l a n d St a t i c f l o w r a t e s PRESSURE (CM H2O) ←INHALATION : EXHALATION→

PRESSURE (CM H2O) ←INHALATION : EXHALATION→

-1

2

0 -0.5 Zero Pressure Line Linear (Static Flow Pressure) 10 per. Mov. Avg. (Tidal Flow Pressure)

Graph 5: AS-CSCP paediatric static vs tidal flow rates.

1.5 0.5 -0.5

-1 zero pressure line Linear (Static Flow Pressure) 2 per. Mov. Avg. (Tidal Flow Pressure)

O C D - A S 1 5 L / m i n V a l ve P r e s s u r e a t T i d a l a n d St a t i c f l o w r a t e s 0.5

Zero Pressure Line Linear (Static Flow Pressure ) 10 per. Mov. Avg. (Tidal Flow Pressure)

Graph 7: ACFV+ paediatric static vs tidal flow rates.

0

Graph 6: VOR-NE paediatric static vs tidal flow rates.

PRESSURE (CM H2O) ←INHALATION : EXHALATION→

2.5

1

-2

AC FV + 1 5 L / m in V a lve P r e s s u r e a t T id a l a n d St a t ic f lo w r a t e s PRESSURE (CM H2O) ←INHALATION : EXHALATION→

0

Graph 4: OCD-AS adult static vs tidal flow rates.

0.5

-1.5

1

-3

A S - C SC P 1 5 L / m in V a lve P r e s s u r e a t T id a l a n d St a t ic f lo w r a t e s

-1

2

0 -0.5 -1 -1.5 -2

Zero Pressure Line Linear (Static Flow Pressure) 10 per. Mov. Avg. (Tidal Flow Pressure)

Graph 8: OCD-AS paediatric static vs tidal flow rates.

243


Drug Delivery to the Lungs 27, 2016 - Investigation Into The Effects Of Valve Pressure On Total Delivered Dose (TDD)

From Vhcs

Table 3 : Average inhalation data for adults and paediatric at both static and tidal breathing Adult TIDAL breathing pressure differential

Paed. TIDAL breathing pressure differential

(cm/H2O)

(cm/H2O)

VHC

Adult STATIC pressure differential (cm/H2O)

Paed. STATIC pressure differential (cm/H2O)

Inhalation

Exhalation

Inhalation

Exhalation

Inhalation Only

Inhalation Only

AS-CSCP

-0.447

0.359

-0.167

0.134

-0.682

-0.706

ACFV+

-0.661

2.568

-0.291

0.785

-0.574

-0.794

OCD-AS

-0.702

1.033

-0.413

0.093

-2.390

-1.459

VOR-NE

-0.727

1.497

-0.284

0.380

-1.386

-1.112

Correlation between valve pressure and total delivered dose can be seen in the graphs below.

Graph 9. Average adult Inhalation Pressure vs TDD

Graph 10: Average Paed. Inhalation Pressure vs TDD

Discussion At adult flow rates the AS-CSCP displays the lowest valve pressure on inhalation. This is also true for paediatric tidal breathing rates and paediatric static rates. The order of the TDD from highest to lowest for adult tidal flow rates was AS-CSCP > ACFV+ > OCD-AS > VOR-NE and ACFV+ > AS-CSCP > VOR-NE > OCD-AS for paediatric flow rates. The AS-CSCP delivers the highest TDD compared with the other 3 devices at adult flow rates while producing the lowest valve pressure. The results also show that even though the AS-CSCP delivers slightly less than the ACFV+ at paediatric flow rate (2.15%), these 2 devices provide the lowest pressure differentials in 3 of the 4 flow rate categories, particularly in the static flow rate categories where the AS-CSCP and ACFV+ produce approximately half the pressures on both inhalation and exhalation. The lower inhalation pressure correlates with the higher TDD with the AS-CSCP and ACFV+ comparing favourably with the OCD-AS and VOR-NE when tested using either adult or paediatric tidal breathing waveforms. Conclusion Total delivered dose and valve pressure differential tests have been conducted and compared for several valved holding chambers under simulated tidal breathing conditions. The results confirm a correlation between lower valve pressure generation and improved delivered dose. Valves which produce lower pressure differentials tend to return the better total delivered dose results. This improved TDD is more pronounced at adult breathing rates. A future, larger study will use a post-actuation delay to alter drug residence time and aim to better determine the factor(s) that cause valve resistance to influence drug dosage. References 1. Jolyon P Mitchella , Mark W Nagelb, Valved holding chambers (VHCs) for use with pressurised metered-dose inhalers (pMDIs): a review of causes of inconsistent medication delivery, Primary Care Respiratory Journal (2007); 16(4). 2. Spacer and holding chambers for use with metered-dose inhalers, CSA Standards, Update No.2, CAN/CSA-Z264.1-02, January 2008. 3. André Schultz, Timothy J. Le Souëf, André Venter, Guicheng Zhang, Sunalene G. Devadason, Peter N. Le Souëf, Aerosol Inhalation From Spacers and Valved Holding Chambers Requires Few Tidal Breaths for Children. Pediatrics, December 2010, VOLUME 126 / ISSUE 6.

244


Drug Delivery to the Lungs 27, 2016 – Mårten Svensson et al. Comparison Between Pharmacopeial Testing and Testing Based on Mixing Inlet Lung Simulator (MILS) Methodology with Application to Nebuliser Testing: Part 1 – Pilot Study to Select Most Feasible MILS Method 1

1

2*

Mårten Svensson , Elna Berg , Jolyon Mitchell & Dennis Sandell

3

1

2

Emmace Consulting AB, Lund, SE-22381, Sweden Jolyon Mitchell Inhaler Consulting Services Inc., London, N6H 2R1, Canada (* corresponding author) 3 S5 Consulting, Blentarp, SE-27562, Sweden

Summary The compendial approach to determine fine particle dose (FPD) for nebulisers during simulated tidal breathing is to determine delivered dose (DD) for tidal breathing, while fine particle fraction (FPF) is established separately using a cooled Next Generation Impactor (NGI) operated at 15 L/min. FPD is then calculated as the product of DD and FPF. An advance in mixing inlet lung simulation (MILS) methodology has made it possible for both measurements to be combined in a single configuration. A comparison of the MILS procedure against the ® combination of pharmacopeial methods using the eFlow Rapid vibrating membrane nebuliser as device representative of this class of orally inhaled products (OIPs) is reported as Part 2 of the overall investigation. However, a preliminary study was needed before this comparison could take place, since the operating flow rate for the NGI when used with the mixing inlet has to be greater than the peak inspiratory flow rate (PIFR) of 24 L/min of the selected adult breathing pattern. The primary purpose of this pilot study was to establish and minimize the influence of a NGI flow rate change from 15 to 30 L/min. At the same time the investigation was extended to examine possible influences of the relative humidity of the pressurized air supply to the mixing inlet as well as whether or not the impactor needed to be cooled. It was found that the best performing configuration is to operate the NGI cooled (5ºC) with humidified air (90-92%RH) supplied to the mixing inlet. Introduction Currently, nebulising systems when operated under simulated breathing conditions are evaluated in the laboratory by two separate procedures in order to determine the FPD, regarded as a critical quality attribute for performance [1] of OIPs . DD is determined by connecting the nebuliser mouthpiece via a filter to a breathing simulator, usually set to deliver an age-appropriate standardized tidal breathing pattern; in the pharmacopeial method this [2,3] . FPF is established by the NGI in a parallel series of measurements; the corresponds to adult use recommended flow rate in the compendia is 15 L/min and the NGI should be cooled before use. FPD is [4] subsequently calculated as the product of DD and FPF. A recent advance in MILS methodology has made it possible for cascade impaction based assessments to be made at the required constant flow rate whilst operating [5] an OIP using a continuously varying flow profile associated with breath simulation , or any inhalation profile. A ® comparison of the MILS procedure against the combination of pharmacopeial methods using the eFlow Rapid vibrating membrane nebuliser as device representative of this class of OIPs is reported as Part 2 of the overall investigation. However, a preliminary study was needed before this comparison could take place, since the operating flow rate for the NGI has to be greater than the PIFR of 24 L/min for the selected breathing pattern. The primary purpose of this pilot study was to establish and minimize the influence of a flow rate change from 15 to 30 L/min. At the same time the investigation was extended to examine possible influences of the relative humidity of the air supply to the mixing inlet as well as whether or not the impactor needed to be cooled. The main study comparing the combination of compendial procedures with the optimised MILS-based methodology is described in Part 2. Materials and Methods ®

Three nebulizer units (eFlow Rapid, PARI GmbH, Starnberg, Germany) were evaluated with a single power unit ® and evaluated salbutamol sulphate nebuliser solution (Ventolin 5 mg/2.5 mL (GSK plc, Middlesex, UK) as test product. This pilot study compared the pharmacopeial method A (Table 1) for preparations for nebulisation, based on an NGI (MSP Corp., St. Paul, MN, USA), equipped with a USP/Ph.Eur. induction port and cooled to 5°C whilst operated at a flow rate of 15 L/min using ambient air with 4 different alternative approaches (test methods B to E in Table 1). Figure 1 illustrates the five test configurations investigated. Note that a “dummy” exhalation filter was added to all test configurations to mimic the set-up in Part 2. The aim was to determine which alternative best matched the data derived using the pharmacopeial approach (A). In these alternate procedures, the nebuliser-ontest was connected to both pressurized air supply via a Nephele mixing inlet (RDD OnLine, Richmond, VA, USA). The NGI was operated at 30 L/min, comprising the sum of the 15 L/min flow rate withdrawn from the nebulizer and a further 15 L/min supplied as supplementary flow from the pressurized air supply to the Mixing Inlet (Table 1). The impactor was either at room temperature (21-24C) or cooled in a refrigerator (5°C) for at least 90 minutes before testing when required. Following sampling, the mass of salbutamol sulphate deposited in the throat, on stages 1-7 and on the micro-orifice collector (MOC) was determined in each instance by a validated HPLC procedure.

245


Drug Delivery to the Lungs 27, 2016 - Comparison Between Pharmacopeial Testing and Testing Based on Mixing Inlet Lung Simulator (MILS) Methodology with Application to Nebuliser Testing: Part 1 – Pilot Study to Select Most Feasible MILS Method Table 1 Pharmacopeial Test Method (A) and Alternative Procedures (B-E) Investigated in the Pilot Study Method A B C D E

NGI Method A

NGI Temperature Cooled to 5C RT (21-24C) RT(21-24C) Cooled to 5C Cooled to 5C

NGI Flow rate (L/min) 15 30 30 30 30

Exhalation filter

eFlow

Air Ambient (25-55% RH) Dry (2-5% RH) Humid (90-92% RH) Dry (2-5% RH) Humid (90-92% RH)

NGI Methods B, C, D and E

Exhalation filter

USP inlet

eFlow

USP inlet

15 L/min

Ambient air: Temp = 21-24C RH = 25-55%

Cooled NGI 15 L/min Vacuum source

15 L/min Ambient air: Temp = 21-24C RH = 25-55%

Mixing inlet Vacuum source

Dry or humid air 15 L/min

NGI (RT or cooled) 30 L/min

Figure 1 Configurations for the Compendial Procedure (A) and for the Methods Using the NGI with Mixing Inlet (B-E) The measurements were performed by the same operator during a 4 day period to the design shown in Table 2. Table 2 Design for Pilot Study Day

Tests and Order*

1

A1, D1, E1, B1, C1

2

D2, E2, A2, C2, B2

3

E3, A3, D3, B3, C3

4

A1, A2, A3, E1, E2, E3

(*) The notation “Xn” denotes nebuliser unit number n tested with method letter code X; X = A-E, n = 1-3

In this arrangement, nebuliser unit number was confounded with day during days 1-3, as this part of the study was used to compare methods B-E to method A. On the other hand, the 4th day was used to assess possible differences between nebuliser units to ensure there were no drifts in performance with any device during the course of the study that might have prejudiced the selection of the optimum mixing inlet-based procedure. The data obtained with methods B-E collected during days 1-3 were used to assess the presence of significant effects by the two factors studied (NGI operating temperature and the relative humidity of the air supply to the mixing inlet). The degree of fit between each of methods B to E and the “reference” method A was investigated as follows: 1.

The match in the following 4 non-overlapping parameters was assessed: G1 = ISM - FPD<8µm G2 = FPD<8µm -FPD<5µm G3 = FPD<5µm -FPD<2µm G4 = FPD<2µm, where FPD<Xµm denotes the amount of drug contained in droplets with aerodynamic diameter < X m, and ISM is the impactor-sized mass (sum of depositions on stage 2 to MOC);

246


Drug Delivery to the Lungs 27, 2016 – Mårten Svensson et al. 2.

For each of methods B to E, each day and each parameter, the relation to method A was determined;

3.

The average relation was then calculated for each parameter and each of methods B-E;

4.

The relative standard deviation (RSD) associated with each parameter was calculated for method A to provide reference measures of variability;

5.

The fit between each method B to E and method A was determined, using the inverse RSD values as weightings (W G1, W G2, W G3 and W G4), scaled to sum in total to unity, by formula [1]: 4

fit ( X , A) WGi  X Gi / AGi  1 2

[1]

i 1

where XGi denotes the mean result for the parameter Gi when using method X; i = 1-4. Results Figure 2 shows the mass of salbutamol sulphate recovered stage-by-stage from the NGI by each method and for each impactor run in accordance with the design shown in Table 2. Within each panel results from method A-E are shown from left to right and identified by colour as indicated in the legend. Since the data using method A were obtained operating the NGI at a flow rate of 15 L/min, while the NGI was operated at 30 L/min for methods B-E, the stage cut-off sizes differ for the two flow rates. For this reasons methods cannot be directly compared using stage groupings, and instead the non-overlapping parameters based on aerodynamic size ranges defined above are assessed as primary end-points.

Figure 2: Individual Stage-by-Stage NGI Data Collected by Methods A-E in the Pilot Study The results for each method were quite reproducible, with the greatest range of data for stages where most of the drug containing droplets are deposited. The results clearly differed depending upon the mixing inlet-based method tested, see for example S5 where more salbutamol sulphate is found with method B. The mean results for the non-overlapping parameters G1-G4 are summarized for each method in Table 3. Table 3 Key Measures (Mass Salbutamol Sulphate, g) Determined from the Pilot Study Method

NGI DD

A B C D E

1945 1851 1861 1825 1833

Induction Port 50 45 42 51 46

G1 283 225 243 245 254

247

Size Range Parameter G2 G3 523 878 402 849 481 836 461 801 476 805

G4 93 211 139 137 125


Drug Delivery to the Lungs 27, 2016 - Comparison Between Pharmacopeial Testing and Testing Based on Mixing Inlet Lung Simulator (MILS) Methodology with Application to Nebuliser Testing: Part 1 – Pilot Study to Select Most Feasible MILS Method DD by the NGI for reference method A (1945 µg) was about 6% greater than this measure obtained using any of the MILS-based procedures, which taken together provided very similar outcomes for this metric in the range 1825 to 1861 µg. As a consequence, the results for G1-G3 were also higher (17, 15 and 7%, respectively) for method A. In contrast, the G4 deposition (finest sized droplets) was smaller for the reference method (93 µg) compared with any of the MILS-based procedures (125 to 211 µg). The mass of salbutamol sulphate recovered from the induction port was comparable with all methods (42 to 51 µg). Discussion: It is clear that none of the methods B to E involving the MILS perfectly matched method A, and that the results for methods C-E were broadly comparable. The question therefore becomes: Which approach provided the closest agreement? Table 4 contains significance outcomes from an analysis of variance (ANOVA) for the MILS-based procedures, as part of the answer. Table 4 Main Effect ANOVA Results (p-values) From Pilot Study (Based on Data From Methods B-E) Factor Nebuliser Number Dry/Humidified Air NGI Cooled/Room Ambient

NGI DD 0.13 0.78 0.40

Throat 0.23 0.62 0.50

G1 0.06 0.24 0.19

G2 0.56 0.0331 0.17

G3 0.09 0.92 0.38

G4 0.41 0.0367 0.0292

This statistical analysis showed no effect by nebuliser unit for any of the 6 parameters studied. The use of dry air/NGI at room temperature (method B) resulted in a significantly greater G4 (FPD<2µm), indicative of enhanced droplet evaporation. However, the influences of either operating the NGI with humidified air (methods C vs. E) or cooling the NGI with either dry or humidified air (methods D vs. E) were marginal. The alternate method with best match to the reference method A was determined by calculation of the fit factor fit(X,A) as described above by equation [1], with the outcomes that are summarized in Table 5 Table 5 Fit Analysis Between Reference Method A and the MILS-Based Procedures B to E Method A vs. Method B Method C Method D Method E

Fit Factor 0.1568 0.0276 0.0267 0.0169

The smaller the fit factor, the better the agreement between methods, so on this basis, mixing inlet method E (cooled NGI using humid air) was found to be the best choice. This method was therefore selected for the main study described in Part 2. Conclusions The pilot study provided guidance as to the most appropriate procedure to be adopted for the main study comparing existing compendial methods for testing nebulisers with the new MILS-based approach. It was found that the best option of the four alternative procedures investigated was when the NGI is cooled before use and the mixing inlet air supply is humidified. This approach was therefore used in the main investigation (Part 2). References 1

Hess D: Nebulizers: Principles and performance, Respir Care 2000; 45(6): pp609-622.

2

European Directorate for Quality in Medicines and Healthcare (EDQM). European pharmacopeia 8.0, monograph 2.9.44. Preparations for nebulisation, EDQM, Strasburg, France; 2014.

3

US Pharmacopeial Convention. United States Pharmacopeia 39/National Formulary 34, Chapter <1601> Products for nebulization, USP, Rockville, MD, USA; 2016.

4

Miller NC: Apparatus and process for aerosol size measurement at varying gas flow rates. US Patent 6,435,004-B1, 2002.

5

Olsson B, Berg E, Svensson M: Comparing aerosol size distribution that penetrates mouth–throat models under realistic inhalation conditions. In: R N Dalby, P R Byron, J Peart, J D Suman, P M Young, (eds): Respiratory drug delivery 2010. Davis Healthcare International Publishing LLC, River Grove, IL, USA; pp225–234, 2010.

248


Drug Delivery to the Lungs 27, 2016 - Mårten Svensson et al. Comparison Between Pharmacopeial Testing and Testing Based on Mixing Inlet Methodology with Application to Nebuliser Testing: Part 2 – Main Study 1

1

2*

Mårten Svensson , Elna Berg , Jolyon Mitchell & Dennis Sandell

3

1

2

Emmace Consulting AB, Lund, SE-22381, Sweden Jolyon Mitchell Inhaler Consulting Services Inc., London, N6H 2R1, Canada (* corresponding author) 3 S5 Consulting, Blentarp, SE-27562, Sweden

Summary The compendial approach to determine fine particle dose (FPD) of preparations for nebulisation when using a simulated patient breathing pattern, is to combine the results from two pharmacopeial methods. The delivered dose (DD) is collected on the inhalation filter during breathing simulation, while the fine particle fraction (FPF) is determined using the cooled Next Generation Impactor (NGI) operated at 15 L/min. FPD is then calculated as the product of DD and FPF. A recent advance in mixing inlet lung simulation (MILS) allows cascade impactor assessments to be made at constant flow rate whilst operating the nebuliser using a varying flow profile associated with breath simulation. The purpose of this study was to investigate the feasibility of the MILS ® approach for direct determination of FPD<5µm using a vibrating membrane nebulizer (eFlow Rapid) delivering aqueous salbutamol sulphate with an adult tidal-breathing pattern (inspiratory:expiratory ratio = 1:1, tidal volume = 500 mL, 15 bpm, peak inspiratory flow rate (PIFR) = 24 L/min) as the model system. FPD<5µm obtained with the MILS was 72% of that obtained using the compendial method, showing that the two procedures were not equivalent for this particular comparison, although MILS methodology may be more pertinent. Whether this conclusion holds for other nebulizer types, drug products or alternate breathing patterns remains to be investigated. Regardless of the outcome of those studies, the present finding raises an important question as to which of the two methods provides the FPD<5µm that is more representative of the lung dose received by a patient. Introduction The pharmacopeial method for the evaluation of preparations for nebulisation to establish the aerodynamic particle size distribution (APSD) in general and fine particle dose in particular is based on sampling the aerosol by [1,2] . When studying the nebuliser performance during breathing the cooled NGI at a constant flow rate of 15 L/min simulation, measurements of drug delivery rate and delivered dose are made separately with the nebuliser-on-test [1,2] . The FPD, with cut-off attached to a breathing simulator, mimicking the selected tidal breathing pattern typically chosen to be 5 m (FPD<5µm), is subsequently calculated as the product of DD and FPF<5µm. The need to undertake two separate tests to establish FPD <5µm, which is the critical quality attribute for product performance, is cumbersome and increases the potential for error. Moreover, one can question whether the value of FPF<5µm determined at a constant flow rate is fully representative of that obtained using breathing simulation when only the inhaled portion of the generated aerosol is entering the impactor. Furthermore, in the context of developing more clinically appropriate testing methods, it would be useful to be able to observe changes in FPD<5µm at the same time as nebuliser performance is being evaluated at different breathing patterns reflective of patient age [3] categories and obstructive lung disease state. A recent advance in MILS methodology has made it possible for cascade impactor assessments to be made at the required constant flow rate whilst operating the nebuliser using [4] a continuously varying flow profile associated with breath simulation . The purpose of the present bench study was to evaluate the potential of MILS methodology as a direct, more robust and potentially clinically more relevant alternative to the current approach combining two pharmacopeial methods for determination of APSD when the nebuliser is operated at a varying flow rate. As model system for the evaluation a vibrating membrane device was selected as a representative of the nebuliser class of orally inhaled products (OIPs). In this initial assessment the standard adult tidal breathing pattern was studied because the focus was on investigating if the new and more realistic MILS method provided equivalent outcomes to the combination of compendial test procedures. It was hypothesized that FPD<5µm obtained by the MILS-based procedure would be statistically equivalent with the corresponding measure derived from the separate determinations of DD and FPF<5µm. Materials and Methods ®

Three nebulizer units (eFlow Rapid, PARI GmbH, Starnberg, Germany) were tested with a single power unit and ® evaluated with salbutamol sulphate nebuliser solution (Ventolin 5 mg/2.5 mL (GSK plc, Middlesex, UK)). Figure 1 illustrates the three test configurations that were studied. For the reference compendial methods, DD was determined capturing the droplets emitted by the nebulizer-ontest when connected to a breathing simulator (model F-SIG 6300, FIA AB, Södra Sandby, Sweden) set to operate with the following adult tidal-breathing waveform: inspiratory:expiratory ratio = 1:1, tidal volume = 500 mL, 15 bpm, PIFR = 24 L/min) [BS-procedure]. A filter (Respirgard II, model 303EU, Vital Signs, US) captured during the inhalation phase the emitted droplets at the mouthpiece of the nebuliser. In a parallel study FPF<5µm was determined, sampling the emitted droplets into an NGI (MSP Corp., St. Paul, MN, USA), equipped with a USP/Ph.Eur. induction port and operated at a nominal flow rate of 15 L/min [NGI-procedure].

249


Drug Delivery to the Lungs 27, 2016 - Comparison Between Pharmacopeial Testing and Testing Based on Mixing Inlet Methodology with Application to Nebuliser Testing: Part 2 – Main Study The NGI and induction port were cooled in a refrigerator at 5°C for at least 90 minutes before testing. Following sampling, the mass of salbutamol sulphate deposited in the throat, on stages 1-7 and on the MOC was determined in each instance by a validated HPLC procedure.

Figure 1 Configurations for the Compendial Procedures (BS for DD and NGI Alone) and for the NGI with MILS Methodology FPD<5µm for the MILS procedure was determined directly from the NGI stage depositions measured with the nebuliser-on-test connected via the Nephele MILS mixing inlet (RDD OnLine, Richmond, VA, USA) to both the breathing simulator and to cascade impactor, as shown in Figure 1. Note that the flow rate through the NGI for the MILS procedure had to be increased to 30 L/min (the lowest apart from 15 L/min, for which archival calibration [5] data are available ), because it was necessary that the constant flow through the impactor be higher than the PIFR of the selected breathing pattern (24 L/min). The need to run NGIs at different flow rates for the compendial and alternate MILS procedures was a cause of concern. For this reason a pilot study (Part 1) was undertaken, investigating four different configurations, each operating the NGI at a flow rate of 30 L/min but either with the impactor cooled or at room ambient temperature, as well as examining the use of dry or humid supplementary air [6] feeding the mixing inlet . This investigation demonstrated that the difference between these alternate configurations and the compendial method was minimized when the impactor was cooled and the pressurized air to the mixing inlet was humidified. Table 1 shows the design and test sequence for the entire 4-day study, balanced in terms of test order within day using each of the three test procedures and undertaken by a single operator. The study uses a paired design so that the MILS/(BS+NGI) FPD<5µm ratio could be calculated for each combination of day and nebuliser unit; these ratios were used for calculation of the desired confidence interval (CI). Table 1 Parallel Study Design Allowing Determination of MILS/(BS+NGI) FPD<5µm Ratio for Each Day and Nebuliser Unit Test Order

Day No.

Nebulizer unit No.

BS

NGI

1

1

2

1

2

4

1

3

2 2 2

Test Order

MILS

Day No.

Nebulizer unit No.

BS

NGI

MILS

3

1

3

1

20

21

19

5

6

3

2

22

23

24

9

7

8

3

3

27

25

26

1

11

12

10

4

1

29

30

28

2

15

13

14

4

2

33

31

32

3

16

17

18

4

3

34

35

36

The protocol stated that equivalence between the investigated MILS method and the combined pharmacopeial BS+NGI method could be claimed if the obtained 90% confidence interval for the MILS/(BS+NGI) FPD<5µm ratio was completely within the acceptance interval 85% - 118%. The study was sized (n = 12 ratios) based on the pilot study so the obtained CI for the mean ratio would have a length of at most 15%.

250


Drug Delivery to the Lungs 27, 2016 - Mårten Svensson et al. Results The key measures obtained from the investigation are summarised in Table 2. Table 2 Performance Measures Obtained from the Vibrating Membrane Nebulisers Evaluated by Compendial (BS and NGI) Procedures and the Alternate MILS Method Parameter DD from BS FPF<5µm from NGI FPD<5µm from BS+NGI FPD<5µm from MILS

Procedure

Separate Compendial Combined Approach

N

Mean

SD

RSD

Range

12

1009 g

57 g

5.7 %

926 – 1127 g

12

47 %

3.0 %

6.3 %

43 – 52 %

12

475 g

50 g

10.6 %

424 – 572 g

12

340 g

29 g

8.5 %

297 – 392 g

The primary end-point for assessing the equivalence between the compendial BS+NGI methods and the MILS method was FPD<5µm. This metric obtained by MILS testing was expressed as a percentage of FPD<5µm calculated as DD x FPF<5µm from the data derived from the compendial BS and NGI methods, respectively. Individual values of this ratio (Figure 2) varied from 59 to 83%, with the overall average being 72%. The 90% CI for the MILS to BS+NGI FPD ratio was 68.9% to 75.2%. The length of the obtained CI was 6.3%, much shorter than the expected length of 15%; this outcome was a consequence of the SD of the ratio being much smaller than in the pilot study (6.6% vs. 15.5%).

Figure 2 Values of FPD<5µm by MILS Procedure Expressed as a Percentage of FPD<5µm by BS + NGI Compendial Procedures for Each Nebuliser Tested on Each Day of the Investigation

Discussion The outcome from the investigation disproved the hypothesis, since the MILS procedure resulted in a significantly smaller FPD<5µm, compared with the existing combination method. Importantly, there were no obvious trends in the data associated with a particular nebulizer, nor did the order of testing indicate any systematic drift in the ratio (Figure 2). The droplet size distributions from the NGI normalised to impactor-sized mass and expressed as cumulative distribution functions (CDFs) were largely equivalent (Figure 3) between the two configurations, indicating that evaporation/condensation-related changes in the MILS configuration compared with the compendial set-up had not taken place to any appreciable extent. However, the raw data for absolute mass salbutamol sulphate recovered (mean ± SD) revealed substantially less drug entered the NGI in the MILS configuration (728 ± 50 µg) than using the compendial method (1724 ± 84 µg). This outcome was expected, since the NGI-MILS only collected the ”inhaled” dose from the nebuliser, whereas all the emitted aerosol was captured by the NGI when operated in the compendial configuration.

251


Drug Delivery to the Lungs 27, 2016 - Comparison Between Pharmacopeial Testing and Testing Based on Mixing Inlet Methodology with Application to Nebuliser Testing: Part 2 – Main Study

Finally, 577 ± 58 µg salbutamol sulphate was recovered from the exhalation filter in the MILS set-up, greater than 449 ± 42 µg when the nebulisers were evaluated for DD. This small but significant difference is explicable because slightly more drug particles exited the measurement apparatus via the induction port during exhalation phase than was the case with the simpler set-up for determination of DD during breathing simulation.

Figure 3 Cumulative Size Distributions for Salbutamol Sulphate as Percentage of Impactor Sized Mass (ISM = Sum of Drug Depositions on S2 to the MOC) by Compendial (NGI) and MILS Configurations for the ® eFlow Rapid Vibrating Membrane Nebuliser Conclusions The MILS-based procedure evaluated, allowing direct measurement of FPD<5µm during breath simulation by inserting a mixing inlet between the nebulizer and cascade impactor, is not equivalent to the determination of this metric by the existing compendial procedures involving separate determinations of DD and FPF<5µm. The FPD<5µm determined with the MILS configuration was found to be 72% of that found using the current pharmacopeial approach, and the associated 90% CI for the FPD ratio was 69% - 75%. Whether similar deviations will be found for other nebuliser types, drug products, FPD size limits, or alternate breathing patterns remains to be investigated. Regardless of the outcome from those studies, the present finding raises the important question, which of the two approaches provides an FPD that is more representative of the lung dose received by a patient. In response, it is suggested that the MILS-based approach, despite its increased complexity, may provide a more accurate answer to this question than the current pharmacopeial methods. References 1

European Directorate for Quality in Medicines and Healthcare (EDQM). European pharmacopeia 8.0, monograph 2.9.44. Preparations for nebulisation, EDQM, Strasburg, France; 2014.

2

US Pharmacopeial Convention. United States Pharmacopeia 39/National Formulary 34, Chapter <1601> Products for nebulization, USP, Rockville, MD, USA; 2016.

3

Miller NC: Apparatus and process for aerosol size measurement at varying gas flow rates. US Patent 6,435,004-B1, 2002.

4

Olsson B, Berg E, Svensson M: Comparing aerosol size distribution that penetrates mouth–throat models under realistic inhalation conditions. In: R N Dalby, P R Byron, J Peart, J D Suman, P M Young, (eds): Respiratory drug delivery 2010. Davis Healthcare International Publishing LLC, River Grove, IL, USA; pp225–234, 2010.

5

Marple VA, Olson BA, Santhanakrishnan K, Mitchell JP, Hudson-Curtis BL: Next generation pharmaceutical impactor: a new impactor for pharmaceutical inhaler testing. Part III. Extension of archival calibration to 15 L/min. J Aerosol Med 2004; 17(4): pp335-343.

6

Svensson M, Berg E, Sandell D, Mitchell J: Comparison between pharmacopeial testing and testing based on Mixing Inlet Lung Simulator (MILS) methodology with application to nebuliser testing: Part 1 – Pilot study to select most feasible MILS method. Drug Delivery to the Lung 27, The Aerosol Society, Edinburgh, UK 2016.

252


Drug Delivery to the Lungs 27, 2016 – A.J. Lexmond et al.

Comparison of bronchoconstrictor responses to AMP and adenosine in sensitised guinea pigs ® using the PreciseInhale system for generation and administration of dry powder aerosols A.J. Lexmond, S. Keir, C.P. Page, B. Forbes King’s College London, Institute of Pharmaceutical Science, 150 Stamford Street, London, SE1 9NH, UK

Summary Inhaled adenosine induces bronchoconstriction and inflammation in asthma and can therefore be used for bronchial challenge testing. Bronchial challenge tests are sometimes performed with nebulised solutions of adenosine 5’-monophosphate (AMP). The nebulised AMP test has several disadvantages, including long administration times and a maximum concentration that does not induce sufficient bronchoconstriction in all patients. Using aerosolised dry powder adenosine instead of nebulised AMP may improve adenosine delivery to the lungs. However, it remains to be elucidated whether comparable effects are obtained when administering the challenge using a powder versus droplet presentation, and when using adenosine versus AMP. The aim of this study was to compare the bronchoconstrictor responses to the different delivery methods in an allergic guinea pig model. ®

The PreciseInhale system was used to generate dry powder aerosols of AMP and adenosine for pulmonary administration, with an MMAD of 1.81 μm and 3.21 μm, respectively. In a small proof-of-principle study, bronchonconstrictor responses to these dry powder aerosols in passively sensitised guinea pigs were compared to the responses to nebulised and intravenously administered AMP and adenosine. Airway obstruction in response to the challenges was measured as an increase in total lung resistance (RL) and decrease in dynamic lung compliance (Cdyn). All six administrations successfully induced bronchoconstriction in this animal model. The time to maximal response was comparable for both inhaled administration methods and the bronchoconstriction produced by adenosine and AMP was similar, but the magnitude of response varied: iv > dry powder > nebulisation. The next step will be to investigate dose-response to the dry powder aerosols.

Introduction Bronchial hyperresponsiveness (BHR) is present in nearly all patients with asthma and in the majority of patients [1] with chronic obstructive pulmonary disease (COPD). BHR refers to an abnormal increase in airflow limitation in response to airway exposure to nonspecific stimuli and can be measured by means of a bronchial challenge [2] test. The gold standard for quantifying BHR is methacholine, which acts directly on airway smooth muscle [3] cells. However, use of the indirectly acting stimulus adenosine may provide significant diagnostic benefits, since [4,5] BHR to adenosine is more relevant to the disease pathology airway inflammation than BHR to methacholine. On this basis it has been argued persuasively that adenosine bronchial challenge testing provides a more reliable non-invasive tool for monitoring disease activity and an improved method for assessing the response to anti[6] inflammatory treatments. Moreover, recent findings suggest that challenging with adenosine may improve [7] diagnostic discrimination between asthma and COPD. Until recently, the only way to deliver adenosine to the lungs was by means of nebulisation. To this end, solutions of adenosine 5’-monophosphate (AMP) rather than adenosine itself have been used because of AMP’s higher aqueous solubility. Following inhalation, AMP is rapidly hydrolysed to adenosine by the ubiquitous enzyme 5’nucleotidase. However, nebulisation of AMP has a number of drawbacks. First, there is a restrictive maximum [7,8] Moreover, the high AMP concentration based on solubility that does not result in BHR in all patients. concentrations required for the test have been shown to greatly affect nebuliser performance, which may have [9] implications for the test outcome. This lack of a suitable delivery method means that bronchial challenge testing with adenosine has yet to be exploited to its full potential. Delivery of adenosine as a dry powder aerosol is a promising strategy to overcome the issues identified [10,11] However, it remains to be elucidated whether comparable effects are obtained when administering above. dry adenosine versus dissolved AMP. A recent study showed that not only adenosine exerts pharmacological responses, but also AMP elicits effects directly (on dendritic cells) by binding to adenosine receptors without [12] These findings stress the importance of thoroughly characterising the biological being hydrolysed first. responses to the physically (powder versus droplet) and chemically (adenosine versus AMP) different aerosols.

To elucidate the biological responses to AMP and adenosine, an animal model is required that mimics the response to these compounds in man as closely as possible. Criteria that have been defined to characterise this

253


Drug Delivery to the Lungs 27, 2016 - Comparison of bronchoconstrictor responses to AMP and adenosine in sensitised guinea pigs using the PreciseInhale® system for generation and administration of dry powder aerosols bronchoconstrictor response include upregulation by the disease process, mast cell involvement, and [13] tachyphylaxis. Work at King’s has previously demonstrated that guinea pigs that are sensitised with ovalbumin [14] then subjected to relatively mild allergen display such effects mediated by the A1 receptor, which is up[15] regulated in biopsies from asthmatic patients. This model is an excellent tool to study the biological responses to AMP and adenosine in vivo. A relatively new technique to administer dry powder aerosols to animals is the PreciseInhale system (Inhalation Sciences, Sweden). This system generates an aerosol cloud in a very efficient way, approaching the primary particle size of the powder. Administered dose is accurately controlled per animal, resulting in lower variability between animals. Comparing guinea pig airway responses to the aerosols generated with this system to nebulised aerosols allows the effect of adenosine versus AMP and the effect of powder versus solution administration to be investigated. The aim of this study was to prepare dry powder formulations of AMP and adenosine, then compare the airway obstruction induced by dry powder AMP and adenosine to nebulised formulations in an allergic guinea pig model.

Methods AMP and adenosine (Sigma-Aldrich, UK) were micronised using an Alpine AS 50 jet mill (Hosakawa, Germany) by applying a nozzle pressure of 3 bar and a milling pressure of 1 bar. The primary particle size distributions (PSDs) of the micronised compounds were measured with a HELOS BR laser diffraction apparatus (Sympatec, Germany) using a 100 mm (R3) lens and the FREE calculation mode based on the Fraunhofer theory. Powder samples were loaded into the PreciseInhale dosing chamber in accurately weighed doses of approx. 1 mg (adenosine) or 0.7 mg (AMP). The particle size distribution of the aerosol was determined by cascade impaction analysis using a Marple cascade impactor by gravimetric measurement. The fractions available for bronchial, alveolar and total lung deposition were calculated using Multiple-Path Particle Dosimetry Model software (ARA, US) and a breathing simulation model for the guinea pig (tidal volume 1.7 mL and 60 breaths/min). [16]

Male Dunkin-Hartley guinea pigs were passively sensitised as described before and subsequently randomised into six treatment groups: iv AMP 1 mg/kg, iv adenosine 0.77 mg/kg, nebulised AMP 10 mg/mL (10 s), nebulised adenosine 7.7 mg/mL (10 s), dry powder AMP 280 µg and dry powder adenosine 215 µg (three to four animals per group). Animals were anaesthetised by isoflurane inhalation and iv propofol to allow for spontaneous breathing and subsequently intratracheally cannulated. Dry powder AMP and adenosine were administered with the PreciseInhale system, whereas for nebulisation a DeVilbiss nebuliser was used. Airway obstruction was measured as an increase in total lung resistance (R L) and decrease in dynamic lung compliance (Cdyn) using a pneumotachograph. Results are expressed as percentage change from baseline.

Results The powders obtained after micronisation had an X50 (X10, X90) of 1.40 µm (0.65 µm, 3.38 µm) for AMP and of 1.42 µm (0.66 µm, 2.93 µm) for adenosine respectively. Adenosine was micronised twice using the same settings to obtain a powder with a PSD comparable to AMP. Dispersion with the PreciseInhale system had a yield of 27% for AMP and 10% for adenosine. Figure 1 shows the aerodynamic PSDs of the two compounds. Although AMP and adenosine had very similar primary PSD, the PSD in the aerosol clouds from the PreciseInhale system were different (Figure 1). AMP exhibited an MMAD of 1.81 µm (GSD 2.18) while for adenosine an MMAD of 3.21 µm (GSD 1.76) was found. This can be explained by the difference in yield. Adenosine had a much lower yield and based on the difference in PSD it can be assumed that it was especially the finest fraction (< 1 µm) that did not come out of the aerosol chamber. Since these very fine particles are also likely to be exhaled, the total lung deposition fraction calculated to be higher for adenosine (48% of the delivered dose) than for AMP (31%), with quite comparable ratios between bronchial and alveolar deposition (70:30 for adenosine and 63:37 for AMP). The system corrects for these deposition fractions allowing precise control over the dose that is administered to individual animals.

254


Drug Delivery to the Lungs 27, 2016 – A.J. Lexmond et al.

Figure 1 - Aerodynamic particle size distribution of the delivered dose of AMP (left) and adenosine (right) generated with the PreciseInhale system obtained with cascade impaction analysis. The dotted line indicates the MMAD. n = 3, mean and SD are shown.

The bronchoconstrictor responses after the different administrations of adenosine and AMP are shown in Table 1. All administrations induced bronchoconstriction, although three individual animals were unresponsive (most likely due to failure in sensitisation). Two of the unresponsive animals received nebulised adenosine and one received adenosine iv. Overall, quite large variations were found in the amount of bronchoconstriction induced (RL), especially in the iv groups. The rank order in the magnitude of response was iv > dry powder > nebulised solution for both compounds. No difference was observed between the responses to adenosine and AMP. The time to maximal response was around 9 min for all inhaled administrations, whereas it took only 3 min after iv administration of either compound.

Table 1 -

Bronchoconstrictor responses in sensitised guinea pigs following exposure to adenosine or AMP administered as dry powder aerosol, nebulised aerosol, or intravenously.

dry powder

Adenosine nebulisation

iv

n=4

n=4

n=4

RL maximal increase (%)

153 ± 77

61 ± 31*

Cdyn maximal decrease (%)

49 ± 7 8.5 ± 1.2

Time to maximal response (min)

dry powder

AMP nebulisation

iv

n=4

n=3

n=3

203 ± 156**

83 ± 27

33 ± 5

171 ± 78

13 ± 12*

61 ± 5**

35 ± 10

30 ± 16

52 ± 5

9.2 ± 0.8*

2.9 ± 0.9**

9.8 ± 2.4

9.0 ± 1.9

3.6 ± 2.0

RL: total lung resistance; Cdyn: dynamic lung compliance. Mean ± SEM shown. * n = 2: two animals excluded that did not respond, ** n = 3: one animal excluded that did not respond

Discussion In this study a new delivery system was used to administer aerosolised compounds to animals. The PreciseInhale system was shown to effectively disperse micronised formulations of AMP and adenosine. The system’s advanced feedback system, based on measurement of light scattering at the outlet of the aerosol chamber, enabled precise control of the dose that is administered to the individual animals, thereby reducing the variation of delivered dose per animal. The difference in time to maximal response between the inhaled administrations and iv administration can be explained by the time it takes for the compounds to distribute in the lung following deposition at the mucosal surface. The differences in response size between the three administration methods are likely due to dosimetry, where the largest dose was delivered by iv, followed by dry powder and the dose delivered by nebulisation was the lowest. The dry powder dose was based on the estimated deposition following from administration by nebulisation, but these values were not corrected for particle size or any exhaled fraction. Dose-response studies for all three methods would be necessary to determine how the delivered doses relate to each other.

255


Drug Delivery to the Lungs 27, 2016 - Comparison of bronchoconstrictor responses to AMP and adenosine in sensitised guinea pigs using the PreciseInhale® system for generation and administration of dry powder aerosols

Conclusions Dry powder aerosols of AMP and adenosine suitable for inhalation were obtained using the PreciseInhale system. Although these aerosols differed slightly in their aerodynamic particle size distribution, their deposition patterns are expected to be similar. The system controls the dose that is inhaled by an individual animal enabling equal lung doses for each individual animal to be obtained by changing the amount that is aerosolised. Both dry powder and nebulised aerosols of adenosine and AMP induced bronchoconstriction in sensitised guinea pigs. The time to maximal response was comparable for both inhaled administration methods. Dose-response studies are required to investigate further the characteristics of the bronchoconstriction produced by dry powder aerosols.

References 1

Postma D, Kerstjens H: Characteristics of airway hyperresponsiveness in asthma and chronic obstructive pulmonary disease, Am J Respir Crit Care Med 1998; 158: ppS187–S192.

2

Sterk PJ, Fabbri LM, Quanjer PH, Cockcroft DW, O’Byrne PM, Anderson SD, Juniper EF, Malo J-L: Airway Responsiveness: Standardized challenge testing with pharmacological, physical and sensitizing stimuli in adults, Eur Respir J 1993; 6(Suppl. 16): pp53–83.

3

Cockcroft DW: Provocation Methods: Direct Challenges, Clin Rev Allergy Immunol 2003; 24: pp19–26.

4

van den Berge M, Meijer RJ, Kerstjens HAM, de Reus DM, Koeter GH, Kauffman HF, Postma DS: PC(20) adenosine 5’monophosphate is more closely associated with airway inflammation in asthma than PC(20) methacholine, Am J Respir Crit Care Med 2002; 165: pp1546–1550.

5

de Meer G, Heederik D, Postma DS: Bronchial responsiveness to adenosine 5-monophosphate (AMP) and methacholine differ in their relationship with airway allergy and baseline FEV1, Am J Respir Crit Care Med 2002; 165: pp327–331.

6

Spicuzza L, Di Maria G, Polosa R: Adenosine in the airways: implications and applications, Eur J Pharmacol 2006; 533: pp77–88.

7

Spicuzza L, Scuderi V, Morjaria JB, Prosperini G, Arcidiacono G, Caruso M, Folisi C, Di Maria GU: Airway responsiveness to adenosine after a single dose of fluticasone propionate discriminates asthma from COPD, Pulm Pharmacol Ther 2014; 27: pp70–75.

8

Cohen J, Postma DS, Douma WR, Vonk JM, de Boer AH, ten Hacken NHT: Particle size matters: diagnostics and treatment of small airways involvement in asthma, Eur Respir J 2011; 37: pp532–540.

9

Lexmond AJ, Hagedoorn P, Frijlink HW, de Boer AH: Challenging the two-minute tidal breathing challenge test, J Aerosol Med Pulm Drug Deliv 2013; 26: pp380–386.

10

Lexmond AJ, Hagedoorn P, van der Wiel E, ten Hacken NHT, Frijlink HW, de Boer AH: Adenosine dry powder inhalation for bronchial challenge testing, part 1: Inhaler and formulation development and in vitro performance testing, Eur J Pharm Biopharm 2014; 86: pp105–114.

11

Lexmond AJ, van der Wiel E, Hagedoorn P, Bult W, Frijlink HW, ten Hacken NHT, de Boer AH: Adenosine dry powder inhalation for bronchial challenge testing, part 2: Proof of concept in asthmatic subjects, Eur J Pharm Biopharm 2014; 88: pp148–152.

12

Panther E, Dürk T, Ferrari D, Di Virgilio F, Grimm M, Sorichter S, Cicko S, Herouy Y, Norgauer J, Idzko M, Müller T: AMP affects intracellular Ca2+ signaling, migration, cytokine secretion and T cell priming capacity of dendritic cells, PLoS ONE 2012; 7: e37560.

13

Fozard JR, Hannon JP: Species differences in adenosine receptor-mediated bronchoconstrictor responses, Clin Exp Allergy 2000; 30: pp1213–1220.

14

Reynolds S, Docherty R, Robbins J, Spina D, Page CP: Adenosine induces a cholinergic tracheal reflex contraction in guinea pigs in vivo via an adenosine A1 receptor-dependent mechanism. J Appl Physiol 2008; 105: pp187–196.

15

Brown RA, Clarke GW, Ledbetter CL, Hurle MJ, Denyer JC, Simcock DE, Coote JE, Savage TJ, Murdoch RD, Page CP, Spina D, O’Connor BJ: Elevated expression of adenosine A1 receptor in bronchial biopsy specimens from asthmatic subjects, Eur Respir J 2008; 31: pp311–319.

16

Keir S, Boswell-Smith V, Spina D, Page C: Mechanism of adenosine-induced airways obstruction in allergic guinea pigs, Br J Pharmacol 2006; 147: pp720–728.

256


Drug Delivery to the Lungs 27, 2016 - Vincent Levet et el Local and systemic pharmacokinetic evaluation of immediate-release and controlled-release cisplatin dry powders for inhalation against lung cancer Vincent Levet, Rémi Rosière, Karim Amighi & Nathalie Wauthoz Laboratory of Pharmaceutics and Biopharmaceutics, Faculty of Pharmacy, Université libre de Bruxelles (ULB), CP 207, Boulevard du Triomphe, 1050 Brussels, Belgium Summary Immediate-release and controlled-release cisplatin dry powder formulations based on solid-lipid microparticles (SLMs) with high drug-content (≥ 50%) and exhibiting high deposition abilities in vitro (fine particle fractions (FPF) between 37% and 52%), were produced by high-pressure homogenization and spray-drying using tristearin and [1] polyethylene glycol (1000) tocopheryl-vitamin E succinate . Dry powder blends (DPB) were realized using a spray-dried Mannitol:L-Leucine (9:1) diluent and mixed with the formulations for accurate and reliable administration to mice. DPB were characterized by their drug-content, uniformity of content, recovered mass, ® recovered cisplatin and particle size distribution by laser diffraction with a Malvern Spraytec through the actuation of a Penn-Century Dry Powder Insufflator™ model DP-4M for mouse. All but one of the DPB were able to be reliably delivered in vitro. The local and systemic pharmacokinetic distributions of cisplatin from formulations were then evaluated in vivo in CD-1 mice vs. IV and vs. a nebulized cisplatin aqueous solution. Quantification of platinum (Pt) content in organs was realized using validated methods by electrothermal atomic absorption spectrometry (ETAAS) in lungs, kidneys, liver, spleen, mediastinum and total blood of mice after a single 1.25 mg/kg administration and dosed over 48 hours. It showed (1) that the inhaled route could effectively lower systemic exposure while increasing lung exposure when compared to IV, (2) that immediate-release formulations were very quickly absorbed in the lungs and (3) that controlled-release formulations promoted higher total exposure in the lungs, but that the presence of PEGylated excipient was needed to avoid active and fast elimination of particles from the lungs. Introduction Non-small cell lung cancer (NSCLC) accounts for 1.6 million deaths per year and has a very poor prognosis at [2] advanced stages, with one of the lowest 5-year survival rates of all cancers . Treatment includes combinations of surgery, radiotherapy and chemotherapy, depending on the stage of the disease. Cisplatin is one of the most potent and the most employed anti-cancer drug against various cancers and is the principal constituent of doublet [3] chemotherapy against NSCLC . It is currently only administered by IV infusion with concomitant hydration with [4] nephroprotective agents because of its acute and chronic dose-limiting renal toxicity . Local administration of cytotoxic agents against lung cancer has been investigated in the past using various agents, mostly by aerosol therapy, but lacked the ability to reach efficacious doses within a suitable timeframe compatible with clinical [5] practice . Dry powder formulations for inhalation are able to deliver large doses of active agent to the deeper lung in a few minutes and are patient actuated, which could help limit environmental and healthcare personnel exposure. They could also help lower systemic exposure while increasing exposure in the lungs. Controlledrelease (CR) formulations could also help lower acute exposure locally in order to limit lung toxicity and diminish the strain of repeated administrations. However, inhaled particles undergo many challenges as they are confronted by elimination processes such as mucocilliary clearance and uptake by alveolar macrophages in the lungs, which could hinder their CR properties. Potential stealth properties of inhaled particles, provided by the addition of PEGylated excipients to formulations, are very much needed in order to promote their local [6] residence . These aspects have to be fully evaluated through preclinical studies using animal models such as [7] rodents, along with adapted endotracheal administration devices . Formulation behavior using those devices has to be fully characterized in vitro in order to ensure the reliability of in vivo results. This study focuses on the development and in vitro characterization of suitable dry powder blends (DPB) for the administration of cisplatin to mice and on the comparative pharmacokinetic (PK) results obtained in vivo with different immediate-release (IR) and CR formulation strategies for pulmonary nebulization and the IV route. Previous work [1]

Cisplatin formulations for human use have been previously produced and characterized . Briefly, cisplatin at 5% w/v was micronized in an isopropanol suspension through high-pressure-homogenization for 40 cycles up to 20 000 psi. The resulting microcrystals, with a d(0.5) of 0.89 ± 0.01 m, were then spray-dried with polyethylene glycol (1000) tocopheryl-vitamin E succinate (TPGS, F1) or with solubilized tristearin (TS) only (F2) or with TS and a PEGylated excipient using TPGS (F3) or distearoyl phosphoethanolamine polyethylene glycol 2000 (DSPEmPEG-2000, F4), all at 2% w/v in solids content (Table 1). These formulations exhibited a high fine particle fractions (FPF) based on the nominal dose and a high fine particle dose (FPD). The inhalable fraction of these formulations also displayed IR for the carrier-free formulation F1, CR over more than 24 h for the 50% TS[8] comprising formulations (F2, F3) and a low burst-effect in vitro in modified simulated lung fluid (mSLF) .

257


Drug Delivery to the Lungs 27, 2016 - Local and systemic pharmacokinetic evaluation of immediate-release and controlled-release cisplatin dry powders for inhalation against lung cancer Table 1. Theoretical compositions, FPFs and FPDs related to a 20 mg-filled capsule of cisplatin formulations ® obtained using a multistage liquid impinger operated at 100 L/min for 2.4s (n = 3) using an Axahaler dry powder inhaler (SMB S.A., Belgium), and the in vitro drug released fractions after 10 min and 24 h from the inhalable fraction (particles below 5 m) of formulations. Formulation

F1 F2 F3 F4

Theoretical composition (% w/w)

FPF (% w/w)

FPD (mg)

Released fractions (%w/w) 10 min 24 h

Cisplatin TPGS

95% 5%

52 ± 3

9.8 ± 0.6

97 ± 5

100 ± 0

Cisplatin TS Cisplatin TS TPGS Cisplatin TS DSPE-mPEG-2000

50% 50%

46 ± 7

4.5 ± 0.7

16 ± 4

79 ± 10

37 ± 2

3.6 ± 0.2

24 ± 3

55 ± 11

50 ± 6

5.0 ± 0.6

25 ± 7

56 ± 13

50.0% 49.5% 0.5% 50.00% 49.75% 0.25%

Materials and methods 1. Production and characterization of dry blends In order to administer dry powder formulations to mice in vivo at a 1.25 mg/kg dose, DPB of formulations had to be realized using a suitable diluent for a repeatable and accurate administration with an inhalation device, a Dry Powder Insufflator™ model DP-4M for mouse (Penn-Century Inc., USA). The diluent was obtained by spray drying an aqueous solution of 1.0% w/v Pearlitol 25C mannitol (Roquette Frères, France) and 0.1% w/v L-Leucine (Merck-Millipore, Germany) with a Mini Spray Dryer B-290 (Büchi, Switzerland). The spray-drying parameters were as follows: solution feed rate 3 g/min, inlet temperature 130°C, 0.7 mm nozzle, 1.5 mm nozzle-cap, spraying 3 air flow 800 L/min, drying air flow 35 m /h and the use of a B-296 dehumidifier (Büchi, Switzerland). The actual outlet temperature was 54°C and the process yield was around 71%. DPB were then prepared by mixing the formulations with diluent using a Turbula 2C three-dimensional motion mixer (Bachofen AG, Switerzland) for 4 hours in a 2-mL glass vial, aiming at a target concentration of 2.0% w/w cisplatin in DPB. F1 was therefore diluted 50-fold, while F2, F3 and F4 were diluted 25-fold. DPB were then characterized by their drug content (mean and [1] uniformity of content) by ETAAS with a 2.0 mg mass for each DPB (n = 10). They were then characterized for their ability to be administered with the DP-4M. The in vitro delivery from the DP-4M was assessed by its efficiency and repeatability of deliverance for each DPB by weighing the device before and after activating the device 5 times (mean ± SD, n = 15). The in vitro accuracy of delivery was established between the theoretical dose calculated from the delivered mass of DPB and the actual emitted dose of DPB, which was measured by activating the device 5 times into an empty 5-mL glass vial, and quantifying the deposited Pt content with ETAAS (mean ± SD, n = 10). Particle size distributions (PSD) of DPB were measured in the plume from the DP-4M [7] (n = 5) using a Spraytec laser diffractometer (Malvern Instruments Ltd., UK) and expressed as d(0.5), d(0.9) and percentage particle volume undersize (particles below 5 m). 2. Administration protocol 200 CD-1 (ICR) female mice aged between 5 and 15 weeks were administered cisplatin at a 1.25 mg/kg dose either in solution via the intravenous route (IV) or by nebulization or with a DPB via inhalation. Inhaled © formulations were administered to mice via the endotracheal route using a MicroSprayer Aerosolizer - Model 1A1C-M for mouse (Penn-Century Inc., USA) or the DP-4M (i.e. for solution or for DPB, respectively), using the [9] protocols described elsewhere . Briefly, the MicroSprayer and the DP-4M are comprised of a reservoir (syringe or sample chamber, respectively) that are connected to a hollow stainless-steel tip. They can deliver nebulized solutions or powders into a plume directly into the lungs of anesthetized mice through endotracheal insufflation ® using the syringe or a pump, respectively. Mice were then euthanized with sodium pentobarbital (Nembutal , Ceva Animal Health, Belgium) at 10 min, 30 min, 1 h, 2 h, 4 h, 8 h, 24 h and 48 h (n = 5). Blood was sampled by retro-orbital bleeding in a lithium-heparin tube (Sarstedt, Germany) and organs (i.e. lungs, kidneys, liver, spleen and mediastinum) were removed, washed in saline and weighed and frozen until analysis using ETAAS. All animal experiments were approved by the ethical committee of the Faculty of Medicine, ULB (CEBEA) under the agreement number 585N. 3. Quantification of the platinum content in organs and in total blood © Organs were thawed and digested in 69% v/v Suprapur nitric acid (Merck-Millipore, Germany) under ultrasonication at 65°C for 3 hours to obtain completely clear yellow solutions. Pt content in total blood was established by digesting 50 L of heparinized blood using 50 L Suprapur nitric acid and 100 L 0.5% w/v Triton X-100 (Merck-Millipore, Germany). Measurements were then realized with a SpectrAA 300 atomic absorption [1] [10] spectrometer using adapted and validated methods consistent with the requirements of the FDA . Calibration was realized by the autosampler from freshly prepared matrix-matched 800 ng/mL cisplatin standards and blank matrix for each organ. Results were all adjusted to the same drug dose based on the emitted mass and drug content of DPB. AUC was calculated using the trapezoidal rule between measured data points.

258


Drug Delivery to the Lungs 27, 2016 - Vincent Levet et el Results and discussion The IR formulation F1 and CR formulations F2, F3 and F4 were diluted to obtain the DPBs (Table 2), showing a smaller PSD for DPB1 and a uniformity of drug content comprised between 7.3 and 8.9% for all formulations except for DPB4, for which a higher variability of 13.4% w/w was observed. In vitro results showed that delivery from the DP-4M endotracheal device was more efficient and more repeatable with DPB1. This is probably attributable to the greater quantity of excipients present in this blend (50-fold dilution vs. 25-fold dilution for all the others formulations). Moreover, DPB1 was the only formulation without lipids, which could have impaired delivery. This is because it could have increased interparticle interactions with the diluent during mixing and within the [1] device, causing more particle aggregates, especially as the initial formulation exhibit larger PSD . Finally, the accuracy of delivery showed that emitted cisplatin levels were in accordance with the expected values calculated from the weighed device before and after actuations for DPB1, DPB2 and DPB3. These values were later on used in vivo. In contrast, a great variability in accuracy was observed for DPB4, probably due to its lesser uniform content and by particle demixing during actuation of the device. DPB4 was therefore discarded for the in vivo part of the study. In vivo, the DPB were administered to mice with lower delivery efficiency and a greater variability than in vitro. This could be caused by a blockage of the device tip or by powder sticking caused by the damper environment in mouse trachea. Table 2. Measured parameters for the DPBs (PSD, drug-content, delivery with the DP-4M device).

Undersize (% below 5 m, mean ± SD, n = 5) 86.5 ± 6.7

Cisplatin content (% w/w, mean ± SD, n = 10) 2.37 ± 0.21

Uniformity of content (% w/w, n = 10) 8.9

Efficiency and repeatability (% w/w, mean ± SD, n = 15) 96.3 ± 7.6

Accuracy (% w/w, mean ± SD, n = 10)

DPB1

d(0.5) (m, mean ± SD, n = 5) 2.4 ± 0.8

In vivo delivery to the lung Efficiency and repeatability (% w/w, mean ± SD, n = 40)

97.2 ± 4.3

72.9 ± 33.7

DPB2

3.7 ± 0.8

69.1 ± 14.0

1.67 ± 0.17

9.9

85.8 ± 17.4

98.0 ± 16.2

67.3 ± 26.1

DPB3

3.6 ± 0.5

69.2 ± 10.3

1.64 ± 0.12

7.3

87.3 ± 16.7

104.8 ± 19.3

82.9 ± 25.5

DPB4

3.4 ± 0.5

73.8 ± 9.0

1.60 ± 0.21

13.4

85.8 ± 19.3

198.3 ± 78.5

Laser diffraction Dry powder blend

Drug content

In vitro delivery

-

PK results (Figure 1) showed that lung exposure was relatively higher with all inhaled experiments compared to IV. It also showed that nebulization and IR DPB1 had similar PK profiles, and that Pt levels fell very rapidly, already before the first dosing time (10 min). PK profiles of CR formulations in vitro (DPB2 and DPB3) showed that DPB2, despite being preserved in larger amounts at the first dosing times, was rapidly eliminated within the first hour. This suggests an active elimination process, probably by alveolar macrophages uptake. However, the PEGylated excipient-comprising formulation F3, showed high and sustained levels of platinum for up to 24 hours, [6] confirming the well-described abilities of PEG chains to delay macrophage recognition and capture of particles . Blood platinum levels, were higher with IV than with the pulmonary route and showed absorption profiles compatible with the slow distribution from the lungs to the systemic compartment for all inhaled experiments.

Figure 1. Local and systemic pharmacokinetic profiles showing Pt content in lungs and in total blood after a single 1.25 mg/kg administration (Mean ± SEM, n = 5 for each dosing time). Pt Cmax which is more representative of the acute toxicity linked to cisplatin exposure, was also reduced in total blood (Figure 2a) for all the inhaled experiments as compared to IV. This was especially the case for CR formulation DPB3 (161 ± 57 ng/mL vs. 791 ± 181 ng/mL for IV). The reduced exposure was also particularly observable in kidneys (Figure 2b) for IR formulation DPB1 and nebulization (0.6 ± 0.2 ng/mg and 1.1 ± 0.5 ng/mg vs. 3.0 ± 0.7 ng/mg for IV). Observed Cmax at 10 min was, however, significantly increased in lungs for DPB2 and DPB3 (Figure 2c), which does not reflect CR properties of these formulations as ETAAS does not differentiate between available Pt, dissolved Pt and undissolved Pt from particles in the lung parenchyma (Kruskall-Wallis test with Dunn’s multiple comparison test, p < 0.01).

259


Drug Delivery to the Lungs 27, 2016 - Local and systemic pharmacokinetic evaluation of immediate-release and controlled-release cisplatin dry powders for inhalation against lung cancer It also showed that IR formulations were quickly absorbed (less than 10 min). Analysis of the area under the curve (AUC) for all PK profiles (Figure 2d) showed that the global exposure of lungs to Pt was increased for inhaled formulations compared to IV injection (very significantly with DPB3 with a more than 10-fold increase, p < 0.001). Interestingly, AUC in kidneys was lowered for all inhaled formulations, as compared with IV (for instance, AUC in -1 -1 kidneys of 475 ± 187 ng.min.mL vs. 2168 ± 149 ng.min.mL for DPB2 and IV, respectively). An exception was -1 observed for DPB3, for which AUC in kidneys was closer to IV with 1625 ± 806 ng.min.mL (mean ± SEM, n = 5 per group per dosing time, 8 times). Further experiments are needed to explain this phenomenon. Exposure was also lowered in liver for all inhaled formulations, as compared with IV. a)

Total blood

b)

Kidneys

c)

Lungs

d)

Figure 2. Cmax in total blood (a), kidneys (b) and lungs (c) (mean ± SEM, n = 5 per group) and total exposure of organs (d) expressed as the AUC in organs for all considered routes and formulations (mean ± SEM, n = 5 per group per dosing time, 8 times). Significance of the difference with the IV control group or the mentioned group was calculated using a Kruskall-Walis test with Dunn’s multiple comparison test (c) or a two-way ANOVA with Bonferroni’s multiple comparison test (d) and is expressed as follow: ** = p < 0.01, *** = p < 0.001). Conclusion The PK of cisplatin is of tremendous interest in order to assess the outcome of cisplatin or its reactive species after lung administration. We showed that a formulation comprised of TS and PEGylated excipients greatly improves Pt exposure in the lungs for more than 8 hours compared with (i) an IV-delivered cisplatin solution, (ii) a nebulized cisplatin solution or (iii) IR formulations. The use of PEGylated excipients seems crucial in order to prolong residency of inhaled SLMs in the respiratory tract. The increased Pt exposure observed with the CR formulation in the kidneys will have to be further assessed using more specific renal parameters for renal toxicity such as blood urea nitrogen or creatinine levels. Lung toxicity will also have to be assessed using broncho[9] alveolar lavage fluid analysis . References 1

Levet V, Amighi K, Wauthoz N: Development of cisplatin-based dry powder for inhalation with controlled release properties for the lung cancer therapy. Respiratory Drug Delivery (2016) 3(549-552). 2 Howlader N, Noone A, Krapcho M, Garshell J, Miller D, Altekruse S, Kosary C, Yu M, Ruhl J, Tatalovich Z, Mariotto A et al: Seer cancer statistics review, 1975-2012. In: National Cancer Institute. Bethesda, MD (2015). 3 Clinical practice guidelines in oncology non-small cell lung cancer. V 3.2014 (2014). http://www.nccn.org/professionals/physician_gls/pdf/nscl.pdf. 4 NHS: Guidelines for intravenous hydration for chemotherapy regimes in adults. In: network sc (Ed) (2012). 5 Phase 2 study of inhaled lipid cisplatin in pulmonary recurrent osteosarcoma (2012). http://clinicaltrials.gov/ct2/show/NCT01650090. 6 Muralidharan P, Mallory E, Malapit M, Hayes D, Jr., Mansour HM: Inhalable pegylated phospholipid nanocarriers and pegylated therapeutics for respiratory delivery as aerosolized colloidal dispersions and dry powder inhalers. Pharmaceutics (2014) 6(2):333-353. 7 Duret C, Wauthoz N, Merlos R, Goole J, Maris C, Roland I, Sebti T, Vanderbist F, Amighi K: In vitro and in vivo evaluation of a dry powder endotracheal insufflator device for use in dose-dependent preclinical studies in mice. European journal of pharmaceutics and biopharmaceutics : official journal of Arbeitsgemeinschaft fur Pharmazeutische Verfahrenstechnik eV (2012) 81(3):627-634. 8 Son YJ, McConville JT: Development of a standardized dissolution test method for inhaled pharmaceutical formulations. International Journal of Pharmaceutics (2009) 382(1-2):15-22. 9 Rosiere R, Van Woensel M, Mathieu V, Langer I, Mathivet T, Vermeersch M, Amighi K, Wauthoz N: Development and evaluation of well-tolerated and tumor-penetrating polymeric micelle-based dry powders for inhaled anti-cancer chemotherapy. International Journal of Pharmaceutics (2016) 501(1-2):148-159. 10 Food and Drug Administration: Bioanalytical method validation guidelines for industry (2013).

260


Drug Delivery to the Lungs 27, 2016 – R. H M Hatley et al. Ensuring the Consistency of Performance of Mesh Nebulizers R. H M Hatley, L. E A Hardaker, A. P Metcalf, J. Parker, F. Quadrelli & J. N Pritchard Respironics Respiratory Drug Delivery (UK) Ltd, a business of Philips Electronics UK Limited, Chichester, West Sussex, UK. Summary The consistency in aerosol performance of the mesh in a mesh nebulizer is important as variability can affect the respirable delivered dose and treatment time. In addition meshes should be replaced periodically as part of routine mesh nebulizer maintenance, here it is particularly important that the output rate and thus treatment time of the new mesh should be similar to the one it replaced as differences could be perceived by the patient as a less effective treatment if there is variability in treatment time. We tested the consistency of performance of 73 randomly selected InnoSpire Go meshes from 7 batches manufactured using new in-process controls intended to minimize variability in aerosol performance. The mesh was placed in an Aerogen fixture and attached through a connector to a Spraytec laser diffraction system (set to 6 L/min extraction with an additional 7 L/min sheath air). Data acquisition was started on the Spraytec and the generator was started. Two priming runs were performed by transferring 50 µL of 0.9% saline to the fixture then tests were performed in triplicate with 250 µL of 0.9% saline. This method was repeated for each of the 73 meshes. Volume median diameter (VMD) and output rate were comparable between the batches; mean VMD was 4.76 µm with a range of ±0.2 µm around the mean, and mean output rate was 0.55 mL/min with a range of 0.03 mL/min below and 0.01 mL/min above the mean. The maximum difference between the batches for VMD was 0.4 µm and for output rate was 0.04 mL/min. VMD and output rate were consistent between the batches of meshes tested; therefore, replacement of the mesh should not cause variation in treatment time or particle size. Introduction Correct nebulizer use is important to ensure that patients receive an effective dose, and factors such as consistent treatment time can indicate to a patient that they are correctly using their nebulizer. The mesh in a mesh nebulizer should be changed periodically as part of routine maintenance, and it is important that the performance of the nebulizer does not alter between mesh batches to avoid changes in treatment times and particle size when a mesh is replaced. New in-process controls have been developed for the InnoSpire Go mesh nebulizer (Respironics Respiratory Drug Delivery [UK] Ltd, Chichester, UK) in order to ensure that treatment time [1] and VMD do not vary between mesh batches. In this study, we aimed to assess the consistency of performance of 73 meshes from 7 batches that have been through the in-process control by assessment of particle size and output rate. Methods Prior to testing, the meshes, test fixture, and connector were washed in warm soapy water, rinsed, and air-dried. All equipment and solutions/reagents were stabilized to ambient laboratory conditions at least 2 hours before use (22-24 °C, 995-1016 mbar atmospheric pressure, and 46-56% relative humidity). The 73 meshes were randomly selected from 7 different manufactured batches (Table 1). Table 1. The number of meshes randomly selected from each batch.

Batch Number of meshes

A 3

B 12

C 20

D 15

E 8

F 8

G 7

A Spraytec laser diffraction system (Malvern Instruments Ltd, Malvern, UK) was set up to perform testing with the open flow cell method with a 6 L/min extraction and an additional 7 L/min sheath air with the flow cell in horizontal [2] A mesh was placed into a fixture orientation, in line with the novel method reported in Slator et al., 2016. (Aerogen Limited, Galway, Ireland), which was then attached to the Spraytec with a connector. Data acquisition was started on the Spraytec, and the controller attached to the generator was started. A priming run was performed by transferring 50 µL of 0.9% saline to the Aerogen fixture, and another prime was performed when the obscuration dropped to background levels. When the obscuration returned to background levels again, a test was performed by transferring 250 µL of 0.9% saline to the Aerogen fixture. The test runs were performed in triplicate; data acquisition was stopped and the controller turned off once the obscuration returned to background levels again after the final test. The test was repeated with each of the 73 meshes. The mean data of the 3 runs were recorded for each mesh. Output rate was calculated based on the nominal charge volume and the number of records produced by the Spraytec, according to Equation 1, as each record is equivalent to 1 second. [Equation 1]

Output rate (mL/min) = (0.25 x 60) / Number of records produced

261


Drug Delivery to the Lungs 27, 2016 – Ensuring the Consistency of Performance of Mesh Nebulizers Results

Figure 1 - Mean volume median diameter for each batch ( denote 1 standard deviation about the mean.

) and overall mean of all the batches (

). Error bars

Figure 2 - Mean output rate for each batch ( ) and overall mean of all the batches ( ). Error bars denote 1 standard deviation about the mean.

262


Drug Delivery to the Lungs 27, 2016 – R. H M Hatley et al. Table 2. Mean particle size distribution span for each batch of meshes.

Batch Span

A 1.53

B 1.59

C 1.56

D 1.56

E 1.53

F 1.54

G 1.56

The volume median diameter (VMD) was comparable between the batches of meshes, with a maximum difference of 0.4 µm between the batch means and a range of ±0.2 µm around the mean (Figure 1). The output rate was also comparable, with a maximum difference of 0.04 mL/min between the batch means and a range of 0.03 mL/min below and 0.01 mL/min above the mean (Figure 2). The relative standard deviations were 7% for the VMD results and 8% for the output rate results. Discussion The VMD and output rate were consistent between the batches of meshes, indicating that the replacement of these meshes as part of routine maintenance would not produce variations in particle size or treatment time. The VMD was below 5 µm for all of the batches of meshes, indicating that the meshes produced aerosol droplets of a [3,4] The output rate of 0.55 mL/min is relatively size that would be likely to deposit in the smaller airways of a lung. fast, and therefore short treatment times could be achieved. This may be a deciding factor for patients selecting an inhalation device, as short treatment times are desirable, and with conventional nebulizers, treatment times are [5,6] The maximum difference in output rate between the batches was 7%, which would equate to a usually longer. 6-second variation in a 3-minute treatment. Such a short variation in treatment time would be imperceptible to the user and could therefore provide them with confidence in the consistency of their treatments. Conclusion The performance of the meshes was consistent in terms of VMD and output rate. The output rate data indicated that short and consistent treatment times can be achieved between meshes, which is beneficial as users could therefore be confident in the consistency of the device’s operation, even when the mesh is replaced as part of routine maintenance. References 1

Hatley R, Hardaker L, Zarins-Tutt J, Quadrelli F, Hogan B, Mac Loughlin R, Pritchard JN: Investigation of optical density for the characterization of nebulizer meshes. (Abstract). Presented at: Respiratory Drug Delivery 2016, Scottsdale, Arizona, April 17-21, 2016: Respiratory Drug Delivery vol. 3: pp489-492.

2

Slator L, Degtyareva Y, Hardaker L, von Hollen D, Pritchard J: Comparison of aerosol particle size measured by 2 methods using 3 brands of mesh nebulizer. (Abstract). Presented at: Drug Delivery to the Lungs 27 2016, Edinburgh, Scotland, UK, December 7-9, 2016 (In press).

3

American Association for Respiratory Care: Aerosol consensus statement – 1991, Respir Care 1991; 36:9; pp916-921.

4

Newman SP: Aerosol deposition considerations in inhalation therapy, Chest 1985; 88:2 Suppl; pp152S-160S.

5

Ari A: Jet, ultrasonic, and mesh nebulizers: an evaluation of nebulizers for better clinical outcomes, Eurasian J Pulmonol 2014; 16:1; pp1-7.

6

Elphick M, von Hollen D, Pritchard JN, Nikander K, Hardaker LEA, Hatley RHM: Factors to consider when selecting a nebulizer for a new inhaled drug product development program, Expert Opin Drug Deliv 2015; 12:8; pp1375-1387.

263


Drug Delivery to the Lungs 27, 2016 - Ayasha Patel et al. Drug Loading of Human Albumin Nanocarriers for Inhaled Anti-Tuberculosis Therapy Ayasha Patel1, Arcadia Woods1, Peter Imming2, Adrian Richter2, Nick Childerhouse3, Lea Ann Dailey1,2 & Ben Forbes1 1

Institute of Pharmaceutical Science, King’s College London, 150 Stamford Street, London SE1 9NH, United Kingdom 2 Department of Pharmacy, Martin Luther University of Halle-Wittenberg, Wolfgang-Langenbeck-Str.4, 06120 Halle, Germany 3 Vectura Group plc, One Prospect West, Chippenham, Wiltshire, SN14 6FH, United Kingdom

Summary Albumin nanoparticles are promising candidates for drug delivery to the lungs. Previous studies on the clearance, biocompatibility and biodistribution of albumin nanoparticles have demonstrated their ability to target drug delivery to the lungs. We have demonstrated that therapeutically relevant compounds can be incorporated into albumin nanoparticles, however a need for an optimised method and further characterisation was identified. In this study we investigated two methods of manufacture: pre-loading and in situ loading of two new antimycobacterial benzothiazinones (IR 20 & IF 274) with albumin nanoparticles. Nanoparticles were successfully manufactured robustly with both methods. In situ loaded nanoparticles offered higher capacity to carry these antimicrobial compounds: around 6 Âľg/mg compared to about 3 Âľg/mg for pre-loaded nanoparticles, despite similar drug encapsulation efficiencies for IR 20 pre-loaded nanoparticles (54% compared to 46% for the in situ method). In situ loaded nanoparticles also produced less drug loss during the purification process. This study has shown the development of novel methods for synthesis of drug-loaded albumin nanoparticles which has demonstrated that potent poorly-soluble, therapeutically-relevant compounds can be loaded into albumin nanocarrier formulations for inhaled drug delivery. Introduction The potential for nanoparticles to provide controlled drug delivery to the lungs provides exciting opportunities to [1] enhance inhaled therapy for lung diseases . Albumin-based nanoparticles in particular exhibit favourable [2] properties for this application as they are non-immunogenic, non-toxic and easily manufactured . The biocompatibility, clearance, and biodistribution profiles of albumin nanoparticles have been investigated previously [3] demonstrating prolonged retention after delivery to the lungs . These studies have provided important proof-ofconcept regarding the fate of albumin nanoparticles in the lungs and their potential for targeting drug delivery to the lungs via inhalation. One application proposed for this technology is as a carrier for anti-tuberculosis drugs, with alveolar macrophages being their therapeutic target. However, few studies have explored whether [4] therapeutic agents can be loaded into albumin nanoparticles for inhaled delivery to the lung . Benzothiazinone (BTZ) compounds are a novel class of hydrophobic anti-tuberculosis drugs which we have investigated previously to establish their interaction with bovine serum albumin (BSA) and potential to fabricate [5] into BTZ-loaded albumin nanoparticles . BTZs are high potency compounds with low aqueous solubility and low [6] oral bioavailability . These studies have shown the propensity of BTZ for molecular interaction with albumin, but identified the shortcoming in terms of drug-loading into albumin nanoparticles. Further investigations into alternative methods for manufacture and optimising drug loading of the nanoparticles are warranted. The aim of this study was to compare two different methods for the manufacture of drug-loaded albumin nanoparticles. Characterisation of the nanoparticles included drug loading and drug recovery. These studies were carried out with human serum albumin (HSA). The model BTZ compounds used in the study and their physicochemical characteristics are shown in Table 1. Table 1. Molecular weight, structure, and aqueous solubility of benzothiazinone compounds under investigation as new chemical entities for the treatment of tuberculosis.

Experimental Methods Preparation of BTZ-loaded HSA nanoparticles using a pre-loading technique

264


Drug Delivery to the Lungs 27, 2016 - Drug Loading of Human Albumin Nanocarriers for Inhaled AntiTuberculosis Therapy HSA solution in tris buffer (1 mL, pH 8.9) was incubated with 0.5 mg of BTZ compound as a film for four days. A validated UV spectroscopic assay was used to determine how much drug had been solubilised by albumin. BTZ[7] loaded albumin nanoparticles were manufactured using a modified desolvation method . In short, 25 µL NaOH was added to the BTZ-saturated HSA solution (1 mL), then 4.0 mL of ethanol was added drop wise to the stirred protein solution. Nanoparticles were cross-linked by addition of 47.2 µL 10% glutaraldehyde in water and overnight stirring. Particles were purified through at least 4 cycles of spin-filtration (30 kDa MWCO) into saline (0.9%) and the wash (“supernatant”) was collected for drug recovery analysis. All used glassware was also retained for “residual” drug analysis. HSA nanoparticles without drug were prepared as controls. Preparation of BTZ-loaded HSA nanoparticles using an in situ loading technique HSA solution in tris buffer (1 mL, pH 8.9) was prepared. BTZ compound was dissolved in dichloromethane and ethanol (1:10, 4.4 mL). After the addition of 25 µL NaOH to the HSA solution, the BTZ mixture was added drop wise to the stirred protein solution. Nanoparticles were then cross-linked and purified in the same manner as described above. Characterisation of BTZ-loaded HSA nanoparticles Particle size and polydispersity index (PDI) values were measured by photon correlation spectroscopy (PCS) using the Zetasizer Nano Series ZS (Malvern Instruments, Malvern, UK). Zeta potential was measured in water and saline and measured at a concentration of 20 µg/mL. Nanoparticle concentration was determined gravimetrically. Determination of drug loading To determine the amount of drug present in nanoparticles, the supernatant obtained during the purification process (see above) was dried and analysed by UV measurements at 347 nm to determine the amount of drug present (unloaded fraction). Purified nanoparticles (1 mL) were added to a vial containing zinc sulphate solution (0.1 M, 850 µL) and ammonia solution (10%, 200 µL). Dichloromethane (5 mL) was added and the vial was sonicated for 20 minutes in a water bath, followed by 60 minutes incubation time. The dichloromethane layer was removed and transferred to a clean vial and left to evaporate. The extraction procedure was repeated two times. Dimethyl sulfoxide was added to the extracted samples and the drug amount measured by UV spectrophotometry as described above. Any residual drug left in glassware that had not been solubilised was also analysed in the same manner. Results BTZ-loaded albumin nanoparticles were manufactured successfully through both the pre-loading & in situ loading methods. No significant differences were observed in size and zeta potential (Table 2) between drug-loaded particles and drug-free controls. Nor were any significant differences observed between the two methods. Visually, BTZ-loaded HSA nanoparticle suspensions appeared similar to the control, remaining colloidally stable as the suspensions remained cloudy, appeared uniform and no precipitation was observed. Particle yield was quantified following manufacture. This was done by determining the HSA nanoparticle concentration gravimetrically and calculating the yield as the fraction of the theoretical particle concentration had all the HSA been formed into nanoparticles. All particle batches were found to have a yield of at least 95%. The drug loading of the nanoparticles manufactured and loading efficiency are also presented in Table 2. In terms of the total drug recovery for the BTZ nanoparticles, “residual” drug refers to drug that was left in vials/glassware and hence not involved in the nanoparticle manufacturing process and “supernatant” refers to the supernatant/wash collected during the purification process after manufacture of the nanoparticles (see above). For the pre-loading method of manufacture, the amount of BTZ solubilised by the albumin before the manufacture process was measured and calculated as a percentage of the total amount of drug added (0.5 mg). BTZ solubilisation was 63.1% and 54.5%, respectively, for IR 20 & IF 274 which was consistent with drug solubility [5] determined in previous work . Unsolubilised drug was expected to be left as residual drug as only the saturated solution was used for nanoparticle manufacture. For both IR 20 & IF 274, the washing process removed unbound drug from the albumin nanoparticles: 13.2% & 6% respectively. Despite IR 20 being solubilised to a greater extent; thereby having more drug available for incorporation into the particles; the loading efficiency for both drugs was similar: 55.3% and 53.7% respectively for IF 274 & IR 20. When BTZ nanoparticles were manufactured by the in situ loading method, more drug was available to be incorporated into the nanoparticles compared to the pre-loading method as there was no incubation with albumin prior to manufacture. Thus, all the drug introduced in the system, 0.5 mg, had the potential to be incorporated into the particles. Drug loading efficiency for this method differed between the drugs, with IF 274 being the more efficient (60.6% compared to 46.4% for IR 20). Table 2. Characteristics of human serum albumin (HSA) nanoparticles and loading with benzothiazinones (BTZ), IR 20 and IF 274. Values reported are the mean and standard deviation of at least n = 3 batches/experiments

265


When BTZ nanoparticles were manufactured by the in situ loading method, more drug was available to be incorporated into the nanoparticles compared to the pre-loading method as there was no incubation with albumin prior to manufacture. Thus, all the drug introduced in the system, 0.5 mg, had the potential to be incorporated into for this method the particles. Drug loading27,efficiency Drug Delivery to the Lungs 2016 - Ayasha Patel et al. differed between the drugs, with IF 274 being the more efficient (60.6% compared to 46.4% for IR 20). Table 2. Characteristics of human serum albumin (HSA) nanoparticles and loading with benzothiazinones (BTZ), IR 20 Drug Delivery to the Lungs 27, 2016 - Ayasha Patel et al. and IF 274. Values reported are the mean and standard deviation of at least n = 3 batches/experiments

Figure 1. Recovery of BTZ compound from human serum albumin nanoparticles manufactured using an albumin solution pre-loaded with IR 20 (A) and IF 274 (B); and recovery of BTZ compound for in situ loaded human serum albumin nanoparticles for IR 20 (C) and IF 274 (D). Expected recovery (%) of each drug from pre-loaded nanoparticles was determined as the proportion of drug from solubilised albumin (determined by UV analysis) and hence the Figure 1. Recovery of BTZ compound humanbyserum albumin nanoparticles manufactured using introduced an albumininto solution manufacturing process. expected recovered for frominthe fraction i.e. drug left pre-loaded with IR 20 The (A) remaining and IF 274fraction (B); andwas recovery of to BTZbecompound situ“residual” loaded human serum albumin behind in glassware. “Supernatant” refers the washrecovery obtained(%)during the drug purification of nanoparticles. Expected nanoparticles for IR 20 (C) and IF 274 (D).toExpected of each from pre-loaded nanoparticles was recovery (%) as of the eachproportion drug fromofindrug situ solubilised loaded nanoparticles 100% with drug. Data represent determined by albuminwas (determined byno UVexpected analysis) residual and hence introduced into the mean and standard deviation of n = 3-5 particle manufacturing process. The remaining fractionpreparations. was expected to be recovered from the “residual” fraction i.e. drug left behind in glassware. “Supernatant” refers to the wash obtained during the purification of nanoparticles. Expected Discussion recovery (%) of each drug from in situ loaded nanoparticles was 100% with no expected residual drug. Data represent [4] particle preparations. mean standard deviation of n = 3-5 , BTZ-loaded bovine serum albumin nanoparticles were successfully As weand have reported previously

manufactured using a modified desolvation method. However, development of these nanoparticles towards drug Discussion carriers for medical applications required the use of human rather than bovine albumin and resolution of [4] shortcomings drug loading into nanoparticles. In thisbovine study, serum we havealbumin produced HSA nanoparticles have , BTZ-loaded nanoparticles were which successfully As we havein reported previously equivalent characteristics to those developed using BSA and report an improved method for loading drug into drug the manufactured using a modified desolvation method. However, development of these nanoparticles towards nanoparticles. This new method shows similar robustness (i.e. reproducibility and ease of manufacture) as carriers for medical applications required the use of human rather than bovine albumin and resolutionthe of pre-loading method, butloading with the added advantages saving time increasing loading. shortcomings in drug into nanoparticles. Inofthis study, weand have produceddrug HSA nanoparticles which have

equivalent characteristics to those developed using BSA and report an improved method for loading drug into the nanoparticles. This new method shows similar robustness (i.e. reproducibility and ease of manufacture) as the pre-loading method, but with the added advantages of saving time and increasing drug loading.

266


Drug Delivery to the Lungs 27, 2016 - Drug Loading of Human Albumin Nanocarriers for Inhaled AntiTuberculosis Therapy

In situ BTZ-loaded albumin nanoparticles have a higher drug loading capacity compared to pre-loaded albumin nanoparticles. The capacity increased 1.5 fold for IR 20 (5.7 µg/mg compared to 3.8 µg/mg) and more than 2 fold for IF 274 loaded nanoparticles (6.1 µg/mg compared to 2.8 µg/mg). This suggests the more hydrophobic compound IF 274 has stronger interactions with albumin. Drug binding sites for albumin are well documented in [8] the literature and it is possible IF 274 may bind more strongly to a hydrophobic binding site on albumin . This also verifies in situ loading of drug into the nanoparticles as a better method, both in terms of time and increasing drug loading capacity, particularly for hydrophobic drugs which sometimes prove troublesome to formulate. The increase in drug carrying capacity of the in situ loading method suggests this can be optimised further perhaps by increasing drug: albumin ratios- this is an area for consideration for future experiments. Despite the increase in drug loading, drug loading efficiencies remained similar for both methods. This is why we found it important to investigate drug distribution in the manufacturing process; hence a recovery study was performed. This looked at residual drug that was not loaded, drug that was washed off the particles during the purification process (supernatant) and finally drug actually loaded in the particles. Total drug recovery for IR 20 & IF 274 was 80% and 70%, respectively. This was due to losses in drug throughout the experimental process that were unable to be quantified. Despite having significantly more drug available, IR 20 pre-loaded HSA nanoparticles had a similar loading efficiency compared to IF 274 i.e. more drug was found in the supernatant. The presence of drug loaded in the nanoparticles as well as in the washed off supernatant suggests different mechanisms of attachment of the drug. It suggests that drugs bind strongly to albumin as even with washing drug remained associated with the particles, but there are also weaker interactions which are disrupted during the purification process. This suggests that drugs that have higher aqueous solubility may tend to interact weakly with albumin resulting in them detaching more easily and resulting in differences in drug release during washing of the particles. Drug loss during purification for in situ loaded particles however was similar for both drugs, and lower than the loss when particles were formed from pre-loaded albumin solution. Future studies will include drug release studies which will provide insight into the release profiles of therapeutic agents from albumin nanoparticles and in vitro assays for drug activity. Despite the low drug:HSA ratio the albumin nanoparticles utilised in this study can be used to deliver highly potent drugs such as BTZs because the amount of drug required to produce the target concentration in the lungs (i.e. in the micromolar to nanomolar range) is deliverable by inhalation. Conclusion This study has demonstrated an improved method for incorporating hydrophobic BTZ compounds into human serum albumin nanoparticles and has provided insights into how these agents interact with albumin during nanoparticulate formation. These studies, together with previous findings demonstrating their enhanced retention in the lungs, constitute promising steps in the developments of albumin nanoparticles as an inhaled nanocarrier formulation. References 1 2 3

4 5 6

7 8

Paranjpe, M. & Müller-Goymann, C., 2014. Nanoparticle-Mediated Pulmonary Drug Delivery: A Review. International Journal of Molecular Sciences, 15(4), pp.5852–5873. Farokhzad, O.C. & Langer, R., 2006. Nanomedicine: Developing smarter therapeutic and diagnostic modalities. Advanced Drug Delivery Reviews, 58(14), pp.1456–1459. Woods, A., Patel, A., Spina, D., Riffo-Vasquez, Y., Babin-Morgan, A., de Rosales, R.T.M., Sunassee, K., Clark, S., Collins, H., Bruce, K., Dailey, L.A. & Forbes, B., 2015. In vivo biocompatibility, clearance, and biodistribution of albumin vehicles for pulmonary drug delivery. Journal of Controlled Release, 210, pp.1– 9. Choi, Seong Ho, et al. "Inhalable self-assembled albumin nanoparticles for treating drug-resistant lung cancer." Journal of Controlled Release 197 (2015): 199-207. , , Patel, A., Woods A., Imming, P., Richter, A., Dailey L. A. & Forbes B., 2015. Formulation of Benzothiazinones in Bovine Serum Albumin Nanoparticles. Drug Delivery to the Lungs. Edinburgh, 2016. Batt, S.M. Jabeen, T., Bhowruth, V., Quill, L., Lund, P.A., Eggeling, L., Alderwick, L.J., Fütterer, K. & Besra, G.S., 2012. Structural basis of inhibition of Mycobacterium tuberculosis DprE1 by benzothiazinone inhibitors. Proceedings of the National Academy of Sciences, 109(28), pp.11354– 11359. Weber, C., Kreuter, J. & Langer, K., 2000. Desolvation process and surface characteristics of HSAnanoparticles. International Journal of Pharmaceutics, 196(2), pp.197–200. Ghuman, J. et al., 2005. Structural basis of the drug-binding specificity of human serum albumin. Journal of Molecular Biology, 353(1), pp.38–52.

267


Drug Delivery to the Lungs 27, 2016 – Dongmei Cun et al. Development and Characterization of Nanocrystal-embedded Microparticles for Pulmonary Delivery of Budesonide 1

1

1

Dongmei Cun , Tingting Liu , Meihua Han , Mingshi Yang 1

1.2

Department of Pharmaceutics, School of Pharmacy, Shenyang Pharmaceutical Univeristy, Wenhua Road No. 103, Shenhe District, Shenyang, 110023, China 2 Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen , Universitetsparken 2, DK-2100 Copenhagen, Denmark

Summary Nanoparticle technology represents an effective approach for formulating poorly water-soluble pulmonary medicines. But unfortunately, directly using nanoparticles for inhalation often suffered from the problem of physical instability if they are applied in liquid form, or nanoparticles are likely to be exhaled before deposition if applied in the form of dry powder. In addition, rapid dissolution of nanoparticles will cause the rapid absorption of medicine, which means the rapid clearance of medicine from local lung tissue. To address these problems, a dry powder inhalation formulation composed of nanocrystal-embedded microparticles with hyaluronic acid as matrix excipients was designed for the pulmonary delivery of budesonide. Nanosuspension of budesonide was prepared by using wet milling method and by varying the rotation rate of milling bowl and milling time, three types of budesonide nanosupension were obtained. The nanosuspension was further spray dried with hyaluronic acid into so-called budesonide nanocrystal-embedded microparticles. The crystalline state of budesonide in nanosuspension or microparticles was confirmed by using XRPD and DSC. The morphology of microparticles was observed with SEM and it was shown that the microparticles were round in shape and appeared as wrinkled or crumpled spheres, with some degree of eccentricity. No significant difference between the formulations with different size of nanocrystals was observed. In vitro inhalation performance of microparticles was examined using the NGI system and demonstrated that the obtained nanocrystal-embedded microparticles has good aerosolization properties and suitable size for inhalation. Introduction Budesonide (BUD), a potent corticosteroid with high topical anti-inflammatory activity, is one of the mainly used corticosteroids to treat asthma and COPD. Pulmonary delivery is a preferred way for the treatment of these lungrestricted diseases since it could provide relatively high local drug concentration and low systemic exposure. Both of MDI and dry powder inhalation formulations of BUD are currently available in the market. However, hydrophobic nature of the drug is an important obstacle in its clinical use. In the case of dry powder inhalation, the low solubility of BUD would cause low drug concentration in lung and subsequently compromise the therapeutic performance of BUD. The strategy of narrowing the size of particle down to nano-scale was promising in improve [1] the dissolution of BUD , and the nebulized BUD nanosuspension was shown to be superior to the micronized [2] dosage form in terms of droplet aerodynamics, cellular uptake and efficacy . But unfortunately nanocrystal used in nebulizers often encounters the problem of physical instability in the form of uncontrolled agglomeration or Ostwald ripening. Processing such suspensions into dry powders could achieve better stability but it can also yield undesirable broad particle size distributions and large portion of powder could be exhaled due to the too [3] small particles size . In addition, for the poorly soluble drug BUD, although rapid dissolution would mean higher local drug level in lung, the maintenance of high local drug level would not last for a long time period due to the high permeability of BUD. BUD is a biopharmaceutical classification system class II drug, and the dissolved drug will be absorbed into systemic circulation rapidly and exert unwanted side effects. To resolve these problems, a system of nanocrystal-embedded microparticles with hyaluronic acid (HA) as matrix materials was designed. HA, a naturally occurring polymer has excellent bioadhesion by anchoring to mucous constituents, which can delay mucociliary clearance in the lung. HA with high molecular weight can form hydrogel [4] in water and provide sustained release profiles for entrapped drug . We hypothesised that by incorporating BUD nanocrystal into HA microparticlesďźŒthe release of BUD could be controlled and the local resident time of BUD is expected to be prolonged. The aim of current study is to obtain an inhalable HA dry powder formulation with BUD nanocrystals embedded. Nanocrystals-embedded microparticles were prepared and characterized, and the invitro inhalation performance of microparticles was also examined. Experimental methods 1. Materials Budesonide was purchased from Hubei Gedian Humanwell Pharmaceutical Co.Ltd. China. Pluronic F-68 was a gift from BASF, China. Hyaluronic acid (Mw~1,470,000 Da) was purchased from Shandong Freda Biochem. Co. Ltd. (Jinan, China). The other reagents and solvents used were of analytical grade. 268


Drug Delivery to the Lungs 27, 2016 - Development and Characterization of Nanocrystal-embedded Microparticles for Pulmonary Delivery of Budesonide 2. Preparation and Characterization of BUD nanosuspension The budesonide nanosuspensions were prepared by a wet milling technique. Briefly, 0.3 g F68 was dissolved in 30 ml of water. Then, 1 g of budesonide was dispersed in the stabilizer solution. The drug suspensions were pipetted into the milling bowl containing 100 g of milling pearls (zirconium oxide, 0.5mm in diameter). Two milling bowls were fixed in a planetary milling machine (Pulverisette 7 Premium, Fritsch Co., Germany). The grinding was performed at various rotation rates for different time to get BUD nanosuspension with different size. The BUD nanoparticles was collected by centrifugation (Thermo fisher) at 16000rpm for 15min and re-dispersed into distilled water to removed F-68 and be ready for size measurement and following processing. The mean particle diameter (Z-average) and polydispersity index (PDI) were determined by photon correlation spectroscopy (PCS, Malvern Nano ZS Malvern Instrument, UK). The crystalline form of BUD nanoparticles was analysed using X-ray powder diffraction (XRPD) analysis and differential scanning calorimetry (DSC). 3.Spray drying of nanocrystals into nanoembedded microparticles Spray drying was applied to incorporate the nanocrystals into nanoembedded using HA as excipient. A total amount of 100 mg BUD nanocrystal were dispersed in 100 ml water with 350 mg HA -290 , Flawil, Switzerland). The process parameters of spray drying were listed in table1. Table 1.The process conditions of B-290 spray dryer Inlet temperature (°C)

Aspiration rate (%)

Feed rate (%)

Q-flow (mm)

Feed volume (mL)

140

100

15

40

200

4. Characterization of nanocrystal-embedded microparticles The Morphology of the obtained dry powder of nanocrystal-embedded microparticles was observed with SEM and the physical state of BUD in microparticles was analyzed with XRPD and DSC. Aerodynamic particle size measurements were carried out according to the European Pharmacopoeia (Apparatus E) using a Next Generation Pharmaceutical Impactor (NGI) (Copley Scientific, Nottingham, UK). 10mg of particles were dispersed ® using an Cyclohaler DPI (Pharmachemie B.V., Netherlands) at a constant air flow rate of 100L /min Results 1. BUD nanocrystals with different size were obtained In order to examine the size effects of nanocrystal on the characteristics of final spray dried microparticles, three kinds of BUD nanocrystals with various diameters (around 260, 400, and 600nm respectively) were obtained by keeping the formulation constant but varying the rate of rotation and milling time (shown in table 2). The size distribution indicated by polydispersity index (PDI) was narrow for all of the three formulations. According to the Xray diffraction patterns, as shown in Fig. 1a,it was confirmed that no substantial crystalline change was found in the nanocrystals compared with raw crystals. However, the differences in the relative intensities of their peaks might be attributed to differences in the degree of crystallinity of the samples. DSC was also performed to analyze the different samples (Fig. 1b). In all cases, the DSC scanning of each sample showed a single sharp endothermic peak ascribed to the melting of the drug, which also indicated that there was no substantial crystalline change.

(b)

(a)

Fig. 1 XRPD (a) and DSC (b) curves of budesonide crystals with different particle sizes, raw BUD materials, and physical mixture of raw BUD materials and F68

269


Drug Delivery to the Lungs 27, 2016 – Dongmei Cun et al. Table2. The size and size distribution of BUD nanocrystals prepared under different process conditions (n=3, Mean±SD)

Formulation No.

Speed (rpm)

Time (min)

Particle size (nm)

PDI

1 2 3

200 200 500

6 31 120

606.77±7.03 406.30±24.98 267.97±6.07

0.202±0.067 0.127±0.054 0.268±0.026

2. Characterization of nanocrystals-embedded micropartlces The dry powder manufactured by spray drying has good flowability and appeared as wrinkled or crumpled spheres, with some degree of eccentricity (shown in Fig 2). It seems like the size of embedded nanocrystal has no significant influence on the morphology of microparticles. Again, the results of XRPD and DSC (Fig. 3) demonstrated that no substantial crystalline change happened during the process of spray drying.

Fig. 2 SEM image of microparticles containing BUD nanocrystals with different size and raw BUD material

Fig.4 SEM for four kinds of bud-HA microspheres (nanocrystals size are 260nm,400nm,600nmand raw material) DSC:

(a)

(b)

Fig.3 XRPD (a) and DSC (b) curves of HA microspheres containing budesonide crystals with different particle sizes, raw BUD materials, HA alone and physical mixture of raw BUD materials and HA Table.3. Aerodynamic performance of microparticles containing three kinds of nanocrystals or raw material No. MS(260nm) MS(400nm) MS(600nm) MS(rm)

MMAD(mm) 5.32 5.35 5.23 3.12

GSD(mm) 1.68 1.64 1.66 6.09

270

FPF% 37.47 30.19 31.53 38.97


Drug Delivery to the Lungs 27, 2016 - Development and Characterization of Nanocrystal-embedded Microparticles for Pulmonary Delivery of Budesonide For further characterization, the in vitro inhalation performance of nanocrystal-embedded microparticles was ® examined using the NGI system with Cyclohaler (Fig. 4). From the deposition profile obtained from the NGI analysis, FPF value of microspheres was calculated to be in the range of 30% -39% (table 3). The microparticles containing the smallest nanoparticles exibited the higher FPF value than the other two nanocrytals-contained microparticles. All of the four types of microparticles evaluated here have the mean mass aerodynamic diameter around the range of 3.12 to 5.35 μm, which are suitable for deposition at trachea and bronchi where the main action site of BUD for the treatment of asthma and COPD.

Fig.4 Aerodynamic behaviour of the microparticles containing three kinds of nanocrystals or raw material using a NGI at a flow rate (Q) of 100 L/min (Mean ± SD, n = 3). Discussion and conclusion In current study, BUD loaded nanoembedded HA microparticles were successfully prepared by using the technique of spray drying. In order to investigate the effects of size of BUD nanocrystal on the performance of nanocrystal-contained microspheres, three kinds of BUD nanocrystals with different particles size was prepared. Three types of nanocrystals appeared the same crystalline form with raw BUD materials and the crystalline form was maintained during the process of spray drying. Importantly, the spray dried BUD loaded nanoembedded HA microparticles exhibited relative high FPF value and suitable aerodynamic particle size for inhalation. Therefore it is interesting to see how is the performance of these microparticles with respect to in vitro release profiles, mucus retention time in vitro and PK/PD study. In addition, the impact of scaling the process up and the stability of the particles need to be assessed in the future. Acknowledgement The authors would like to acknowledge the support from the National Nature Science Foundation of China (No. 81302720) and Liaoning Provincial Education officer’s Excellent Talents Supporting Plan. References 1. C. Jacobs, R.H. Muller, Production and characterization of a budesonide nanosuspension for pulmonary administration, Pharm Res, 19 (2002) 189-194. 2. S. Britland, W. Finter, H. Chrystyn, D. Eagland, M.E. Abdelrahim, Droplet aerodynamics, cellular uptake, and efficacy of a nebulizable corticosteroid nanosuspension are superior to a micronized dosage form, Biotechnology progress, 28 (2012) 1152-1159. 3. N. El-Gendy, E.M. Gorman, E.J. Munson, C. Berkland, Budesonide nanoparticle agglomerates as dry powder aerosols with rapid dissolution, J Pharm Sci, 98 (2009) 2731-2746. 4. D. Pliszczak, S. Bourgeois, C. Bordes, J.P. Valour, M.A. Mazoyer, A.M. Orecchioni, E. Nakache, P. Lanteri, Improvement of an encapsulation process for the preparation of pro- and prebiotics-loaded bioadhesive microparticles by using experimental design, Euro J Pharm Sci, 44 (2011) 83-92.

271


Drug Delivery to the Lungs 27, 2016 – L. Slator et al. Assessment of Delivered Dose Rate from 2 Mesh Nebulizers tested with an Adult Breathing Pattern and 2 Drugs L. Slator, J. Parker, L. Hardaker & R. Potter Respironics Respiratory Drug Delivery (UK) Ltd, a business of Philips Electronics UK Limited, Chichester, West Sussex, UK Summary Rapid delivery of inhaled treatment is a desirable attribute of most aerosol inhalation devices. We aimed to compare the delivered dose output rate of production equivalent InnoSpire Go nebulizers with an existing mesh nebulizer using 2 drugs at their respective fill volumes with an adult breathing pattern, and to establish typical treatment times with each volume of drug. Three of each nebulizer brand (InnoSpire Go, and MicroAir U22) were tested in triplicate delivering 2.5 mL of 5 mg/2.5 mL salbutamol sulphate and 2 mL of 0.5 mg/2 mL ipratropium bromide to a filter with a breathing simulator set with an adult breathing pattern (500 mL tidal volume, 1:1 inhalation:exhalation ratio, 15 breaths per minute). At treatment completion, elutions from filters and nebulizers were analyzed by high performance liquid chromatography. Delivered dose output rates to filter for the InnoSpire Go were 0.511 mg/min and 0.063 mg/min for salbutamol sulphate and ipratropium bromide, respectively, and 0.256 and 0.038 mL/min, respectively, for the MicroAir U22. Residual volumes were 0.02 mL and 0.01 mL, respectively, for the InnoSpire Go nebulizer, compared with 0.27 and 0.24 mL, respectively, for the MicroAir U22 nebulizer. Treatment times were 4.7 minutes and 3.7 minutes, respectively, for the InnoSpire Go nebulizer and 7.0 and 4.6 minutes, respectively, for the MicroAir U22. The dose delivery rates, residual volumes, output to filters, and treatment times found using the InnoSpire Go nebulizer compared favourably with the other mesh nebulizer. Aerosol particle size output also needs to be assessed to determine overall delivery efficiency. Introduction The InnoSpire Go nebulizer has been developed to rapidly deliver drugs commonly used to treat asthma and chronic obstructive pulmonary disease (COPD). Rapid delivery of inhaled treatment is a desirable attribute of most aerosol inhalation devices, and as mesh nebulizer design is improved it is expected that the time taken to deliver an aerosol treatment will reduce. In this study, we aimed to compare the delivered dose output rate and treatment time of production equivalent InnoSpire Go mesh nebulizers with those of an existing model of mesh nebulizer currently used to deliver asthma and COPD drugs to patients. The nebulizers were tested with a simulated adult breathing pattern and 2 different drugs with different fill volumes. Methods Prior to testing, the nebulizers (Table 1) were disassembled, washed in warm soapy water, rinsed, and air-dried. All equipment and solutions/reagents were stabilized to ambient laboratory conditions at least 2 hours before use (22-24 °C, 1000-1028 mbar atmospheric pressure, and 47-55% relative humidity). Table 1. The nebulizers and fill volumes of drugs tested.

Fill volume of salbutamol sulphate (Salamol; IVAX Pharmaceuticals, Castleford, UK) Fill volume of ipratropium bromide (Atrovent; Boehringer Ingelheim GmbH, Ingelheim, Germany)

InnoSpire Go (production equivalent; Respironics Respiratory Drug Delivery [UK] Ltd, Chichester, UK)

MicroAir U22 (Omron Healthcare, Kyoto, Japan)

2.5 mL of 5 mg/2.5 mL

2.5 mL of 5 mg/2.5 mL

2 mL of 0.5 mg/2 mL

2 mL of 0.5 mg/2 mL

An InnoSpire Go nebulizer was weighed before and after being filled with approximately 2.5 mL of salbutamol sulphate from the respule. The nebulizer was connected to a Harvard respirator (Harvard Apparatus, Harvard College, Holliston, MA) via a filter (3M, St. Paul, MN) and 22 mm connector (Intersurgical Ltd, Wokingham, UK), all housed within a fume cabinet so that aerosol emitted during the exhalation phase was not inhaled by laboratory personnel. The breathing simulator was set to produce an adult breathing pattern (tidal volume = 500 mL, inhalation:exhalation ratio = 1:1, breaths per minute = 15). The nebulizer mouthpiece was connected to the filter and sealed with parafilm. End of treatment was indicated by audible and visual signals for Innospire Go nebulizer, and by visual assessment of cessation of aerosol exiting the nebulizer during exhalation for the MicroAir U22 nebulizer.

272


Drug Delivery to the Lungs 27, 2016 - Assessment of Delivered Dose Rate from 2 Mesh Nebulizers tested with an Adult Breathing Pattern and 2 Drugs Treatment time was assessed using a timer, at the end of treatment the nebulizer and filters were eluted with suitable volumes of assay diluent and analyzed by HPLC. The nebulizer was washed in warm soapy water, rinsed, and air-dried. The test was performed in triplicate with the nebulizer, and then performed in triplicate with the other 2 InnoSpire Go nebulizers. The test was repeated in triplicate with the 3 InnoSpire Go nebulizers with ipratropium bromide, and then with the MicroAir U22 nebulizers with both drugs. The results were reported as; 

delivered dose output rate (total drug deposited onto the inhalation filters divided by the treatment time in minutes),

residual volume (volume left in the medication cup after treatment completion),

total output to filter (amount of drug deposited onto the filters),

nebulizer deposition (amount of drug recovered from the nebulizer [excluding the medication cup] after treatment completion),

treatment time.

Results

Figure 1 - Mean delivered dose output rate for each nebulizer when tested with 2.5 mL salbutamol sulphate ( 2 mL ipratropium bromide ( ). Error bars show 1 standard deviation about the mean.

) and

Table 2. Mean total output to filter, residual volume, and nebulizer deposition for each of the nebulizer and drug combinations, shown with ±1 standard deviation about the mean.

Total output to filter (mg) Residual volume (mL) Nebulizer deposition (mg)

InnoSpire Go (production equivalent) 2.5 mL 2 mL salbutamol ipratropium sulphate bromide 2.370 0.234 (± 0.061) (± 0.005) 0.016 0.012 (±0.005) (±0.004) 0.694 0.068 (± 0.078) (± 0.008)

273

MicroAir U22 2.5 mL salbutamol sulphate 1.694 (± 0.096) 0.270 (±0.027) 0.402 (± 0.076)

2 mL ipratropium bromide 0.171 (± 0.008) 0.240 (±0.016) 0.046 (± 0.013)


Drug Delivery to the Lungs 27, 2016 – L. Slator et al.

Figure 2 - Mean treatment time for each nebulizer when tested with 2.5 mL salbutamol sulphate ( ipratropium bromide ( ). Error bars show 1 standard deviation about the mean.

) and 2 mL

The fastest delivered dose output rate for both drugs (Figure 1) and lowest residual volume (Table 2) were found with the InnoSpire Go nebulizer. Delivered dose output rate was 100% faster for the InnoSpire Go nebulizer compared with the MicroAir U22 nebulizer when tested with salbutamol sulphate and 65% when tested with ipratropium bromide (Figure 1). Consequently treatment time was shortest for the InnoSpire Go nebulizer (Figure 2). Total output to filter was higher for the InnoSpire Go nebulizer, but combined residual volume and nebulizer deposition was lower for the InnoSpire Go nebulizer (Table 2). Discussion The InnoSpire Go nebulizer produced the fastest delivered dose output rates, lowest residual volume, shortest treatment times, and the largest total output to filter. These data indicate that the InnoSpire Go nebulizer can deliver these drugs rapidly. Short treatment times may be a deciding factor for patients selecting an inhalation [1,2] the treatment times from these mesh nebulizers were not as long as those found with conventional jet device, [3,4] and continued improvement in mesh devices by reducing treatment time may reduce the nebulizers, perception of treatment time as a negative factor when comparing treatment via nebulizer with that of an inhaler device. Although rapid and efficient dose delivery are desirable characteristics of aerosol treatment the overall efficiency of a nebulizer cannot be assessed without reference to the particle size distribution of the aerosol [5] output, as this determines the efficiency of delivery into the lungs, and is approximated by the respirable dose. Conclusions When tested for the rate of dose delivery into simulated adult breathing using two solution based drugs the performance of the InnoSpire Go nebulizer compared favourably with another popular model of mesh nebulizer, with a rapid dose delivery rate, low residual, high output to filter, and short treatment time. In order to determine the overall aerosol delivery efficiency of the InnoSpire Go nebulizer further tests need to be conducted into aerosol particle size output to determine the respirable dose output rate. Acknowledgements The authors would like to acknowledge S. Cowley and S. Hinton (Respironics Respiratory Drug Delivery [UK] Ltd, a business of Philips Electronics UK Limited, Chichester, West Sussex, UK) for preparation of the samples.

274


Drug Delivery to the Lungs 27, 2016 - Assessment of Delivered Dose Rate from 2 Mesh Nebulizers tested with an Adult Breathing Pattern and 2 Drugs

References 1

Elphick M, von Hollen D, Pritchard JN, Nikander K, Hardaker LEA, Hatley RHM: Factors to consider when selecting a nebulizer for a new inhaled drug product development program, Expert Opin Drug Deliv 2015; 12:8; pp1375-1387.

2

Dolovich MB, Ahrens RC, Hess DR, Anderson P, Dhand R, Rau JL, Smaldone GC, Guyatt G: Device selection and outcomes of aerosol therapy: Evidence-based guidelines, Chest 2005; 127:1; pp335-371.

3

Loffert DT, Ikle D, Nelson HS: A comparison of commercial jet nebulizers, Chest 1994; 106:6; pp1788-1792.

4

Ari A: Jet, ultrasonic, and mesh nebulizers: an evaluation of nebulizers for better clinical outcomes, Eurasian J Pulmonol 2014; 16:1; pp1-7.

5

Arzu A, Hess D, Myers TR, Rau JL. A guide to aerosol delivery devices for respiratory therapists; 2009. Available from: www.aarc.org/education/aerosol devices

275


Drug Delivery to the Lungs 27, 2016 - M. Abadelah et al. ®

Effect of inhalation manoeuvre parameters on the dose emitted from Onbrez Breezhaler using inhalation profiles of patients with chronic obstructive pulmonary disease (COPD) 1

2

M. Abadelah , H. Chrystyn , H.Larhrib 1

1

Department of Pharmacy and Pharmaceutical Science, University of Huddersfield, Queensgate, Huddersfield, HD1 3DH, UK 2

Inhalation Consultancy Ltd, Yeadon, Leeds, LS19 7SP, United Kingdom

Summary Dry powder inhalers (DPIs) are commonly prescribed for asthma and chronic obstructive pulmonary disease (COPD). During each inhalation manoeuvre, the metered dose in a DPI needs to be de-aggregated to ensure that emitted dose contains drug particles with the potential for lung deposition. Three important inhalation profile parameters maximum inhalation flow (MIF), acceleration rate (ACIM), inhalation volume (Vin) were identified to [1] affect dose emission and drug deposition from DPIs . However, which of these three parameters is the most important it is still unclear. In this work, the COPD patient inhalation profiles were modified to change only one variable at a time from the above mentioned inhalation parameters, whilst maintaining the two other parameters constant. 9 inhalation profiles were used. They were classified into three groups. The inhalation profiles were replayed using a breath simulator (BRS) connected to an Andersen Cascade Impactor (ACI) via a mixing inlet providing supplementary air ® to identify the aerodynamic characteristics of the emitted dose from an indacaterol 150 µg Onbrez Breezhaler . The Total emitted dose (TED), fine particle dose (FPD) and residual amount (RA) were all significantly (p<0.05) affected by each of the inhalation parameters (ACIM, MIF, and Vin) but the ACIM has less significant effect (p>0.05) on the mass median aerodynamic diameter (MMAD) than the MIF. The study illustrates the importance of identifying the effect of the main inspiratory parameters on the aerodynamic characteristics of indacaterol ® 150µg dose emission from Onbrez Breezhaler using altered COPD patient inhalation profiles. Introduction The pulmonary route of drug administration is considered and widely known as the best route of drug [2] administration for the treatment of asthma and chronic obstructive pulmonary disease (COPD) . The dose emitted from the DPIs is generated as a result of the inhalation manoeuvre, which generates a pressure drop in the metering cup and the inhalation channel of each device; the pressure drop is proportional to the patient’s [3] inspiratory flow and the intrinsic resistance of the inhaler device . The difference in the pressure generates turbulence energy, which leads to the dispersion of the powder formulation resulting in small drug particles with a greater tendency for lower airways deposition. There are different inspiratory parameters that could affect the dose emission during each inhalation. These are the patient’s maximum inspiratory flow (MIF), inhaled volume [1, 2] . The patient inspiratory flow rates effect (Vin), acceleration rate (ACIM) and the degree of airway obstruction on the dose emission and lung deposition and the consistency of delivered dose from DPIs have been shown [4, 5] . However, the study of each inhalation manoeuvre especially for those devices with high intrinsic resistance parameter separately will illustrate the effect of each parameter on dose emission and drug deposition. The aim of the study was to investigate the effect of each inhalation manoeuvre parameter ex-vivo when the other two parameters remained fixed. Therefore altered inhalation profiles were generated and used to evaluate the ® impact of each inhalation parameter on the dose emitted from 150µg indacaterol Onbrez Breezhaler under inhalation conditions that mimic patient use. Materials and method Simulated COPD patient’s inhalation profiles The inhalation profiles of 38 patients with different COPD severity aged 55-79 with a mean age of 66 were ® measured. The profiles were recorded using an empty (Placebo) Onbrez Breezhaler inhaler device, it is a low 1/2 resistance device with an intrinsic resistance of 0.063 [CmH2O / (L/min)], and the patients were able to generate an average maximum inhalation flow rate of 88 L/min. The patients were given the PIL to read and trained to [5] inhale as hard and fast as they can from the start of the inhalation manoeuvre . The inspiratory parameters (e.g., MIF, ACIM and Vin) were measured for each profile; the maximum flow rate (MIF) for each profile is the highest flow the patient can achieve from the start to the finish of the inhalation time. The inhalation volume (Vin) was calculated from the area under the curve of the inhalation flow against the time profile. The acceleration rate (ACIM) was determined by firstly calculating the slope of the flow values against the 2 inhalation time, then the value with a high linear regression (r ) was chosen. In the case of Onbrez Breezhaler the 2 linear regression was almost r =1 for all the values, therefore acceleration rate was determined as the values with high slope.

276


Drug Delivery to the Lungs 27, 2016 - Effect of inhalation manoeuvre parameters on the dose emitted from ÂŽ Onbrez Breezhaler using inhalation profiles of patients with chronic obstructive pulmonary disease (COPD) The inhalation profiles with a MIF close to 40, 65 and 85 L/min were chosen. For group one, the MIF of 40L/min 2 and Vin of 2L were kept constant, while ACIM was modified to obtain 2, 4 and 8 L/s . For group 2, the MIFs of 40, 65 and 85 L/min were generated by patients. However, for group three, the Vin was altered to achieve 1, 2 and 3L inhaled volume, whilst maintaining the ACIM and the MIF constant. The inhalation time was either increased or decreased to achieve the desired Vin. The ACIM was modified by increasing the steepness of the slope, whereas the MIF was the one generated by the patient. The 9 Inhalation profiles used for this study shown in figure 1. The profiles were modified as explained above to fix two inspiratory parameters to study one variable at a time. The altered profiles were classified according to the inspiratory factor that will be investigated into; group 1= Acceleration (ACIM) study, group 2= Maximum inhalation flow (MIF) study, group 3= inhalation volume (Vin) study.

100 Flow rate( L/min)

Flow rate (L/min)

50

80

40 30

60

20

40

10

20

0 0

2

Time (s)

4

0

6

1 Time(s) 2

0

MIF=40 L/min, Vin=2 L, ACIM =2 L/S2

3

MIF=40 L/min, Vin =1 L, ACIM =2 L/S2

MIF=40 L/min, Vin =2 L, ACIM =4 L/S2

MIF=65 L/min, Vin =1 L, ACIM =2 L/S2

MIF=40 L/min, Vin =2 L, ACIM =8 L/S2

MIF=85 L/min, Vin =1 L, ACIM =2 L/S2

MIF= 65 L/min, Vin = 1 L, ACIM = 4 L/S2 MIF= 65 L/min, Vin = 2 L, ACIM = 4 L/S2 MIF= 65 L/min, Vin = 3 L, ACIM = 4 L/S2

Figure 1: Three Groups of inhalation profiles classified to study one inhalation parameters at a time, Group 1: ACIM study, Group 2= MIF study, Group 3=Vin study Experimental procedure to measure the aerodynamic dose emission characteristics of Onbrez ÂŽ Breezhaler A schematic diagram is shown in figure 2. The experimental set-up was changed according to the profiles that were investigated such that for profiles less than 60L/min, the ACI with the Alberta throat was set at 60 L/min and for profiles above 60 L/min; the ACI was set at 90 L/min. The ACI was set-up with the valve of supplementary air through the mixing inlet closed in order to set the flow using the vacuum pump and flow controller either at 60 or 90 L/min. The supplementary air was then introduced through the side arm of the mixing inlet by opening the valve connected to a compressed air cylinder and this flow was adjusted to achieve (0 L/min) through the mouth piece. The flow through the Andersen cascade impactor was maintained at either 60 or 90 L/min depending on the (MIF) of the profile that was to be used. All the settings were maintained for each determination. Three Capsules were aerosolised into the ACI for each profile (n=3). The residual amount (RA) retained in the capsule and device were recovered. The amounts of drug deposited in each stage, Alberta throat and mixing inlet were collected using washing solution of the mobile phase, then quantified using a validated High Performance Liquid Chromatography (HPLC) method. The aerodynamic characteristics of the emitted dose were calculated using the Copley Inhaler Testing Data Analysis Software (CITDAS version 2.0)

277


Drug Delivery to the Lungs 27, 2016 - M. Abadelah et al.

Figure 2: Schematic diagram of ACI methodology with Mixing Inlet using breath simulator

Result and discussion A summary of the mean (SD) of aerodynamic characteristics of indacaterol 150 µg from Onbrez Breezhaler® for Group 1 ACIM study, Group 2 MIF study ,Group 3 Vin study inhalation profiles is shown in Tables 1,2 and 3. The results of Group 1 showed that the increase in the acceleration rate has a significant effect (p<0.05) on the fine particle dose (FPD) and the residual amount (RA) left in capsule and device, but it has no significant effect on the MMAD. It has been suggested that the acceleration rate during an inhalation is important in the generation of [6] the FPD as shown in the present work. The increase in the FPD is more probably attributed to the increase in the TED. Group 2 results showed that TED, FPD, and MMAD were all affected by the MIF. Hence there is a flow rate dependent dose emission. Gamma scintigraphy has demonstrated that the total lung deposition, for some [7] and it has been shown that lung deposition for some DPIs generally DPIs, can vary with inhalation flow [8] increased with the inhalation flow rate . This is consistent with the higher FPD and lower MMAD with increasing the MIF identified in this work. When a patient inhales through a DPI containing doses stored inside the device (either in a reservoir or as single-dose blisters), de-aggregation of the powder formulation occurs almost [9] immediately as the dose leaves the device , however, capsule based devices would require a prolonged inhalation to empty the dose. This was demonstrated in the present work (Group 3, Table 3). The most significant increase in TED occurred between the 2 L and 3 L inhaled volume. Prolonged inhalation would provide the drug with a longer time of flight to maximize drug deposition as shown from the increase in the FPD. Table1: Group 1; mean (SD) aerodynamic characteristics of emitted dose of Indacaterol at three different ACIM 2, 2 4 and 8 L/s at fixed MIF (40 L/min) and Vin (2 L) Group 2; mean (SD) aerodynamic characteristics of emitted 2 dose of Indacaterol at three different MIFs 40, 65 and 85 L/min at fixed ACIM (2 L/s ) and Vin (1 L) Group 1 a). MIF=40 L/min, VIN=2 L, 2 ACIM =2 L/S TED (µg) FPD (µg) MMAD (µm) Residual amount (µg) b). MIF=40 L/min, VIN=2 L, 2 ACIM =4 L/S TED (µg) FPD (µg) MMAD (µm) Residual amount (µg) C). MIF=40 L/min, VIN=2 L, 2 ACIM =8 L/S TED (µg) FPD (µg) MMAD (µm) Residual amount (µg)

Indacaterol ( nominal dose :150µg) 114.66 (1.17) 34.27 (1.76) 3.35 (0.06) 26.87 (1.75) 116.09 (4.55) 35.80 (1.92) 3.30 (0.1) 22.07(1.00) 122.11 (2.91) 39.30 (0.50) 3.35 (0.06) 18.91 (0.36)

Group 2 a). MIF=40 L/min, VIN=1 L, 2 ACIM =2 L/S TED (µg) FPD (µg) MMAD (µm) Residual amount (µg) b). MIF=65 L/min, VIN=1 L, 2 ACIM =2 L/S TED (µg) FPD (µg) MMAD (µm) Residual amount (µg) C). MIF=85 L/min, VIN=1 L, 2 ACIM =2 L/S TED (µg) FPD (µg) MMAD (µm) Residual amount (µg)

278

Indacaterol ( nominal dose :150µg) 113.46 (2.83) 32.43 (0.68) 3.45 (0.12) 24.83 (3.36) 115.87 (5.35) 37.93 (1.32) 2.90 (0.1) 21.94 (0.90) 118.77 (1.71) 40.70 (1.05) 2.80 (0.06) 16.89 (3.03)


Drug Delivery to the Lungs 27, 2016 - Effect of inhalation manoeuvre parameters on the dose emitted from ® Onbrez Breezhaler using inhalation profiles of patients with chronic obstructive pulmonary disease (COPD) Table 2: Group 3; mean (SD) aerodynamic characteristics of emitted dose of Indacaterol at three different Vins 1, 2 2 and 3 L at fixed MIF (65 L/min) and ACIM (4 L/s ) Group 3 2

a). MIF= 65 L/min, VIN= 1 L, ACIM = 4 L/S TED (µg) FPD (µg) MMAD (µm) Residual amount (µg) 2 b). MIF= 65 L/min, VIN= 2 L, ACIM = 4 L/S TED (µg) FPD (µg) MMAD (µm) Residual amount (µg) 2 C). MIF= 65 L/min, VIN= 3 L, ACIM = 4 L/S TED (µg) FPD (µg) MMAD (µm) Residual amount (µg)

Indacaterol ( nominal dose :150µg) 115.95 (1.43) 38.13 (0.60) 2.90 (0.06) 24.26 (1.72) 119.03 (1.95) 39.90 (0.53) 2.85 (0.06) 20.83 (3.26) 122.82 (3.77) 41.73 (0.80) 2.80 (0.06) 17.24 (1.12)

Conclusion The use of altered COPD patient inhalation profiles provided a clear view on the impact of each inhalation parameter on the aerodynamic dose emission characteristics of Indacaterol. The TED, FPD, and RA were all affected by each of the inhalation parameters (ACIM, MIF, and Vin). The ACIM has less effect on the MMAD than the MIF. Slow inhalation resulted in the formation of drug particles with the largest MMAD. An ongoing study on the combination of inhalation parameters looks promising to maximising drug emission from indacaterol Onbrez Breezhaler. The results highlight how standard compendial dose emission methods can be adapted to replace the traditional vacuum pump with inhalation profiles collected from patients. These provide an insight into the real life dose the patient would receive. Reference 1.

Laube, B.L., Janssens, H.M., de Jongh, F.H., Devadason, S.G., Dhand, R., Diot, P., Everard, M.L., Horvath, I., Navalesi, P., Voshaar, T: What the pulmonary specialist should know about the new inhalation therapies. Eur. Respir. J 2011. 37, 1308-1417.

2.

Price, D., & Chrystyn, H: Concept review of dry powder inhalers: correct interpretation of published data. Multidisciplinary respiratory medicine 2015, 10 (1), 1.

3.

Clark, A. R., & Hollingworth, A. M: The relationship between powder inhaler resistance and peak inspiratory conditions in healthy volunteers—implications for in vitro testing. J. aerosol. Med 1993, 6(2), 99-110.

4.

Pedersen, S. O. R. G., Hansen, O. R., & Fuglsang, G : Influence of inspiratory flow rate upon the effect of a Turbuhaler. Archives of disease in childhood 1990, 65(3), 308-310.

5.

Pavkov, R., Mueller, S., Fiebich, K., Singh, D., Stowasser, F., Pignatelli, G., Walter, B., Ziegler, D., Dalvi, M., Dederichs, J: Characteristics of a capsule based dry powder inhaler for the delivery of indacaterol. Curr. Med. Res. Opin 2010. 26, 2527-2533.

6.

Newman, S.P., Morén, F., Trofast, E., Talaee, N., Clarke, S.W: Terbutaline sulphate Turbuhaler: effect of inhaled flow rate on drug deposition and efficacy .Int. J. Pharm 1991. 74, 209–213.

7.

Everard, M.L., Devadason, S.G., Souëf, P.N.L: Flow early in the inspiratory manoeuvre affects the aerosol particle size distribution from a Turbuhaler. Respir. Med 1997. 91, 624–628.

8.

Malmberg, L. P., Everard, M. L., Haikarainen, J., & Lähelmä, S: Evaluation of in vitro and in vivo flow rate dependency of budesonide/formoterol Easyhaler®. J.aerosol .Med. pulmon. drug delivery 2014, 27(5), 329-340. Haughney, J., Price, D., Barnes, N.C., Virchow, J.C., Roche, N., Chrystyn, H: Choosing inhaler devices for people with asthma: current knowledge and outstanding research needs. Respir. Med. CME 2010. 3, 125-131.

9.

279


Drug Delivery to the Lungs 27, 2016 – Amy Worle et al. Development of dry powder formulations combining both a biologic therapeutic entity and a small molecule drug substance Amy Worle & Harriet Bridgwater Vectura Group plc, Vectura House, Bumpers Way, Chippenham, Wiltshire, SN14 6FH Summary An inhalation product combining a biological therapeutic entity with a small molecule drug substance is still a very novel concept in inhalation. A combination product would potentially have significant compliance and patient benefits via simplifying and reducing the time of treatment. This study assessed the combination of an immunomodulatory protein (Omalizumab) with a co-prescribed corticosteroid (Fluticasone Propionate). The objective was to produce a biologically, chemically and physically stable powder formulation as evidenced by the stability data presented. Small molecules and biologics are typically formulated and delivered in different ways, with biologics commonly developed as liquid formulations for injection. For this study the formulation was designed for delivery via a dry powder inhaler (DPI). Biologics can be rendered more stable in the solid state via co-formulation with specific excipients therefore a dry powder format can offer improved stability and potentially remove the necessity for refrigerated storage and transport. The development of a combination dry powder formulation presented a number of challenges that were addressed through the combined use of spray drying and low intensity blending techniques. Excipients were identified to provide protein stability and improve aerosol performance. The physical stability (via aerosol performance testing) and biological integrity of the formulation was assessed for 6 months at 30⠰C/65% relative humidity. The study demonstrated that a small molecule and a biologic can be combined successfully to produce a novel model dry powder formulation with good homogeneity and stability.

Introduction Currently, the most investigated approach to combining a biologic and small molecule is antibody drug conjugates (ADCs). ADCs are produced using complex manufacturing methods and can restrict dose ratios [1], therefore manufacturing techniques were identified to simplify and improve the flexibility of a combination formulation whilst maintaining protein stability. A number of small molecule drug products are co-prescribed with a biologic therapy , for example the combination of mucolytics and bronchodilators for the treatment of cystic fibrosis [2]. Current treatment may require timeconsuming nebuliser regimens or parenteral administration, therefore reducing the need for these will simplify treatment and potentially improve patients’ quality of life. Providing more than one method of treatment in a single device also has the potential to increase patient compliance[3]. Biologics are generally more stable in the solid state[4] therefore biomolecules in a dry powder format can offer improved stability. As well as increasing the shelf life, the dry powder format may also remove the requirement for refrigerated storage and transport. There is also the potential for therapeutic benefit in applying the molecule directly to the site of action as opposed to delivering systemically. Manufacturing a combination formulation The aim of this work was to produce a proof-of-concept formulation combining a biologic with a co-prescribed small molecule. The study assessed the combination of a monoclonal antibody (Omalizumab) with a corticosteroid (Fluticasone Propionate), both prescribed for the treatment of severe asthma. These molecules were selected as models to demonstrate capability as they pose the additional challenge of cold chain storage of Omalizumab and low dosage of FP. The small molecule and biologic were initially formulated separately with suitable excipients to produce optimum pre-blend formulations independently. Common blending excipients include sugars such as lactose or force control agents (FCAs) such as magnesium stearate. Reducing sugars may not be compatible with proteins due to the Maillard reaction therefore a taurine carrier system was selected over lactose for this study. The mono formulations were developed to improve aerosol performance, particle size, homogeneity and protein stability. Producing separate pre-blends and coating the particles with excipients reduces the interactions between the two drugs which may lead to variability in aerosol performance and differences between the mono and combination blends [5]. API concentrations of the final blend were based on delivering standard doses of each drug using a dry powder inhaler.

280


Drug Delivery to the Lungs 27, 2016 - Development of dry powder formulations combining both a biologic therapeutic entity and a small molecule drug substance The objective was to produce stable combination formulations as evidenced by 6 months stability data. Physicalchemical analysis, protein stability and aerosol testing were performed over 6 month’s stability and results were compared to the mono biologic formulation.

Materials and Methods Spray drying and LabRAM acoustic blending Spray drying, mechano chemical bonding (MCB) and LabRAM ResonantAcoustic® Mixer (RAM) techniques were selected for rapid manufacture of the combination product. Spray drying was used to produce the initial biologic ‘pre-blend’; combining the biologic with a stabilising excipient and FCA. Spray drying provides control over the powder particle size, morphology and density via adjustment of process parameters. The FP particles were coated with FCA in ‘pre-blend’ 2 using a mechano chemical bonding method. The LabRAM acoustic blending technique was chosen for combining the spray dried pre-blend with the small molecule pre-blend as it is an efficient low shear mixing process which reduces the stress applied to the formulation and ensures homogeneity [6]. Spray dried particles are particularly friable and may fracture, therefore this low shear process can reduce particulate damage. Three ‘pre-blends’ were manufactured using the components and manufacturing methods detailed in Table 1. The ‘pre-blends’ were not assessed for homogeneity as the processes used are considered intrinsically uniform. These pre-blends were subsequently combined using the LabRAM to produce a Omalizumab / FP combination formulation. Table 1: Pre-blend components of the Omalizumab / Fluticasone propionate combination formulation

Pre-blend # 1 2

Material

Blending method

Omalizumab / stabilising excipient / FCA Fluticasone Propionate / FCA

Spray drying – bespoke spray dryer Mechano chemical bonding (MCB) Spray drying – bespoke spray dryer

3

Taurine / FCA

%w/w in combination formulation

PSD (D50)

32

2.6

2

1.8

66

1.7

Analytical requirements In general, analysis of a biologic entity requires additional complexity compared to small molecule analytical methods therefore a combination of both requires method assessment and development [7]. SDS-PAGE, size exclusion and activity assay methods were assessed and deemed acceptable to analyse the protein in the presence of the small molecule. Additionally particle size, water content and glass transition temperature (T g) were also measured. Content uniformity was analysed by RP-HPLC, using a separate method for each API due to diluent solubility incompatibility. A fast screen impactor (FSI) was used to assess aerodynamic performance using a Vectura unit dose inhaler at a 60L/min flow rate and a 5µm insert.

Results and Discussion Process Development Initial process development investigated the effect of LabRAM blending intensity on the homogeneity of the combination formulation at a 1g scale. Content uniformity results (figure 1) show optimum assay results for both molecules were achieved at the highest intensity (80%). However the Omalizumab %RSD result remains high suggesting wall losses during the blending process. Adjustments were put in place to reduce this issue including increasing batch size, increasing the particle size of the carrier and assessing the effect of order of addition in the final blending step. The main improvement in homogeneity was identified when the final blend was increased from a 1.0g to a 5.0g scale. This resulted in an Omalizumab content of 99% of nominal and 1.5% RSD in the final formulation (see table 2, initial data).

281


50

80

40

60

30

40

20

20

10

0

0

% Content (solid lines)

100

20

35

50

65

80

%RSD (dashed lines)

Drug Delivery to the Lungs 27, 2016 – Amy Worle et al.

LabRAM % Intensity FP (%content)

Omalizumab (%content)

Figure 1: Process optimisation – content homogeneity assessment at a range of LabRAM blending intensities (mean of 5 content uniformity replicates)

Stability assessment Following development a final 5g combination blend was manufactured and set down for a 6 month stability assessment at 30˚C/65%RH. Figure 2 shows an SEM image of the formulation placed on stability.

Aerosol performance results Aerosol resultsblister (figureevacuation 3) show consistent blister evacuation and FP %FPF across time points. FP(table content (figure 3)performance show consistent and %FPF across time points. content uniformity results 2) uniformity resultsthroughout (table 2) are stable throughoutcontent the study. Omalizumab uniformity results are however are also stable thealso study. Omalizumab uniformity resultscontent are however inconsistent which may inconsistent which may be due to protein blister retention. Size exclusion data (table 2) shows no however clear trends be due to protein blister retention. Size exclusion aggregation dataaggregation (table 2) shows no clear trends an however additional degradation peak is at observed at thetime 2 week andan shows an increase at 24 weeks. additionalandegradation peak is observed the 2 week pointtime andpoint shows increase at 24 weeks.

Evacuation / %FPF of nominal

2: SEM image of/ FP the combination final Omalizumab/FP combination blend. re 2: SEM imageFigure of the final Omalizumab blend.

100 90 80 70 60 50 40 30 20 10 0

0

2

4

12

24

Time point (weeks) %FPF

Blister clearance

Figure 3: Gravimetric FSI stability results- evacuation and percentage fine particle fraction (%FPF)

282


Drug Delivery to the Lungs 27, 2016 - Development of dry powder formulations combining both a biologic therapeutic entity and a small molecule drug substance Table 2: Physical properties, content uniformity and size exclusion stability results of the combination blend

Test Particle size (D90, µm) Moisture content (%w/w) Glass transition temperature (Tg, ˚C) Content Uniformity: Omalizumab %w/w of nominal %RSD Content Uniformity: FP %w/w of nominal %RSD Size exclusion Aggregation (% of total peak area) Degradation (% of total peak area) *No result generated due to methodology reasons

0 4.7 1.2 81

2 5.0 1.0 80

99 1.5

*

102 5.4 0.2 0.0

Time point (weeks) 4 5.0 2.2 79

12 4.8 1.0 79

24 4.6 1.1 77

88 5.8

91 2.3

83 4.7

*

96 4.4

97 1.2

97 2.4

0.0 0.3

0.3 0.0

0.2 0.0

0.1 1.4

Overall, results demonstrate physical properties, aerodynamic performance and protein stability data (table 2) are within the project assigned acceptable limits for up to 6 months stability at 30˚C/65%RH. Stability results of the combination formulation are comparable to the mono Omalizumab formulation (data not presented), supporting the concept of a combination product. Following the successful combination formulation, the manufacturing methods were applied to an alternative model combination; an enzyme and formoterol. These results are not presented in this summary however the study also showed positive results following a 3 month stability study at accelerated storage conditions (40˚C/75%RH).

Conclusion The overall project objectives were met; a small molecule and biologic were successfully combined to produce a homogenous dry powder inhalable product. While a number of potential challenges around mixing uniformity and analytical assessment were encountered, positive stability trends were produced following storage for 6 months at 30˚C/65%RH. Although further development of the formulation may be possible, the study demonstrates feasibility for the use of spray drying and LabRAM blending technology to produce a DPI combination product of both small molecule and biologic drug substance.

1

Bakhtiar R: Antibody drug conjugates. Biotechnol Lett 2016 Oct; 38(10): 1655-1664

2 Konstan M, VanDevanter D, Rasouliyan L, Pasta D, Yegin A, Morgan W, Wagener J: Trends in the use of routine therapies in cystic fibrosis: 1995 – 2005, Pediatr Pulmonol. 2010 Dec; 45(12): 1167-1172 3 Nair A, Clearie K, Menzies D, Meldrum K, McFarlane L, Lipworth B J: A novel breath-actuated integrated vortex spacer device increases relative lung bioavailability of fluticasone/salmeterol in combination, Pulm Pharmacol Ther. 2009 Aug; 22(4): 305-310

Parry M, Solomon D, Hammond M, Ward D, Wake A: Formulation of biologics for inhaled and nasal delivery. (Abstract). Presented at: DDL26 Drug Delivery to the Lungs 26, Edinburgh, Scotland, December 9-11, 2015

4

Wang W, Nema S, Teagarden D: Protein aggregation—Pathways and influencing factors. International Journal of Pharmaceutics. 2010 May; 390(2): 89–99

5

6 Childerhouse N, Green, M: Acoustic mixing – Highly efficient new technology for dry powder inhalers. (Poster). Presented at: International Society of Aerosols in Medicine (ISAM), Munich, Germany, May 31 – June 3, 2015 7 Western K, Childerhouse N, Munro S, Sefton S: The application of biological product analytical methodologies to support the development of biologic dry powder inhalation products. (Abstract). Presented at Presented at: DDL26 Drug Delivery to the Lungs 26, Edinburgh, Scotland, December 9-11, 2015

283


Drug Delivery to the Lungs 27, 2016 – Wachirun Terakosolphan et al. Does glycerol interact with dipalmitoylphosphatidylcholine membranes? 1

1

2

Wachirun Terakosolphan , Precious Akhuemokhan , Richard Harvey & Ben Forbes

1

1

2

Institute of Pharmaceutical Sciences, King’s College London, London, SE1 9NH, UK Institute of Pharmacy, Martin-Luther-University Halle-Wittenberg, Halle (Saale), 06099, Germany

Summary In some solution-based pressurised metered dose inhalers (pMDIs), glycerol is used as a non-volatile excipient in order to adjust the aerosol droplet size into the micron range. After actuation, the volatile components of the pMDI formulation evaporate, leaving the drug and glycerol to form the respirable particles. Although the purpose of glycerol is to modify the aerosol particle size distribution, evidence is emerging that it may affect drug kinetics after the particles deposit on the surface of the respiratory tract. Liposomes or monolayers formed from dipalmitoylphosphatidylcholine (DPPC) can be employed as simple models of the epithelial cell membrane, and are used in many pharmaceutical studies to examine the drug absorption or membrane interaction. DPPC also predominates the composition of lung surfactant which coats the mucosa at the glycerol-containing drug particles deposition site. The purpose of this investigation was to conduct a preliminary screen to determine whether glycerol interacts with biomimetic DPPC models as a precursor to more detailed studies into mechanisms that may explain differences in drug permeability in human airways. A fluorimetric assay using laurdan as a fluorescent probe was designed to measure glycerol-DPPC interactions which alter the lipid phase transition temperature. Increasing the concentration of glycerol solution (0, 1.0, 5.0, 10.0, 20.0, and 30.0% w/w) in the DPPC liposome system produced incremental increases in phase transition temperature from 42.6 to 44.2°C. Thus, it can be inferred that glycerol has the potential to modify the molecular rigidity of membrane components which may influence drug permeability in the respiratory tract. Introduction According to the Montreal protocol in 1989, chlorofluorocarbon (CFC) propellants were banned from industrial and [1] household products in over 165 countries throughout the world owing to environmental concerns . As a result, CFCs in pressurised metered dose inhalers (pMDIs) were replaced with the alternative propellants, hydrofluoroalkanes (HFAs), which are more appropriate because they are less ozone depleting and in common with CFCs are non-flammable. Conversion to HFA-based pMDI, required additives to the formulations for a variety of purposes. For example, ethanol or isopropanol are used as co-solvents in order to enhance drug [2,3] . Several pMDIs were solubility in solution-based formulations due to the inferior solubilizing ability of HFAs formulated to contain non-volatile components such as glycerol or polyethylene glycol to bulk out the drug particle after dose emission in order to produce a final aerosol droplet in a micron size range similar to that of CFC-based [2–4] . After actuation, all of volatile ingredients including propellant and co-solvent suspension formulations evaporate and leave the drug and glycerol to form the aerosol particles. Evidence is emerging that glycerol may [5,6] . affect drug kinetics after the particles deposit on the surface of the respiratory tract Drug dissolution and absorption in the respiratory tract occurs after particle deposition on the mucosa of the airway or alveolar region. The respiratory mucosa is coated in surface active agents, forming lung surfactant, [7,8] . To examine the effect of glycerol on physiological barriers to drug uptake, it is which lines the air space relevant to study its effect upon lung surfactant and the epithelial cell membrane. Liposomes or air/water interface lipid monolayers formed from dipalmitoylphosphatidylcholine (DPPC) constitute alluringly simple biomimetic [9,10] . DPPC is also a models of the cell membrane and as such have been used in many pharmaceutical studies [11,12] . Thus, DPPC was selected in this study as a model for investigating the major component of lung surfactant effect of glycerol at the particle deposition site. The aim of this study was to use a simple method to produce DPPC liposomes with various concentrations of glycerol in solution in order to measure their effect on the lipid phase transition temperature using a fluorimetric assay. This is a preliminary study to establish putative biophysical influences of glycerol on model DPPC systems, as a precursor for more detailed investigations into the molecular mechanisms that may explain differences in drug permeability in human airways. These studies will develop our understanding of biological phenomena observed using in vitro models of the site in respiratory tract on which aerosolised particles from glycerolcontaining pMDIs have been deposited. Materials 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (dipalmitoylphosphatidylcholine; DPPC) was purchased from Avanti Polar Lipids (Birmingham, AL), 6-dodecanoyl-N,N-dimethyl-2-naphthylamine (laurdan) was obtained from Molecular Probes (Eugene, OR), chloroform from Fisher Scientific (UK), and glycerol from Sigma-Aldrich (London, UK). Ultrapure water with 18.0 MΩ·cm residual specific resistance was obtained using an Elgastat Maxima purifier (Elga, UK). All other reagents were obtained from standard sources.

284


Drug Delivery to the Lungs 27, 2016 - Does glycerol interact with dipalmitoylphosphatidylcholine membranes? Experimental methods DPPC liposomes formation Liposomes were formed by dissolving 10 mg of DPPC and adding a small amount of the fluorescent probe laurdan in chloroform solution to give a total lipid/laurdan molar ratio between 300:1 and 400:1. For the remaining steps after this point, the samples were protected from light to limit photo bleaching of the laurdan by covering the vials and equipment with aluminium foil. The chloroform was removed by evaporation at room temperature in a vacuum desiccator to obtain thin dry lipid films which deposited on the inside walls of the vials. The dried lipid films were resolvated using ultrapure water and various concentrations of glycerol solution (1.0 – 30.0% w/w) to make the final concentration of DPPC in the dispersion at 4 mg/mL. The liposomes were ultrasonicated using a Soniprep 150 ultrasonic disintegrator (MSE, UK) at an amplitude of 10 microns for 5 minutes. Fluorimetric measurement Steady-state laurdan fluorescence emission was measured using a Varian Cary Eclipse spectrofluorimeter (Agilent, USA) using a 104F-QS semi-micro dual light path (10 × 4 mm) fluorescence cell (Hellma Analytics, Germany). Fluorescence spectra were recorded over a temperature range from 25°C to 65°C increasing in 5°C increments. The temperature of the sample compartment was adjusted using a water-circulating thermostatcontrolled Varian Cary Single-Cell Peltier Accessory (Agilent, USA). Additionally, to get a precise reading, the temperature inside the cuvette was directly measured with a K-thermocouple thermometer HI 935005 (Hanna, USA). Emission intensities of all samples were acquired with an excitation wavelength of 340 nm, adjusted for optimal spectral intensity with an excitation slit width of 10 nm and an emission slit width of 5 nm, and scanned from 360 nm to 560 nm. Data analysis The generalised polarisation (GP) was calculated using the fluorescence intensities at the maximum emission wavelengths of laurdan exhibited when mixed with lipids in either their gel (Ig) or liquid-crystalline (Ic) phases, in this case 440 and 490 nm are used respectively, according to the following formula.

The generalised polarisation as determined using equation (1) was plotted against temperature. Then, the main phase transition temperatures of each sample were obtained from the interception on the X-axis of the GP versus temperature curves. Results The fluorescence emission spectra shown in figure 1 are from for laurdan in DPPC multilamellar vesicles at different temperatures, from 25°C (gel phase) to 65°C (liquid crystalline phase), measured in ultrapure water. During the phospholipid phase transition, the spectra show a systematic shift to longer wavelengths as temperature increases; the laurdan emission spectrum in the liquid crystalline phase (high temperature) shows a decrease in intensity at longer wavelengths with respect to the spectrum observed in the gel phase (low temperature). Accordingly, the fluorescence intensities at the maximum emission wavelength of laurdan in the gel and liquid crystalline phases of the lipids, in this case 440 and 490 nm, respectively, were used for the calculation of the GP using equation (1).

Figure 1. Fluorescence emission spectra for laurdan in dipalmitoylphosphatidylcholine multilamellar vesicles at different temperatures, measured at excitation wavelength 340 nm at temperatures between 25-65°C, measured in ultrapure water.

285


Drug Delivery to the Lungs 27, 2016 – Wachirun Terakosolphan et al. The temperature dependence of laurdan GP values for DPPC MLVs dispersed in various concentrations of glycerol solutions was determined and presented in Figure 2. As expected, the value of laurdan GP decreased from about 0.5 at temperatures corresponding to gel phase lipid to about -0.35 for the liquid crystalline phase. The melting temperature of the glycerol-free system lies between 40°C and 45°C. In the presence of glycerol, the measurements showed a similarity in GP values in the temperature range at 25°C - 35°C and 45°C - 65°C at all glycerol concentrations. However, an increase in GP values in the glycerol-containing liposomes over the range of the phase transition temperature, approximately at 35°C - 45°C, was detected. This difference in GP measurement, especially at the temperature immediately prior to a steep drop in GP, provides evidence for a glycerol-induced modification of membrane phase behaviour. The estimated phase transition temperatures, which were obtained from the intercept on the X-axis of each curve (the temperature at the GP value = 0), of DPPC MLV in 0%, 1.0%, 5.0%, 10.0%, 20.0%, and 30.0% w/w glycerol solution were 42.6°C, 42.9°C, 43.3°C, 43.6°C, 43.7°C, and 44.2°C, respectively (Figure 3).

Figure 2. Temperature dependence of the effect of glycerol 0-30% w/w on the of laurdan GP values in dipalmitoylphosphatidylcholine multilamellar vesicles.

Figure 3. Phase transition temperature of dipalmitoylphosphatidylcholine multilamellar vesicles containing glycerol 030% w/w. Data were obtained from the temperature at the generalised polarisation value = 0 at each glycerol concentration (Figure 2) and expressed as mean + SD (n=3).

Discussion The phase transition temperature (Tm) of DPPC liposomes was estimated using a fluorimetric technique. Although this biophysical study could not determine directly what happens in the airway when emitted glycerolcontaining drug particles deposit there, the alteration in the Tm demonstrates that a modification of membrane [13] phase properties can be caused by glycerol . The transition temperature, which is a specific parameter of each phospholipid, depends on the length of fatty acyl chain, degree of saturation, charge, and the type of head group [14] . According to the literature, the typical value reported for calorimetric assessment of pure DPPC species [15] vesicles is about 41.5°C . From our observations, the estimated Tm of DPPC alone MLVs is 42.6°C, whereas the presence of glycerol raised the phase transition temperature incrementally from 42.9°C to 44.2°C with increasing glycerol content from 1.0 – 30.0% w/w. Factors which can affect the Tm relate to phospholipid aggregate structure and intermolecular interactions, implying that glycerol interacts at some part(s) of DPPC molecules to interfere with intermolecular forces and thus modify the membrane structure.

286


Drug Delivery to the Lungs 27, 2016 - Does glycerol interact with dipalmitoylphosphatidylcholine membranes? One hypothesis is that glycerol might induce interdigitation of the DPPC bilayers. The interdigitated bilayer is an unusual arrangement of the hydrocarbon chain of phospholipids which exhibits altered physicochemical [16] . In an interdigitated properties including, size, shape, surface behaviour and phase transition temperature DPPC bilayer, the closer association between the hydrocarbon chains of each leaflet might cause greater rigidity, [17] resulting in an increase of the transition temperature . More detailed studies using a wider range of analytical techniques will be required to test this hypothesis. Conclusion The results of this study demonstrate that glycerol, which is currently used as non-volatile agent in some pMDI formulations, interacts with DPPC bilayers. This provides the first evidence towards the hypothesis that the effect of glycerol in the emitted particles from pMDI on drug kinetics observed in vitro may be mediated in part through changes in rigidity of epithelial membranes. We speculate that glycerol may increase rigidity in DPPC liposomes by inducing interdigitation of the bilayers. These observations warrant more extensive studies using a wider spectrum of techniques to investigate the effect of glycerol on DPPC membranes in other aspects, for example, structural interaction, alterations in the physicochemical properties of DPPC membranes, and the effect on DPPC monolayers. References 1. Ozone Secretariat. The montreal protocol on substances that deplete the ozone layer | ozone secretariat [Internet]. Motreal Protoc. 1989 [cited 2016 Apr 14]. Available from: http://ozone.unep.org/en/treaties-anddecisions/montreal-protocol-substances-deplete-ozone-layer 2. Wright P. Inhalation dosage forms. In: Gibson M, editor. Pharm. Preformulation Formul. A Pract. Guid. from Candidate Drug Sel. to Commer. Dos. Form. Florida: CRC Press; 2004. p. 355–78. 3. Taylor KMG. Pulmonary drug delivery. In: Aulton ME, Taylor KMG, editors. Aulton’s Pharm. Des. Manuf. Med. London: Elsevier Ltd.; 2013. p. 638–56. 4. Young PM, Traini D, Edge S. Advances in pulmonary therapy. In: Williams III RO, Taft DR, McConville JT, editors. Adv. Drug Formul. Des. to Optim. Ther. Outcomes. Florida: CRC Press; 2007. p. 1–52. 5. Grainger CI, Saunders M, Buttini F, Telford R, Merolla LL, Martin GP, et al. Critical characteristics for corticosteroid solution metered dose inhaler bioequivalence. Mol. Pharm. 2012;9:563–9. 6. Haghi M, Bebawy M, Colombo P, Forbes B, Lewis DA, Salama R, et al. Towards the bioequivalence of pressurised metered dose inhalers 2 . Aerodynamically equivalent particles (with and without glycerol) exhibit different biopharmaceutical profiles in vitro. Eur. J. Pharm. Biopharm. 2014;86:38–45. 7. Hastedt JE, Bäckman P, Clark AR, Doub W, Hickey A, Hochhaus G, et al. Scope and relevance of a pulmonary biopharmaceutical classification system AAPS/FDA/USP Workshop March 16-17th, 2015 in Baltimore, MD. AAPS Open [Internet]. AAPS Open; 2016;2:1. Available from: http://aapsopen.springeropen.com/articles/10.1186/s41120-015-0002-x 8. Derendorf H, Hochhaus G, Meibohm B, Möllmann H, Barth J. Pharmacokinetics and pharmacodynamics of inhaled corticosteroids. J. Allergy Clin. Immunol. [Internet]. 1998;101:S440–6. Available from: http://www.sciencedirect.com/science/article/pii/S0091674998701563 9. Matyszewska D, Bilewicz R. DPPC monolayers as simple models of biological membranes for studies of interactions with perfluorinated compounds. Ann. Univ. Mariae Curie, Chem. 2008;63:201–10. 10. Wójtowicz K. Effect of some pharmaceuticals on DPPC membranes phase transition — Ultrasound Absorption Study. Instrum. Sci. Technol. 1999;24:287–93. 11. Goerke J. Pulmonary surfactant : functions and molecular composition. Biochim. Biophys. Acta. 1998;1408:79–89. 12. Bhatia SK. Biomaterials for clinical applications. New York: Springer; 2010. 13. Harris FM, Best KB, Bell JD. Use of laurdan fluorescence intensity and polarization to distinguish between changes in membrane fluidity and phospholipid order. Biochim. Biophys. Acta. 2002;1565:123–8. 14. Eze MO. Phase transitions in phospholipid bilayers: lateral phase separations play vital roles in biomembranes. Biochim. Biophys. Acta. 2010;19:204–8. 15. Parasassi T, Stasio G De, Ravagnan G, Rusch RM, Gratton E, Sperimentale M, et al. Quantitation of lipid phases in phospholipid vesicles by the generalized polarization of Laurdan fluorescence. Biophys. J. 1991;60:179–89. 16. Slater JL, Huang CH. Interdigitated bilayer membranes. Prog. Lipid Res. 1988;27:325–59. 17. Kranenburg M, Vlaar M, Smit B. Simulating induced interdigitation in membranes. Biophys. J. [Internet]. Elsevier; 2004;87:1596–605. Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1304565&tool=pmcentrez&rendertype=abstract

287


Drug Delivery to the Lungs 27, 2016 - Maria Palha et al. Breathing Simulators: One step closer to representative deposition profiles? 1

1

1

Maria Palha , Sofia Silva , Isabel S. Lopes & Eunice Costa 1

1

Hovione FarmaCiencia SA, Sete Casas, 2674-506 Loures, Portugal

Summary Cascade impactors are widely used for aerodynamic performance testing of dry powder inhalers (DPIs). Impactors are useful and efficient tools for quality control of inhalation products as well as for formulation development as a simple in vitro assessment of the deposition of the drug in the respiratory tract. However, the current USP pharmacopeia method considers a non-physiological representation of the human mouth-throat, the induction port, and a constant flow during testing. The flow achieves a 4 kPa pressure drop and an inhalation volume of 4 L, which does not account for the variability inherent in the patient population when evaluating product performance [1]. The main goal of this work was to compare the recently developed breathing pattern simulator and the Alberta Idealized Throat (AIT) [2], with the current pharmacopeia standard in the aerodynamic performance assessment. Three different setups operating under pharmacopoeia conditions, mimicking a square wave breathing profile, were tested: Fast Screening Impactor (FSI) + Induction Port (IP); FSI + AIT; FSI + AIT + Breathing Simulator (BRS). The two parameters evaluated, Emitted Dose (ED) and Fine Particle Dose (FPD), were relatively insensitive to test conditions when comparing the three setups, with the FSI + AIT + BRS improving reproducibility. In addition, three different breathing patterns were evaluated: medium, strong and weak, as defined in the literature [1]. These patterns represent inhalation profiles derived from patient measurements and capture different target patient populations. It was observed that BRS reduced method variability and that the square and medium profiles yielded the same results. In relation to the weak profile, a reduced FPD was observed, most probably due to insufficient powder dispersion. In relation to the strong inhalation profile, the ED was higher but no differences were observed in terms of the FPD in comparison to the medium and square waves profiles. Introduction The aerodynamic particle size distribution (APSD) of dry powder inhalers is typically measured via cascade impaction techniques. These apparatus fractionate the product into discrete aerodynamic size ranges, providing information on the potential product deposition in the respiratory tract. These methodologies rely on the fact that a an average adult generates approximately 4 kPa pressure drop across a given forced inhalation and inhales a total volume of 4 L of air. These parameters are used to define a constant test flow and duration and give origin to a square wave profile that is applied during dose uniformity and cascade impactor testing [3]. In addition, the actual acceleration of the flow from the inhalation manoeuvre is lower than the acceleration of the flow provided by a vacuum pump; hence, most patients may not be able to achieve the inhalation volume of 4 L [1]. The performance assessment of DPIs, as per the USP method, is not, therefore, representative of some of the patient populations. Patients with lower inspiratory capacity or with severely impaired lung function may not be able to generate a 4 kPa pressure drop upon device actuation, especially geriatric and paediatric users. On the other hand, healthy patients who may have been using DPIs for systemic treatments (rather than for the treatment of pulmonary disease) might have a much higher inspiratory capacity. In addition, there are also those patients that operate the device itself incorrectly (sub-standard manoeuvre technique) and that have sub-standard deposition profiles. This leads to differences observed between the products’ particle deposition prediction (in vitro results) and the real particle deposition results (in vivo results), amongst other factors such as lung anatomy and physiology [1]. Based on this, there is the need to test DPIs under conditions that mimic the breathing pattern of the target patient population. In recent years researchers have been developing breathing simulators in order to test the products with more representative inhalation profiles, mimicking the variable lung functions of certain patient groups [1, 4]. Additionally, and after recognizing that the standard Induction Port does not provide accurate data on the deposition in the upper respiratory tract, the Alberta Idealized Throat (AIT) was developed based on typical patient population geometries provided by Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) scans [3], enabling a better simulation of both the aerosol deposition and the flow across the human mouth-throat [3]. There has also been development work to adjust models to the different types of patients and their respective anatomy (neonates, infants, adolescents and adults). The current work compares results obtained using the standard pharmacopoeia test conditions while using different breathing patterns and the AIT to evaluate the product performance and prediction of delivered lung doses. This assessment was performed using a composite particle formulation, with a commercially available device and an abbreviated impactor measurement (AIM) method, the Fast Screening Impactor (FSI), in order to determine Emitted Dose and Fine Particle Dose parameters (the key properties to evaluate the formulation’s performance).

288


Drug Delivery to the Lungs 27, 2016 – Breathing Simulators: One step closer to representative deposition profiles?: One step closer to representative deposition profiles? Experimental methods The characterized material comprises 80% w/w Trehalose and 20% w/w Leucine and was produced by spray drying as described in [5]. The PSD of the obtained powder was Dv10 = 0.49 µm; Dv50 = 1.49 µm and Dv90 = 3.97 µm (the complete powder characterization can be found in [5]). The obtained product was characterized in terms of aerodynamic performance using one commercially available capsule based device: Plastiape, model RS01 operating at 60 L/min at 4 kPa. HPMC size #3 capsules were filled using a MG2 FlexaLab capsule filler unit, with a target fill weight of 17 mg per capsule. The product performance was evaluated by gravimetric FSI, where the weight changes detected in the capsule, device and filter stage were determined by an analytical balance with an accuracy of ±0.01mg. The ED parameter corresponds to the fraction that exits both the capsule and device; and the FPD parameter corresponds to the fraction that is collected with a cut-off diameter of 5 µm. 

Aerodynamic product performance by AIM 

Standard pharmacopoeia set up: Induction Port + FSI

The ED and FPD were determined by FSI coupled with the Induction Port (USP throat) using a gravimetric method. A total of 3 replicates were carried out. The total inhalation volume considered was 4 L and the total inhalation time was 4 s [6]. 

AIT + FSI

The ED and FPD were determined by FSI coupled with the AIT using a gravimetric method coupled. A total of 3 replicates were carried out. The total inhalation volume considered was 4 L and the total inhalation time was 4 s. 

BRS + AIT + FSI

The ED and FPD were determined by FSI, using a gravimetric method, coupled with a mixing inlet and the AIT and applying different breathing patterns using a BRS. The patterns evaluated were: i) square wave pattern and ii) three user defined patterns taking into account real inhalation profiles described in the literature (where a different device was evaluated but presenting the same airflow resistance as the Plastiape device) [1] that th th th represent the 10 , 50 and 90 percentiles of a patient population (mimicking weak, medium and strong inhalation profiles) [1]. A weak inhalation profile represents the impaired population lung function and additionally may represent paediatric and geriatric patients (who either have lower inhalation capacity or might have an incorrect inhalation manoeuvre). A strong inhalation profile represents healthy patients that use inhalation products for the delivery of systemic therapies. A total of 3 replicates were carried out for each pattern. The total inhalation volume and time considered for the medium, weak and strong profiles were 1.4 L and 2.2 s, 1.8 L and 5.5 s, 2.3 L and 2.6 s, respectively. (A)

(C)

(B)

Figure 1 – Set up equipment used during the course of this work: (A) – IP + FSI (image from Copley Scientific website), (B) AIT + FSI, (C) – BRS + AIT + FSI (image from Copley Scientific website) Results and Discussion The key properties of the product performance were compared in two different stages, one where the pharmacopoeia testing conditions were maintained and the comparison was performed using the standard equipment and the state of the art tools (AIT and BRS); another where only AIT and BRS were used applying different breathing patterns.

289


Drug Delivery to the Lungs 27, 2016 - Maria Palha et al. 

Pharmacopoeial testing conditions

Three set ups were tested: FSI coupled with IP, FSI coupled with AIT (mimicking a human like upper airway geometry) and FSI coupled with AIT and BRS (also generating a square wave profile). The ED and FPD results are presented in Figure 1. (B)

(A)

Figure 2 - (A) ED and FPD results obtained by gravimetric FSI with the Plastiape device, where the error bars correspond to one standard deviation; (B) Square wave profile (representative of a vacuum profile). In terms of ED, the FSI + AIT setup yields similar aerodynamic performance as the conventional apparatus. Although slightly higher values of ED are observed for the FSI + AIT set up, the high variability presented does not enable definitive conclusions. FSI + AIT + BRS set up reduces the variability observed, which can be explained by the fact that the BRS (and adjacent mixing inlet) allows steady state conditions during testing and that a turbulence-free mixing of the sample air stream occurs before the sample enters the impactor [3]. In terms of FPD, it is observed that the introduction of the AIT leads to a reduced FPD, which is in accordance with expectations [3]. As the AIT was designed to represent a more physiological airway geometry, it is expected that this design leads to a higher entrapment of the particles. It has been recognized that the standard Induction Port tends to overestimate the lung deposition [3]. Once again, the steady state conditions promoted by the BRS lead to a lower variability in the FPD obtained. The results demonstrate that the FSI + AIT + BRS setup can be used in replacement of the standard equipment. 

Breathing profiles

Three different breathing patterns were tested and compared with the square wave profile using BRS. The ED and FPD results are presented in Figure 3. (A)

(B)

Figure 3 – (A) ED and FPD results obtained by gravimetric FSI with the Plastiape device, where the error bars correspond to one standard deviation; (B) Different inhalation profiles: square, medium, weak and strong. Once again, it was observed that BRS improves reproducibility of the aerodynamic product performance. When comparing the ED and FPD parameters of the square and medium profiles, although the 4L inhaled volume is not achieved when mimicking a medium profile, no significant differences are observed between them, which might indicate that the formulation is mostly dispersed in the initial part of the inhalation. When comparing to the weak profile, the ED presents higher variability that might be related to a deficient powder dispersion when exiting the capsule/device, leading also to a reduced FPD. In relation to the strong inhalation profile, the dose that leaves the capsule/device is higher, which could be related to a higher dispersion energy at the peak inhalation flow rate during the inhalation peak flow rate. However, the respirable fraction is similar to the one observed in a medium / square inhalation profile.

290


Drug Delivery to the Lungs 27, 2016 – Breathing Simulators: One step closer to representative deposition profiles?: One step closer to representative deposition profiles? Conclusions With the pharmacopoeial standard conditions, no significant differences are observed when applying either the conventional set up of FSI or FSI + AIT and FSI + AIT + BRS, for this particular formulation and device (considering the three replicates performed for each set up). It is known that one of the advantages of carrier-free composite particles formulations is that they tend to be less sensitive to shifts in aerodynamic performance at different inhalation conditions; therefore, as future work, carrier-based systems will be evaluated using the three different set-ups and benchmarked against composite systems. The use of breath simulators enables a better understanding of the product performance, specifically for patients with impaired lung function where it was shown that the product performance is patient dependent. In addition, it was also seen that the use of the AIT yielded lower results in the ED parameter than the standard Induction Port, which could be related to the fact that it better mimics the human throat geometry, enabling a better prediction of the upper respiratory tract drug deposition. These tools can help to predict in vivo performance of DPIs and expedite product development, especially the development of generic products, where the characterisation of flow rate dependency in the different patient populations must be presented and the demonstration of bioequivalence is a mandatory requirement [7]. References [1] Chrystyn H, Safioti G, Keegstra JR, Gopalan G: Effect of inhalation profile and throat geometry on predicted lung deposition of budesonide and formoterol (BF) in COPD: an in vitro comparison of Spiromax with Turbuhaler, International Journal of Pharmaceuticals 2015, 491: 268 – 276. [2] Copley M: Using breathing simulators to enhance inhaled product testing. Copley Scientific, REF COP/JOB/251 [3] Copley M: Improving Inhaled Product Testing: Methods for Obtaining Better In vitro–In vivo Relationships, Pharmaceutical Technology 2013, 37 (2). [4] Chapman KR, Fogarty CM, Peckitt C, Lassen C, Jadayel D, Dederichs J, Dalvi M, Kramer B: Delivery characteristics and patients’ handling of two single-dose dry-powder inhalers used in COPD, International Journal of COPD 2011, 6: 353 – 363. [5] Moura C, Vicente J, Palha M, Neves F, Aguiar-Ricardo A, Costa E: Screening and Optimization of Formulation and Process Parameters for the Manufacture of Inhalable Composite Particles by Spray-drying, Drug Delivery to the Lungs 25. The Aerosol Society, Edinburgh, UK: 2014 [6] USP chapter <601> Inhalation and nasal drug products: Aerosols, sprays and powders – Performance quality tests [7] EMEA guideline CPMP/EWP/4151/00 Rev.1: “Guideline on the requirements for clinical documentation for orally inhaled products (OIP) including the requirements for demonstration of therapeutic equivalence between two inhaled products for use in the treatment of asthma and chronic obstructive pulmonary disease (COPD) in adults and for use in the treatment of asthma in children and adolescents”.

291


Drug Delivery to the Lungs 27, 2016 –Daryl L. Roberts Further Investigation of the Maldistribution of Aerosol Deposits on the Upper Stages of the Andersen 8Stage Non-Viable Cascade Impactor (ACI): Evidence for Pre-Classification of Incoming Aerosol 1

Daryl L. Roberts and Jolyon P. Mitchell

2

1

2

MSP Corporation, 5910 Rice Creek Parkway, Suite 300, Shoreview, Minnesota 55126, USA Jolyon Mitchell Inhaler Consulting Services Inc., 1154 St. Anthony Road, London, Ontario, N6H 2R1, Canada

Summary In a previous study of the Andersen 8-stage non-viable cascade impactor (ACI), we observed maldistribution of monodisperse particles of known aerodynamic diameter between the four concentric rings of nozzles associated with stages 0 and 1 of an ACI (operated at 28.3 L/min without pre-separator). In the current study, we investigated this behavior by recovering the calibration particles from each ring of deposits separately. The performance of stages 0 and 1 were close to expectations when deposits from all four rings were assessed as a single entity. However, the fraction of the total deposits in the individual rings was not only different than the ideal value of 25% but changed with the size of the calibration particles. For example, the fraction of the total deposit in the outermost ring of both stages increased as the calibrant particles decreased in size. This outcome can happen only if the smaller particles are preferentially dispersed to the outermost ring, indicating an airborne preclassification of large from small particles before the incoming aerosol enters into the nozzles. This undesirable airborne size classification is consistent with the focused recirculating dispersion of a laminar jet in the Andersen cone predicted by its large expansion angle (110º). We conclude that when a polydisperse aerosol enters the ACI, the particle masses collected on stages 0 and 1 bear no known relationship to the aerodynamic particle size of this aerosol in the size range where these two stages are operational. Introduction The Andersen 8-stage non-viable cascade impactor (ACI), although developed originally for room air quality [1] assessment , has become one of several cascade impactors (CIs) recognized in the pharmacopeial compendia as appropriate apparatuses for the assessment of the aerodynamic particle size distribution (APSD) of aerosols [2,3] . CIs size-fractionate particles on the principle of differing particle inertia in from orally inhaled products (OIPs) [4] a laminar flow regime , enabling size measurements to be related directly to the aerodynamic diameter size [5] scale , accepted as being predictive of likely deposition of drug particles in the human respiratory tract from the [6] OIP being assessed . In its simplest form with no pre-separator present in the flow path, there is an underlying tacit assumption that the flow entering the CI is reasonably uniform (ideally laminar) and with a uniform distribution of aerosol particles and that all 96 nozzles at the first size-fractionating stage receive the same input [7] aerosol . Even if a pre-separator is used to remove coarse particles incapable of being size-fractionated properly by the CI, it is still assumed that the flow leaving the pre-separator will rapidly become uniform and remain so [8] before passing through the nozzles of the impactor stages . In a previous study, we evaluated an ACI operated at 28.3 L/min in the configuration without pre-separator, as [9] would be the case for the assessment of aerosols from pressurized metered dose inhalers . We reported that each ring of the four concentric rings on stages 0 and 1, each comprising 24 nozzles, has a different collection efficiency profile for monodisperse calibration particles of differing sizes, even though we also found that these [9] stages, considering all nozzles together, do exhibit the efficiency curves typical of published literature . Our evidence for maldistribution among the rings was essentially visual, and the effect could be seen most clearly when the calibration particles were smaller than the average 50% efficiency size, such as when 8-µm particles were collected beneath stage 0 (Figure 1). The purpose of the present study was to investigate the reason(s) that each ring had a different efficiency for the calibration particles by separately determining the mass of these calibration particles on EACH of the four rings of nozzles for stages 0 and 1. This process was intended to allow efficiency curves for each ring to be determined, curves that logically, when added together, would yield the efficiency curve for each stage as a whole. Materials and Methods We generated airborne, fluorescent, monodisperse microspherical particles of oleic acid containing 1% w/v uranine dye with a Flow-Focusing Monodisperse Aerosol Generator (Model 1520 FMAG™, MSP Corp.) in separate experiments in which the size of calibrant particles was varied. We introduced room air containing calibrant aerosol at a constant flow rate of 28.3 L/min into and through an ACI (Copley Scientific Ltd., Nottingham, UK) with a control valve and vacuum source located downstream. Particle size was confirmed by means of an ® Aerodynamic Particle Sizer aerosol spectrometer (TSI Corp., St. Paul, MN). Aerosolization times were chosen to provide sufficient deposit for subsequent assay. Rather than determine the collection efficiency curves for stages [9] 0 and 1, as was done in the previous investigation , we separately and manually recovered the mass of calibrant on each of the four concentric rings of deposits located on the collection surface of each of these stages by [10] means of Q-tips wetted with 2-propanol. Particle mass from each sample was quantified fluorometrically .

292


Drug Delivery to the Lungs 27, 2016 – Further Investigation of the Maldistribution of Aerosol Deposits on the Upper Stages of the Andersen 8-Stage Non-Viable Cascade Impactor (ACI): Evidence for Pre-Classification of Incoming Aerosol

Figure 1 — 8.0-µm Particles Deposited Non-Uniformly on Stage 0; Average Cut-Point Size (D50) is 9.0 μm

Results Table 1 summarizes the findings, showing the fraction of the total deposit recovered from each ring (1=innermost ring, 4=outermost ring). Repeatability was checked by running three replicates for the 9.0-µm particles (both stages) and 5.0-µm particles (stage 1); the coefficients of variation were < 10%. Ideal behavior would be that 25% of the particle mass would be found on each ring. Table 2 shows the total collection efficiency for each plate, measured separately by washing the entire plate. We reported previously that the size at which the collection efficiency of each stage was 50% (D50) was the same as that reported in the pharmacopeial compendia (9.0 µm [9] for stage 0 and 5.8 µm for stage 1) . In this previous study we showed that the plate collection was somewhat less than the entire stage collection (which includes the wall losses), and the plate collection efficiencies shown in Table 2 are consistent with these earlier results. Table 1: Capture of Various Sizes of Monodisperse Microsphere Calibration Aerosols under Each Concentric Ring for Collection Plates 0 and 1 of an ACI Operated at 28.3 L/min (n=1 at each condition)

Diameter of Calibrant (µm aerodynamic diameter)

Stage

8.0 9.0 10.0

0

11.0 5.0 9.0 10.0 11.0

1

Mass of Calibrant Collected Under Individual Rings of Nozzles as Fraction of Total Mass on All Rings (%) Ring 1 Ring 2 Ring 3 Ring 4

53.5 67.0 77.7 74.4 47.2 40.1 48.6 64.8

Ideal Behavior Irrespective of Calibrant Size

14.1 14.0 11.6 16.4 9.7 24.3 25.4 20.4

13.1 5.7 3.6 3.9 12.7 16.6 15.9 8.1

19.2 13.3 7.0 5.4 30.4 18.9 10.1 6.7

25.0

Table 2. Overall Collection Efficiency for Collection Plates 0 and 1 of the ACI as a Function of Calibrant Particle Size

Diameter of Calibrant (µm aerodynamic diameter) 11 10 9 8 5

Overall Collection Plate Efficiency (%) Stage 0 Stage 1 76 88 74 90 60 94 36 not measured not measured 8

293


Drug Delivery to the Lungs 27, 2016 –Daryl L. Roberts Discussion In the previous investigation, we concluded from the visual data that the cut-point size (D50) at which a particular [9] nozzle ring collects incoming particles must be different from one ring to another . This conclusion was (and remains) the logical explanation of the observed findings from that study – and this conclusion means that the velocity of air approaching and passing through each nozzle is NOT the same but rather depends on its radial distance from the central axis of the nozzle plate. Indeed, and referring to the 8-µm aerodynamic diameter particles used to create Figure 1, the velocity through the nozzles of ring 1 must be the highest, implying the collection efficiency curve for this individual ring has the smallest D50. This value is therefore smaller than the mass-average D50 for the stage, which is approximately 9.0-µm aerodynamic diameter. The D50 value for ring 4 must also be smaller than the mass-average D50, but larger than that of ring 1; the D50 values for the individual rings 2 and 3 must then be larger than the mass-average D50 for the stage, to complete the assessment. These qualitative observations are important for understanding the data shown in Table 1. On the basis of these qualitative observations alone, one would expect the fraction of the collected material to be different on each ring. However, because the air flow velocity through each nozzle is independent of the size of the calibration particles, we also expected the mass fraction to remain constant and independent of particle size. We conclude that an alternative physical process, something other than maldistribution of the air flow itself, is responsible for the Table 1 data. Stated another way, it is one thing that each ring did not collect 25% of the mass; a constant mass fraction could be explained by air flow maldistribution alone. However, these fractions were not constant. The key issue we believe lies in the expansion of the laminar jet formed when the air flows into the cone of the ACI and expands into the air space before entering the nozzles of stage 0. The flow Reynolds number at the entrance of the cone is 1670 at the chosen volumetric flow rate of 28.3 L/min, suggesting laminar flow exists at that location. In a regular expanding cone-shaped geometry, it is well understood that smooth expansion of a [11] laminar jet takes places if the expansion angle is smaller than 15º . As the angle increases over 20 degrees, flow separation at the wall begins, leading to recirculation zones for angles over 40º (Figure 2). However, the interior angle of the ACI cone is 110º, ensuring a focused jet flow and flow recirculation zones are formed, as in the right-most diagram of Figure 2. As the incoming air flow turns back around toward the entrance of the cone at the base of the recirculation, particle trajectories in flight differ on the basis of aerodynamic size because of the differing inertia of the particles. This process results in a simultaneous focusing of large particles toward the central axis and of smaller particles towards the periphery before they enter the nozzles of stage 0. Under such circumstances, ring 4 will preferentially collect the smaller particles, as observed in Table 1.

11

Figure 2 — Air Patterns in Laminar Jet Expansion (simplified from Table 7-6 of reference [ ]) Something similar appears to occur with stage 1, because of the short-circuiting of the air flow through the central large hole in the collection plate of stage 0 as this air flows to the nozzles of stage 1. There is a greater tendency than on plate 0 for the small particles to prefer collection at ring 4, consistent with the data in Table 1, because air also flows around the periphery of plate 0 on its way to stage 1. Therefore we conclude that stage 1 is also not operating as a proper size fractionator in the ACI configuration studied, because pre-classification is occurring in the space between the collection surface of stage 0 and the entry to the nozzles of stage 1. The key consequence from the foregoing is that for a polydisperse aerosol, such as those associated with almost all OIPs, our data indicate that airborne size fractionation will occur in the gas phase before the air flows through the nozzles of both stages 0 and 1. So the size distribution of particles ENTERING each ring of nozzles will differ from the inlet distribution and will differ from ring to ring. We already know from our previous work with monodisperse calibrant aerosols that the D50 values of each ring are different. Hence it follows that the mass captured on the collection plate of either stage will bear no known relationship to the aerodynamic particle size of the airborne particles entering the ACI cone (which of course is the aerosol that the patient would inhale), thereby invaliding the common interpretation of the size of the particles captured on the collection plate.

294


Drug Delivery to the Lungs 27, 2016 – Further Investigation of the Maldistribution of Aerosol Deposits on the Upper Stages of the Andersen 8-Stage Non-Viable Cascade Impactor (ACI): Evidence for Pre-Classification of Incoming Aerosol It is important to note that the non-ideal behavior reported here cannot be found by standard impactor calibration methods, in which the entire deposit on a stage or plate is recovered as one entity. This limitation may therefore explain why the effects we describe have hitherto been unreported. The impactor designer can avoid the problem altogether if the first stage of the impactor is a single nozzle on axis with the flow entering the apparatus, as in the [12] [13] and the Next Generation Impactor . Marple-Miller Impactor Conclusions We have established that differences exist in the cut-point sizes of individual nozzle rings associated with stages 0 and 1 of an ACI operated at 28.3 L/min and have further provided evidence that undesirable airborne preclassification of large from small particles is occurring before the incoming aerosol enters the nozzles of either stage. These findings indicate that the ACI is not reliable as an apparatus for the accurate assessment of aerosol aerodynamic particle size distribution in the size range where these two stages are operational. Nevertheless, the implication for the testing of OIPs may be slight because most formulations have only a small fraction of their mass larger than about five microns. Our findings may indeed make it more important for regulatory agencies to focus on the mass of particles smaller than five microns aerodynamic diameter. But in any case, users should be aware of these limitations and in consequence be cautious of aerodynamic particle size distribution assessments made by the ACI if their product contains a significant portion of its emitted mass in the critical range from about 5 to 10 µm aerodynamic diameter, where these stages are operative. Acknowledgement The authors wish to thank Mr. Hongxu Duan of MSP Corporation for operating the Model 1520 Flow-Focusing Monodisperse Aerosol Generator and the TSI APS for these studies. References 1

Andersen A: A sampler for respiratory health assessment, Am Ind Hyg Assoc J 1958; 27:pp160-165.

2

European Directorate for Quality in Medicines and Healthcare (EDQM): European pharmacopeia 8.0, monograph 2.9.18. Preparations for inhalations: Aerodynamic assessment of fine particles, EDQM, Strasburg, France; 2014.

3

US Pharmacopeial Convention: United States Pharmacopeia 39/National Formulary 34, Chapter <601> Aerosols, nasal sprays, metered-dose inhalers, and dry powder inhalers, USP, Rockville, MD, USA; 2016.

4

Marple V A, Liu B Y H: Characteristics of laminar jet impactors, Environ Sci Technol 1974; 8: pp648-654.

5

Mitchell J P, Nagel M W: Cascade Impactors for the Size Characterization of Aerosols from Medical Inhalers: Their Uses and Limitations, J Aerosol Med 2003; 16(4): pp 341-77.

6

Heyder J, Svartengren M U: Basic principles of particle behavior in the human respiratory tract, In: H Bisgaard, C O’Callaghan and G C Smaldone, (eds): Drug Delivery to the Lung. Marcel Dekker, NY, USA; pp 21-45, 2002.

7

Marple V A, Willeke K: Impactor Design, Atmos Environ 1976; 10: pp 891-96.

8

Roberts D L, Ngo D, Doub B: Improving Flow Distribution and Analytical Repeatability with a Novel Pre-Separator Base for the Andersen Eight-Stage Non-Viable Cascade Impactor (ACI), The Aerosol Society, Drug Delivery to the Lungs 26, Edinburgh, UK, December 9-11, 2015, pp 14-17.

9

Roberts D L, Mitchell J P: Deposit Maldistribution in the Andersen Non-Viable Mark-II Cascade Impactor (ACI): An Initial Investigation, The Aerosol Society, Drug Delivery to the Lungs 25, Edinburgh, UK, December 10-12, 2014, pp154-58.

10

Marple V A, Olson B A, Santhanakrishnan K, Mitchell J P, Murray S C, Hudson-Curtis B L: Next Generation Pharmaceutical Impactor; Part II: Archival Calibration, J Aerosol Med 2003; 16(3): pp 301-24.

11

12

13

Blevins R D: Theoretical diffuser performance: In: R D Blevins (ed): Applied Fluid Dynamics Handbook, Van Nostrand Reinhold, NY, USA 1984; pp.144-161. Marple V A, Olson B A, Miller N C: A low-loss cascade impactor with stage collection cups: Calibration and pharmaceutical inhaler applications, Aerosol Sci Technol 1995; 22: pp 124-34. Marple V A, Roberts D L, Romay F J, Miller N C, Truman K G, Van Oort M, Olsson B, Holroyd M J, Mitchell J P, Hochrainer D: Next Generation Pharmaceutical Impactor - Part 1: Design, J. Aerosol Med. 2003; 16: pp 283-99.

295


Drug Delivery to the Lungs 27, 2016Nina Warner et al. Ring Opening Metathesis Polymerization of a ciprofloxacin-conjugated copolymer for pulmonary intracellular antibiotic delivery 1

1

2

1

Nina Warner , Maryann Zhao , Daniel M. Ratner , & Daniel J. O'Leary 1

2

Department of Chemistry, Pomona College, Claremont, CA 91711, USA Department of Bioengineering, University of Washington, Seattle, WA 98195, USA

Summary The utility of Ring Opening Metathesis Polymerization (ROMP) is explored in the context of aerosolized pulmonary drug delivery. A poly(oxanorbornene) random copolymer loaded with the fluoroquinolone antibiotic, Ciprofloxacin, and sulfobetaine pendant groups (P1) is synthesized to target intracellular Burkholderia pseudomallei, the causative agent of respiratory meliodosis and an attractive candidate for bioterrorism. This system is compared to a previoulsy reported Reversible Addition-Fragmentation Chain-Transfer Polymerization (RAFT) synthesized poly(methacrylate) random copolymer loaded with Ciprofloxacin and PEG pendant groups (P2). In both cases, Ciprofloxacin is conjugated to the polymer via a labile phenolic ester linkage. However, we expect the enhanced hydration of the hydrophillic poly(oxanorbornene) backbone of P1 to encourage hydrolytic cleavage of the drug to a greater extent than the hydrophobic P2 poly(methacrylate) backbone. Substitution of the P2 PEG moiety with a sulfobetaine pendant is predicted to enhance intracellular targeting due to the betaine's affinity for the plasma membrane. Control over the antibiotic release rate and target will allow for more precise, effective treatment of respiratory meliodosis and other pulmonary infections caused by intracellular bacterial pathogens. Additionally, this work will inform future design of aerosolized polymers for pulmonary drug delivery by elucidating distinct advantages/disadvantages associated with the polymerization approach (ROMP v. RAFT), pendant group chemistry (PEG v. betaines) and backbone properties (poly(oxanorbornene) v. poly(methacrylate)) of aerosolized systems. Introduction Burkholderia pseudomallei is a particularly non-fastidious, soil-dwelling bacterium endemic in Southeast Asia and [1] [2] When inhaled, the bacterium can localize in the lung, causing respiratory meliodosis, an Northern Australia. [3] infectious disease clinically presented most commonly by pneumonia. Treatment of meliodosis is still largely unoptimized, with mortality reaching 40% even with treatment. Unsettlingly, this clinical shortcoming, together with [4] the bacterium's capacity for latency, make Burkholderia pseudomallei an attractive candidate in bioterrorism. Optimization of pulmonary antibiotic delivery could significantly shrink disease regression rates following [4] treatment and diminish Burkholderia pseudomallei's utility in biological warfare. Polymeric carriers have proved widely successful in optimizing antibiotic targeting, water solubility, pharmacokinetics, stimuli-responsiveness, [5] biodistribution, and biocompatibility in previous studies. Recently, Das and colleagues demonstrated tunable release rates for Reversible Addition-Fragmentation chain-Transfer (RAFT) polymerized prodrug monomers of the fluoroquinolone, Ciprofloxacin (Cip). The lung-targeted poly(methacrylate) copolymers with Cip and polyethylene glycol (PEG) pendant groups demonstrated successful bactericidal activity against Burkholderia thailandensis in a -1 [4] bacteria-macrophage co-culture model of meliodosis, while remaining nontoxic up to 20 mg mL . Like RAFT, Ring Opening Metathesis Polymerization (ROMP) is a widely-used approach to industrial-scale living polymerization known for its compatibility with organic and protic solvents, high degree of functional group [6] tolerance, mild conditions, and readily synthesized, commercially available alkylidene initiators. Moreover, ROMP polymers are characterized by high degrees of drug loading, absolute linearity, utility in forming controlled architectures, monodispersity, stereoregularity, good biocompatibility and targeting abilities. The mechanism by which ROMP proceeds concerns the redistribution of highly strained cyclic olefins, driven by enthalpy release. [7] [8] [9] [Figure 1]

A

M

M

M

B Figure 1. Simplified mechanism of Ring Opening Metathesis Polymerization (ROMP). A is metal alkylidene catalyst. B is substituted or un-substituted cyclic olefin with ring strain >5 kcal/mol, e.g. cyclobutene, cyclopentene, ciscyclooctene, or norbornene. [8]

Industrial applications of ROMP have rapidly arisen over the past two decades, with ROMP polymerized materials being employed in the manufacture of wind turbines (Telene, Metton, Pentam, Proxima), tires (Norsorex), camera lenses and prisms (Zeonex, Zeonor, Arton), candles (Nature Wax), asphalt modifiers and rubber viscosity [10] reducers (Vestenamer) and other products. ROMP's utility in the biomedical and pharmaceutical spheres has also been noted, namely for the relative ease with which polymers of narrow dispersity, complex architecture,

296


Drug Delivery to the Lungs 27, 2016 - Ring Opening Metathesis Polymerization of a ciprofloxacin-conjugated copolymer for pulmonary intracellular antibiotic delivery [9]

diverse functional groups and high drug loading capacity can be synthesized under mild conditions. Distinct physical and chemical properties of poly(norbornene)s make ROMP polymers advantageous in many biomedical contexts as compared to their RAFT poly(methacrylate) analogs. Among these advantages are superior thermal stability, flexibility, hydrophilicity, anti-fouling character, mechanical durability, optical transparency, and reduced [10] [11] [12] [13] Moreover, RAFT monomers are limited to 1-substituted and 1,1-disubstituded substrates; permittivity. in poly(methacrylate)s, this has been shown to induce aggregation of adjacent pendant groups, resulting in [12] [13] [14] generally undesirable crystalline domains. Achieving non-hydrolyzable conjugation of pharmaceutically inert pendant groups, such as solubilizing and targeting functionalities, to the polymer backbone is synthetically simple via ROMP given the wide variety of [11] Contrarily, robust methacrylate linkages tend to be more readily obtainable norbornene derivatives. synthetically laborious, and are thus often substituted with labile ester bonds, even when in vivo hydrolysis is not desired. Recently, carbo-, phospho- and sulfobetaines have arisen as an attractive class of stimuli-responsive [12] [13] Unlike net charged moieties, betaines are characterized by excellent biocompatibility, solubilizing agents. likely due to their resemblance to endogenous zwitterions such as amino acids and phosphatidylcholines (PCs). Consequential of this resemblance, betaine loaded polymers serve as promising candidates for intracellular drug delivery due to their affinity for the lipid bilayer. In a pulmonary context, betaines such as [14] dipalmitoylphosphatidylcholine are also known to behave as surfactants. In this work we explore the synthesis and ROMP copolymerization of Cip prodrug monomers with sulfobetaine functionalized oxanorbornenes, which, unlike their carboxy-counterparts are known not to influence ROMP [13] kinetics. Cip is coupled to the polymer via a labile phenolic ester linkage resembling that employed in the P2 system previously described. We expect the system in development to result in both more precise intracelullar targeting capacity–– due to the presence of the sulfobetaine moiety–– and more a rapid Cip release profile–– on [4] account of the poly(oxanorbornene) hydrophillicity–– compared to the previoulsy reported RAFT system. Experimental Materials All materials were purchased from Sigma Aldrich. Synthesis of 7-(4-(tert-Butoxycarbonyl)piperazin-1-yl)-cyclopropyl-6-fluoro-4-oxo-1,4-dihydroquinoline-3carboxylic acid. (1a) [4]

1a was synthesized via methods previously described. An additional ether precipitation was employed for further 1 purification. Yield = 12.57 g (96.48%). H NMR (400 MHz, CDCl3 δ 1.29 (m, 2H), 1.39 (m, 2H), 1.51 (m, 51), 3.31 (t, 4H), 3.56 (m, 1H), 3.68 (t, 4H), 7.36 (d, 2H), 7.99 (d, 1H), 8.74 (s, 1H). Synthesis of 2-(4-hydroxybenzyl)-3a,4,7,7a-tetrahydro-1H-4,7-epoxyisoindole-1,3(2H)-dione. (1b) 1b was synthesized by allowing exo-7-oxanorbornene-5,6-dicarboxylic anhydride, 4-hydroxybenzylamine, and triethylamine (1:1:1 molar ratio) to stir at RT in ethanol overnight. Purification of product, 1b, was achieved via Amberlite IR-120 ion exchange chromatography in a mobile phase of 0.5 N Sodium Borate buffer solution. The product could be visualized via thin layer chromatography. (80:20 v/v MeOH: Acetic Acid) Successful synthesis was confirmed via liquid chromatography-mass spectrometry. (m/z: 271.1) Synthesis of 4-1,3-dioxo-1,3,3a,4,7,7a-hexahydro-2H-4,7-epoxyisoindol-2-yl)methyl)phenyl 7-(4-(tertbutoxycarbonyl)piperazin-1-yl)-1-cyclopropyl-6-fluoro-4-oxo-1,4-dihydroquinoline-3-carboxylate. (1c) Synthesis of 1c is currently underway. Various explored synthetic routes are reported in scheme 1. Synthesis of exo-[2-(3,5-Dioxo-10-oxa-4-azatricyclo[5.2.1.0]-dec-8-en-4-yl]dimethylpropylsulfobetaine. (2a) [16] [14]

A final wash in excess tetrahydrofuran (THF) was 2a was synthesized via previously described methods. 1 performed to enhance purity. Yield = 4.9306g (65%). H NMR (400 MHz, D2O) δ 6.52 (2H, s), 5.24 (2H, s), 3.89 (2H, t), 3.45 (12H, m), 2.88 (2H, t), 2.15 (2H, m). Results and Discussion [4] [14] [16]

Various synthetic approaches to 1c Syntheses of 1a and 2a were achieved by methods formerly reported. are under present investigation. (Scheme 1) Conditions were chosen as to maximize the reactivity of the 1a carboxylic acid while minimizing acidity as to protect the integrity of the BOC group.

297


Drug Delivery to the Lungs 27, 2016Nina Warner et al.

Scheme 1. Synthetic approaches to prodrug monomer synthesis.

[17] [18] [19]

ROMP copolymerization will follow successful synthesis of 1c. (Figure 2)

Figure 2: ROMP random copolymer synthesis. G3 is Grubb's Third Generation catalyst.TFE is 2,2,2-trifluoroethanol. Ethyl vinyl ether is used to quench the reaction. Black ribbon is exo-7-oxanorbornene-5,6-dicarboximide backbone. Blue spheres are Ciprofloxacin. Red hexagons are sulfobetaine pendant groups. Polymer morphology is presented in an aqueous environment.

Due to the significantly lower MW of 2a as compared to the poly(ethylene glycol) methyl ether methacrylate solubilizing agent utilized in the RAFT system reported by Das et al., we expect our polymer to have a significantly higher drug loading capacity (wt %). (Figure 3) Drug release kinetics, serum protein binding affinity, and toxicity will be assessed and reported in due course. We predict more rapid Cip hydrolysis due to hydration of poly(oxanorbornene) backbone. With respect to serum binding affinity, we expect the ROMP polymer to resist adhesion of endogenous proteins due to the known antifouling character of both the sulfobetaine and [13] [14] [15] poly(oxanorbornene). Finally, toxicity is not anticipated given the well cited biocompatibility of betaines [12] and poly(oxanorbornene) systems. B

A

a O

O O

O

O

b

O

a O

b

O O

N

O

N

O

O

Ciprofloxacin = HO

O O N

F N

NH

N

O HN

O

O

O

19

O O3S

O O

Figure 3. RAFT and ROMP polymer systems for intracellular pulmonary delivery of Cip. A is previously reported poly(methacrylate) based polymer synthesized via RAFT.[4] B is poly(oxanorbornene) based polymer synthesized by ROMP, as described herein. B is predicted to present more rapid in vivo drug cleavage.

298


Drug Delivery to the Lungs 27, 2016 - Ring Opening Metathesis Polymerization of a ciprofloxacin-conjugated copolymer for pulmonary intracellular antibiotic delivery Conclusion ROMP was employed in the synthesis of a Cip-loaded poly(oxanorbornene) for the treatment of respiratory meliodosis. Pendant sulfobetaine functional groups were included to encourage optimal solubility, [13] [14] [16] Polymer characterization, release kinetics, serum binding biocompatibility, and membrane penetration. affinity, toxicity, and pharmaceutical efficacy against a model pathogen, Burkholderia thailandensis were determined. These values were compared to those previously reported for the methacrylate-based RAFT polymer [4] comprising phenolic ester drug linkages and solubilizing PEG pendant groups. The work performed herein is intended to inform future studies concerning pulmonary intracelullar delivery of antibiotics. The broad-spectrum therapeutic efficacy of Cip, including its particular clinical success against Pseudomonas aeruginosa, a leading cause of exacerbated morbidity in CF, bronchiectasis, and pneumonia, suggests possible wider applications of the system elucidated. Substitution of Cip with alternative fluoroquinolone antibiotics should also be facile given [20] This is particularly valuable the highly comparable MWs, reactivities and logP values among class members. given the unique capacity of fluoroquinolones to treat intracellular pulmonary pathogens, where other popular [21] Unfortunately, however, many of the same antibiotic classes, such as aminoglycosides and β-lactams fail. chemical characteristics that allow for fluoroquinolone intracellular pharmaceutical activity also encourage liver localization during systemic circulation. In fact, bioassays following oral administration of Cip reveal hepatic tissue [22] Further study of inhaled systems for to be more concentrated with the antibiotic than any other tissue. pulmonary delivery is necessary to reduce fluoroquinolone toxicity and enhance therapeutic potency. This work will inform the development of future polymer systems for pulmonary drug delivery by elucidating distinct pros/cons associated with polymerization approach (ROMP v. RAFT), pendant group chemistry (PEG v. betaines) and backbone properties (poly(oxanorbornene) v. poly(methacrylate). References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

Wiersinga, W. J., Currie, B. J. & Peacock, S. J. Melioidosis. N. Engl. J. Med. 367, 1035–1044 (2012). Aldhous, P. Tropical medicine: melioidosis? Never heard of it… Nature 434, 692–693 (2005). Chiang, K.-H., Cheng, K.-C. & Yang, H.-H. Severe Septicemic Melioidosis in a Patient With Lung Adenocarcinoma Following Cytotoxic Chemotherapy. J Med Cases 6, 14–16 (2014). Das, D. et al. RAFT polymerization of ciprofloxacinCip prodrug monomers for the controlled intracellular delivery of antibiotics. Polym. Chem. 7, 826–837 (2016). Langer, Robert, and Nicholas A. Peppas. Advances in Biomaterials, Drug Delivery, and Bionanotechnology AIChE Journal 49, no. 12 (1 December 2003): 2990–3006. Scherman, Oren. Enhancing Materials through Controlled Architectures with Ring-Opening Metathesis Polymerization. Doctor of Philosophy Thesis. California Institute of Technology, 2004. Grubbs, Robert H., Anna G. Wenzel, Daniel J. O'Leary, and Ezat Khosravi, eds.Handbook of Metathesis. 2nd ed. Vol. 3. Weinheim: WILEY-VCH GmbH, 2003. Print. Grubbs, Robert H, and Christopher W. Bielawski. Living ring-opening metathesis polymerization. Progress in Polymer Science 32 1-29 (2007). Slugovc, C. in Olefin Metathesis (ed. Grela, K.) 329–333 (John Wiley & Sons, Inc., 2014). Smith, D., Pentzer, E. B. & Nguyen, S. T. Bioactive and Therapeutic ROMP Polymers. Polymer Reviews 47, 419–459 (2007). Tew, G. & Colak, S. Dual-Functional ROMP-Based Betaines: Effect of Hydrophilicity and Backbone Structure on Nonfouling Properties. Langmuir, 2012, 28 (1), pp 666–675 Rankin, D. A. & Lowe, A. B. New Well-Defined Polymeric Betaines: First Report Detailing the Synthesis and ROMP of Salt-Responsive Sulfopropylbetaine- and Carboxyethylbetaine-exo-7-oxanorbornene Monomers. Macromolecules 41, 614–622 (2008). Xie, N. et al. Water-soluble copolymeric materials: switchable NIR two-photon fluorescence imaging agents for living cancer cells. J. Mater. Chem. B 2, 502–510 (2014). Jones, B. G., Dickinson, P. A., Gumbleton, M. & Kellaway, I. W. Lung surfactant phospholipids inhibit the uptake of respirable microspheres by the alveolar macrophage NR8383. Journal of Pharmacy and Pharmacology 54, 1065– 1072 (2002). Lalezari, I. & Lalezari, P. United States Patent Application: 0140142279 - Method for peptide synthesis. (A1). Salewska, N. & Milewska, M. J. Efficient Method for the Synthesis of Functionalized Basic Maleimides. Journal of Heterocyclic Chemistry 51, 999–1003 (2014). Lima, E. L. S., Rodrigues, R., & Barros, I. Mild one-pot conversion of carboxylic acids to amides or esters with Ph3P/trichloroisocyanuric acid. Tetrahedron Letters 46, 5945–5947 (2005).21 Wagle, D. R. & K. Venkataraman. Cyanuric chloride: A useful reagent for converting carboxylic acids into chlorides, esters, amides and peptides. Tetrahedron Letters. 32, 3037-3040 (1979). Mills, J. & J. P. Human. Action of thionyl chloride on carboxylic acids in presence of pyridine. Nature. 4024, 877 (1946). Becnel Boyd, L. et al. Relationships among Ciprofloxacin, Gatifloxacin, Levofloxacin, and Norfloxacin MICs for Fluoroquinolone-Resistant Escherichia coli Clinical Isolates. Antimicrob Agents Chemother 53, 229–234 (2009). Ong, H. X., Traini, D., Bebawy, M. & Young, P. M. Ciprofloxacin Is Actively Transported across Bronchial Lung Epithelial Cells Using a Calu-3 Air Interface Cell Model. Antimicrob. Agents Chemother. 57, 2535–2540 (2013). Dan, M., Verbin, N., Gorea, A., Nagar, H. & Berger, S. A. Concentrations of Ciprofloxacin in human liver, gallbladder, and bile after oral administration. Eur J Clin Pharmacol 32, 217–218

299


Drug Delivery to the Lungs 27, 2016 – L. Slator et al. Comparison of Aerosol Particle Size Measured by Two Methods Using Three Brands of Mesh Nebulizer 1

1

1

2

1

L. Slator , Y. Degtyareva , L. Hardaker , D. von Hollen & J. Pritchard 1

Respironics Respiratory Drug Delivery (UK) Ltd, a business of Philips Electronics UK Limited, Chichester, West Sussex, UK 2 Respironics, Inc., a Philips Healthcare Company, Murrysville, PA, USA

Summary We compared the particle size and fine particle fraction (FPF, % ≤5 µm) of 3 mesh nebulizers using cascade impaction with a Next Generation Impactor (NGI) and a modified laser diffraction method with a Spraytec particle sizer. Three of each nebulizer brand (InnoSpire Go [production equivalent], Aeroneb Go, and MicroAir U22) were tested with 1 mL of 5 mg/2.5 mL salbutamol sulphate in triplicate with a 15 L/min constant flow rate. NGI tests: the nebulizer was connected to the NGI and run until treatment completion; samples from the NGI were analyzed via high performance liquid chromatography. Spraytec tests: the Spraytec was set up with the inhalation cell fitted with the sheath flow collar with ports open, the flow straightener component fitted and connected to a vacuum pump such that there was 15 L/min flow at the inlet adapter ferrule. The nebulizer was connected to the inlet adapter ferrule; the nebulizer was run and stopped after 60 seconds of data measurements following a 10-second prime. The differences between the results of the NGI and Spraytec particle sizer were within 0.5 µm for particle size and 5% for FPF. The InnoSpire Go produced aerosols with the smallest particle sizes (mean ± standard deviation; 3.99 ± 0.26 µm MMAD and 4.17 ± 0.22 µm VMD) and largest FPFs (64.4 ± 4.42% and 63.4 ± 3.74%, respectively); the MicroAir U22 produced the largest particle sizes and smallest FPFs (5.83 ± 0.43 µm MMAD and 6.30 ± 0.19 µm VMD, and 41.3 ± 3.83% and 36.7 ± 1.62%, respectively). Span and geometric standard deviation were lowest for the InnoSpire Go. The modified Spraytec method has potential for use as a routine method in the future. Introduction Different methods can be used for determination of particle size; cascade impaction has traditionally been the most commonly used method of assessing aerosol droplet sizes and is recommended in regulatory guidances, but laser diffraction is also a popular method due to the speed and ease of analysis compared with cascade impaction. We compared the particle size and fine particle fraction (FPF) measurements of 3 different mesh nebulizers determined with a standard cascade impaction method using a Next Generation Impactor and a modified laser diffraction method using a Spraytec, when performed at the same constant extraction flow rate to confirm the acceptability of the new method. Methods (A)

Three of each of the InnoSpire Go , Aeroneb Go, and MicroAir U22 nebulizers (Figure 1) were tested in triplicate with 1 mL of 5 mg/2.5 mL salbutamol sulphate (Salamol; IVAX Pharmaceuticals, Castleford, UK). Each nebulizer was tested with a Next Generation Impactor (NGI) and a Malvern Spraytec laser diffraction system, each set to a 1,2 Before each of the tests, the nebulizers were cleaned in warm soapy 15 L/min constant extraction flow rate. water, rinsed, and air-dried, and all equipment and solutions were stabilized to ambient conditions for at least 2 hours before use. For the NGI tests, a vacuum pump attached to the NGI was set to an extraction flow of 15 L/min. An (A) InnoSpire Go nebulizer was filled with 1 mL of salbutamol sulphate solution, the mouthpiece was attached to the NGI via a custom connector, and the join was sealed with Parafilm (Alcan Packaging, Neenah, WI). The nebulizer and timer were started simultaneously. At the end of nebulization, the nebulizer and timer were stopped simultaneously, and the vacuum pump was stopped after a further 10 seconds. Samples were eluted from the NGI induction port, collection cups, custom connector, nebulizer, and back-up filter using volumes of diluent appropriate for analysis by high performance liquid chromatography. The nebulizer was disassembled, cleaned in (A) warm soapy water, and air-dried. The test was performed in triplicate and with the other 2 InnoSpire Go nebulizers. The test was also performed in triplicate with 3 Aeroneb Go and 3 MicroAir U22 nebulizers. NGI data analysis was performed using Copley Inhaler Testing Data Analysis Software (CITDAS; Copley Scientific Ltd., Nottingham, UK), and the results reported were mass median aerodynamic diameter (MMAD), geometric standard deviation (GSD), and FPF (% ≤5 µm). The CITDAS summary data was checked to ensure the data was unimodal. The Spraytec was set up with the inhalation cell fitted with the sheath flow collar with ports open, the flow straightener component and the inlet adapter ferrule. A vacuum pump provided an extraction of 22 L/min through the inhalation flow cell as measured at the inlet adapter ferrule. The inhalation flow cell was then adjusted so as to allow a combined secondary sheath air flow of 7 L/min (±0.3 L/min) into the inhalation cell from either side, resulting in a 15 L/min extraction at the inlet adapter ferrule, which provided the same flow rate through the nebulizers as used in the NGI method. The sheath air flow of 7 L/min ensured that the aerosol was transported

300


Drug Delivery to the Lungs 27, 2016 - Comparison of Aerosol Particle Size Measured by Two Methods Using Three Brands of Mesh Nebulizer across central portion of the flow cell. The InnoSpire Go nebulizer was filled with 1 mL of salbutamol sulphate solution and connected to the Spraytec via a connector. The nebulizer and timer were started simultaneously, and data acquisition on the Spraytec was started after 10 seconds. Data acquisition, nebulizer, and timer were stopped after 70 seconds of nebulization. The nebulizer was disassembled, cleaned in warm soapy water, and (A) air-dried; the test was performed in triplicate and with the other 2 InnoSpire Go nebulizers. The test was also performed in triplicate with 3 Aeroneb Go and 3 MicroAir U22 nebulizers. The results reported were volume median diameter (VMD), FPF, and span of delivered aerosol calculated using Spraytec software version 3.2.

Figure 1 - Nebulizers and laboratory equipment used in the study.

Results The results for the MMAD and VMD for the 3 nebulisers with the NGI and Malvern Spraytec are shown in Figure 2. The results for the FPF for the 3 nebulisers with the NGI and Malvern Spraytec are shown in Figure 3. Median mass balance for the NGI tests was 97.6% (range 94.6 to 99.4%). The mean results for MMAD from the Next Generation Impactor and VMD from the Spraytec were similar for each of the nebulizers tested (Figure 2); the mean FPF for each nebulizer was also similar for the 2 methods (Figure 3). (A) Mean particle size was smallest for the InnoSpire Go nebulizers and largest for the MicroAir U22 nebulizers (A) (Figure 2). Mean fine particle fraction was largest for the InnoSpire Go nebulizers and smallest for the MicroAir U22 nebulizers (Figure 3).

301


Drug Delivery to the Lungs 27, 2016 – L. Slator et al.

Figure 2 - Mean particle size from the NGI shown as mass median aerodynamic diameter (MMAD) (white) and mean particle size from the Spraytec shown as volume median diameter (VMD) (grey) for each of the nebulizers tested (n = 9). Error bars show the standard deviation about the mean.

Figure 3 - Mean fine particle fraction (% ≤5 µm) as tested via the Next Generation Impactor (white) and Spraytec laser diffraction system (grey) for each of the nebulizers tested (n = 9). Error bars show the standard deviation about the mean.

302


Drug Delivery to the Lungs 27, 2016 - Comparison of Aerosol Particle Size Measured by Two Methods Using Three Brands of Mesh Nebulizer Table 1. Span and geometric standard deviation results for each of the nebulizers tested.

InnoSpire Go (production equivalent)

Aeroneb Go

MicroAir U22

Span (SD)

1.39 (± 0.02)

1.59 (± 0.04)

1.73 (± 0.06)

Geometric standard deviation (SD)

1.82 (± 0.01)

2.10 (± 0.03)

2.07 (± 0.07)

The results for span and geometric standard deviation were lowest for the InnoSpire Go

(A)

nebulizer (Table 1).

Discussion The differences between the mean particle size measurements (Figure 2) and the mean FPF measurements (Figure 3) produced by the 2 methods varied between the nebulizers but the differences were all relatively small; the differences between the particle size results were all within 0.5 µm, and the differences between the FPF results were all within 5%. This relative consistency between the VMD and MMAD results, and the FPF results, indicates that the Spraytec method is suitable for use as an indicator of aerodynamic particle size distribution when large numbers of samples need to be tested. Furthermore, these different methods of droplet sizing at a constant flow rate could be used to produce comparable results when testing mesh nebulizers, as found in a [3,4,5] number of other studies that used various types of nebulizer and measurement systems. (A)

nebulizers compared with The small particle size and high FPF of the aerosol produced by the InnoSpire Go the other nebulizers suggest that it would produce more particles likely to penetrate deeper into the lungs [6,7] Furthermore, the smaller span and GSD in the results for the compared with the other nebulizers tested. InnoSpire Go indicated a particle size distribution closer to a mono dispersion; mono disperse aerosols have the [8] potential for preferential deposition to specific areas of the lungs. Conclusion The droplet size and FPF results for each of the nebulizers when tested with the NGI and Spraytec methods were generally comparable for the formulation tested; therefore, the modified laser diffraction method could potentially be used as a routine method for the determination of mesh nebulizer particle size characteristics in the future. The (A) InnoSpire Go nebulizers produced the smallest particle size and largest FPF compared with the Aeroneb Go and MicroAir U22 nebulizers.

A

The InnoSpire Go nebulizers used in the tests were production equivalent devices.

References 1

European Pharmacopeia 7.3 (2012): General Chapter 2.9.44. Preparations for nebulization: Characterisation

2

United States Pharmacopeia 35 (2012): General Chapter <1601>, Products for nebulization – characterization tests.

3

Waldrep J C, Berlinski A, Dhand R: Comparative analysis of methods to measure aerosols generated by a vibrating mesh nebulizer, J Aerosol Med 2007; 20:3; pp310-319.

4

Vecellio None L, Grimbert D, Becquemin M H, Boissinot E, Le Pape A, Lemarié E, Diot P: Validation of laser diffraction method as a substitute for cascade impaction in the European Project for a nebulizer standard, J Aerosol Med 2001; 14:1; pp107-114.

5

Ziegler J, Wachtel H: Comparison of cascade impaction and laser diffraction for particle size distribution measurements, J Aerosol Med 2005; 18:3; pp311-324.

6

American Association for Respiratory Care: Aerosol consensus statement – 1991, Respir Care 1991; 36:9; pp916-921.

7

Newman S P: Aerosol deposition considerations in inhalation therapy, Chest 1985; 88:2 Suppl; pp152S-160S.

8

Biddiscombe M F, Meah S N, Underwood S R, Usmani O S: Comparing lung regions of interest in gamma scintigraphy for assessing inhaled therapeutic aerosol deposition, J Aerosol Med Pulm Drug Deliv 2011; 24:3; pp165-173.

303


Drug Delivery to the Lungs 27, 2016 – Zarif Mohamed Sofian et al. Polyamine ion-pairs to target drugs to the lung 1

1

1

Zarif Mohamed Sofian , Arcadia Woods , Paul G. Royall & Stuart A. Jones 1

1

Institute of Pharmaceutical Science, King’s College London, 150 Stamford Street, London, SE1 9NH, UK

Summary Polyamines can act as targeting moieties during drug delivery to the lungs as they are actively taken up by pulmonary artery endothelial cells and they contain charged functionalities, which provide an opportunity to form links to therapeutic agents through ion-pairing. In the present study, the ability of polyamines to form ion-pair complexes with theophylline (THE) and the physicochemical consequences of the ion-pair process was characterised. The THE-amine association Fourier Transform Infrared spectroscopy (FTIR) experiments showed that THE formed ion-pairs with all the tested polyamine counterions when the counterions were present in a molar excess. THE showed the strongest binding association with spermine at pH 9.4 in water (pKapp = 1.96) followed by spermidine (pKapp = 1.93) then ethylenediamine (pKapp = 1.43). All of the polyamines showed a stronger binding affinity with THE compared to the monoamine, ethylamine (pK app = 1.32). The ion-pair complexes displayed a significantly (p<0.05) lower lipophilicity and increases aqueous solubility compared to the parent drug and these physicochemical effects were greater at pH 9.4 compared to pH 7.4. As the THE-spermine showed the greatest changes in physicochemical properties followed by THE-spermidine > THE-ethylenediamine > THEethylamine, it appeared that the number of amines in the counterion structure was critical in modifying the THE properties. Introduction Polyamines are ubiquitous in all living cells. They are actively transported via the polyamine transporter system 1 (PTS). This transporter shows a high expression in the lung compared to other major organs and therefore xenobiotics that contain a ‘polyamine like’ moiety are actively sequestered into the lung from the pulmonary circulation. At physiological pH, polyamines are charged and this allows them to associate with other molecules through intermolecular electrostatic interactions and hydrogen bonding. This interactive potential of polyamines makes it possible for them to form ionic complexes with ionizable xenobiotics in a manner that could facilitate their uptake into the lung. One agent that could benefit from improved targeting to the lung is THE. It shows limited uptake into the lung and 2 it is rapidly cleared from the lung tissues when administered intravenously . These biopharmaceutical properties result in THE displaying a narrow therapeutic index (5-20 µg/mL) and poor side effect profile. However, as THE displays a negative charge when formulated as an IV infusion (it is currently formulated at alkaline pHs) it has the potential to form ion-pairs with positively charged polyamines. The formation of ion-pairs could modify the THE to improve its uptake into the lung. During administration, the drop in pH to 7.4 will cause THE-polyamine dissociation because the THE will lose its charge. However, if the ion-pair association is strong enough, the kinetics of the dissociation process should still allow the drug to be presented to the tissue as a complex, which 3,4 would break down once absorbed to allow THE to elicit its pharmacological activity . Ion-pair formation has been shown to increase the affinity of the parent drug with biological tissues, but it has not been frequently 5 employed during drug delivery to the lungs . The aims of this study were to characterise the strength of association between THE and a selection of polyamine counterions and to understand how the THE-polyamine interactions influenced the lipophilicity and aqueous solubility of the drug. A series of polyamines, increasing in the number of amine functional groups were selected including ethylenediamine, spermidine and spermine. A monoamine, ethylamine, was used as control. The interactions and physicochemical properties were characterised at the proposed formulation pH, 9.4, and the pH of the blood, 7.4, to mimic the drug environment pre and post intravenous injection. Methods A universal liquid cell system (Omni-Cell, Specac Ltd, UK) fitted with CaF2 windows and 0.025 mm mylar spacer (Specac Ltd, UK) was used for the Fourier transform infrared absorbance (FTIR) measurements. Deuterated water (D2O) was employed as the solvent for the measurements as it showed less interference with the IR absorption of THE. The pH of the solutions was buffered at 9.4 + 0.2, adjusted using HCl/NaOH. The absorbance spectra of THE (15 mM) at 1:0 – 1:20 THE-counterion molar ratio mixtures were recorded within the IR range of -1 1725 – 1500 cm . All spectra were baseline corrected and subtracted with the spectra of the blank solutions. -1 -1 Changes in the 1535/1551 cm peak absorbance ratio were calculated where the peak at 1551 cm was -1 assigned as the uncomplexed THE and the peak at 1535 cm as the complexed THE. The change in the IR ratio was used to determine the percentage of THE bound as a function of counterion concentration and the percentage of THE bound vs –log [counterion]free were plotted and fitted with a regression model (SigmaPlot 13.0) to determine the conditional association (pKcond) constant of the THE-counterion complex. All spectra were recorded using a Spectrum One spectrometer (Perkin Elmer Ltd, UK) and spectral analysis was performed with -1 Spectrum version 10 software (Perkin Elmer Ltd, UK). The resolution was set at 4 cm and 12 scans were performed for each measurement.

304


Drug Delivery to the Lungs 27, 2016 – Polyamine ion-pairs to target drugs to the lung The apparent distribution coefficients (DB) of THE when dissolved in aqueous amine solutions (1:0-1:100, drugamine molar ratios) were determined using a classical octanol-water partitioning experiment. The concentration of THE in the aqueous phase ([THE]aq) was determined using a UV spectrophotometer (Lambda 2S). THE solubility was tested by adding the drug to a series of amine solutions (0.0 – 2.0 M). After 48 h of mixing, the mixtures were filtered and analysed using the UV assay. The partitioning and solubility experiments were performed at two pH 0 conditions 7.4 + 0.2 (to mimic human physiological pH) and 9.4 + 0.2 (to mimic the normal formulation pH), 37 C (to mimic normal body temperature). Results and discussion -1

The IR spectra of THE (Figure 1) displayed characteristic peaks of the C=O (ca. 1618, 1647, 1670 and 1698 cm ) -1 3 and C-N (ca. 1535 and 1551 cm ) stretches . The addition of excess amine counterions to the THE led to an -1 increase in the absorbance of two C=O peaks at ca. 1618 and 1670 cm and the disappearance of the two C=O -1 peaks at 1698 and 1647 cm . Increasing the counterion concentration increased the absorbance ratio between -1 the two C-N stretching of THE peaks at ca. 1535 and 1551 cm . The counterions did not show a signal in this range. The observed spectral changes were thought to be due to the formation of the hydrogen bonds and electrostatic interactions between THE and the counterions. The THE-amine association curves derived from the 2 IR data were fitted with a sigmodial model (R > 0.97). From the model the concentration at which 50% of the THE was bound at pH 9.4 + 0.2 was calculated. The THE-spermine ion-pair showed the strongest association (pKcond = 1.96) followed by THE-spermidine ion-pair (pKcond = 1.93) > THE-ethylenediamine ion-pair (pKcond = 1.43) > THE-ethylamine ion-pair (pKcond = 1.32) (log scale) (Figure 2). The magnitude of the affinity constant between spermine and THE suggested that the complex was held together more tightly than one might expect from a single electrostatic interaction (represented by THE-ethylamine). Since, spermine has more hydrogen bonding sites (4 hydrogen donor-4 hydrogen accepter) than spermidine (3-3), ethylenediamine (2-2) and ethylamine (1-1), this could explain the strongest aqueous binding constant for spermine compared to other polyamines.

0.10

H 2N

B

O

theophylline theophylline-ethylamine

0.08

Absorbance

0.10

A H3C

0.06

1

3

2

O

1618

6

5

4

7

theophylline theophylline-ethylenediamine

0.08

H N

1 H 2N O H 3C

1

8

N

0.06

9

1618

O

2

CH3

0.04

5

6 3

4

1530

N 9

1530

0.02 1550

0.00

0.00 1700

1650

1600

1700

1550

1650

1600

1

H 2N

O

C

H 3C

1

O

2

0.08 theophylline

theophylline-spermidine

1550

1

H 2N

Absorbance

NH2

8

1670

0.04

1670

1550

0.06

7

CH3

0.02

0.10

2

H N

0.10

6

5

H N

3

4

N

7

2 HN

8

3 NH 2

9

NH

O H 3C

1

O

2

6

5

H N

3

4

N

0.08

CH3

1618

2

D

0.06

theophylline theophylline-spermine

1618

7

8

3

HN

9

CH3

NH2

4

1670

1670

0.04

0.04

1530

1530

0.02

1550

0.02

1550

0.00

0.00 1700

1650

1600

1700

1550

1650

1600

Wavenumber (cm-1)

Wavenumber (cm-1)

Figure 1 Liquid IR spectra of THE-amine mixtures (1:20 drug-amine molar ratio) in D2O pH 9.4 + 0.2.

305

1550


6

A

100

B

[THE-SP] pKcond = 1.96 R2 = 0.97

5

THE bound (%)

Peak abs. ratio (ca. 1535/1551 cm-1) of THE

Drug Delivery to the Lungs 27, 2016 – Zarif Mohamed Sofian et al.

4

3

2

ethylamine ethylenediamine spermidine spermine

1

0 0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

[THE-SPd] pKcond = 1.93 R2 = 0.98 50

[THE-EA] pKcond = 1.32 R2 = 0.98

[THE-EDA] pKcond = 1.43 R2 = 0.98

0 1.0

Total con. of amines (Mole/L)

1.5

2.0

2.5

- Log (con. of free amines (Mole/L))

Figure 2 (A) Peak absorbance ratio (ca. 1535/1551 cm-1) of THE vs total concentration of amines (mM) and (B) THEamine association curves. pKcond refers to the THE-amine conditional stability constant.

Formation of THE-amine complexes in water led to a decrease in lipophilicity of the drug (Figure 3), which suggested that the charges on the amines were not being totally shielded by complexation with THE. The magnitude of the decrease was greatest for the THE-spermine ion-pair, followed by THE-spermidine then the THE-ethylenediamine and the THE-ethylamine ion-pairs. The log D changes for THE were found to be greater at pH 9.4 compared to 7.4 across all the ion-pairs tested, which was attributed to the ion-pairing effect being more pronounced in basic conditions that encouraged a greater degree of theophylline deprotonation, i.e., 97.43 % deprotonation of THE at pH 9.4 compared only 27.73 % at pH 7.4 (Figure 4). Table 1 summarizes the experimental log D of THE and the respective THE-amines ion pairs at 1:100 drug:counterion molar ratio.

0.1

0.0

-0.2

EA EDA SPd SP

-0.1

Log D of THE

Log D of THE

B

0.0

A

-0.2

-0.3

-0.4

-0.6

-0.8

-1.0

-0.4

-0.5

-1.2 0

1

2

3

4

5

6

0

Concentration of amine (mmol/L)

1

2

3

4

5

6

Concentration of amine (mmol/L)

Figure 3 Log D profiles of THE in different THE-amines molar ratios at (A) pH 7.4 + 0.2 and (B) 9.4 + 0.2, 37 0C (n =3+SD). Table 1 Experimental log D of theophylline (THE) and the THE mixtures (1:100 drug:counterion molar ratio) at pH 7.4 and 9.4, 37 0C. Values represent n=3 + SD. THE mixture

Log D at pH 7.4 + 0.2

Log D at pH 9.4 + 0.2

Parent drug

-0.028 + 0.02

-0.179 + 0.01

THE:ethylamine

-0.29 + 0.09

-0.74 + 0.05

THE:ethylenediamine

-0.35 + 0.017

-0.93 + 0.011

THE:spermidine

-0.39 + 0.03

-0.95 + 0.02

THE:spermine

-0.43 + 0.02

-0.99 + 0.04

The solubility results showed that increasing the concentration of counter-ions mixed with theophylline led to a significant increase (p < 0.05) in the drug solubility, presumably due to the increase in hydrophilicity of the drug (Figure 5). The trend of change across the different counterions was in agreement with the partitioning studies with THE-spermine demonstrating the greatest increase in theophylline solubility.

306


Drug Delivery to the Lungs 27, 2016 – Polyamine ion-pairs to target drugs to the lung

B 0.15

* *

**

**

** ** *

**

** **

Con. of theophylline (mM)

0.20

**

**

0.10 0.05 0.00

5

**

4 3

**

*

*

** **

**

**

2 1

**

**

**

0

TH

FR

E E A FR 0 EE E A . 25 M EA 0 . 5 M 1 E . ED A 0 M A 2.0 E D 0. 2 M A 5 ED 0 M A .5 ED 1 M . SP A 0 M D 2.0 S P 0. 2 M D 5 S P 0. M D 5 S P 1. M D 0 SP 2 . M 0 0 M S P . 25 M SP 0 . 5 M 1 SP . 0 2. M 0 M

E EA E T 0 HE EA . 2 5 M EA 0 . 5 M 1 E . ED A 0 M A 2. 0 E D 0. 2 M A 5 ED 0 M A .5 ED 1 . M SP A 2 0 M D .0 S P 0. 2 M D 5 SP 0 . M D 5 SP 1 . M D 0 S P 2. M 0 0 M SP . 2 5 M SP 0 . 5 M 1 SP . 0 2. M 0 M

Con. of theophylline (mM)

A

Figure 5 The aqueous solubility profiles of THE in increasing concentration of different amines solutions at 37 0C, (A) pH 7.4 + 0.2 and (B) 9.4 + 0.2. THE-theophylline, EA-ethylamine, EDA-ethylenediamine, SPD-spermidine and SPspermine.*Statistically significant (p < 0.05)** (p < 0.001) (One-way ANOVA) when compared to the solubility of free theophylline.

Conclusions THE was found to form ion-pairs with all the test amine counter-ions in water at pH 9.4 when the counterions were present in excess. Spermine showed the greatest binding affinity towards THE followed by spermidine, ethylenediamine and ethylamine. Forming polyamine ion-pairs with THE was shown to significantly decrease the drug’s lipophilicity and increase its aqueous solubility. When administered as an intravenous formulation the THEspermine ion pair should display a polar surface and this could lead to recognition and uptake by the PTS. The next phase of this research will continue to assess the uptake of THE-amines in in-vitro and in-vivo models in order to investigate how the changes seen in the physicochemical properties of THE when formed ion-pair with amines could effect its biopharmaceutical properties when deliver as an intravenous formulation. References 1. 2. 3.

4.

5.

Hoet, P. H. M., Nemery, B. (2000) Polyamines in the lung:polyamine uptake and polyamine-linked pathological or toxicological conditions. American Journal of Physiology. 278. 417-433. Kroll, F., Karlsson, J., Nilsson, E., Ryrfeldt, A., Persson, G. A. (1990) Rapid clearance of xanthines from airway and pulmonary tissues. The American Review of Respiratory Disease. 141. 1167-1171. Xi, H., Wang, Z., Chen, Y., Li, W., Sun, L., Fang, L. (2012) The relationship between hydrogen-bonded ion-pair stability and transdermal penetration of lornoxicam with organic amines. European Journal of Pharmaceutical Sciences. 47. 325-330. Song, W., Cun, D., Xi, H., Fang, L. (2012) The relationship between hydrogen-bonded ion-pair stability and transdermal penetration of lornoxicam with organic amines. European Journal of Pharmaceutical Sciences. 47. 325-330. Samiei, N., Mangas-Sanjuan, V., Gonzalez-Alvarez, I., Foroutan, M., Shafaati, A., Bermejo, M. (2013) Ion-pair strategy for enabling amifostine oral absorption: Rat in situ and in vivo experiments. European Journal of Pharmaceutical Sciences. 49. 499-504.

307


Drug Delivery to the Lungs 27, 2016 – Beatriz Fernandes et al. Paddle over disk as a dissolution test for orally inhaled drugs: discriminating composite from carrierbased formulations 1,2

1

1

2

Beatriz Fernandes , Filipa M. Maia , A. Mafalda Paiva , Maria Luisa Corvo & Eunice Costa

1

1

2

Hovione FarmaCiencia SA, Sete Casas, Loures, 2674-506, Portugal iMed.ULisboa, Faculdade de Farmácia, Universidade de Lisboa, Avenida Prof. Gama Pinto, Lisboa, 1649-003, Portugal

Summary Dissolution testing can be a way of discriminating between different pharmaceutical formulations. Even though dissolution testing of oral drugs is widespread and routinely used for both quality control and R&D to identify the influence of critical manufacturing variables on dissolution profiles and for in vitro-in vivo correlations, none of the standard USP apparatus are readily adapted for investigating inhaled products. Son et al. developed a potential standardized test method, the paddle over disk apparatus, which can be conjugated with the Next Generation Impactor (NGI) allowing for a dissolution profile assessment of a respirable fraction of the aerosol. The aim of this work was to investigate the discriminatory power of the newly developed paddle over disk apparatus when analysing dry powder formulations aimed for pulmonary delivery. The obtained results show that the apparatus can be successfully employed to discriminate between different formulations of dry powders, although further research is required to optimize the parameters used. Introduction One of the most important steps with in vitro performance testing of inhalation products is the characterization of the delivery of a given API from a specified inhaler using a pharmaceutical impactor/impinger, to estimate the actual dose that can potentially deposit on the target site of the lung. However, aerodynamic characterization does not completely describe the particles behaviour once inhaled. It completely misses the assessment of the drug absorption profile, which depends in great extent on the dissolution of the pharmaceutical dosage form. An ideal dissolution test procedure for inhaled formulations would involve particle classification followed by an evaluation of the dissolution behaviour for the classified drug particles that may deposit at various sites in the respiratory tract. Although dissolution testing for oral drugs is widespread and routinely used for both quality control and R&D to identify the influence of critical manufacturing parameters on dissolution profiles and for establishing in vitro-in vivo correlations, none of the standard USP apparatus 1, 2, 3 and 4 are readily adapted for assessing inhaled products. Few reports of powder or nanoparticle dissolution testing are based on apparatus suited for oral drugs, [1,2] , as the dry powders are difficult to disperse and studies failed to discriminate between formulations homogenously and tend to adhere to the walls and paddles. Several authors have attempted to develop systems [3-6] [7,8] capable of overcoming the mentioned drawbacks . Son et al. developed a potential standardized test method applicable to various formulations, as a variation of USP Apparatus 2 (paddle), designated paddle over disk apparatus. A stainless steel support disc is placed under the paddle to hold the test article at a precise distance from the bottom edge of the paddle. This method is amenable for inhaled products as particles with a known aerodynamic diameter can be collected using a NGI and can be directly positioned inside the vessel, guaranteeing that the tested powder is within the respirable fraction. The present work aims to utilize the paddle over disk apparatus to characterize dry powder inhaler (DPI) formulations of composite particles produced by spray drying and blend formulations of active pharmaceutical ingredient (API) and lactose aimed for pulmonary delivery, specifically. Experimental methods Aerosol formulation and characterization To produce two different composite powders, C1 and C2, solutions with 2% w/w of solids in a water/ethanol (50/50, % w/w) solvent mixture were prepared, according to Table 1. The powders were spray dried at an outlet temperature (Tout) of 95 ºC, a solution feed flow of 7 g/min, atomization pressure of 8 bar and atomization gas flow at 50 mm in the rotameter, using a Büchi model B-290 unit. The inlet temperature of the drying gas was adjusted to obtain the target outlet temperature (Tout). The particle size distribution (PSD) of the particles were analysed by laser diffraction (Sympatec). To obtain three different carrier based blend powders, B1, B2 and B3, homogeneous mixtures of 100 g of coarse (SV003) and fine (LH300) lactose with 1% w/w of micronized Fluticasone Propionate (FP) where prepared in a TURBULA® Shaker-Mixer, following geometric dilution of FP. Content of fine excipient in the prepared blends was of 5, 10 and 15% w/w for B 1, B2 and B3 respectively. For all formulations (C1, C2, B1, B2 and B3), HPMC size #3 capsules were hand filled with a target fill weight of 12.5 mg, with acceptance limits of +/- 0.5 mg. The aerodynamic performance was assessed by NGI. Composite particles were actuated with a Plastiape Monodose inhaler (40L/min at 4 kPa pressure drop) and its deposition profile determined with a gravimetric analysis of collected NGI deposits - gravimetric NGI - (n=3), using one capsule per replicate. Carrier based blend aerodynamic profile was assessed by chemical NGI (n=3) with 10 capsules per replicate, using PowdAir® inhaler (40L/min at 4 kPa pressure drop). Each formulation was also analyzed by scanning electron microscopy (SEM).

308


Drug Delivery to the Lungs 27, 2016 - Paddle over disk as a dissolution test for orally inhaled drugs: discriminating composite from carrier-based formulations Table 1 – Composition of the spray dried solutions.

C1 (g)

Water/Ethanol (50/50% w/w) Trehalose di-hydrated, from Pfanstiehl Raffinose pentahydrate, from amresco L-Leucine, from Merck Fluticasone Propionate

392 6.34 1.58 0.08

C2 (g) 392 6.34 1.58 0.08

Dissolution profile determination The DPI formulations were actuated in a gravimetric NGI following the conditions previously described in order to obtain particle separation and collect a known amount of API and a powder with a narrow and known aerodynamic particle size distribution. The collection was conducted by placing a dissolution cup assembled with an impactor insert in stage 4 of the NGI (Error! Reference source not found., left) - stage with greater amount of powder – and actuating 3 and 10 capsules for composite and blend formulations, respectively. Following the impaction, the stainless steel collector was removed from the NGI dissolution cup and covered with a pre-soaked polycarbonate membrane (Copley Scientific), which was sealed in place with the securing ring of the membrane holder, to be finally placed inside a dissolution vessel (Error! Reference source not found., right). To assess the dissolution profiles, a Tablet Dissolution Tester DIS 6000 from Copley Scientific was employed as USP apparatus 2 with and without a membrane holder. The apparatus consisted of vessels containing 350 mL of dissolution medium (0.01 M phosphate buffer saline from Sigma-Aldrich, with a pH of 7,4, containing 0.4% w/v of an anionic surfactant), maintained at 32 ± 0.5 ºC and stirred with paddles at 75 rpm, placed 25 ± 2 mm above the bottom of the vessel or the membrane holder. The described disposition allows for a complete submersion of the paddle, securing an effective drug dispersion after membrane release and continuous circulation of the medium in the vessel. For the collected powder, the membrane holder was placed carefully, as its orientation may influence the dissolution rate, and pre-warmed medium was added to the vessel. The dissolution profile of the hand filled capsules was also assessed by opening and pouring a pre-punched capsule (12.5 ± 0.5 mg) into a pre-warmed medium. Aliquots of 3 mL of dissolution medium were withdraw manually, filtrated and centrifuged at 120 000 rpm for 3 min, then replaced with pre-warmed medium. All experiments were done in duplicate and concentrations were determined using a Waters HPLC system with UV detection.

Figure 1 – Left: Modified NGI. A – Securing ring of the membrane holder; B – stainless steel collector and membrane holder; C – Dissolution cup; D – NGI with dissolution cup after actuation. Right: Dissolution vessel with membrane holder.

Results and Discussion The PSD analysis of the composite aerosols C1 and C2 showed a similar particle size distribution, with 90% of particles having a diameter below 3.3 and 3.2 µm, and a span of 2.0 and 2.1, respectively. Additionally, the SEM analysis (Figure 2, G-J) show a similar morphology for these formulations, as well as an alike particle agglomeration for the produced blends (Figure 2, A-F) – smaller agglomerates of fines (≈10 µm), and fines attached to the carrier lactose particles. The aerodynamic performance of the formulations is illustrated in Figure 3. Blends show an increase in fine particle fraction (FPF) with the percentage of fines, and composite particles present a similar FPF. Plate 4 was selected for powder collection for all formulations as it consistently displayed the maximum deposition of API and the collected particles are within the respirable fraction (aerodynamic cutpoint, D50 ≈ 3.5 < 5 µm). Considering the amount of FP collected in stage 4 for each formulation, the number of actuations was defined in order to obtain approximately 100 µg of FP in stage 4 (Table 2).

309


Drug Delivery to the Lungs 27, 2016 – Beatriz Fernandes et al.

A

B

C

D

E

F

G

H

I

J

60

A

50

5% fines (B1) 10% fines (B2) 15% fines (B3)

40 30 20 10 0

FP mass (µg/capsule)

FP mass (µg/capsule)

Figure 2 – SEM micrographs for blends with 5, 10 and 15% fines, B1 (A and B), B2 (C and D) and B3 (E and F), x300 and x1000 respectively; and for composites containing trehalose and raffinose, C1 (G and H) and C2 (I and J), x1000 and x 10 000 respectively; at 20 kV.

60

B

50

Raffinose (C2)

40

Trehalose (C1)

30 20 10 0

Stages

Stages

Figure 3 – API deposition at dose plate for blends B1, B2 and B3 (A) after 3 repetitions of 10 actuations at 40 L/min using PowdAir inhaler; for composites C1 and C2 (B) after 3 repetitions of one actuation at 40 L/min using Plastiape Monodose inhaler. MPA is mouth piece adapter and induction port; PS is pre-separator and MOC is micro-orifice collector. Table 2 – Amount of API loaded in plate 4 in one actuation (mstage4) and respective amount (mdisk) after Nactuations, in µg.

Disk C1 C2 B1 B2 B3

mstage4 (µg)

Nactuations

34.0 34.2 11.495 9.708 9.940

3 3 10 10 10

mdisk (µg) 102.0 102.6 115.0 97.1 99.4

Following dose collection, the release profiles of the collected powders was assessed (Figure 4, left). To better compare the diffusion phenomenon through the membrane in each formulation, the dissolution profile of capsules containing 12.5 mg of powder was also assessed (Figure 4, right). The paddle over disk results do not show a meaningful discrimination between the two composite formulations when calculating the similarity factor, f2=50.5>50, however, it shows an improvement when compared with the capsules’ profile obtained with the paddle apparatus, pointing to a possible optimization of the system in order to obtain discrimination. This might be achieved by varying operational parameters such as paddle speed, or changing the dissolution media. A significant difference is observable between the dissolution profiles of the blend formulations. These results might be due to the interaction between the fine particles of lactose and API, as this is the varying factor in the formulations tested, although differences in interaction between fines and carriers are not clearly visible in the [9] SEM micrographs. The influence of fine particles in formulation performance has been largely studied with two hypotheses being prominent in the literature.

310


Drug Delivery to the Lungs 27, 2016 - Paddle over disk as a dissolution test for orally inhaled drugs: discriminating composite from carrier-based formulations

100

100

80

80

60 40 20

Fluticasone Dissolved (%)

Fluticasone Dissolved (%)

The first one states that the fine lactose particles occupy the stronger binding sites of the carrier particles, and the second that the fines form agglomerates with the API and these agglomerates are more easily dispersed and disaggregated. The present results could be a consequence of a particle behaviour described by the second theory: agglomerates of lactose and API may be formed as the percentage of fines increases in the formulation, decreasing the API solubility due to local saturation in FP. Further studies would be required to understand fully dissolution mechanism of these blends. A third difference that can be analysed in the obtained results is the profile variation between composites and blend formulations. Looking at the B 1 profile containing only 5% of fine lactose (Figure 4, right), with apparent similar equilibrium solubility of the composites, it differs mostly on the dissolution rate. This can be explained by the structural state of the particles, as the crystalline state of the API contained in the blends is expected to be more stable than the amorphous state in the composite formulations, [10] which has been known to influence solubility .

60 40 20

0 0 100 200 300 200 300 Time (min) Time (min) Figure 4 – Left: Dissolution profile in the paddle over disk apparatus of composites C1 () and C2 (), and of blends B1 (▲), B2 () and B3 (●); right: dissolution profile in the paddle apparatus of hand filled capsules filled with of composites C1 () and C2 (), and of blends B1 (▲) and B2 (). Each profile is given by two replicates. 0

0

100

Conclusions The paddle over disk apparatus can be successfully employed to discriminate between different formulations of dry powders, and although the present results do not show the discriminating power regarding composite particles with similar morphology, an improvement was reached when comparing to the paddle apparatus, inspiring a design of experiments to achieve system optimization. The dissolution results of the blend formulations point towards significant differences between formulations with different percentages of fines; however, the exact dissolution mechanism is not yet fully understood and requires further research. References 1

Heng D, Cutler DJ, Chan H-K, Yun J, Raper JA. What is a Suitable Dissolution Method for Drug Nanoparticles? Pharm Res 2008;25(7):1696–701.

2

Sievens-Figueroa L, Pandya N, Bhakay A, Keyvan G, Michniak-Kohn B, Bilgili E, et al. Using USP I and USP IV for Discriminating Dissolution Rates of Nano- and Microparticle-Loaded Pharmaceutical Strip-Films. AAPS PharmSciTech 2012;13(4):1473–82.

3

Davies NM, Feddah MR. A novel method for assessing dissolution of aerosol inhaler products. Int J Pharm 2003;255(12):175–87.

4

Cook RO, Pannu RK, Kellaway IW. Novel sustained release microspheres for pulmonary drug delivery. J Control Release 2005;104(1):79–90.

5

Sdraulig S, Franich R, Tinker RA, Solomon S, O’Brien R, Johnston PN. In vitro dissolution studies of uranium bearing material in simulated lung fluid. J Environ Radioact 2008;99(3):527–38.

6

Kwon MJ, Bae JH, Kim JJ, Na K, Lee ES. Long acting porous microparticle for pulmonary protein delivery. Int J Pharm 2007;333(1-2):5–9.

7

Son YJ, McConville JT. Development of a standardized dissolution test method for inhaled pharmaceutical formulations. Int J Pharm 2009;382(1-2):15–22.

8

Son YJ, Horng M, Copley M, McConville JT. Optimization of an in vitro dissolution test method for inhalation formulations. Dissolution Technol 2010;17(2):6–13.

9

Jones MD, Price R. The influence of fine excipient particles on the performance of carrier-based dry powder inhalation formulations. Pharm Res 2006;23(8):1665–74.

10

Murdande SB, Pikal MJ, Shanker RM, Bogner RH. Solubility advantage of amorphous pharmaceuticals: I. A thermodynamic analysis. J Pharm Sci 2010;99(3):1254–64.

311


Drug Delivery to the Lungs 27, 2016 - Xiangyin Wei1 et al. Aerosol Performance and Stability Characteristics of Spray-Dried Tobramycin Excipient Enhanced Growth Inhalation Powder Formulations 1

1

3

Xiangyin Wei , Anubhav Kaviratna , Ruba S. Darweesh , P. Worth Longest

1,2

and Michael Hindle

1

1

Department of Pharmaceutics, Virginia Commonwealth University, Richmond, VA, 23298, USA Department of Mechanical and Nuclear Engineering, Virginia Commonwealth University, Richmond, VA, 23284, USA 3 Department of Pharmaceutical Technology, Jordan University of Science and Technology, Irbid, 22110, Jordan

2

Summary TM

TM

Despite advances in aerosol delivery with devices such as the TOBI Podhaler , there is still room to improve the antibiotic dose delivered to the lungs, especially if powder inhalers are to be used for pediatric patients. The excipient enhanced growth (EEG) technology enables high efficiency delivery of powder aerosols even at low inhalation flow rates. This study sought to characterize the aerosol performance and stability of new tobramycin EEG (TOB-EEG) formulations. TOB-EEG powders were spray dried and their chemical stability evaluated following storage at ambient temperature and low humidity conditions for 10 months. TOB-EEG formulations were aerosolized following exposure to ambient temperature and humidity for 15 minutes and following storage at ambient temperature and low humidity for 6 months. The primary particle size of TOB-EEG powders was assessed following storage at ambient temperature and low humidity for up to 15 months. There was no significant change in the tobramycin content (98.8% of the initial content) for a TOB-EEG mannitol formulation following 10 months storage. Results also showed that 15 minute exposure to ambient conditions for a TOB-EEG sodium chloride formulation did not affect its aerosolization characteristics. A small improvement in the dispersion properties was observed following 6 months storage, perhaps indicating it is important to remove residual moisture before packaging. The geometric primary particle size of TOB-EEG powders were similar (x50~1 µm) when tested following storage, suggesting the powders are stable over long-term when maintained in low humidity environments, as could be achieved with foil-wrapped capsules or other packaging methods. Introduction Chronic bacterial infection in the airways of cystic fibrosis (CF) patients is believed to cause progressive injury of (1, 2) . Repeated and intensive antibiotic treatment targeting the their respiratory tissues and loss of lung function major bacterial pathogen P. aeruginosa is recommended and has shown significant clinical benefits by increasing (1, 3, 4) the median survival of CF patients in the last decades . Inhaled antibiotics are desired as they better target drugs to the site of pulmonary infection and minimize their systemic exposure and side effects. Among the existing antibiotics, tobramycin has long post-antibiotic effect and low systemic toxicity, which makes it suitable to (5) be developed into high-dose inhalable antibiotic products . ®

Presently, tobramycin is commercially available as both inhalation solution (e.g. TOBI , Novartis AG, Switzerland) TM TM and inhalation powder (TOBI Podhaler , Novartis AG, Switzerland). While the powder formulation uses TM PulmoSphere technology to improve aerosol dispersion, which increases drug deposition in adults’ lung to 34% (6) (three times higher than the inhalation solution) , TOBI Podhaler requires CF patients to inhale at least 30 L/min (7) and is not recommended for children under 6 years of age . As P. aeruginosa infection may occur early in CF (2) patient’s life , a high efficiency tobramycin formulation - aerosol delivery device combination designed particularly for paediatric CF patients aged between 2-10 years may be beneficial for early treatment of P. aeruginosa infection and prevent pulmonary exacerbation. In this study, the initial development of new tobramycin powder formulations using the excipient enhanced growth (8) (EEG) technique is described . This EEG technique allows drug aerosols to be delivered with an initial small particle size to minimize device and upper airway deposition, while aerosol size increases in the lung through condensational growth to facilitate aerosol deposition and lung retention. The chemical and physical stability together with the aerosol performance of the tobramycin EEG (TOB-EEG) formulations were assessed to evaluate the robustness of the new formulations. Materials and Methods Preparation of TOB-EEG Powder (9)

TOB-EEG inhalation powders were prepared using the spray drying technique described in Son et al. . A spray drying solution was prepared by dissolving approximately 200 mg tobramycin (Spectrum Chemicals, New ® Brunswick, NJ), 190 mg sodium chloride (Fisher Scientific, Waltham, MA) or Pearlitol PF-Mannitol (Roquette Pharma, Lestrem, France), 100 mg L-leucine (Sigma-Aldrich, St Louis, MO) and 10 mg poloxamer 188 (BASF Corporation, Florham Park, NJ) in 100 mL 20% ethanol / 80% deionized water (v/v). The solution was spray dried using Buchi Nano Spray Dryer B-90 (BUCHI Corporation, New Castle, DE) with a 4 µm vibrating mesh spray nozzle. Liquid feed rate was set to 100%, drying airflow rate was controlled at 120 L/min and inlet temperature at 70 °C. After spray drying, the formulation was collected from the electrostatic precipitator and stored in a sealed amber vial inside a desiccator (22±2 °C and <10% RH).

312


Drug Delivery to the Lungs 27, 2016 - Aerosol Performance and Stability Characteristics of Spray-Dried Tobramycin Excipient Enhanced Growth Inhalation Powder Formulations

Assessment of Chemical Stability of TOB-EEG Powder (10)

A previously established LC-MS method was used to quantify tobramycin content and screen for additional degradation peaks in the TOB-EEG formulations immediately following spray drying and after storage at ambient temperature and low humidity conditions (22±2 °C and <10% RH) for 10 months. 1 mg TOB-EEG powder was dissolved in 1 mL deionized water and injected into the LC-MS. For reference purposes, a freshly prepared solution containing a mixture of the drug and excipients was prepared at a similar concentration and used as a control. Analyses was performed using Alliance e2695 HPLC (Waters, Milford, MA) coupled with Micromass Quattro micro API Mass Spectrometer (Waters, Milford, MA). Chromatographic separation was performed using a Hypersil Gold C18 column (150×4.6mm, 3μm; Thermo Scientific, Waltham, MA) and mobile phase consisting acetonitrile:0.1% trifluoroacetic acid (TFA) (30:70). Isocratic elution was used at a flow rate of 0.3 mL/min and sample injection volume was set at 10 µL. Electrospray ionization was used with a positive mode. An ion scan was performed at a range from 100 to 900 for each sample. For tobramycin quantification, single ion monitoring at 468.5 daltons was employed. Other parameters were: capillary voltage 3.78 kV; cone voltage 20 V; extractor 6 V; Rf Lens 0.1 V; source block temperature 130 °C; desolvation temperature 350 °C; nitrogen desolvation flow 600 L/min. Mass Lynx 4.0 (Waters, Milford, MA) was used for HPLC-MS control and data processing. Aerosol Characterization and Assessment of Physical Stability of TOB-EEG Powder Physical stability of TOB-EEG powder was first evaluated by characterizing its aerodynamic particle size ® distribution (APSD) delivered from Handihaler device (Boehringer Ingelheim, Ingelheim, Germany) using Next Generation Impactor (NGI; MSP Corporation, Shoreview, MN). Experiments were designed to assess (a) stability of TOB-EEG powder inside the capsule after piercing and being exposed to ambient temperature and humidity (22±2 °C and 40±5%RH) for 15 minutes; this was designed to evaluate the robustness of the TOB-EEG formulation when used in a practical scenario; and (b) stability of TOB-EEG formulations after being stored at room temperature in a desiccator for 6 months (22±2 °C and <10% RH; similar conditions to aluminium-sealed packaging). Prior to each experiment, the pre-separator and NGI collection plates were coated twice using Molykote® 316 Silicone Release Spray (Dow Corning, Midland, MI) to avoid particle re-entrainment. Approximately 2 mg TOB-EEG powder was filled into a size 3 hydroxypropylmethyl cellulose (HPMC) capsule (Capsugel, Peapack, NJ) and placed into the Handihaler device. The 2 mg dose was selected for these initial development and stability studies with the Handihaler, although it is recognized that higher doses are required for therapeutic use. The capsule was then pierced and device inserted into a mouth-piece adapter designed to ensure air tightness between the inhaler and pre-separator/NGI. The USP inlet was omitted in all cases to determine the size distribution of the total emitted aerosol. Airflow rate was controlled at 45 L/min (equivalent to 4 kPa pressure drop across the inhaler) and pulled for 5.3 sec following procedures described in the United States (11) Pharmacopeia (USP) . Tobramycin deposition on the device, capsule, mouth-piece adapter, pre-separator and NGI plates was collected using known volumes of acetonitrile:0.4% trifluoroacetic acid (TFA) (80:20) and assayed using a HPLC-MS method described above. Single doses were used and each protocol was performed for at least three replicates. Emitted dose (ED) was calculated by subtracting drug mass remaining inside the device and capsule from the total recovered mass. Fine particle fraction (FPF5µm/ED) and submicrometer fine particle fraction (FPF1µm/ED) were calculated for the fraction of particles smaller than 5 µm and 1 µm, respectively, using linear interpolation from the cumulative percent under size vs. NGI stage cut-off diameter profiles and normalized to ED. Mass median aerodynamic diameter (MMAD) was also calculated using the linear interpolation approach. Geometric particle size distributions of TOB-EEG formulations were also determined using Sympatec laser diffraction instrument installed with ASPIROS/RODOS dry dispersing unit and HELOS laser diffraction sensor (Sympatec GmbH, Clausthal-Zellerfeld, Germany). The powders were sized immediately following spray drying or stored at room temperature in a desiccator for 6 or 15 months (22±2 °C and <10 %RH). This technique was used to evaluate the long-term stability of the TOB-EEG formulations. The lens and measuring range was selected as R1:0.1 / 0.18-35 µm, and the disperser was initially set at 4 bar with a feed rate of 60 mm/s. Background was removed by conducting a reference run each time and three measurements were performed for each powder. WINDOX 5 software (Sympatec GmbH) was used for instrument control and data evaluation. Results and Discussion Chemical Stability of TOB-EEG Powder There was no significant change in the tobramycin content uniformity for a TOB-EEG mannitol formulation following 10 month storage, with the mean (SD) content of 98.8 (9.3) % of the initial content. However, it was observed that there was greater variability in the content at the 10-month time point compared to the initial time point. For comparison, a similar study with two albuterol-EEG formulations with mannitol as the hygroscopic excipient showed mean (SD) drug contents (% of initial) of 97.8 (1.6) % after three years and 100.6 (0.5) % after two years storage. Examination of the mass spectral chromatographic profile following 6 months storage of the TOB-EEG formulation with sodium chloride as the hygroscopic excipient revealed no additional peaks when + searching for [M+H] ions at m/z 484, 527, 323 and 307 corresponding to known tobramycin potential degradation (12) , respectively. While the present products kanamycin B, 6’’-O-carbamoylkanamycin B, neamine and nebramine HPLC-MS method is sufficient for quantifying tobramycin (and possibly its potential degradation products), the formulation contains excipients including sodium chloride or mannitol, L-leucine and poloxamer 188. More robust assay methods will be needed in the future to characterize both tobramycin and the excipients to fully evaluate the chemical stability of TOB-EEG powder over shelf-life.

313


Drug Delivery to the Lungs 27, 2016 - Xiangyin Wei1 et al. Aerosol characterization of TOB-EEG Powder The aerodynamic properties of a TOB-EEG formulation with sodium chloride as the hygroscopic excipient were evaluated. The results for the three testing conditions are summarized in Figure 1 and Table 1: (a) 5 days storage at ambient temperature and low humidity (22±2 °C and <10 %RH; Control); (b) Exposure to ambient temperature and RH for 15 minutes (22±2 °C and 40±5%RH) immediately prior to aerosolization following storage under control conditions (a) for 5 days (to evaluate the effect of ambient humidity during inhaler use); (c) 6 months storage at ambient temperature and low humidity (22±2 °C and < 10 %RH; to evaluate the effect of 6 months storage);

Figure 1. Effects of storage conditions on the aerosolization characteristics of TOB-EEG sodium chloride formulation, shown as the % tobramycin (% total recovered dose) when tested using Handihaler. Data presented as mean and standard deviation (n≥3). “Device” includes Handihaler and capsule; “PS”, pre-separator, includes mouth-piece adapter and pre-separator; “S1-S7”, NGI stages 1-7; “F”, filter.

All three testing conditions produced highly dispersed particles with relatively good device emptying (ED > 75%) and notably large fine particle fractions (FPF5µm/ED > 72%) as shown in Table 1. Statistical analyses using Student’s t test showed that exposure to ambient conditions for 15 minutes had no effect on aerodynamic properties of TOB-EEG powder; the distribution of tobramycin on Handihaler device, pre-separator and NGI stages produced no significant differences compared to the control group (Figure 1; p>0.05). Surprisingly, the TOB-EEG powder stored for 6 months produced slightly smaller particles than the control group (MMAD = 1.7 µm and 1.9 µm, respectively; Table 1), and statistically significant differences were observed for tobramycin distribution on the pre-separator and NGI stages 1, 6 and 7 between the 6 months group and control group (Figure 1; Student’s t-test, p<0.05). It is hypothesised that the freshly spray-dried TOB-EEG formulation may contain residual moisture that slightly affects the powder dispersity. Moisture content in the spray-dried formulations and its potential effects on powder dispersity may need to be quantified in future experiments. In addition, residual moisture may need to be removed from formulation powders before storage. Table 1. Effects of storage conditions on aerodynamic properties of TOB-EEG sodium chloride powders. ED, emitted dose (% total recovered dose); FPF5µm/ED, fine particle fraction of drugs smaller than 5 µm (% ED); FPF1µm/ED, fine particle fraction of drugs smaller than 1 µm (% ED); MMAD, mass median aerodynamic diameter. Data are presented as mean and standard deviation (n≥3).

Storage Conditions Ambient temp/low RH – 5 days Ambient temp/low RH - 5 days Ambient temp/low RH - 6 month

Exposure to Ambient Conditions No 15 min No

ED (%)

FPF5µm/ED (%)

FPF1µm/ED (%)

MMAD (µm)

75.8 (2.0) 77.5 (5.1) 75.0 (0.7)

72.8 (6.7) 73.0 (8.0) 83.7 (5.0)

12.2 (3.4) 14.4 (7.1) 21.0 (2.1)

1.9 (0.2) 1.8 (0.3) 1.7 (0.1)

Geometric particle size distributions of TOB-EEG powders initially and after 6 months and 15 months storage are summarized in Table 2. While these powders are from mixed batches and contain different hygroscopic excipients (mannitol or sodium chloride), data shows that the powders produced consistently small particles with the volume (mass) median diameters between 1.0 to 1.1 µm and similar polydispersity. This suggests that the spray drying techniques used for produce TOB-EEG powders were robust and the formulations appear to be stable over longterm storage.

314


Drug Delivery to the Lungs 27, 2016 - Aerosol Performance and Stability Characteristics of Spray-Dried Tobramycin Excipient Enhanced Growth Inhalation Powder Formulations

Table 2. Geometry diameter of TOB-EEG powders after storage at ambient temperature and low humidity (<10% RH) for 6 or 15 months. Data are presented as mean and standard deviation (n=3).

TOB-EEG mannitol TOB-EEG sodium chloride TOB-EEG mannitol

Storage Period

X10 (µm)

X50 (µm)

X90 (µm)

Initial size

0.5 (0.0)

1.1 (0.0)

2.4 (0.0)

6 Months

0.5 (0.0)

1.0 (0.0)

2.0 (0.0)

15 Months

0.5 (0.0)

1.0 (0.0)

2.1 (0.0)

Conclusions The preliminary data showed TOB-EEG powders are well dispersed and generally stable when exposed to 15 minutes ambient conditions and after long-term storage in the desiccator. The powder has potential to be used for treating paediatric CF patients aged between 2-10 years, and further studies will be needed to optimize the formulation and storage conditions. More robust HPLC-MS methods will also be needed to fully characterize the formulation and facilitate detection of potential degradation products from both tobramycin and the excipients. Acknowledgements The authors acknowledge Shinal Chandrakant Patel for measuring the geometric particle size distributions of TOB-EEG powders. References 1. 2. 3. 4.

5. 6. 7. 8. 9. 10. 11. 12.

Doring G, Flume P, Heijerman H, and Elborn JS. Treatment of lung infection in patients with cystic fibrosis: current and future strategies. Journal of Cystic Fibrosis : Official Journal of the European Cystic Fibrosis Society. 2012;11: 461-479. Doring G, Conway SP, Heijerman HG, Hodson ME, Hoiby N, Smyth A, and Touw DJ. Antibiotic therapy against Pseudomonas aeruginosa in cystic fibrosis: a European consensus. The European Respiratory Journal. 2000;16: 749-767. Mogayzel PJ, Jr., Naureckas ET, Robinson KA, Mueller G, Hadjiliadis D, Hoag JB, Lubsch L, Hazle L, Sabadosa K, and Marshall B. Cystic fibrosis pulmonary guidelines. Chronic medications for maintenance of lung health. American Journal of Respiratory and Critical Care Medicine. 2013;187: 680-689. Mogayzel PJ, Jr., Naureckas ET, Robinson KA, Brady C, Guill M, Lahiri T, Lubsch L, Matsui J, Oermann CM, Ratjen F, Rosenfeld M, Simon RH, Hazle L, Sabadosa K, and Marshall BC. Cystic Fibrosis Foundation pulmonary guideline. pharmacologic approaches to prevention and eradication of initial Pseudomonas aeruginosa infection. Annals of the American Thoracic Society. 2014;11: 1640-1650. Smith AL. Inhaled antibiotic therapy: What drug? What dose? What regimen? What formulation? Journal of cystic fibrosis : Official Journal of the European Cystic Fibrosis Society. 2002;1: 189-193. Geller DE, Weers J, and Heuerding S. Development of an inhaled dry-powder formulation of tobramycin using PulmoSphere technology. Journal of Aerosol Medicine and Pulmonary Drug Delivery. 2011;24: 175-182. Vandevanter DR and Geller DE. Tobramycin administered by the TOBI® Podhaler® for persons with cystic fibrosis: a review. Medical Devices (Auckland, N.Z.). 2011;4: 179-188. Hindle M and Longest PW. Condensational growth of combination drug-excipient submicrometer particles for targeted high-efficiency pulmonary delivery: evaluation of formulation and delivery device. The Journal of Pharmacy and Pharmacology. 2012;64: 1254-1263. Son YJ, Worth Longest P, and Hindle M. Aerosolization characteristics of dry powder inhaler formulations for the excipient enhanced growth (EEG) application: effect of spray drying process conditions on aerosol performance. International Journal of Pharmaceutics. 2013;443: 137-145. Li M. Tobramycin disposition in the lung following airway administration. (Doctoral Dissertation), Virginia Commonwealth University: Richmond, VA; 2013. U.S. Pharmacopeial Convention. General chapter <601> aerosols, nasal sprays, metered dose inhalers, and dry powder inhalers. United States Pharmacopeia and National Formulary (USP 36-NF 31). Vol 1. U.S. Pharmacopeial Convention: Rockville, MD; pp. 242-262, 2013. Li B, Van Schepdael A, Hoogmartens J, and Adams E. Characterization of impurities in tobramycin by liquid chromatography-mass spectrometry. Journal of Chromatography. A. 2009;1216: 3941-3945.

315


Drug Delivery to the Lungs 27, 2016 - Niklas Renner et al. A deeper insight into the impact of chemical surface properties on inhalation performance 1

Niklas Renner , Hartwig Steckel², Nora A. Urbanetz³ & Regina Scherließ 1

1

Department of Pharmaceutics and Biopharmaceutics, Kiel University, Grasweg 9a, 24118 Kiel, Germany ²Deva Holding A.S., Istanbul, Turkey; ³Daiichi Sankyo, Tokyo, Japan

Summary Glass beads (GBs) are ideal to be used as model carriers in dry powder inhalers as they can be selectively surface modified without altering other physico-chemical properties. In this study, GBs in the size range of 400600 µm were silanised with agents varying in their functional groups thus conferring a broad range of hydrophobicity to the carrier surface, which was determined via contact angle (CA) measurements. Additionally the surface energy (SE) of the carrier particles was measured. Modified and untreated GBs were blended with spray dried budesonide (BUD), serving as a hydrophobic model drug to form interactive mixtures for inhalation. This study expands knowledge gained from the previous study by correlating aerodynamic performance and drug loading to carrier surface characteristics. On the one hand, surface modification had a substantial effect on the actual surface coverage (ASC) as it showed a direct positive correlation to the measured CA. On the other hand, aerodynamic performance was altered by the chemical surface properties of the carrier as a high degree of hydrophilicity (low CA) led to the highest fine particle fraction (FPF). The SEs of not only the carrier but also the API proved to have an impact on both output parameters. Introduction Drug delivery by inhalation is distinctively influenced by interparticle forces between carrier and active [1] pharmaceutical ingredient (API) . Those forces are governed by various factors concerning the carrier including [2] [3] and surface topography . The authors have already presented the effect of drug loading on particle size [4] aerodynamic performance using modified GBs . The present study extends previous investigations in order to provide a profound understanding of how carrier surface characteristics influence aerodynamic performance. Therefore, GBs which have been proven to be suitable as model carriers were subjected to a treatment with different silanes to provide a broad range of hydrophobicity. BUD was selected as hydrophobic model API. Since the respirable fraction has been shown to be drug loading dependent and, in addition, the ASC varied with carrier surface modification, the aerodynamic performance of blends containing different modified glass beads cannot be directly compared amongst each other in an appropriate manner. To overcome this issue, blends with three different theoretical surface coverages were prepared for every type of GB and investigated via impaction analysis. Based on data gained from these trials and ASCs, data was normalized by calculating the resulting FPF at a defined ASC of 30 %. Those results can subsequently be correlated to carrier surface characteristics. Experimental methods Materials Crystalline budesonide was purchased from Minakem SAS, Dunkerque, France. Glass beads (SiLibeads® Type S) in the size range of 400-600 µm were kindly provided by Sigmund Lindner GmbH, Warmensteinach. Trimethoxy(3,3,3-trifluoropropyl)silane (FPTS), Chloro(methyl)diphenylsilane (CDPMS) and Chlorotriphenylsilane (TPCS) were purchased from Sigma Aldrich, St. Louis, USA. Spray drying Budesonide (BUD) was dissolved in methylene chloride and processed using a Büchi Mini Spray Dryer B290 (Büchi Labortechnik AG, Switzerland) equipped with a high performance cyclone and a Büchi B295 Inert Loop as [4] published previously . Silanisation of glass bead surfaces Prior to the actual silanisation, the GBs were pretreated with Piranha solution (H2SO4:H2O2 3:1) to assure a comparable cleanness of the surfaces. GBs were then treated with an ethanolic solution of the respective silane for 30 min. After incubation, they were filtered and rinsed several times with purified water to wash off residual, unreacted amounts of the silane. Determination of contact angle and surface energy Contact angle (CA) measurements were conducted at room temperature with an OCA 20 contact angle meter (DataPhysics Instruments GmbH, Filderstadt, Germany). The instrument was coupled to a USB CCD-camera. Inverse gas chromatography (iGC) was used to determine the SE. Samples were filled into a commercially available glass pipe of 4 mm diameter and analysed with an SMS Inverse Gas Chromatograph (Surface Measurement Systems, London, UK).

316


Drug Delivery to the Lungs 27, 2016 - A deeper insight into the impact of chemical surface properties on inhalation performance Preparation of API/carrier blends Interactive mixtures were produced for calculated surface coverages of 25 %, 50 % and 100 % according to [5] Zellnitz et al. . In brief, 15 g of glass beads and the calculated quantity of BUD were weighed in a stainless steel mixing vessel via the double sandwich method. The blending was performed with a Turbula Blender T2C (Willy A. Bachofen AG Maschinenfabrik, Muttenz, Switzerland) for 30 min at 20 rpm. Only homogeneous blends (RSD of blends homogeneity <5%) were used for further experiments. Aerodynamic assessment The performance of interactive mixtures was evaluated based on impaction analysis using the Next Generation Pharmaceutical Impactor (NGI) (Copley Scientific, Nottingham, United Kingdom) described as Apparatus E in the ® European Pharmacopoeia 8.0 and the Cyclohaler as inhalation device with a fixed flow rate of 100 L/min. 250 mg of the respective blends were filled into hard gelatin capsules of size 3 (Capsugel, Colmar, France) manually using a spatula. For blends with a calculated coverage of 50 % and 100 % the content of three capsules was released in one run, while for blends with 25 % coverage 5 capsules were used in order to ensure sufficient drug for quantification. All NGI experiments were performed in triplicate. Gained data was evaluated using the CITDAS software 3.1 (Copley Scientific, Nottingham, United Kingdom). All trials were conducted in a conditioned environment (20 °C, 35 %RH). HPLC analysis Budesonide quantification was done by HPLC on a Waters 600 (Waters Corporation, Milford, USA) and a ® LiChrospher RP-18 column (Merck KGaA, Darmstadt, Germany). The mobile phase consisted of 75% (v/v) methanol and 25% (v/v) purified water. The flow rate and column temperature were set to 0.8 ml/min and 20°C, respectively. A sample volume of 100 µl was injected and double determination was conducted. The peaks were detected at 220 nm. Prior to sample analysis a calibration of seven points was created confirming linearity in the range of 0.5 µg/ml to 100 µg/ml. Determination of actual surface coverage (ASC) The actual mass ratio of budesonide and glass beads was quantified via HPLC. By comparing these results to the mass ratio for the calculated surface coverages, the ASC could be determined. Normalisation of FPF As already mentioned, the theoretical FPF at 30 % ASC was determined for the different GBs. This was done by plotting the FPF against its corresponding ASC for every type of glass bead. Afterwards the resulting FPF was determined with the help of a regression line. Results and Discussion Spray drying produced spherical API particles with a median particle size (x50) of 3.1 µm (x10=0.8 µm; x90=7.8 µm). Data obtained from X-ray Powder Diffraction showed an amorphous state for the utilised batch and its stability at 20 °C/35 %rH was proven by Dynamic Vapor Sorption (data not shown). The silanisation process of GBs created surfaces of varying hydrophobicity as illustrated by the respective contact angle. The selected GBs and their corresponding contact angle can be taken from Fig. 1 (left).

Figure 1 - Actual surface coverages for calculated coverages of 100 %, 50 % and 25 % plotted against contact angle (left) and surface energy (right)

317


Drug Delivery to the Lungs 27, 2016 - Niklas Renner et al. Influence of chemical properties on drug loading Fig. 1 also illustrates the dependence of actual (effective) surface coverage (ASC) on contact angle (left) and dispersive surface energy (right). Here, the ASC is correlated to the CA in positive manner, as indicated by the ascending arrow. This observation is conclusive in terms of chemical properties of carrier and the hydrophobic API. The relationship between coverage and SE appears prima facie to be non-existent. Including the SE of BUD, a large difference between SEs of carrier and API seems to be favourable to reach high drug loading. While blends prepared with TPCS 0.05M exhibited an ASC of 87.3 %, GB_UT only gained 54.8 %. These findings are also supported by SEM micrographs (Fig. 2). Another study which employed formoterol fumarate as API showed results contrary to those findings. Here, a small discrepancy between SEs of the two components was necessary to obtain optimal drug loading displaying the relevance of physico-chemical nature of the API (data not shown). This is under further investigation.

Figure 2 - SEM photographs of blends containing BUD and (I) untreated glass beads or (II) glass beads modified with TPCS 0.05M with different surface coverages of 25 % (a), 50 % (b) and 100 % (c)

Aerodynamic assessment Fig. 3 (left) displays the effect of CA on inhalation performance, where the FPF decreased with increasing CA illustrated by the descending arrow. Here again, chemical properties provide a logical explanation. The pronounced hydrophilicity of GB_UT facilitates drug detachment and consequently leads to the highest FPF, while with increasing contact angle (increasing hydrophobicity) attractive forces between GBs and BUD are extended. This leads to the lowest respirable fraction for the most hydrophobic carriers (FPTS 0.05M). As seen in Fig. 3 (right) similarity in SEs of carrier and budesonide as API obviously benefits aerodynamic performance as it is the case for GB_UT.

Figure 3 - Resulting FPFs at 30% ASC plotted against CA (left) and SE (right)

318


Drug Delivery to the Lungs 27, 2016 - A deeper insight into the impact of chemical surface properties on inhalation performance Conclusion Silanisation proved to be a suitable technique to alter chemical surface properties of the model carrier glass beads as it was possible to obtain varying contact angles and dispersive surface energies. Both factors have a substantial effect not only on the effective drug loading but also on aerodynamic performance. While the contact angle is correlated in a positive and negative way directly to ASC and FPF, respectively, the influence of SE has shown to be more complex. Here, the relation between SEs of API and carrier seems to be more important than the absolute values. Acknowledgements The authors would like to thank the DFG for funding this project within the priority program SPP 1486 “Particles in contact” and Ann-Kathrin Muhs for her technical support. The authors also want to express their gratitude towards Zinaida Todorova from Otto-von-Guericke University Magdeburg for conducting the silanisation process step.

References 1

Kou X, Chan L W, Steckel H, Heng P: Physico-chemical aspects of lactose for inhalation, Adv Drug Deliv Rev 2012; 64: pp 220-232.

2

Steckel H, Mueller B W: In vitro evaluation of dry powder inhalers II: influence of carrier particle size and concentration on in vitro deposition, Int J Pharm 1997; 154: pp 31-37.

3

Zeng X M , Martin G P, Marriott C, Pritchard J:.The influence of carrier morphology on drug delivery by dry powder inhalers, Int J Pharm 2000; 200: pp 93–106.

4

Renner N, Scherließ R, Steckel H: Modified glass beads as model carriers to understand the performance of interactive powder blends (Abstract). Presented at: Drug deliver to the lungs 26,Edinburgh, Scotland, December 9-11, 2015; J Aerosol Med.

5

Zellnitz S, Schroettner H, Urbanetz N A: Influence of surface characteristics of modified glass beads as model carriers in dry powder inhalers (DPIs) on the aerosolization performance, Drug Dev Ind Pharm 2015; 41: pp1710-1717

319


Drug Delivery to the Lungs 27, 2016 - Alex Slowey et al. 60 Years of the MDI – A History of Innovation 1

Alex Slowey & Steve Stein 1

2

3M Drug Delivery Systems, Morley Street, Loughborough, Leicestershire, LE11 1EP, UK. 2 3M Drug Delivery Systems,3M Center,St. Paul, MN 55144-1000, USA.

Summary th

2016 marks the 60 anniversary of the launch of the pressurised metered dose inhaler (MDI). In April of 1955, a young girl named Susie Maison, unhappy with her squeeze bulb nebulizer that she used to treat her asthma, asked her father, George Maison, MD., President of Riker Laboratories, the question, “Daddy, why can’t they put my asthma medicine in a spray-can like they do hair spray?” This simple question led to a serendipitous convergence of circumstances that resulted in the development of an important new therapy, the MDI. On January 12, 1956 New Drug Applications were filed for Medihaler Epi (containing epinephrine) and Medihaler Iso (containing isoproterenol). With the approval of Medihaler Epi, and Medihaler Iso on March 9, 1956, both products were launched that same month. Since its introduction, the MDI has developed to include breath-actuation to ensure better patient co-ordination, reformulation following the link between CFC’s and ozone layer depletion. More recently, with the expiration of relevant patents, the generics market has been opened, allowing more cost effective products to be launched. The introduction of high-tech inhalers with enhanced capabilities offers a number of benefits from both a compliance and patient outcome point of view. The MDI has developed significantly since its introduction in 1956. Over the 60 years since its launch, the MDI has continued to evolve. Today and into the future, as an established dosage form, it remains a major platform in the delivery of drug to the lungs. Introduction th

2016 marks the 60 anniversary of the launch of the pressurised metered dose inhaler (MDI). In April of 1955, a young girl named Susie Maison, unhappy with her squeeze bulb nebulizer that she used to treat her asthma, asked her father, George Maison, MD., President of Riker Laboratories, the question, “Daddy, why can’t they put my asthma medicine in a spray-can like they do hair spray?” This simple question led to a serendipitous [1] convergence of circumstances that resulted in the development of an important new therapy, the MDI . This review poster aims to look at the long history of the MDI and presents some key points in its evolution to the present day and beyond. In order to prepare this review article, a literature review was performed and the results summarised below. Results Within two months of Susie’s question, Riker Laboratories began clinical testing on MDI formulations of isoproterenol and epinephrine using solution formulations (developed by Irving Porush of Riker) containing a mixture of CFCs Freon 12™ and Freon 114™ with 35% w/w ethanol. On January 12, 1956 New Drug Applications were filed for Medihaler Epi (containing epinephrine) and Medihaler Iso (containing isoproterenol). With the approval of Medihaler Epi, and Medihaler Iso on March 9, 1956, both products were launched that same month. Figure 1 shows a photograph of the first MDI.

Figure 1: Image of Medihaler-Iso

320


Drug Delivery to the Lungs 27, 2016 - 60 Years of the MDI – A History of Innovation Acceptance of the new delivery system by physicians and patients was rapid and in December of 1956 the first clinical evaluation of the MDI was published. Freedman evaluated 42 asthmatics, most of whom had failed to respond to other therapeutic measures, and reported that 30 of the patients using MDIs obtained good to [2] excellent relief of symptoms and five others obtained fair relief . [3]

Over the following years, various drugs were developed for the MDI platform , including MDIs containing drugs for systemic delivery, for example, Medihaler Ergotamine was approved for the treatment of migraine headache in 1959. The first combination MDI for the treatment of asthma, Medihaler Duo, delivering isoproterenol [3] hydrochloride and phenylephrine bitartrate and was approved in 1962 . Through the 1970s and 1980s, MDIs continued to grow in popularity with the introduction of several important MDI products. Some key products are summarised in Table 1. Year 1972 1973 1981 1982 1984 1986

Developer Allen & Hanburys Boehringer Ingelheim Glaxo Wellcome Schering Corporation Glaxo Wellcome Schering Corporation Forest Boehringer Ingelheim

Drug Beclomethasone dipropionate Metaproterenol sulfate Salbutamol sulfate Beclomethasone dipropionate Flunisolide Ipratropium bromide

Brand Name Becotide® Alupent® Ventolin® Proventil® Becloven ® Vanceril® Aerobid® Atrovent®

Table 1: Summary of Key CFC Product Launches in the 1970s and 1980s The Hole in the Atmosphere and The Montreal Protocol: MDIs became the preferred dosage form for the delivery of therapeutic aerosols to the lung. However, the extremely inert nature of CFC propellants used in MDIs (and in many other industrial applications at the time) [4] enabled these molecules to diffuse over time into the upper stratosphere. In 1974 Molina and Rowland demonstrated that CFC propellants break down and release chlorine radicals when exposed to sunlight in the upper stratosphere. These chlorine radicals were shown to have the ability to break down very large numbers of ozone molecules. In the mid-1980s, evidence mounted that stratospheric ozone levels were decreasing at an alarming level and that CFCs were contributing significantly to this depletion. The Montreal Protocol was signed in 1987 and called for CFC propellants to be eliminated by January 1996. Although orally inhaled MDIs were exempt from this ban until ‘medically acceptable alternatives’ were available, the impact on the pharmaceutical aerosol industry of the Montreal Protocol was dramatic. The Development of HFA MDIs: Pharmaceutical companies began to search for alternative propellants to replace CFCs. Non-chlorinated hydrofluoroalkanes (HFAs) were identified as promising candidates and HFA-134 and HFA-227 were quickly identified as potential replacements. Several industry consortia (eg IPAC-I and IPAC-II) were established to collaboratively generate the data needed to demonstrate the safety of these new propellants to the level expected [5] based on heightened regulatory requirements . [6]

The transition from CFC to HFA propellants involved significant investment in research and development , not only to develop new formulations of existing APIs, but to develop new products too. The renewed investment required to transition to HFA propellants resulted in improvements to the MDI device components as well. New valve elastomers were developed, such as EPDM, which were cleaner and resulted in [7] significantly reduced leachables levels . Coated canisters were developed to minimize drug deposition that is prevalent for ethanol-free HFA suspension formulations and to reduce chemical degradation for solution MDI [7, 8, 9, 10] . formulations The first HFA MDI product to reach the market was a salbutamol sulfate suspension in HFA-134a developed by 3M. The product was approved in the UK in 1994 and commercialised in 1995 as Airomir®. The product was approved and launched in the US in 1996 as Proventil® HFA. The product was designed to provide similar drug delivery to the CFC product. The first HFA corticosteroid to reach the market was Qvar®, an HFA-134a formulation of beclomethasone dipropionate commercialised in 2000. In 2001, GSK received approval from the FDA of Ventolin® HFA which was a suspension of salbutamol sulfate and free of surfactant and ethanol. The first HFA-227 MDI to receive approval from the FDA was Astra Zeneca’s Symbicort® (budesonide and formoterol fumarate) which was approved in 2006. A significant number of other HFA MDI products are approved in other regions of the world. In the UK there are over 30 MDIs licenced and marketed. MDIs remain the most widely used delivery system for treating respiratory diseases with more than 75 [11] billion doses sold in 2014 .

321


Drug Delivery to the Lungs 27, 2016 - Alex Slowey et al. In 2004, GSK launched Seretide™ Evohaler™ (salmeterol and fluticasone propionate) in Europe. This was the first MDI with an integrated dose counter incorporated into the device to help patients know when their device has exceeded or is near the last available dose. The Benefit of Breath Actuation Traditionally, one of the key limitations of MDIs is the difficulty that some patients have coordinating inspiration and the actuation of the device. Even back in 1956, Freedman indicated that the primary cause of poor response with Medihaler was the failure of physicians to stress to the patient “the importance of synchronization of [2] inspiration with the administration of the dose” . Developments have been introduced over the years to incorporate breath actuation of the MDI, the use of valved-holding chambers and also devices to enable better patient training. Figure 2 shows an example breath actuated inhaler and the In-Check M slow inspiratory flow (SIF) training device developed by Clement Clarke. Breath actuated devices have been shown to significantly [12] improve dose delivery for patients who struggle with coordination .

[13]

Figure 2: Image of a breath actuated inhaler, 3M’s Autohaler, and Clement Clarke’s In-Check M

training device

Looking to the Future MDI’s have been a mainstay treatment of asthma and chronic obstructive pulmonary disease (COPD) for 60 years, in part due to their relatively low cost, patient familiarity and portability. Looking to the future, generic inhalers and next generation high-tech inhalers with enhanced functionality will contribute to changes in the MDI market. In 2015, Sirdupla™, a generic version of Seretide® Evohaler® was developed and launched by Mylan [14,15] . and 3M

Figure 3 Image of Sirdupla™ The introduction of high-tech inhalers with enhanced capabilities offers a number of benefits from both a compliance and patient outcome point of view. A number of high-tech add-on devices for MDI, DPI, and Nebulizer products have been commercialized for MDI products (Eg MD Turbo developed by Respirics, Propeller developed [16] by Propeller Health, Smartinhaler developed by Adherium Ltd) . Benefits to the patients of some of these products include providing reminders to take a dose or order a new inhaler, providing breath actuation, training the patient on the appropriate inhalation manoeuvre, and recording the time and location of each dose to evaluate adherence to the prescribed dosing regimen. Some systems, such as the system offered by Propeller Health, have Bluetooth functionality to sync with apps on mobile devices and allow the patient to share treatment data with others so that family members or physicians can monitor adherence or even control of the disease state remotely. In 2016 3M introduced a novel device, the Intelligent Control Inhaler (ICI), designed to bring together several aspects of novel technology, including breath actuation, patient flow control, patient training and compliance and provide real time patient information and diagnostics to health care providers. Figure 4 presents images of hightech devices.

322


Drug Delivery to the Lungs 27, 2016 - 60 Years of the MDI – A History of Innovation

Figure 4 Image of Propeller

[17]

and ICI

MDIs containing new chemical entities (NCEs) or off-patent molecules, coupled with the inclusion of devices such as the ICI may benefit patients and lead to improved patient outcomes. By increasing patient compliance through inspiratory optimisation, coupled with up to date inhalation profile data being made available to health care practitioners, there is the potential to monitor and intervene if a medical condition worsens, enhancing the quality of life of asthma and COPD patients. If such systems can be demonstrated to improve pharmacoeconomic outcomes, they may gain widespread market acceptance in spite of their higher cost and substantially change the interaction between the patient and their inhaler in the future. Conclusions The MDI has developed significantly since its introduction in 1956. Over the 60 years since its launch, the MDI has continued to evolve. Today and into the future, as an established dosage form, it remains a major platform in the delivery of drug to the lungs. References [1] Thiel CG : From Susie’s question to CFC-free: an inventor’s perspective on forty years of MDI development and regulation. Respir Drug Deliv 1996;1:115-123 [2]

Freedman T : Medihaler® therapy for bronchial asthma. Postgraduate Medicine 1956;20(6):667-673..

[3] Stein SW, Theil CG : The History of Therapeutic Aerosols – A Chronological Review, accepted for publication in J of Aerosol Med and Pulm Drug Deliv. [4] Molina M, Rowland F : Stratospheric sink for chlorofluoromethanes: chlorine atom-catalysed destruction of ozone. Nature 1974; 249(5460):810-12. [5]

Atkins PJ : Transition to HFA MDIs: how did we get here and where are we going?, Respir Drug Deliv 2006;1:83-90.

[6]

Leach C : The CFC to HFA transition and its impact on pulmonary drug development. Respiratory Care 2005;50(9):1201-1206

[7]

Stein SW, Sheth P, Myrdal PB : Advances in metered dose inhaler technology: hardware development. AAPS PharmSciTech 2013;1-13. (published online 20 December 2013; DOI 10.1208/s12249-013-0062-y).

[8]

Vervaet C, Byron PR : Drug-surfactant-propellant interactions in HFA-formulations, Int J Pharm 1999;186(1):13-30.

[9]

Wu Z, Thatcher ML, Lundberg JK, Ogawa MK, Jacoby CB, Battiste JL, Ledoux KA : Forced degradation studies of corticosteroids with an alumina-steroid-ethanol model for predicting chemical stability and degradation products of pressurized metered-dose inhaler formulations, J Pharm Sci 2012;101(6):2109-22. [10]

Jinks, P., 2008. A new high performance dual-layer coating for inhalation hardware Conference, Drug Delivery to the Lungs XIX. [11]

IMS Health (MIDAS January 2016)

[12] Newman SP, Weisz A, Talaee N, Clarke S. Improvement of drug delivery with a breath actuated pressurised aerosol for patients with poor inhaler technique. Thorax. 1991;46(10):712-716. [13]

Image provided by Clement Clarke International. Used by permission.

[14] Slowey, A : The Formulation Development of Sirdupla™ pMDI, a Generic Version of Seretide ® Evohaler® DDL feasibility publication, Drug Delivery to the Lungs 2015; 291-294. [15] Sessions V, Patel N, Slowey A, Harrison L : In Vitro and In Vivo Development of Sirdupla™ (Salmeterol/Fluticasone Propionate) pMDIs for Europe, Respir Drug Deliv, 2016;2:241-244. [16]

Singh S, Kanbar-Agha F, Sharafkhaneh A : Novel aerosol delivery devices. Semin Respir Crit Care Med 2015;36:543-551.

[17]

Image provided by Propeller Health. Used by permission.

323


Drug Delivery to the Lungs 27, 2016 - Lars Karlsson et al. Performance Indicating Acoustic Emission Measurements on a Dry Powder Inhaler 1

1

1

2

1

Lars Karlsson , Lisa Holmstén , Mats Josefson , Roland Greguletz , Kyrre Thalberg & Staffan Folestad 1

1

Pharmaceutical Technology & Development, AstraZeneca R&D Gothenburg, S-43183 Mölndal, Sweden 2 Technology Department, AstraZeneca Sofotec, D-61352 Bad Homburg, Germany

Summary In this paper Acoustic Emission (AE) technology is used to assess inhaler performance. The aim of the AE approach is to 1) perform accurate measurements of timings/events related to device performance, 2) improve metadata capacity for analysis, and 3) capture deviations or trends in inhaler performance. By applying this technology consistently and long-term, it is envisioned that AE will provide data as an aid for designing in robustness in the technical and mechanical performance of the inhaler. In this study, two types of Acoustic Emission measurements were performed on a marketed Dry Powder Inhaler (DPI). In the univariate mode, AE was used for measuring timings of inhaler (and analysis robot) operations. The parameter of interest in this particular application was the Time to Trig (TtT), defined as the time from start of the flow until the inhaler triggers. TtT, which is linked to device performance, was easily and accurately determined for 9 device units, and was found to correlate well to trigger flow. In another mode of AE assessment, a chemometrics approach was used. Here, the raw sound data was Fourier transformed to frequency spectra, which were fed into a multivariate analysis software. Principal Components Analysis (PCA), as well as Orthogonal Projections to Latent Structures (OPLS), with unit variance scaled data were used for spectral analysis. Some applications are shown, focusing on how and to what degree the acoustic pattern changes with different actuations within the same device and between devices filled with different formulations. It was concluded that the passive Acoustic Emission chemometric approach described will detect also subtle differences in device performance or formulation characteristics. If the AE data is interpreted correctly, it is envisioned that this technology will be aiding future product and device development projects. Introduction Acoustic Emission (AE) monitoring has successfully been developed for various applications in the areas of [1] [2] Material Sciences , Process Engineering , and, manufacturing of pharmaceuticals (drug compounds and drug [3] products) . AE has also been used in monitoring applications to medical inhalation devices. Examples include [4-5] and characterization of inspiratory flow through the device generated by the patient adherence assessment [6] patient . We have extended the exploration of AE into the use in inhalation device development by monitoring [7-8] . The aim of our approach is that the the acoustic data of units being analysed on an automatic platform technology will 1) significantly improve metadata capacity for automated inhalation unit analysis, 2) capture potential deviations in inhaler performance from “normal” indicative of a possible future use in production, and 3) assist in designing in robustness in the technical and mechanical performance of the inhaler. Acoustic technology is very versatile and can be used for several different applications in device analysis. In this paper the flexibility is illustrated in two different measurement applications on a marketed dry powder inhaler (DPI). In the univariate mode AE was used as a simple, fast and accurate tool for measuring timings of inhaler (and analysis robot) operations. The parameter of interest, known to be indicative of inhaler performance, was Time to Trig (TtT), which is defined as the time from start of the flow until the inhaler triggers. Time to Trig was accurately determined for 9 device units, and was found to correlate well to trigger flow. In the multivariate mode of Acoustic Emission analysis, the complex but information rich acoustic signal is assessed and interpreted using a multidimensional, chemometrics based approach. Here, the raw sound data is Fourier transformed to frequency spectra, which are [9-10] , as well as other algorithms fed into a multivariate analysis software. Principal Components Analysis (PCA) [11-12] , with unit variance scaled data are used for spectral analysis. Some applications are shown, focusing on how and to what degree the acoustic characteristics changes 1) for different actuations within the same device, 2) between two devices within the same batch, and, 3) between devices filled with different formulations.

Experimental The device used in this study is a multi-dose, breath actuated DPI, which delivers a metered dose from an internal bulk reservoir (cartridge). The dose is deagglomerated in the dispersion unit which consists of a cyclone and a mouthpiece. The inhaler has multiple feedback mechanisms for successful inhalation, both an audible click and a colour indicator. The audible click, or triggering of the inhaler is a flow dependent mechanical feature within the inhaler. The inhaler typically triggers at between approximately 30 to 40 lpm. The time to trig (TtT) parameter will in this paper be defined as the time from start of flow to the triggering action.

324


Drug Delivery to the Lungs 27, 2016 - Performance Indicating Acoustic Emission Measurements on a Dry Powder Inhaler An automated measuring station was used for dose withdrawal at 65 lpm. A special rig was used to incorporate the AE system in the robot environment with the microphone positioned close to where device actuation takes place. A microphone, AKG C 417PP 20-20000 Hz omni-directional, phantom powered condenser type was employed. The microphone was coupled to the A/D converter (M-Audio 610 Profire, 24-bit/192kHz digital audio) which in turn was connected to the computer via a firewire interface. A LabView based software, specifically designed to record 3-5 seconds time segments of interest, i.e., in connection with the audible actuation sound, was specifically designed for this purpose. The sound was recorded for all dose withdrawals at 96-192 kHz, 1 channel, using the software Audacity v2.0.6 in Windows 7. Raw sound data was stored in Free Lossless Audio Codec (FLAC) file format. Consecutive sound time windows were made using a Python library (scikits.audiolab v 0.11.0). Frequency spectra were then calculated from the time windows by Figure 1. Schematic of the Acoustic Fourier transformation (FT). Multivariate data analysis of spectra was Emission analysis process. performed FT = Fourier Transform in SimcaP (v13.0, MKS Umetrics, Sweden). Overlapping time windows in 0.1 seconds increments were extracted from the sound file. For each inhaler actuation, spectra were collected throughout the whole collected time chunk. Each window was multiplied by a Hann function to reduce frequency aliasing before the FT was applied. Spectra from the consecutive windows were put as observations in a table for further multivariate processing. Here, Principal [9-10] with unit variance Components Analysis (PCA) scaled spectra was used for unsupervised overview. Also, Orthogonal Projections to Latent Structures [11-12] was used to Discriminant Analysis (OPLS-DA) Figure 2. Acoustic trace from dose withdrawal. See text for peak find spectral differences between e.g. different assignments. formulations. The system was set up for both univariate and multivariate analysis, as shown schematically in Figure 1. Results and Discussion Univariate analysis. Here, 9 device units were analysed, either empty or filled with lactose, and prepared for the purpose of this experiment with a variation in trigger flow (37 – 50 lpm). Acoustic data were automatically collected in time segments of approximately 3 seconds in length. A typical acoustic trace is shown in Figure 2. The different acoustic peaks can be allocated to either robot events or device operations. Peaks #1 to #3 were confirmed to be robot mechanical events. Peak #4 can be identified as the flow start since the background change, i.e. the baseline is broader, after this signal. The trig operation is easily determined since it is a very distinct and audible sound, giving a sharp peak. Hence, the time between peak #4 and the trig peak is determined to be time to trig. It was observed that by varying the flow rate and rise time, TtT changes, Figure 3. Time to trig data plotted versus the trig flow rate which is to be expected. Also, a filled inhaler determined for the individual devices. delivering inhalation powder slightly reduces the pressure drop as long as particles are circulating inside the cyclone (classifier), hence giving a slightly longer TtT when compared to corresponding empty inhalers. In Figure 3, time to trig data is presented as a function of trig flow. The TtT values correlate well to the trig flow and the measurements are reproducible.

325


Drug Delivery to the Lungs 27, 2016 - Lars Karlsson et al. Multivariate analysis. Here, acoustic data were collected for a large number of inhalers filled with different formulations. The collection was done in time segments typically in the range 3-5 seconds. In this mode, the entire acoustic signal, with a multitude of frequencies being recorded for every time unit, is utilised. To a degree, every actuation sound is unique, with its own acoustic fingerprint type of pattern. As a consequence, it is of value to assess what type of differences or changes in the sound that reflects a true change in quality or performance. Also, equally important is to ascertain whether a sound change is related to the device being analysed or to the analysis station, i.e. a potential performance issue with the robot. AE has been shown to be a quick and accurate indicator of device [7-8] . Hence, for the DPI performance investigated in this work, acoustic signals have been analysed by collecting sound data from a huge Figure 4. PCA (scores 1, 2) of AE data from actuations numbers 1, 59 and 60, number of drug development respectively, from the same device. analyses. Thus, a database of acoustic data is gradually created. This will form the basis for determining a normal sound pattern versus a signal that indicates a change that is truly related to the quality and the performance of the inhaler. For the device investigated in this work, which is available in a 30 and 60 actuations configuration, the sound pattern from the individual actuations, at least if they are close to each other in the device life, are very similar. For actuations far from each other in device life, there will be a small difference in the acoustic profile, as illustrated in Figure 4. Here, in the PCA plot there is a clear separation between shots early versus late in device life. This difference is due to that the cartridge is emptied as the end of inhaler life is approaching, thus creating another acoustic resonance behaviour. Depending on the aim of the AE investigation and the underlying analytical task, the focus can shift from looking at differences within Figure 5. Orthogonal Projections to Latent Structures - Discriminant Analysis (OPLSthe same device to DA) plot of AE data from inhalers filled with 3 different carriers. Further understanding comparisons between devices of the differences can be done by spectral analysis of the OPLS loadings. or batches of devices. An example of this is shown in [12] plot of Figure 5. Here, an Orthogonal Projections to Latent Structures - Discriminant Analysis (OPLS-DA) sound data from 3 identical devices but filled with different commercially available carriers. The plot demonstrates a separation between the batches in relation to their respective sound characteristics. These two examples clearly demonstrate the capability of the technology to probe for very subtle differences in sound and potentially, device and performance characteristics.

326


Drug Delivery to the Lungs 27, 2016 - Performance Indicating Acoustic Emission Measurements on a Dry Powder Inhaler Conclusions Earlier work has shown that for a prototype DPI, AE is a technology capable of e.g., distinguishing between filled and empty cavities in the different prototype DPI design variants, while also providing very useful metadata for quality control of the analysis since the technique is sensitive to changes in e.g., flow rate [7-8]. The concept has now been extended to another type of DPI, where the dose is loaded from a bulk reservoir instead of from separate cavities. Acoustic emission analysis, in the univariate mode, has also been used for determining timings of events of significance to inhaler performance, e.g. Time to Trig. Also, it is noted that the cyclone action, which is designed for effective dispersion of inhalation powders such as adhesive mixtures, in the investigated device creates a very dynamic acoustic profile that can be linked with the fluid dynamic particle and air flow activity. Experiments with different configurations have shown that also subtle differences in device or formulation characteristics often result in clear separations with regards to the AE chemometric profile. The passive AE chemometric approach described here will develop as the database of acoustic data for different inhaler/formulation configurations grows. However, many aspects of this efficient, yet very affordable concept, can be investigated and developed further. As an example, future work will include directed experiments with many different formulations, varying widely in characteristics. References 1.

2.

3.

4. 5.

6.

7.

8.

9. 10.

11. 12.

Ríos-Soberanis C R: Acoustic Emission Technique, an Overview as a Characterization Tool in Materials Science, Journal of Applied Research and Technology 2011; 9: pp 367-379. Boyd J W R, Varley J: The Uses of Passive Measurement of Acoustic Emissions from Chemical Engineering Processes, Chemical Engineering Science 2001; 56: pp 1749-1767 Whitaker M, Baker G R, Westrup J, Goulding P A, Rudd D R, Belchamber R M, Collins M P: Application of acoustic emission to the monitoring and end point determination of a high shear granulation process, International Journal of Pharmaceutics 2000; 205: pp 79-91. Sensohaler, Inhaler Compliance Monitor, www.sagentia.com Scuri M, Alfieri V, Giorgio A, Pisi R, Ferrari F, Taverna M, Vezzoli S, Chetta A: Measurement Of The Inhalation Profile Through A Novel Dry Powder Inhaler (Nexthaler®) In Asthmatic Patients Using Acoustic Monitoring, American Journal of Respiratory and Critical Care Medicine 2013; 187: pp A1931. Taylor T E, Holmes M S, Holmes, B E, Sulaiman I, Costello R W, Reilly R: Monitoring Inhaler Inhalations Using an Acoustic Sensor Proximal to Inhaler Devices J. Aerosol. Med. Pulmon. Drug Delivery, 2016; 29: pp 1-8. Karlsson L, Josefson M, Andersson P, Folestad S: Acoustic Emission Characterization of Device Performance, Oral Presentation DDL 26, 2015. Karlsson L, Josefson M, Andersson P, Folestad S: Acoustic Emission Characterization of Inhaler Performance, Inhalation Magazine, 2016; August; pp 26-30. Jackson J E: A user’s guide to principal components. New York: Wiley; 1991. Geladi, P, Kowalski B R: Partial Least-Squares Regression: A Tutorial, Analytica Chimica Acta 1986; 185: pp 1–17. Trygg J, Wold S: Orthogonal projections to latent structures (O-PLS). J Chemom 2002; 16: pp 119–128. Bylesjö, M, Rantalainen M, Cloarec O, Nicholson J K, Holmes E, Trygg J: OPLS Discriminant Analysis: Combining the Strengths of PLS-DA and SIMCA Classification, J. Chemom. 2006; 20: 341–351.

327


Drug Delivery to the Lungs 27, 2016 – Mireille Hassoun1, et al. ®

Effect of using bio-relevant media in the DissolvIt system to measure dissolution of fluticasone propionate from Flixotide 50 µg Evohaler 1

2

1

1

2

2

Mireille Hassoun , Maria Malmlöf , Abhinav Kumar , Sukhi Bansal , Mattias Nowenwik , Per Gerde , Ben 1 Forbes 1

Institute of Pharmaceutical Science, King’s College London, SE1 9NH, UK 2 Inhalation Sciences Sweden AB, Stockholm, Sweden

Summary This study evaluated the effect of using different bio-relevant formulations as the dissolution medium in the ® ® DissolvIt system. The fluids investigated were polyethylene oxide with lipids (PEO), Survanta and an in-house ® developed simulated lung lining fluid (sLLF). A Flixotide 50 µg Evohaler was used to generate an aerosol, which ® was deposited on cover slips using PreciseInhale . The fluticasone propionate (FP) particles dissolved into the bio-relevant medium and drug was transferred into a flow-though perfusate within the DissolvIt system. FP as 2 quantified by LC-MS/MS following solid phase extraction (SPE) of the samples. Excellent linearity (R = 0.999) was observed. The inter-day and intra-day precision data complied with the validation guidance for bioanalytical methodologies. Differences in FP concentration in the perfusate when using PEO and sLLF were observed at 20 min; 0.29 ± 0.09 and 0.12 ± 0.02 %/mL respectively (p≤0.05). The FP concentration in the perfusate was similar when using PEO and Survanta. It was concluded that the LC-MS/MS assay was validated successfully and provides a sensitive method to quantify FP in dissolution studies. The hypothesis that dissolution of FP in the perfusate would be enhanced in sLLF was not supported by this study. These findings may be attributable to FP residing preferentially in lipid structures in sLLF, limiting FP availability for transfer into the perfusate. Further studies are required to assess whether using a more bio-relevant media provides a more predictive measure of inhaled particle dissolution. Introduction In vitro dissolution testing is well established for solid oral dosage forms, both for quality control (QC) purposes [1,2] . As with solid oral dosage forms, and to seek in vitro to in vivo correlation (IVIVC) for drug pharmacokinetics the therapeutic effect of an inhaled particulate aerosol is only realised after its dissolution, thus investigating the dissolution of orally inhaled products (OIPs) as a potential determinant of drug bioavailability is currently attracting [3-6] interest . In order to develop IVIVC, a simple in vitro system should reproduce the critical features of the complex in vivo environment in which drug solubilises. Of the various methods for in vitro dissolution developed for OIPs, most systems tend to require large volumes of dissolution medium. Consequently, in dissolution studies [7] to date, the media have been simple aqueous fluids, sometimes with the addition of various surfactants . Therefore, the development of a physiologically representative methodology including a bio-relevant medium is required to address the current unmet need for a system that is more characteristic of the in vivo environment. ® The DissolvIt system from Inhalation Sciences was developed as an in vitro dissolution tool for OIPs that utilises a low volume of dissolution medium and simultaneously allows particle disintegration to be studied visually while [8] drug dissolution is quantified chemically in a dynamic flow-past model . It has been used to characterise the dissolution of fluticasone propionate (FP), generating profiles which closely resemble those of FP in the rat [9] isolated perfused lung . [10]

and represents a challenge in dissolution tests since it is difficult to FP is a poorly soluble inhaled drug maintain sink conditions and it is not easy to assay FP in low concentrations. In order to quantify sub-microgram [11, 12] concentrations, a highly sensitive assay is required with an efficient extraction method . Therefore, the aim of the present study was to validate a new rapid and sensitive LC-MS/MS method to quantify FP in samples from an investigation into the use of bio-relevant media in the DissolvIt system. The effect of dissolution medium on FP aerosol particle dissolution was investigated using three different media: (i) 1.5% polyethylene oxide including ® 0.4% L-alphaphosphatidyl choline, (ii) Survanta , and (iii) an in-house developed simulated lung lining fluid (sLLF) [13] , synthesised based on accurate measurements of human lung fluid composition. Materials Flixotide 50 µg Evohalers were obtained from GSK. Polyethylene oxide and L-alphaphosphatidyl choline required for preparation of PEO solvent, and all components necessary for the preparation of sLLF were provided by Sigma-Aldrich Company Limited (Dorset, UK). Survanta was obtained from Abbvie Ltd (Berkshire,UK). All chemicals and equipment necessary for the dissolution experiments were provided by Inhalation Sciences (Stockholm, Sweden). Chemicals required for SPE validation included micronized FP (USP grade, purity 98%), supplied by LGM Pharma Inc (Boca Raton, USA), FP-d5 (USP grade, purity 97%) supplied by Insight Biotechnology Limited (Wembley, UK) and rabbit serum, purchased from Sigma-Aldrich Company Limited (Dorset, UK). The chemicals needed for solid phase extraction were zinc sulphate powder, supplied by VWR

328


Drug Delivery to the Lungs 27, 2016- Effect of using bio-relevant media in the DissolvIt® system to measure dissolution of fluticasone propionate from Flixotide 50 µg Evohaler International Limited (Lutterworth, UK) and HPLC-gradient grade acetonitrile, 35% v/v ammonium hydroxide solution and Analytical-Reagent grade dichloromethane, all purchased form Fischer Chemical (Loughborough, UK). Experimental methods Validation of the LC-MS/MS assay Primary stock solutions of FP and FP-d5 (internal standard) were prepared, 100 µg/mL in acetonitrile, and stored at 20°C. A 1 µg/mL FP working solution was prepared and calibration standards (156, 313, 625, 1250, 2500, 5000 and 10,000 pg/mL) were prepared by serial dilution of the working solution with acetonitrile. Method validation was conducted in terms of linearity, intra-day and inter-day precision (%CV), accuracy, limit of detection (LOD) [11] and limit of quantification (LOQ), based on FDA guidance . ®

Deposition and dissolution of FP aerosol in the DissolvIt system The Flixotide pMDI canister was connected to the US Pharmacopeia Induction Port No 1 (standardised simulation of the throat) of the PreciseInhale® aerosol system (Figure 1). The inhaler was dosed by 16 actuations (90 s exposure time) into an air flow of 15 L/min. The aerosol particles were deposited on 9 circular microscope glass cover slips (13 mm in diameter) placed in the bottom of the glass coating holding chamber, as described by Borjel [8, 9] et al . The biorelevant medium for dissolution, 5.7 µL, was applied to one side of the polycarbonate membrane ® of a single-use DissolvIt cell. The simulant, together with the polycarbonate membrane provide a diffusion barrier (Figure 1). On the other side of the membrane, the perfusate (consisting of phosphate buffer with 4% albumin) was streamed past at a flow rate of 0.4 mL/min. Particle disintegration was studied from the ‘luminal’ side using optical microscopy and from the ‘vascular’ side by chemical analysis of FP dissolved in the flow-past perfusion medium.

DissolvIt

®

Figure 1 - A schematic diagram of the PreciseInhale

®

and

FP quantification Samples were prepared for analysis using SPE. Briefly, 325 µL of sample was loaded into a deep-well plate followed by 50 µL of working internal standard solution (FP-d5), 300 µL of 0.1 M zinc sulphate and 75 µL of 10% ammonium hydroxide then mixed using an orbital shaker for 30 min. The samples were centrifuged at 3700 rpm ® and transferred to a pre-conditioned Evolute Express SPE 96-well plate (Biotage, Sweden) and washed using 200 µL of water, then 200 µL of 25% v/v methanol in water. The analytes were eluted twice with 200 µL of pure acetonitrile followed by 100 µL dichloromethane and vacuum centrifuged to dryness. Samples were reconstituted with 30 µL of 55% v/v acetonitrile in water and injected into the LC-MS/MS system. Chromatographic separations were carried out on a 100A Acquity BEH (2.1 x 50 mm) analytical HPLC column, packed with 1.7 µm C-18 (Waters, Eltree, UK). Data analysis Peak integration was performed using MassLynx 4.1 computer software. The relationship between peak area ratio (y) and analyte concentration (x, pg/mL) was calculated using the LINEST function in Microsoft Excel. Data was expressed as the mean ± standard deviation (SD) of replicate determinations, where n ≥ 3. For the FP dissolution, the %FP in the perfusate was expressed as a percent of the amount deposited on the glass slide. One-Way ANOVA was applied to the data, using the IBM SPSS version 22 software and statistically significant when p ≤ 0.05.

329


Drug Delivery to the Lungs 27, 2016 – Mireille Hassoun1, et al. Results Excellent linearity between the mean peak area ratio of FP/FP-d5 and the concentration of FP in the samples was 2 observed (R value = 0.999). The inter-day and intra-day precision data complied with the validation guidance, with all CV being < 20%, except for 156 pg/mL and 313 pg/mL. The accuracy data for all FP standard concentrations passed the accepted criteria of 85-115% (Figure 2). The LOD and LOQ were 106 pg/mL and 312 pg/mL, respectively. FP concentration in the perfusate was highest at all time points when FP dissolved in PEO and lowest in sLLF (Figure 3a). The FP concentration profile in perfusate was very similar between PEO and Survanta, both reaching a Cmax at 20 min. Differences in the FP concentration values in all three lung fluids at most time points however, were not statistically significant (One-Way ANOVA, p<0.05), except the difference in FP concentration in PEO and sLLF at 20 min. The cumulative percent of FP transferred into the perfusate over time is shown in Figure 3b, with similar profiles in each medium reflecting the ranking observed in the perfusate concentrations. a)

Figure 2 - Validation of the solid phase extraction and LC-MS/MS assay of fluticasone propionate (FP): a) Linearity of the mean peak area ratio vs concentration; b) FP concentration, precision and accuracy. Data expressed as mean Âą SD (n=9).

b)

330


Drug Delivery to the Lungs 27, 2016- Effect of using bio-relevant media in the DissolvIt® system to measure dissolution of fluticasone propionate from Flixotide 50 µg Evohaler

**

Figure 3 - Dissolution of FP. a) Concentration of FP in the perfusate over time following dissolution in PEO, sLLF and Survanta® normalised to mass deposited on glass cover slips. **Difference in FP concentration in PEO and sLLF is statistically significant (One-Way ANOVA, p<0.05). b) Cumulative % of FP transferred into the perfusate over time, following its dissolution in PEO, sLLF and Survanta®. Data expressed as mean ± SD (n=3).

a)

b)

Discussion SPE offers an improved extraction method over liquid phase extraction since it is less time-consuming and [12] requires minimal sample preparation and solvent use . The published methods for FP analysis are difficult to replicate with adequate reproducibility and sensitivity, hence this study developed and validated a new SPE, LCMS/MS quantification assay for FP. The 156 pg/mL FP standard fell outside the accepted CV (<20%), which was attributed to the concentration being close to the LOD (106 pg/mL). However, the FP concentrations in the dissolution experiments fell within the upper range of the assay, which was fit for purpose. ®

The PEO medium used as standard in the DissolvIt system possessed a lipid content of 4 mg/mL, which was lower than the lipid content in sLLF, 5.4 mg/mL. Thus, it was hypothesised that dissolution of FP in sLLF would be enhanced as the greater lipid content may facilitate drug solubilisation. However, the results showed less FP transfer to the perfusate when SLLF was used compared to PEO, and it is speculated that FP may preferentially reside or become trapped within lipid/lamellar structures in sLLF. sLLF contains cholesterol, unlike other lung fluid simulants, and studies have shown that cholesterol can form tight nanodomain complexes with DPPC, stabilising [14] the DPPC in lipid structures such that once the FP is solubilised within, it is less likely to leave such structures . Conclusion A SPE / LC-MS/MS assay for FP was established successfully and able to quantify low concentrations (pg/mL) of the FP in lung fluids. The hypothesis that the dissolution and transfer of FP in the perfusate would be enhanced by using sLLF was not supported by this study. FP may reside preferably in the lipid structures, limiting the transfer into the perfusate. Limited differences in perfusate FP concentration and the overall dissolution profiles were observed, although the difference in Cmax at 20 min was significant for PEO vs sLLF. Further studies are required to evaluate more fully the impact of the medium composition on dissolution profile and whether more biorelevant media can provide data more predictive of inhaled particle dissolution.

331


Drug Delivery to the Lungs 27, 2016 – Mireille Hassoun1, et al.

References [1]

[2] [3] [4] [5] [6] [7] [8] [9] [10]

[11] [12] [13]

[14]

US FDA CDER. Guidance for industry: dissolution testing of immediate release solid oral dosage forms, 1997, May 15. Accessed from: http://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidance/ucm070237.pdf Cardot, J.M.; Davit, B.M. In vitro-In Vivo Correlations: Tricks and Traps. AAPS J. 2012, 14 (3), 491-499 Arora, D.; Shah, K.A.; Halquist, M.S.; Sakagami, M. In vitro aqueous fluid-capacity-limited dissolution testing of respirable aerosol drug particles generated from inhaler products. Pharm. Research, 2010, 27, 786-795 Son, Y.J; Horng, M.; Copley, M.; McConville, J.T.; Optimization of an in vitro dissolution test method for inhalation formulations. Dissolut. Technol, 2010, 13, 46-54 May, S. Dissolution testing of powders for inhalation. phD Thesis, 2013. Accessed from: http://d-nb.info/1053681895/34. Davis, N.M.; Feddah, M.R. A novel method for assessing dissolution of aerosol inhaler products. Int. J. Pharm, 2003 Wiedmann, T., Bhatia, R., Wattenberg, L. Drug solubilisation in lung surfactant. J. Contr. Release, 2000, 65, 43-47 Börjel M, S.R., Gerde P. The DissolvIt: An In Vitro Evaluation of the Dissolution and Absorption of Three Inhaled Dry Powder Drugs in the Lung. Respiratory Drug Delivery, 2014, Puerto Rico. Börjel M; Ewa Selg.; S.R., Gerde P. In Vitro- Ex Vivo Correlation of Fluticasone Propionate Pharmacokinetic Profiles. DDL, 2015, Edinburgh Hastedt, J.E., Backman, P., Clark, A.R., Doub, W., Hickey, A., Hochhaus, G., Kuel, P.J., Lehr, C., Mauser, P., McConville, J., Niven, R., Sakagimi, M., Weers, J.G. Scope and relevance of a pulmonary th biopharmaceutical classification system AAPS/FDA/USP Workshop March 16-17 , 2015 in Baltimore, MD. AAPS, 2016, 1-20 US FDA CDER. Guidance for industry: bioanalytical method validation, 2001, Feb 16. Accessed from: http://www.fda.gov/downloads/Drugs/Guidances/ucm070107.pdf Lombardi, C. Solid phase extraction. Chemistry in New Zealand, 2015, 88-90 Kumar, A., Bicer, E.M., Morgan, A.B., Pfeffer, P.E., Monopoli,M., Dawson, K.A., Eriksson, J., Edwards, K., Lynham, S., Arno, M., Behndig, A.F., Blomberg, A., Somers, G., Hassell, D., Dailey, L.A., Forbes, B., Mudway, I.S. Enrichment of immunoregulatory proteins in the biomolecular corona of nanoparticles within human respiratory tract lining fluid. Nanomedicine: Nanotechnology, Biology, and medicine, 2016, 12, 10331043 Kim, K., Choi, S.Q., Zell, Z.A., Squires, T.M., Zasadzinski, J.A. Effect of cholesterol nanodomains on monolayer morphology and dynamics. PNAS, 2013, E3054 – E3060

332


Drug Delivery to the Lungs 27, 2016 – Wachirun Terakosolphan et al. Solubility of fluticasone propionate and beclomethasone dipropionate in simulated lung lining fluids 1

1

1

Wachirun Terakosolphan , Mireille Hassoun , Abhinav Kumar & Ben Forbes 1

1

King’s College London, Institute of Pharmaceutical Science, London, SE1 9NH, UK

Summary There is burgeoning interest in the dissolution of inhaled medicines in the lungs and speculation that a biopharmaceutics classification system (BCS) concept, analogous to that currently used for oral drug dosage forms, may be extended to pulmonary drug products. This exposes the need for reliable and relevant methods for measuring the solubility in the lungs, especially for poorly soluble drugs for which dissolution is thought to be an important determinant of their pharmacokinetics and pharmacodynamics in the lungs. Inhaled corticosteroids (ICS) are a widely used class of anti-inflammatory agents for the treatment of respiratory disease which includes drugs with low aqueous solubility, i.e. fluticasone propionate (FP) and beclomethasone dipropionate (BDP). The solubility of such drugs has been reported in a variety of simulated lung lining fluids of different compositions. The purpose of this study was to compare drug solubility in solvents typically used to represent lung lining fluid. The solvents investigated in this study were based on a physiological salt solution (Gamble’s solution), a licensed lung ® surfactant product (Survanta ) and a synthetic simulated lung lining fluid (sLLF) based on the measured composition of human lung lining fluid. Drug solubility was compared in five solutions in total: ultrapure water, Gamble’s solution, sLLF, an aqueous dilution of Survanta, and Survanta itself. The solubility of FP in these solvents was 1.92, 1.99, 2.04, 3.89 and 20.28 μg/mL, respectively. BDP solubility in the same solvents was 2.17, 1.03, 16.79, 5.78, to 37.16 μg/mL, respectively. These data illustrate that simulated lung fluids that contain surfactant have enhanced potential to dissolve inhaled steroid drug particles that have deposited in the lungs.

Introduction Inhaled corticosteroids (ICS) are mainstay therapy for several respiratory diseases such as asthma and chronic [1] obstructive pulmonary disease (COPD) , and in recent guidelines were recommended for use for all patients [2] except those with mild, intermittent symptoms of chronic asthma (GINA 2016) . Fluticasone propionate (FP) and beclomethasone dipropionate (BDP) are widely used ICS for the treatment of asthma in children and adults. In terms of pharmacokinetics, dissolution of drug particles appears to be a rate-limiting step in the disposition of these drugs after they deposit in the lungs. Particles dissolution in the lungs depends predominately on the [3,4] physicochemical properties of the active pharmaceutical ingredient . Studies in pulmonary drug product research that have focused on dissolution have often studied ICS due to their poor water-soluble profiles which [5] potentially affect the drugs’ therapeutic efficacy . In order to simulate the dissolution of drugs in airways, artificial lung fluids were used as solvents in this solubility study. Lung lining fluid is a complex mixture of glycoproteins, proteins, and lipids. It is produced and secreted by alveolar epithelial type-II cells. It consists of lipid-rich lipoproteins with the lipid composition dominated by phosphatidylcholine, with a high dipalmitoyl content. Apart from phospholipids, many proteins are also present. In addition to the most abundant protein, albumin, there are four non-serum apoproteins (SP-A, SP-B, SP-C and SP[6] ® D) . The most widely used simulated lung lining fluids used in pharmaceutical research, namely Survanta , Gamble’s solution, and an in-house synthetic simulated lung lining fluid (sLLF) were the solvents utilised in this study (Table 1). Survanta is a marketed artificial lung fluid, which is a bovine lung extract containing similar substances to those found in normal human lung surfactant. It is indicated for the treatment of respiratory distress syndrome in newborn premature infants with deficiency of lung surfactant and acts by replenishing and restoring [7] the surfactant and its activity . In this study, Survanta was used as an undiluted solvent and as a 21.6% v/v diluted solvent with Hank’s balanced salt solution (HBSS). The latter was used to match the total lipid concentration in human lung lining fluid. Another simulated lung fluid, which is designed more to mimic the [8] interstitial fluid in the airways, is Gamble’s solution which consists of a wide range of electrolytes . Finally, sLLF is an in-house developed simulated epithelial lung lining fluid, composed of a number of lipids and proteins, and [9] designed to reflect the human lung fluid composition . This study aimed to investigate the solubility of FP and BDP in the respiratory biological milieu using simulated lung lining fluids compared to ultrapure water as control.

333


Drug Delivery to the Lungs 27, 2016 – Solubility of fluticasone propionate and beclomethasone dipropionate in simulated lung lining fluids Materials 6α,9-Difluoro-17-[[(fluoromethyl)sulfanyl]carbonyl]-11β-hydroxy-16α-methyl-3-oxoandrosta-1,4-dien-17α-yl propanoate (fluticasone propionate: FP) was purchased from Adooq Bioscience (Irwin, CA), 9-Chloro-11βhydroxy-16β-methyl-3,20-dioxopregna-1,4-diene-17,21-diyl dipropanoate (beclomethasone dipropionate; BDP) was purchased from Medchem Express (US), and Survanta® from AbbVie Ltd. (UK). Ultrapure water with 18.0 MΩ·cm residual specific resistance was obtained using an Elgastat Maxima purifier (Elga, UK). All other reagents were obtained from standard source. Table 1. The composition of the in vitro lung lining fluid simulants

Survanta® - Phospholipids 25 mg/mL (including 11.0 – 15.5 mg/mL disaturated phosphatidylcholine) - Triglycerides 0.5 – 1.75 mg/mL - Free fatty acids 1.4 – 3.5 mg/mL - Protein less than 1.0 mg/mL

-

Gamble’s solution Magnesium chloride 0.095 mg/mL Sodium chloride 6.019 mg/mL Potassium chloride 0.298 mg/mL Disodium hydrogen phosphate 0.126 mg/mL Sodium sulfate 0.063 mg/mL Calcium chloride dehydrate 0.368 mg/mL Sodium acetate 0.574 mg/mL Sodium hydrogen carbonate 2.604 mg/mL Sodium citrate dihydrate 0.097 mg/mL

-

sLLF DPPC 4.8 mg/mL DPPG 0.5 mg/mL Cholesterol 0.1 mg/mL Albumin 8.8 mg/mL IgG 2.6 mg/mL Transferrin 1.5 mg/mL Ascorbate 140 µM Urate 95 µM Glutathione 170 µM

Experimental methods Solubility measurement The solubility of FP and BDP was investigated separately in ultrapure water, Gamble’s solution, sLLF, diluted Survanta, and undiluted Survanta by placing excess drug powder (approximately 0.5 mg) in a microcentrifuge tube with 0.5 mL of the solvent. The tubes were closed firmly and agitated to assist mixing using a vortex mixer for 5 min before being placed in a bath sonicator at 37°C for 30 min. Thereafter, they were placed into a shaking water-bath at 37°C for 48 h to allow the solutes to dissolve completely and reach their equilibrium solubility. The microcentrifuge tubes containing the drug suspensions in different solvents were centrifuged at 13000 rpm for 10 minutes to pellet any undissolved solute. After centrifugation, 0.2 mL of the supernatant was transferred to fresh tubes and re-centrifuged as described above. After the second centrifugation, 0.1 mL of the supernatant was transferred to fresh tubes and diluted 10 times with methanol, followed by further centrifugation to sediment any precipitates from the dilution. Eventually, 0.2 mL of the supernatant was transferred to HPLC vials for analysis of each drug in specific HPLC condition (table 2). All measurements were performed in triplicate. It was assumed that the addition of methanol had dissolved the micelles from either several surfactants or phospholipids in tested solvents, so that all the solutes were detectable.

HPLC assay The saturated solution of FP and BDP were analysed using the following conditions: Table 2. Chromatographic condition for HPLC analysis

Chromatographic condition Column Mobile phase Flow rate Column temperature Sample temperature Detection wavelength Injection volume Retention time Run time

Fluticasone propionate (FP) Beclomethasone dipropionate (BDP) Luna C18, 3 μm, 150 x 4.6 mm I.D. or equivalent Methanol: 0.6% w/v aqueous Acetronitrile:Water (65:35 %v/v) ammonium acetate solution (75:25 %v/v) 1.0 mL/min 1.5 mL/min 40°C 65°C 15°C 25°C UV 240 nm UV 254 nm 60 μL 75 μL FP ~ 5.5 minutes BDP ~3.2 minutes 7 minutes 6 minutes

334


Drug Delivery to the Lungs 27, 2016 – Wachirun Terakosolphan et al. Results and Discussion Different solubility profiles were observed for FP and BDP in different solvents (Figure 1). As expected, when FP and BDP were added to ultrapure water and Gamble’s solution they produced the lowest concentrations, approximately 1.0 – 2.2 μg/mL, since both corticosteroid drugs are poorly water-soluble. For the solubility of FP in the rest of simulated lung fluids, it gradually increased from 2.0, 3.9, to 20.3 μg/mL in sLLF, diluted Survanta, and undiluted Survanta®, respectively. On the other hand, BDP showed higher concentrations than those from FP results. Indeed, 16.8 μg, 5.8 μg, and 37.2 μg of BDP was solubilised in 1 mL of these three solvents, respectively.

37.16

40

16.79

20.28

30

5.78

2.04

1.03

1.99

2.17

10

FP BDP

3.89

20

1.92

Drug concentration (μg/mL)

50

0 Ultrapure water

Gamble's solution

sLLF

Diluted Survanta®

Survanta®

Figure 1. Solubility profile of fluticasone propionate (blue) and beclomethasone dipropionate (orange) in ultrapure water, Gamble’s solution (a commonly used artificial lung lining fluid), synthetic human lung lining fluid, diluted Survanta (total lipid concentration matches the total lipid concentration in human lung lining fluid) and undiluted Survanta. Data expressed as mean ± SD (n=3).

The solubility of lipophilic drugs in respiratory tracts is proposed as a critical parameter for pulmonary drug [5] product development under the proposed inhaled BCS concept . The saturated solutions of FP and BDP, which were prepared by placing excess amount of the drugs in the promising solvents were quantified using HPLC [10] , with concentrations of analysis. FP and BDP have been reported to be highly water-insoluble drugs [11,12] approximately 0.1 – 0.2 μg/mL in water . In our measurement, the concentration of FP and BDP in water are higher than those from earlier studies, 1.92 and 2.17 μg/mL, respectively. This may be attributed to sample preparation, where a longer time period was utilised (48 h) to allow for equilibrium of saturated solutions to be reached. Similarly, 1.99 μg of FP and 1.03 μg of BDP were dissolved in 1 mL of Gamble’s solution, which is a [8] simple electrolyte solution . As anticipated these results were similar to the solubility profile of both drugs in water. Higher concentrations of BDP (more than 10 μg/mL) were shown in the three solvents; sLLF, diluted Survanta, and undiluted Survanta. Several previous studies have revealed that lung surfactants can enhance the solubility of small, lipophilic drug molecules, such as gluco-corticosteroids and cationic compounds because of their composition and the ability to [13,14] . This explains the higher solubility in form liposomal structures which effectively entrap these particles within the 3 phospholipid-containing fluids. However, although the solubility of FP increased when placed into sLLF, diluted Survanta and undiluted Survanta, its concentration in those fluids was significantly lower than that of BDP. The concentrations of FP and BDP in sLLF and diluted Survanta fell between the lower value in water and the higher value in undiluted Survanta, which was attributed to the fact that the solvents had the same lipid concentration of phospholipid, 5.4 mg/mL. It was anticipated FP solubility in sLLF would be higher (closer to that of FP in Survanta rather than the observed solubility of 2.04 µg/mL, which is closer to the solubility values in water and Gamble’s solution) because of the presence of the liposomal structures and specifically the presence of cholesterol, which can potentially form tight nanodomain complexes with DPPC, stabilising lamellar structures in [15] the fluid . However, the lower solubility value may be due to the presence of albumin, which studies have [16] , and hence reduce the stability of the lamellar phase and shown, has the ability to solubilise the cholesterol reduce the extent to which FP particles are effectively entrapped and solubilised.

335


Drug Delivery to the Lungs 27, 2016 – Solubility of fluticasone propionate and beclomethasone dipropionate in simulated lung lining fluids Conclusion In this study, we investigated the solubility of drugs in simulated lung lining fluids which represent the biological milieu. Gamble’s solution, which mimics the interstitial fluid deep within the lung, has a low solubilising power for FP and BDP. However, other lung surfactant simulant fluids, including sLLF, diluted Survanta, and undiluted Survanta, manifested up to 10-40x higher drug solubility, particularly for BDP. It can be summarised that poorly water-soluble drugs, including lipophilic molecules and some steroids, can be solubilised effectively in the lungs when those respirable drug particles are inhaled and deposited on a mucosal surface covered by a surfactantcontaining liquid film. References 1. Adams NP, Lasserson TJ, Cates CJ, Jones P. Fluticasone versus beclomethasone or budesonide for chronic asthma in adults and children. In: Group. CA, editor. Cochrane Database Syst. Rev. 2007 [Internet]. John Wiley & Sons, Ltd.; 2007. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=coch&NEWS=N&AN=00075320-10000000002192 2. Asthma GI for. Global Strategy for Asthma Management and Prevention [Internet]. 2016. Available from: www.ginaasthma.org 3. Derendorf H, Hochhaus G, Meibohm B, Möllmann H, Barth J. Pharmacokinetics and pharmacodynamics of inhaled corticosteroids. J. Allergy Clin. Immunol. [Internet]. 1998;101:S440–6. Available from: http://www.sciencedirect.com/science/article/pii/S0091674998701563 4. Olsson B, Bondesson E, Borgström L, Edsbäcker S, Ekelund K, Gustavsson L, et al. Controlled Pulmonary Drug Delivery [Internet]. Control. Pulm. Drug Deliv. 2011. Available from: http://link.springer.com/10.1007/978-14419-9745-6 5. Hastedt JE, Bäckman P, Clark AR, Doub W, Hickey A, Hochhaus G, et al. Scope and relevance of a pulmonary biopharmaceutical classification system AAPS/FDA/USP Workshop March 16-17th, 2015 in Baltimore, MD. AAPS Open [Internet]. AAPS Open; 2016;2:1. Available from: http://aapsopen.springeropen.com/articles/10.1186/s41120-015-0002-x 6. Hillery AM, Lloyd AW, Swarbrick J, editors. Drug Delivery and Targeting for Pharmacists and Pharmaceutical Scientists. New York: Taylor & Francis; 2001. 7. Corporation A. SURVANTA® Product monograph. Quebec: AbbVie Corporation; 2012. p. 1–33. 8. Marques MRC, Loebenberg R, Almukainzi M. Simulated biological fluids with possible application in dissolution testing. Dissolution Technol. 2011;18:15–28. 9. Kumar A, Bicer EM, Morgan AB, Pfeffer PE, Monopoli M, Dawson KA, et al. Enrichment of immunoregulatory proteins in the biomolecular corona of nanoparticles within human respiratory tract lining fluid. Nanomedicine Nanotechnology, Biol. Med. [Internet]. Elsevier B.V.; 2016;12:1033–43. Available from: http://dx.doi.org/10.1016/j.nano.2015.12.369 10. Anhydrous Beclometasone Dipropionate - British Pharmacopoeia [Internet]. Available from: https://www.pharmacopoeia.com/bp-2016/monographs/anhydrous-beclometasone-dipropionate.html?publisheddate=2015-08-03&text=anhydrous+beclometasone 11. Tokumura T, Miyazaki E, Isaka H, Kaneko N, Kanou M. Solubility of fluticasone propionate in aqueous solutions measured by a method avoiding its adsorption to experimental tools. Int. Res. J. Pharm. Appl. Sci. [Internet]. 2014;4:19–24. Available from: http://www.irjpas.com/File_Folder/IRJPAS 4(4)19-24.pdf 12. Sahib MN, Abdalwahed S, Abdulameer, Darwis Y, Peh KK, Tan YTF. Solubilization of beclomethasone dipropionate in sterically stabilized phospholipid nanomicelles (SSMs): Physicochemical and in vitro evaluations. Drug Des. Devel. Ther. 2012;6:29–42. 13. Wiedmann TS, Bhatia R, Wattenberg LW. Drug solubilization in lung surfactant. J. Control. Release. 2000;65:43–7. 14. Liao X, Wiedmann TS. Solubilization of Cationic Drugs in Lung Surfactant. Pharm. Res. 2003;20:1858–63. 15. Kim K, Choi SQ, Zell ZA, Squires TM, Zasadzinski JA. Effect of cholesterol nanodomains on monolayer morphology and dynamics. Proc. Natl. Acad. Sci. U. S. A. [Internet]. 2013;110:E3054-60. Available from: http://www.pnas.org/content/110/33/E3054.short 16. Kim SH. Adsorption and interactions of lung surfactant lipids and proteins at air/aqueous interfaces and in aqueous solution. Purdue University; 2007.

336


Drug Delivery to the Lungs 27, 2016 – Georgina E. Marsh et al. Using 3D printed standards to isolate the effect of surface morphology and surface chemistry in DPIs 1

1

2

3

4

Georgina E. Marsh , Morgan R. Alexander , Ricky D. Wildman , Matt J Bunker , Mark Nicholas & Clive J. 1 Roberts 1

2

School of Pharmacy, The University of Nottingham, Nottingham, UK Department of Chemical and Environmental Engineering, The University of Nottingham, Nottingham, UK 3 AstraZeneca R&D, Macclesfield, UK 4 AstraZeneca R&D, Mölndal, Sweden

Summary The particulate interactions that occur in the manufacture and use of dry powder inhalers (DPIs) can be important factors in their efficiency and potential issues for drug delivery. The aim of this work is to produce 3D printed structures of well-defined morphology and surface chemistry on the micron scale to aid our understanding of particle-surface interactions. This has been achieved using a novel two-photon 3D printing technique. This work shows the production of semi-circular test geometries on the micron scale, which have been coated with model coatings (plasma polymerised hexane and acrylic acid) producing a consistent well-defined surface chemistry. Initial particle-substrate adhesion measurements was acquired using atomic force microscopy force-volume mapping and demonstrated and the differences observed using different surface chemistries and different AFM colloidal probes discussed. The work aims to isolate and investigate the effect of morphology and surface chemistry on particulate adhesion and assist in the development of better DPI formulations and devices. Introduction In order to achieve drug delivery via the respiratory route, an understanding of particulate interactions is of importance. For successful delivery to the distal airways, an aerodynamic diameter of less than 5 µm must be achieved. However, using micronisation to produce particles of this size presents a difficult formulation challenge, [1] due to the inherent cohesiveness between particles and adhesion to the device, due to the high surface to volume ratio of such small particles, causing the particles to aggregate. This tendency can cause a reduction in dispersion, aerosolisation and device efficiency. Previous attempts to evaluate the interaction between pulmonary [2] particles and surfaces, have generally been restricted to model probes and inhalation surfaces , or by comparing [3] adhesion of different inhalation particles to a inhalation surface , both with no ability to alter the morphology of the inhalation surface. Research that has investigated morphology generally have used destructive techniques to [4] alter the roughness , which is likely to alter other factors such as surface energy and provide limited control for optimisation. We aim to overcome these issues by producing bespoke and easily modified well-defined microscale morphologies with a consistent surface chemistry using two photon polymerisation (TPP) 3D printing. TPP is a novel 3D printing technique which involves the curing of, usually acrylate containing, polymer resins by the absorption of two infra-red photons at the focus of a laser beam. TPP has been shown to produce a sub[5] diffraction limit lateral resolution of 120 nm , allowing the production of both controlled nano- and micro-scale roughness. As the polymer resin, is unaffected by individual IR photons, this effect is extremely localised in all [6] directions, eliminating the need for structures to be built in layers , allowing true 3D structure capability. The aim of this work is to produce 3D printed structures of well-defined morphology and surface chemistry on the micron scale. This is a scale which is at least two orders of magnitude improvement on current state of the art 3D inkjet printers. These structures will then be used to further understand the importance of morphology and surface chemistry on powder interaction in DPIs and powder blends in general. The objectives of this work is to reproducibly produce micron scale structures, ensuring a consistent surface chemistry, allowing adhesion measurements with colloidal probe microscopy. Method TM

A Nanoscribe photonic professional GT two photon printer was used to polymerise a commercial photoresin, IPL (Nanoscribe), on glass to produce controlled 3D surfaces. Surface imaging was conducted using atomic force microscopy, in tapping and quantitative nanomechanical measurement modes. Plasma polymerisation coating was used to produce a consistent surface chemistry and surface energy on the 3D printed surfaces. Hexane and acrylic acid monomers were used to form a polymeric coating of plasma polymerised hexane (ppHex) and plasma polymerised acrylic acid (ppAAc) respectively. Contact angle measurements were obtained using a CAM200. Colloidal probe microscopy was used to measure adhesion between the probe (polystyrene 10µm beads or Respitose (Lactose) SV003) and well defined surfaces. Force curves were collected in force-volume mode using an ICON AFM (BrukerNano) mapping was conducted on five 20 x 20 µm areas on two different coatings; ppHex and ppAAc, at room temperature and ambient relative humidity.

337


Drug Delivery to the Lungs 27, 2016 - Using 3D printed standards to isolate the effect of surface morphology and surface chemistry in DPIs Using 3D printed standards to isolate the effect of surface morphology and surface chemistry in DPIs Results and Discussion Curved edge ridges of a range of dimensions have been successfully printed. AFM was used to evaluate the shape and determine accurate height measurements for the fabricated structures. Figure 1A is an AFM 3D topography image showing the ridges with a width of 2.8 µm and a height of 1.3 µm. These ridges can then be successfully printed with reproducible spacing (see Figure 1B), in order to produce a well-defined morphology.

Figure1: A) AFM image of a nanoprinted ridges and B) a 3D topography image of multiple ridges with 3 µm spacing.

Figure 2 shows a comparison of the water contact angle obtained on ppHex and ppAAc coatings on glass coverslips. An uncoated standard glass coverslip possesses a static contact angle of 52° ± 5°. In comparison after ppHex was deposited to a 9 nm average thickness, a hydrophobic surface was generated with a contact angle of [7] 96° ± 4° as expected due to the hydrocarbon structure and in line with previous findings . A coating with ppAAc, with its added carboxyl functionality produced a more hydrophilic surface with an average contact angle of 64° ± [8] 3°, as expected .

Figure 2: A comparison of contact angle measurements of surfaces. (left) a ppHex coated glass coverslip and (right) a ppAAc coated coverslip.

Force-volume AFM mapping was conducted using an ideal probe (polystyrene 10 µm bead) and a lactose carrier particle (Respitose SV003 particle) and tested against each model coating. For each area of 20 x 20 µm evaluated, a low resolution topography map (Figure 3A) and corresponding adhesion map (Figure 3B) was created. These maps were produced by collecting 256 force curves (Figure 3C) across the area, the adhesion energy for each pixel was then calculated as the area under the hysteresis in the retract trace, as identified by the red marker, on Figure 3C. The respective topography data was then also recorded at these locations to produce an equivalent pixel image resolution. An example topography map showing a surface absent of any significant morphology, and its respective adhesion map for one such area are shown in Figure 3, along with a representative force curve.

338


Drug Delivery to the Lungs 27, 2016 – Georgina E. Marsh et al.

C)

Figure 3: A representative ppHex height image collected in force volume mapping, along with (below) the corresponding adhesion map, and (right) an example of a processed force-distance curve.

500

200

400

150

300

Count

Count

Figure 4 shows the resultant histogram plots obtained for the maximum adhesion pull-off force for a polystyrene 10 µm bead measured on a control ppHex and ppAAc coated surface. Figure 4A represents the adhesion between a polystyrene 10 µm bead and the ppHex coating, and is essentially unimodal as expected for a consistent surface chemistry, with a peak around 25 nN. However, in contrast the histogram in Figure 4B representing the same bead’s adhesion to a ppAAc coating, indicates a bimodal distribution with a peak around 4 nN and a peak around 23 nN. As ppAAc is more hydrophilic a lower unimodal adhesion to the hydrophobic polystyrene bead would be expected. The second higher peak is proposed to be due to capillary forces from moisture forming on the hydrophilic ppAAc chemistry, with some areas wetting faster than others. These forces [9] would lead to an increase in adhesion pull-off forces .

200

100 50

100

0

0 0

10

20

30

40

0

50

10

20

30

40

Adhesion (nN)

Adhesion (nN) (A)

(B)

Figure 4: Histograms showing the bin frequency of pull-off adhesion (dots) for five 20 x 20 µm areas of a polystyrene 10 µm bead from A) ppHex and B) ppAAc, the continuous interpolation of these points is then superimposed (lines).

In Figure 5, a Respitose SV003 particle was tested against five areas of the same ppHex and ppAAc coated surfaces. Figure 5A shows that the adhesion pull off for Respitose is reasonably unimodal for the ppHex surface, although displaying a broader peak than for the polystyrene at around 30 nN. The greatly increased size of Respitose SV003 and less well controlled contact geometry in comparison to the polystyrene bead, is likely to cause an increase in adhesion, with asperities present on the surface causing a greater spread of adhesion. Finally, in Figure 5B the adhesion histogram between a Respitose SV003 particle and a ppAAc coated surface is shown. As before this shows a bimodal distribution with less of an adhesion than the ppHex counterpart. However, this bimodal distribution possesses a narrower spacing with the two peaks at around 20 nN and 26 nN. If the occurrence of the two peaks in polystyrene 10 µm adhesion are due to capillary forces then the presence of the nanoscale asperities on the surface of the Respitose particle could be the reason for this difference. As [9] Hooton et al. showed, a nanoscale asperities can result in a change in the adhesion force response to humidity due to different modes of wetting that occur on highly curved nanoscale architectures.

339


120

300

100

250

80

200

Count

Count

Drug Delivery to the Lungs 27, 2016 - Using 3D printed standards to isolate the effect of surface morphology and surface chemistry in DPIs Using 3D printed standards to isolate the effect of surface morphology and surface chemistry in DPIs

60 40 20

150 100 50

0 0

10

20

30

40

0

50

0

Adhesion (nN)

10

20

30

40

Adhesion (nN)

(A)

(B)

Figure 5: Histograms showing the bin frequency of pull-off adhesion (dots) adhesion for five 20 x 20 µm areas of a Respitose SV003 particle from A) ppHex and B) ppAAc, the continuous interpolation of these points is then superimposed (lines).

Conclusions In conclusion, we have demonstrated the ability to print well defined geometries of an appropriate micron scale size range for particle adhesion testing, which opens up the opportunity to accurately and easily vary the morphology of a surface, keeping the surface chemistry homogeneous by plasma coating. Also, this work shows the beginning of adhesion measurements acquired using force-volume mapping and the differences obtained using different surface chemistries and different colloidal probes. Future work for this project will proceed on multiple fronts. Firstly, AFM measurements will be conducted at a low humidity to avoid capillary effects. This will then extend to measuring adhesion on the well- defined 3D printed topographies, with consistent surface chemistry. In parallel to this an investigation into designs and production of surfaces that are more relevant to the respiratory drug delivery is needed, as current surfaces are “ideal” models and therefore any trends seen will be difficult to extrapolate. References 1.

Zeng XM, Martin GP, Marriott C, Pritchard J. The influence of carrier morphology on drug delivery by dry powder inhalers. Int J Pharm 2000;200(1):93–106.

2.

Louey MD, Mulvaney P, Stewart PJ. Characterisation of adhesional properties of lactose carriers using atomic force microscopy. J Pharm Biomed Anal 2001;25:559–67.

3.

Young PM, Price R, Tobyn MJ, Buttrum M, Dey F. The Influence of Relative Humidity on the Cohesion Properties of Micronized Drugs Used in Inhalation Therapy. J Pharm Sci 2004;93(3):753–61.

4.

Zellnitz S, Schroettner H, Urbanetz NA. Influence of surface characteristics of modified glass beads as model carriers in dry powder inhalers (DPIs) on the aerosolization performance. Drug Dev Ind Pharm [Internet] 2015;41(10):1710–7. Available from: http://www.tandfonline.com/doi/full/10.3109/03639045.2014.997246

5.

Kawata S, Sun HB, Tanaka T, Takada K. Finer features for functional microdevices. Nature 2001;412(6848):697–8.

6.

Maruo S, Nakamura O, Kawata S. Three-dimensional microfabrication with two-photon-absorbed photopolymerization. Opt Lett [Internet] 1997 [cited 2014 Dec 11];22(2):132–4. Available from: http://www.opticsinfobase.org/abstract.cfm?URI=ol-22-2-132

7.

Zelzer M, Albutt D, Alexander MR, Russell N a. The role of albumin and fibronectin in the adhesion of fibroblasts to plasma polymer surfaces. Plasma Process Polym 2012;9(2):149–56.

8.

Alexander MR, Duc TM. A study of the interaction of acrylic acid/1,7-octadiene plasma deposits with water and other solvents. Polymer (Guildf) 1999;40(20):5479–88.

9.

Hooton JC, German CS, Allen S, Davies MC, Roberts CJ, Tendler SJB, et al. An Atomic Force Microscopy Study of the Effect of Nanoscale Contact Geometry and Surface Chemistry on the Adhesion of Pharmaceutical Particles. Pharm Res [Internet] 2004;21(6):953–61. Available from: http://link.springer.com/10.1023/B:PHAM.0000029283.47643.9c

340


Drug Delivery to the Lungs 27, 2016 - Thomas Kopsch et al. Modelling Drug Entrainment in a Dry Powder Inhaler: Benchmarking and Sensitivity Analysis of a Multiphase CFD Approach 1

1

Thomas Kopsch , Digby Symons & Darragh Murnane 1

2

University of Cambridge, Trumpington Street, CB2 1PZ, UK 2 University of Hertfordshire, College Lane, AL10 9AB, UK

Summary Dry powder inhalers (DPIs) are used to deliver drug powder to the lungs. When a patient inhales through a DPI, air flows through the drug compartment and entrains drug. The rate of entrainment depends on a number of factors, including the inhalation flowrate, the geometry of the inhaler and physicochemical properties of the drug formulation. While the influence of each of these factors has been studied experimentally, a robust computational method to predict drug entrainment accurately for real world combinations of these parameters would be useful in the design and optimization of DPIs. The objective of this study was to investigate the applicability of a multiphase computational fluid dynamics (CFD) simulation approach: Firstly, results of two different CFD solvers (ANSYS Fluent and OpenFOAM) were compared with experimental data. Secondly, the sensitivity of the drug release rate to variations of drug properties and CFD solver settings was investigated. Particularly, (a) the size of drug particles, (b) the density of particles, (c) the initial volume fraction, α, (d) the drag model, (e) the packing limit, αmax, and (f) the solids pressure were varied. This investigation showed that the two different CFD solvers produced similar results and that these results were consistent with experimental data. The sensitivity analysis indicated that the simulated entrainment is not strongly dependant on either particle size or the choice of drag and solids pressure model. However, drug density, the initial volume fraction of drug, and the choice of the packing limit, do significantly influence the results. Introduction When designing dry powder inhalers (DPIs) it is desired to achieve a number of medical, manufacturing and commercial objectives. Possible objectives of optimizing DPIs include: (A) independence of the drug emission [1, 2] . To optimize DPIs in accordance with profile for different patients and (B) achieving a desired emission profile these objectives, it is necessary to predict how well a DPI can achieve them, as a function of the entrainment geometry and physical properties of the drug powder. Computational fluid dynamics (CFD) simulations are often [3, 4] . However, most of these approaches are either single phase simulations or used to assess DPIs particle-tracking (Eulerian-Lagrangian) approaches. Only a few studies exist where an Eulerian-Eulerian (EE) [1, 5] . In EE approaches all phases are treated as continuous phases. When applied to the approach was applied simulation of DPIs, one phase may represent air, while the other phase represents formulation components. In order to simulate a granular phase, like drug powder, with an EE approach, additional models are required to describe the interaction between the fluid (air) and the granular phase (drug formulation). The kinetic theory of granular flow (KTGF) was developed to describe the constitutive behaviour of granular material. In order to apply KTGF in EE simulations, sub-models have to be selected. For example, these include sub-models for the drag of air on particles (drag coefficient), for the probability of particle-particle collisions (radial distribution function) and [6] for detailed for the solids pressure (the additional pressure due to the presence of the granular phase). See description of these sub-models. Zimarev et al. applied EE CFD simulations to model the entrainment of drug in a [5] DPI entrainment geometry and to optimize the entrainment part of a DPI . The CFD solver used was ANSYS [6] [5] Fluent . In this EE-model Zimarev et al. compared CFD predictions with experimental results from Tuley et [7] [1] [8] as a CFD solver and al. . Kopsch et al. used a similar approach to optimize DPIs , but used OpenFOAM improved objective functions. In both studies the DPI geometry was optimized for a given drug powder, with defined particle size, density and porosity. However, a direct comparison of these two CFD packages for DPI entrainment simulations has not been conducted. It is obviously desirable that simulation results should be independent of the particular CFD solver used. There are two objectives of this study: Firstly, to investigate whether two different CFD solvers, i.e. ANSYS [6] [8] Fluent and OpenFOAM , produce similar results and to compare these results to experimental data. Secondly, to analyse the sensitivity of the results when various settings were varied. In particular, the mass of entrained drug M(t) was plotted as a function of time when six different powder properties and CFD solver settings were changed. Powder properties that were varied include (a) the size of drug particles, (b) the density of particles and (c) the initial volume fraction α. CFD solver settings that were varied include; (d) the drag model, (e) the packing limit αmax and (f) the solids pressure.

341


Drug Delivery to the Lungs 27, 2016 - Modelling Drug Entrainment in a Dry Powder Inhaler: Benchmarking and Sensitivity Analysis of a Multiphase CFD Approach Methods CFD Eulerian-Eulerian approach [7]

The entrainment geometry considered was the 90° geometry chosen by Tuley et al. . This geometry was drawn and meshed, Figures 1 & 2. An Eulerian-Eulerian CFD case was prepared with both ANSYS Fluent and OpenFOAM. In both CFD packages the boundary conditions of the problem had to be specified. The inlet boundary condition was a constant atmospheric pressure boundary condition. The outlet boundary condition was a transient pressure boundary condition. In ANSYS Fluent the time-dependant pressure boundary condition was [9] was installed to specify the boundary specified with tabular data. In OpenFOAM the library swak4FOAM condition. The initial conditions of the problem had to be also specified. This included the location of the drug powder and the initial volume fraction of the drug. Other physical conditions that were specified were the density of air and drug, the size of drug particles and parameters for the KTGF sub-models. Comparison of the CFD approaches with experimental results [7]

The entrainment of lactose (16% fines) and glass powder was modelled using the powder properties found in . Initially, the KTGF models shown in Table 1 were chosen. Note that not all models were available in both Fluent and OpenFOAM. Model

OpenFOAM

ANSYS Fluent

Viscosity Model / Granular Viscosity

Gidaspow

Gidaspow

Frictional Stress Model / Frictional Viscosity

Johnson Jackson

Johnson et al.

Conductivity Model / Granular temperature

Gidaspow

Algebraic

Granular Pressure Model / Solids Pressure

Lun

Lun et al.

Radial Distribution

Lun Savage

Lun et al.

Drag Coefficient

GidaspowErgunWenYu

Gidaspow

Table 1 - Initial selection of KTGF models

Sensitivity analysis The sensitivity of drug entrainment to three drug properties ((a) size of drug particles, (b) particle density, (c) initial particle volume fraction) and three CFD model settings ((d) drag model, (e) packing limit, (f) solid pressure) was studied. This was performed in Fluent by varying one of these settings while keeping the other settings fixed.

Figure 1 Comparison of experimental results from Tuley [7] et al. with CFD simulation results from ANSYS [6] Fluent and [8] OpenFOAM . This comparison is for glass powder (mean particle d = 45 Âľm). Time before the beginning of emptying excluded.

342


Drug Delivery to the Lungs 27, 2016 - Thomas Kopsch et al.

2

4

Released drug in 10-5 kg

Released drug in 10-5 kg

Figure 2 - Comparison of experimental [7] results from Tuley et al. with CFD [6] simulation results from ANSYS Fluent . This comparison is for lactose (16% fines) powder. Time before the beginning of emptying excluded.

1.5 1 0.5 0 0

0.05

0.1

0.15

0.2

3 2 1 0 0

time in s d=35 μm

d=70 μm

Released drug in 10-5 kg

Released drug in 10-5 kg 0.05

0.1

0.15

2 1 0.5 0 0

0.2

α=0.75

0.05

αmax=0.55

c) Initial volume fraction in powder bed α

Released drug in 10-5 kg

Released drug in 10-5 kg

1.5 1 0.5 0 0.1

time in s

WenYu e) Model for drag coefficient

0.1

0.15

0.2

0.15

αmax=0.8

αmax=1.0

d) Packing limit αmax

2

symmetric

ρ=1550 kg/m^3

time in s

α=0.48

0.05

0.2

1.5

time in s

0

0.15

b) Solid density of drug ρ

2.5 2 1.5 1 0.5 0

α=0.3

0.1

time in s

ρ=800 kg/m^3 ρ=3100 kg/m^3

d=140 μm

a) Size of particles d

0

0.05

2 1.5 1 0.5 0 0

0.2

0.05

0.1

time in s

0.15

0.2

ma-ahmadi Lun et al. syamlal-obrien

Gidaspow SyambalObrien

f) Model for solids pressure

Figure 3 - Drug emission profiles M(t), when various particle properties or CFD settings were varied while all other [6] properties and settings were kept the same. The CFD solver was ANSYS Fluent . -3

The default settings are: size of particles d = 70 µm, solid density of drug ρ = 1550 kg m , initial volume fraction in powder bed α = 0.48, packing limit αmax =0.8, model of drag coefficient: Gidaspow, model for solids pressure: Lun et al.

343


Drug Delivery to the Lungs 27, 2016 - Modelling Drug Entrainment in a Dry Powder Inhaler: Benchmarking and Sensitivity Analysis of a Multiphase CFD Approach

Results and discussion Figure 1 shows good agreement between experimental data and the CFD predictions from Fluent and OpenFOAM when glass powder was used. Not all KTGF sub-models were available in both Fluent and OpenFOAM and thus some differences in the results from these two solvers are to be expected. Similarly, Figure 2 compares experimental data and predictions from Fluent. These predictions can be compared to [2] and are in good agreement. Figure 3 shows the mass of drug OpenFOAM results for the same geometry in released M(t), from the compartment as a function of time, t, when various settings were varied (sensitivity analysis with ANSYS Fluent). It was observed that; (a) the size of drug particles d, (e) the choice of the drag model, and (f) the choice of the solids pressure model had little influence on the release rate. Note that the EE approach used in this study is not capable of modelling cohesive forces between particles. For particles of small [10] size (d < 15 µm) it is known that cohesive forces may significantly influence entrainment behaviour . For this reason the EE approach tested in this study is only relevant for large particles d > 15 µm. Comparison of Figures 1 & 2 shows that the entrainment rate, in terms of (visible) volume fraction, is higher for the lower density lactose powder. This may appear to contradict the results shown in Figure 3b. However, in all of Figure 3 it is the mass of drug released that is plotted as a function of time. In the case of Figure 3b the increase in powder density dominates the lower volumetric release rate and therefore the rate of mass entrainment is actually higher for higher density powders. In Figure 3c the final total released mass reflects the different initial volume fraction in the powder bed (i.e. all powder has been evacuated). However, the results show that if the initial volume fraction α is close to the packing limit αmax then the release is somewhat delayed during the initial stages. Figure 3d shows that the mass release rate is only insensitive to the packing limit αmax when it is significantly higher than the initial volume fraction in the powder bed (α = 0.48). Conclusions In summary, it has been shown that two different CFD solvers can predict the rate of drug entrainment in the chosen geometry with similar accuracy and a preliminary sensitivity analysis has been conducted. It is hoped that these results may be useful in supporting the use of CFD to predict entrainment of drug to aid the design and optimization of DPIs. References 1.

Kopsch T, Symons D, Murnane D. Design-Optimization of Dry Powder Inhalers: Selecting an Objective Function. In: DDL 26, drug delivery to the lungs 26. 2015.

2.

Kopsch T, Murnane D, Symons D. Optimizing the entrainment geometry of a dry powder inhaler: Methodology and preliminary results. Pharm Res. 2016;

3.

Ruzycki C a, Javaheri E, Finlay WH. Review: The use of computational fluid dynamics in inhaler design. Expert Opin Drug Deliv. 2013;10:307–23.

4.

Wong W, Fletcher DF, Traini D, Chan HK, Crapper J, Young PM. Particle aerosolisation and break-up in dry powder inhalers 1: Evaluation and modelling of venturi effects for agglomerated systems. Pharm Res. 2010 Jul;27(7):1367–76.

5.

Zimarev D, Parks G, Symons D. Computational Modelling and Stochastic Optimisation of Entrainment Geometries in Dry Powder Inhalers. In: DDL 24, drug delivery to the lungs 24. 2013.

6.

ANSYS I. Ansys ® fluent ® 12.0 Theory Guide. 2009.

7.

Tuley R, Shrimpton J, Jones MD, Price R, Palmer M, Prime D. Experimental observations of dry powder inhaler dose fluidisation. Int J Pharm. 2008 Jun 24;358(1-2):238–47.

8.

The OpenFOAM Foundation. OpenFOAM 2.4 http://www.openfoam.org/.

9.

Https://openfoamwiki.net/index.php/Contrib/swak4Foam, 2016. swak4Foam (accessed 15th February).

10.

Finlay WH. The Mechanics of Inhaled Pharmaceutical Aerosols - An Introduction. Academic Press; 2001.

344


Drug Delivery to the Lungs 27, 2016 - Cathy McKenna et al. Tracheal Inhalation Comparison across Three Different Inhalation Methods - Normal Adult vs. an Asthmatic Adult Cathy McKenna, Andrew O’Sullivan, Patrick Kelly, Louise Sweeney, Conor Duffy, Ronan MacLoughlin Aerogen, IDA Business Park, Dangan, Galway, Ireland. Summary The following study investigates and compares the effect of aerosol delivery through three different inhalation methods (nose only, mouth only & nose/mouth simultaneously). Typically testing is carried out with inhalation through the nose and the mouth simultaneously, however we were interested in discovering whether a larger quantity of drug was being inhaled via the mouth, or the nose separately using a mesh nebuliser/adapter combination in line with a facemask. In this study Albuterol was used as a tracer aerosol to measure the drug delivery to both adult patients, with and without asthma. During an asthma attack the muscles around the airways tighten (bronchospasms) and are common in patients with asthma. Albuterol is used as a bronchodilator, relaxing the muscles around the airways [3] which eases chest tightness . An adult head model (LUCY) was used to simulate both normal adult and asthmatic adult breathing patterns. The nose and mouth tracheal doses were measured separately by blocking the mouth or nose in order to block any aerosol affecting the result. Results: On completion of testing it was discovered that as the supplemental gas flow increased, the (%) Tracheal dose delivered of the medicinal product decreased. The preferred supplemental gas flow for both normal adult and adult with asthma was 2 LPM. The greatest (%) Tracheal dose delivered for a normal adult was the mouth inhalation method (39.2% ± 0.498%). The inhalation method for the adult with asthma that gave the greatest (%) Tracheal dose recovery was inhalation through the nose (28.305% ± 0.33%) Introduction Inhalation methods can have a substantial impact when treating a patient. The airflow to the lungs can have an impact on the amount of drug inhaled e.g. an increase in inspiratory flowrate causes a decrease in drug [5] delivery . The difference between three different inhalation methods (nose only, mouth only and nose/mouth simultaneously) was investigated. The intent of this study was to identify the optimal inhalation technique for tracheal dose delivery when using a facemask, with an adapter and vibrating mesh nebuliser. Gas flow rates can [1] have a substantial impact on aerosol drug delivery , three different supplemental gas flow rates (2, 4 & 6 LPM) [2,4] . Tracheal dose (%) were tested across two different breathing parameters (normal adult & asthmatic adult) was calculated and statistical analysis was performed to ascertain if there were significant differences between the inhalation methods across the matrix of head models and breathing patterns analysed. Experimental methods and materials A breathing simulator (Ingmar ASL 5000) was used to simulate two different breathing patterns, normal adult (15 BPM, Vt 500 mL, I:E 1:1) and asthmatic adult (22 BPM, Vt 290 mL, I:E 1:2). Tracheal dose (%) at each supplemental gas flow rate under test (2, 4 and 6 LPM) was recorded (n=3) for each inhalation method/breathing pattern combination, which were connected to a breathing simulator via an absolute filter (RespirGard II 303, Baxter). This study incorporated an adult head model with three different inhalation methods, see Figure 1; (nose only with mouth taped closed, mouth only with nose taped closed and nose/mouth simultaneously). A 2.5 mL dose of albuterol sulphate (1 mg/mL) was nebulised as a tracer aerosol using a vibrating mesh nebuliser (VMN) (Aerogen Solo, Aerogen, Ireland), with an average volume median diameter (VMD) of 4.4 µm, used in combination with the Aerogen Ultra adapter for testing with a valved facemask. The drug delivery was measured by means of UV Spectroscopy at a wavelength of 276 nm. The (%) Tracheal dose recovered was then calculated. Finally, a paired t-test analysis was carried out to determine whether there were significant differences between the inhalation methods.

345


Drug Delivery to the Lungs 27, 2016 - Tracheal Inhalation Comparison across Three Different Inhalation Methods - Normal Adult vs. an Asthmatic

Figure 1: Adult head model breathing simulation.

Results

Adult Normal % Recovery results 45.000 40.000 35.000 30.000 25.000 20.000 15.000 10.000 5.000 0.000

2 LPM

4 LPM

6 LPM

Nose Inhalation

39.200

34.933

34.324

Mouth Inhalation

39.200

30.248

33.105

Unblocked inhalation

34.324

31.086

30.400

Nose Inhalation

Mouth Inhalation

Unblocked inhalation

Figure 2: Normal adult breathing simulation recovery results (%) across all supplemental gas flows tested.

346


Drug Delivery to the Lungs 27, 2016 - Cathy McKenna et al.

Adult Asthma % recovery 35 30 25 20 15 10 5 0

2 LPM

4 LPM

6 LPM

Nose Inhalation

28.305

24.267

21.410

Mouth Inhalation

27.467

24.381

20.343

Unblocked inhalation

28.838

20.152

16.648

Nose Inhalation

Mouth Inhalation

Unblocked inhalation

Figure 3: Asthmatic adult breathing simulation recovery results (%) across all supplemental gas flows tested.

Discussion and Conclusion As intended, a particular means of inhalation and a supplemental gas flow could be chosen for maximum inhalation for each type of breathing simulation. Normal adult (see Figure 1): Based on the results presented, the mouth inhalation & 2 LPM supplemental gas flow rate facilitated the highest tracheal dose (39.2% Âą 0.498%) for the normal adult breathing pattern. Although the mouth inhalation gave the same (%) Tracheal dose recovered as the nose inhalation (39.2%), it had greater deviation between results (1.814), making the mouth inhalation the preferred method of inhalation for a normal adult. Asthmatic adult (see Figure 2): The preferred supplemental gas flow rate that gave the greatest percentage Tracheal delivery was 2 LPM. Although the nose and mouth breathing simultaneously (unblocked) gave the greatest Tracheal dose recovered (28.838%), it also gave the largest standard deviation (2.441%). The nose inhalation however, gave a (%) Tracheal recovery with no significant difference to the unblocked inhalation (P=0.386, 28.305%), and with less deviation (0.33%) and is therefore the preferred method of testing for an asthmatic adult. Although 2 LPM supplemental gas flow gave the largest (%) Tracheal delivered for asthmatic adult (28.305%, 27.467% & 28.838%), it was significantly lower (P=0.009, P=<0.001, P=0.046) than that of a normal adult at 2 LPM supplemental gas flow (39.2%, 39.2% & 34.324%).

References (1)

Officer TM, Pellegrino R, Brusasco V, Rodarte JR. Measurement of pulmonary resistance and dynamic compliance with airway obstruction. J Appl Physiol 1998;85(5):1982-1988.

(2)

Woolcock AJ, Read J. The static elastic properties of the lungs in asthma. Am Rev Respir Dis 1968;98(5):788-794.

(3)

GlaxoSmithKline, GSK, 2014, Ventolin HFA (albuterol sulfate) Inhalation Aerosol, accessed: 28 th July 2016, URL: https://www.gsksource.com/pharma/content/dam/GlaxoSmithKline/US/en/Prescribing_Information/Ventolin_HFA/pdf/ VENTOLIN-HFA-PI-PIL.PDF

(4)

Pilbeam SP. Chapter 2: Basic terms and concepts of mechanical ventilation. In:Cairo JM. Pilbeam’s mechanical ventilation physiological and clinical applications, 5th edition. St Louis, MO: Mosby|Elsevier; 2012:15-30.

(5)

Fernandes Tena. A, Casan Clara . P, 2012, Elsevier, accessed: 28th July 2016, URL: http://www.archbronconeumol.org/en/deposition-inhaled-particles-in-lungs/articulo/S1579212912000845

347


Drug Delivery to the Lungs 27, 2016 – Neritan Alizoti et al. A New Size Facemask for Use with a Valved Holding Chamber (VHC) by Adults with Smaller Facial Features 1

1

1

1

Neritan Alizoti , Noel Gulka , Jason A Suggett , Robert Morton , Jolyon P. Mitchell

2

1

2

Trudell Medical International, 725 Third Street, London, Ontario, N5V 5G4, Canada Jolyon Mitchell Inhaler Consulting Services Inc., 1154 St. Anthony Road, London, Ontario, N6H 2R1, Canada

Summary Facemasks are widely used with valved holding chambers (VHCs) that are prescribed for use with pressurized metered dose inhalers for patients who have difficulty coordinating inhalation with inhaler actuation. Adults also require this type of interface, but there is a wide variety of facial profiles that must be accommodated to minimize dead space and the risk of leakage of ambient air at the facemask-to-face seal. We describe several key steps that were taken to design a slightly smaller size of facemask to accompany the current large adult facemask supplied with the AeroChamber Plus* Flow-Vu* VHC. The desired facemask size lies between the sizes of the medium (child) and large (adult) facemask sizes. The new facemask is primarily intended for adults having smaller facial features. We describe the use of anthropomorphic data pertinent to this class of user that, in conjunction with the identification of key dimensions, enabled the development of prototype facemask designs for optimisation via user evaluation. Each prototype was evaluated in terms of of a) inhalation flow rate as a measure of face-to-facemask fit, and b) user preference for comfort / fit. The prototype selected for progression performed best overall in terms of flow rate and fit. The aim of this abstract was to describe the process followed in order to design and optimize, with respect to fit and comfort, a new VHC facemask developed for adults with smaller facial features. Introduction VHCs are widely prescribed for adult patients with obstructive lung disease as they avoid the need to coordinate [1] actuation of the pressurized metered dose inhaler (pMDI) with the onsert of inhalation , and virtually eliminate coarser particles that would otherwise deposit in the upper airway where they may result in undesirable systemic [2,3] . These add-on devices may be absorption and / or local symptoms such as dysphonia and candidiasis [4] supplied with a mouthpiece or facemask as the patient interface . However, the facemask option is more [5] appropriate for patients having co-morbidities affecting motor skills or with limited cognitive ability , and is therefore more likely to be required by the older patient. Regardless of patient age, it is essential that the facemask seals properly with the face to avoid any ingress of ambient air via leakage pathways, particularly via [6] the chin and nose-bridge that can dilute the aerosol delivered to the lungs . Typically, only one size of facemask is made available by the VHC manufacturer to be used with VHCs intended for a particular sub-population (i.e. infant, child, adult). However, in the case of the adult sub-population, significant differences in facial size and [7] shape associated with gender and ethnic origin make it difficult to ensure leak-tightness for all in this group with just one size of facemask. We therefore foresaw a need to better serve the adult population by introducing a new size of facemask for adults having smaller facial features that would be a good companion to our current large facemask intended for this sub-population. The purpose of the investigations reported herein is to describe the process followed in order to design and optimize, with respect to fit and comfort, a new VHC facemask developed for adults with smaller facial features.

Methodological Approach to Development. – Including Results th

The target for the adult with smaller facial features patient population was determined to be the 50 percentile [8,9] . The facemask adult female, and primary ergonomic anthropometric source data were utilized as reference design was intended to be suitable across a range of facial sizes below and above the target anthropometric data chart sub-population. This intention is helped to be made possible by the large surface area coming in contact [4] with the face associated with the ergonomic ComfortSeal* contour cushion, forming the mask periphery .

348


Drug Delivery to the Lungs 27, 2016 - A New Size Facemask or Use with a Valved Holding Chamber (VHC) by Adults with Smaller Drug Delivery to theFacial LungsFeatures 27, 2016 - A New Size Facemask

Figure 1 – Key Dimensions Showed on 3D Female Model Head [3]

The 3D CAD (Computer Aided Design) model female head, shown in Figure 1 for illustrative purposes, identifies three key dimensions critical for face to facemask fit and they are further described below: N = bridge of nose width X = maximum width of mouth (K) plus 15 mm / each side Y = distance from bridge of nose to bottom of chin, minus 60% of lower lip to bottom of chin distance th

The magnitude of each key dimension was targeted based upon the 50 percentile female anthropomorphic data set, and was defined in relation to the facemask contact points, that were in turn defined by the centre-line of the ComfortSeal* contour, Figure 2.

Figure 2 - Key Dimensions of the Facemask

Front and side views of the protopype 1 facemask overlaid on the model female head are illustrated in Figure 3.

Figure 3 – Front and Side Views of the Mask Overlaid on a 3D Model Head

An iterative process was subsequently followed to refine the design further. Recognising that theory is good to set the target, but real life user testing is required to optimize further, the development process incorporated user studies to confirm suitability of the final facemask configuration.

349


Drug Delivery to the Lungs 27, 2016 – Neritan Alizoti et al. th

th

An initial assessment of fit with 12 volunteers, representative (between 25 and 75 percentiles for key [9] dimensions) of the target patient population , identified that, at least in terms of comfort, slightly larger dimensions of X and Y may be preferred. Hence, two additional designs (prototypes 2 and 3) were developed with incrementally larger X and Y dimensions. The three prototype facemasks are shown in Figure 4.

Figure 4 – Three Small Adult Facemask Prototypes Tested

The three prototypes were tested in terms of a) inhalation flow rate, a measure of face to facemask fit, and b) patient preference for comfort / fit. Flow rate performance The test set-up is shown in Figure 5. Each of the 12 volunteers inhaled (at a comfortable inhalation rate for them) through each prototype facemask in turn, with the facemask attached to a VHC (adult antistatic AeroChamber Plus* Flow-Vu* ,Trudell Medical International, London, Canada). The port of the VHC normally used to couple the pMDI was instead coupled to a thermal mass flowmeter (model 4040, TSI Inc., St. Paul, MN, USA) by means of corrugated 22 mm laboratory tubing with a protecting in-line filter.

Figure 5 – Schematic Showing Arrangement for Flow Test of VHC-Facemask with Model Face

The outcome from these tests is summarized in Table 1. A higher flow rate was indicative of a better facemaskto-face seal, whereas a lower flow rate would be indicative of air leakage within the system. Given that each individual user had a different inherent inhalation flow rate, the results for the different prototype masks were expressed simply as a ranking from high to low flow rate, for each user. A ‘sense check’ was performed if the numerical values were within +/- 10% and in that situation no difference was noted in the rank. User preference: comfort and fit Each of the prototype facemasks were also evaluated for perceived fit, by the 12 volunteers. They were asked to rank the three masks in preference order, with the outcome summarized in Table 1. Prototype 2 performed best overall for both flow rate and fit. It was ranked at least joint highest by all users in terms of flow rate (facemask to face seal) and highest ranking for comfort / fit by over 50% of users (in top two ranking positions for all). Hence, this prototype mask design was selected to progress into production indicative injection moulding. A confirmation user study was finally completed using the molded prototype and the results for flow rate and comfort/fit were consistent with those from the prototype evaluation.

350


Drug Delivery to the Lungs 27, 2016 - A New Size Facemask or Use with a Valved Holding Chamber (VHC) by Adults with Smaller Drug Delivery to theFacial LungsFeatures 27, 2016 - A New Size Facemask Table 1: Evaluation of Facemask Relative Performace by Inhalation Flow Rate and User Perceived Fit

Prototype #1

1 = Best 3 = Worst

Inhalation Flow Scores

Subject 1

Prototype #2

User Fit Scores

Inhalation Flow Scores

1

3

Subject 2

2

Subject 3

Prototype #3

User Fit Scores

Inhalation Flow Scores

User Fit Scores

1

2

1

1

2

1

1

2

3

2

3

1

1

2

2

Subject 4

1

3

1

1

1

2

Subject 5

1

3

1

1

1

2

Subject 6

1

1

1

2

1

3

Subject 7

1

2

1

1

1

3

Subject 8

3

3

1

2

1

1

Subject 9

3

3

1

2

1

1

Subject 10

2

3

1

1

2

2

Subject 11

1

2

1

1

1

3

Subject 12

3

3

1

2

2

1

Prototype Ranking (number of users) Prototype #1

Prototype #2

Prototype #3

1st

6

1

12

7

8

4

2nd

3

3

0

5

4

4

3rd

3

8

0

0

0

4

Conclusion Development of a new small adult facemask using anthropometric data as a starting point, followed up by refining the design through user evaluations, proved to be an effective approach, based upon both theory and real life application. The result is a new facemask that will better fit the smaller facial features of some adult patients, giving the prescribing clinician more options to ensure optimum inhaled drug delivery via the pMDI-VHC therapeutic modality. Further investigational work will be performed with this new facemask, including in vitro drug delivery performance using the ADAM adult face model and upper airway. References 1

Mitchell J P, Dolovich M B: Clinically Relevant Test Methods to Establish In Vitro Equivalence for Spacers and Valved Holding Chambers Used with Pressurized Metered Dose Inhalers (pMDIs), J Aerosol Med Pulmon Deliv 2012; 25(4): pp217-242.

2

Rachelefsky GS, Liao Y, and Faruqi R: Impact of inhaled corticosteroid-induced oropharyngeal adverse events: re- sults from a meta-analysis, Ann Allergy Asthma Immunol 2007;98: pp.225–238.

3

Barnes NC: The properties of inhaled corticosteroids: similarities and differences, Prim Care Resp J 2007;16: pp149–154.

4

Dolovich MB, Ahrens RC, Hess DR, Anderson P, Dhand R, Rau JL, Smaldone GC, Guyatt G: Device selection and outcomes of aerosol therapy: evidence-based guidelines, Chest 2005;127: pp335–371.

5

Mitchell J P, Morton R W: Design of Facemasks for Delivery of Aerosol-Based Medication via Pressurized Metered Dose Inhaler with Valved Holding Chamber: Key Issues that Affect Performance, J Aerosol Med. 2007;20S1: pp. S29-S45.

6

Esposito-Festen JE, Ates B, Van Vliet FLM, Verbraak AFM, de Jongste JC, Tiddens HAWM: Effect of a facemask leak on aerosol delivery from a pMDI-spacer system, J Aerosol Med 2004;17: pp1–6.

7

Zhuang Z, Landsittel D, Benson S, Roberge R, Shaffer R: Facial Anthropometric Differences among Gender, Ethnicity, and Age Groups, Ann Occup Hyg 2010;54(4): pp. 391–402.

8

Humanscale 6A Head and Vision by Henry Dreyfuss Associates The MIT Press, New York, USA: 1981, ISBN: 0262540274 (v.1 :0-262-04059-X : v.2-3).

9

Young Joseph W.: Head and Face Anthropometry of Adult U.S. Citizens, Civil Aeromedical Institute, OK, USA: 1993, Report No. DOT/FAA/AM-93/10

351


Drug Delivery to the Lungs 27, 2016 – Ronan Mac Loughlin Can Aerosol Therapy Keep Pace with Innovations in Patient Care? A Review. Ronan Mac Loughlin Aerogen, IDA Business Park, Dangan, Galway, IRELAND Summary The opportunity for aerosol mediated therapy exists across all patient types and all patient settings. Patients ranging from very low birth weight premature infants to the elderly and obese may benefit from aerosol therapy, and whilst it is well understood that the associated range of breathing patterns has a significant bearing on the ultimate efficiency of delivery of aerosol to the lung, so too does the interface that facilitates connection between the aerosol generator and the patient airway. The aim of this review was to assess the effect of selected interfaces on the quantity of aerosol being delivered to the lung. For the purposes of this review, patient interventions were broadly divided between those for use with spontaneously breathing patients and those for use with patients requiring ventilatory support. The aerosol generators included in the review were Vibrating Mesh Nebulisers (VMN), Jet Nebulisers (JN) and pressurised Metered Dose Inhalers (pMDI). The patient interventions and associated interfaces assessed included facemasks, mouthpieces, tracheostomy tubes, high flow nasal therapy, endotracheal tubes and mask-mediated non-invasive ventilation. Following a review of the literature it is clear that the different aerosol generators can be significantly affected by choice of interface and published data is provided to support the contention that each interface introduces the risk of altered aerosol therapy. This work should allow end users and drug developers make more informed decisions around interface and paired device selection early in their respective decision making and development processes.

Introduction The opportunity for aerosol mediated therapy exists across all patient types and all patient settings. Patients ranging from very low birth weight premature infants to the elderly and obese may benefit from aerosol therapy, and whilst it is well understood that the associated range of breathing patterns has a significant bearing on the ultimate efficiency of delivery of aerosol to the lung, so too does the interface that facilitates connection between the aerosol generator and the patient airway. Beyond patient type and patient setting, the default choice of patient interface, even in the critical care environment, is now more than likely to be non- or minimally invasive. For example, where invasive intubation was previously used in the majority of cases, high flow nasal therapy (HFNT) is becoming increasingly prevalent in the treatment of any patient requiring ventilatory support. Fitting it to the patient requires little training, or can be done by the patient themselves, and the equipment is cheap and readily available. HFNT is even finding increasing use in the treatment of critically ill newborns over the traditional nasal CPAP intervention. Further, with the increased burden of high value capital equipment, i.e. mechanical ventilators, exceptional circumstances such as the SARS epidemic, have led to the adoption of cheaper and more accessible means of ventilatory support, i.e. non-invasive ventilators (NIV). All these have a multitude of specific patient interfaces that may be used, and so with each there is a risk of altered aerosol therapy. Whilst many users assume that the aerosol dose delivered to the lung is agnostic to the patient intervention and interface, the simple fact is that it is not. The aim of this review is to provide an overview of the effect of selected interfaces on the amount of aerosol being delivered to the lung. This work should allow end users and drug developers make more informed decisions around interface and paired device selection early in their respective decision making processes. For the purposes of this review, patient interventions were broadly divided between those for use with spontaneously breathing patients and those for use with patients requiring ventilatory support. The aerosol generators included in the review were Vibrating Mesh Nebulisers (VMN), Jet Nebulisers (JN) and pressurised Metered Dose Inhalers (pMDI).

Spontaneously breathing patients Facemasks are commonly used for increased patient compliance and the reduced requirement for coordinated breathing, for example by an obligate nasal breathing infant. Different facemask designs lend themselves to use

352


Drug Delivery to the Lungs 27, 2016 -Can Aerosol Therapy Keep Pace with Innovations in Patient Care? A Review. with various aerosol generator types, e.g. open, non-valved for use with JN, and valved for use with VMN and pMDI. Several reports describe the difference in lung dose using the combination of device and open or valved mask. Bench comparisons between JN and VMN indicate that there can be large difference in delivered dose, for example at ~2% and ~29% respectively in a healthy model and ~1.6 and 12.6% respectively in a model of [1] COPD . PMDI delivery rates in combination with a valved spacer and facemask in vivo are reported to be [2] ~28% . In real use, the reasons for variability in delivered dose include the obvious fit issues that may result from an ill-sized or poorly designed mask but also the open vented facemask designs that prevents the between breath build-up of a bolus of aerosol within the facemask. Features such as valves or filters can increase the holding capacity of the facemask, but the likely greatest contributor to losses is the supply of carrier gas, at the gas flow rates required by JN. JN are not generally supplied with or recommended for use with valved or filtered masks.

Mouthpieces are used with patients able to hold the aerosol generator in hand and coordinate their breathing. Again these are supplied valved and non-valved. Mouthpieces are commonly used with both JN and VMN, and are an inherent part of MDI design. Across the publications reviewed, none were identified that directly assessed mouthpiece design. However several device comparisons were reported. In these, VMN consistently outperformed JN with scintigraphic studies [3,4] . No single study was suggesting that JN achieved ~4-5% lung dose whilst VMN achieved 22-32% in adults identified that compared VMN, JN and pMDI performance, however, pMDI deposition rates were reported to be [2] ~15% .

Tracheostomy tubes (TT) are short tubes that are inserted into the patient airway via an incision in the windpipe. Typically chosen for rapid instrumentation of the airways in case of an emergency, but may remain long term and go home with the patient as they are discharged from the clinical setting. Often in-dwelling, the patient breaths through the tracheostomy. Aerosol is administered via a direct connection or more usually a loose fitting, open mask. VMN, JN and pMDI are commonly used devices that interface with the tracheostomy tube. Bench studies suggest that aerosol delivery varies between devices with approximately 19%, 6% and 47% of the dose being delivered [5,6] . Of note however, despite the high deposited fraction of the pMDI, the mass of drug delivered respectively (0.20 mg) was not the highest with VMN recording 0.49 mg and JN 0.15 mg. The variability in dosing across the devices may be explained in part by the low internal diameters of the tracheostomy equipment and the leaks around the masks. As aerosol impacts within that equipment, the greater the likelihood of subsequent impactional losses. The passive nature of the VMN aerosol is associated with less likelihood of occlusion of the tracheostomy, a critical risk factor in this patient population. Further, jet nebulisers continue to supply aerosol at velocity to the vicinity of the tracheostomy interface during inhalation and exhalation. As such, larger fractions of JN aerosol are lost than VMN aerosol with open tracheostomy mask interfaces. The pMDI delivers the largest percentage fraction for the same reason as the passive VMN aerosol, however, as noted, the drug mass delivered is less.

High Flow Nasal Therapy (HFNT) involves the use of nasal cannula connected to a gas supply and source of humidification. Becoming increasingly popular as an intervention used in the stabilisation of patients prior to further treatment. Can be administered either in the home or clinical setting. Aerosol is typically delivered by VMN due to sensitivity of HFNT to additional gas flows introduced by JN. Anecdotal reports of pMDI use suggest that it is sometimes used to administer drug to the nasal cavity and upper airways. Recent reports describing the use of both VMN and JN in models of newborn infants and 9 month paediatric patients indicate that aerosol deposition can be achieved, albeit at low levels. At 8 LPM of total gas flow, both JN and VMN achieved ~0.1% lung dose in the newborn model, whilst the VMN achieved ~0.9% at 2 LPM. It was not [7] possible to test the JN at 2 LPM . These compare well to previous reports. Further, the higher dose of ~0.9% was considered likely to be associated with a clinical bronchodilatory effect. In adults, one scintigraphy study describes higher delivery rates for VMN across a range of gas flow rates, with [7, 8] . lung doses ranging from ~2.2 to 11% Beyond the patient’s breathing regimen, the relatively low levels of aerosol being delivered can be explained by losses throughout the HFNT system itself, with varying amounts of drug being deposited in the humidifier and tubing and also the fact that HFNT is continuously delivering gas, and aerosol, throughout the respiratory cycle.

353


Drug Delivery to the Lungs 27, 2016 – Ronan Mac Loughlin

Patients requiring ventilatory support Endotracheal tubes (ETT) are inserted in the patient trachea via oropharynx in order to facilitate mechanical ventilation of the unconscious patient. Internal diameters range from 2.5 mm for premature infants to 12.0 mm for adults. ETTs are used in both intensive care and operating room. A small number of scintigraphic studies have been published describing the performance of JN and VMN during invasive mechanical ventilation. In a model of a newborn infant, JN and VMN achieved ~0.5 and 12.6% lung [8] deposition respectively . In adults, VMN deposition rates are seen to vary between ~10 and 15% depending on [9] ventilatory mode, another consideration which is beyond the scope of this review . Differences in delivered dose during ETT-mediated ventilation have been shown to be predominately associated [10-12] . Furthermore, the ETT itself with the positional placement of the aerosol generator in the ventilator circuit [13] has been shown to effect droplet size and so deposited dose .

Tracheostomy tubes (TT) as well as being used in spontaneously breathing patients as described previously, are used as interfaces for mechanical ventilation. Bench studies of JN and VMN performance in this use scenario indicates that both choice of nebuliser and placement position in the circuit can have a significant effect on the [14] delivered dose, with ranges of 15-23% and 38-50% for JN and VMN recorded depending on placement .

Non-invasive ventilation (NIV) is increasingly being used as a means of ventilating a critical patient without the need for the complicated, risky procedure of endotracheal intubation nor expensive capital equipment. As with invasive ventilation, JN, VMN and MDI are often used for aerosol administration. The interface used is a tightly fitting mask that has a controlled leak included in the plastics. Performance of these devices again was again seen to vary during use on a NIV intervention with a single [15] scintigraphic study suggesting that lung doses varied between ~1.5 and 5.5% for JN and VMN respectively . This trend is further substantiated by several bench studies that record JN performance ranging from ~6.8-7.2%, [16,17] . in comparison with a VMN performance range of ~14.6-18.7% [18]

Aerosol losses are explained in these studies, and others , as being due to losses in the circuit, and humidifier but also importantly due to the position of the nebuliser relative to the leak port.

Future opportunities for mitigating the differences in delivered dose Whilst choice of device, patient type and disease setting have the largest effect on the delivered dose from an aerosol generator, it is clear that patient intervention/interface also have a significant effect. In an effort to normalise the delivered dose, it may be necessary to develop more sophisticated aerosol generator technologies that can help mitigate drug losses through means such as low velocity aerosols, smaller droplet/particle size, breath-initiated aerosol generation, small footprint devices that allow placement close to the patient etc. Another opportunity however is the development of patient interfaces that are more aerosol-friendly. Several innovations in aerosol generators and interfaces including aerosol generators are currently described in the patent [19records, with new devices being released by manufacturers that have taken consideration of aerosol therapy 21] . This is a relatively new and welcome occurrence given that traditionally aerosol generator designers did not also design interfaces, and visa versa.

Concluding remarks It is evident from this review that different interventions and consequent interface choice have a significant effect on the aerosol dose being delivered to the patient’s lung. Appropriate selection of the interface/aerosol generator combination is critical in ensuring safe and efficient patient care, but also aids in de-risking drug development and clinical research programs in ensuring that the researcher can deliver appropriate levels of drug to the lung as cost effectively as possible. This data should be used to inform both end users and researchers alike of the importance of the choice of an appropriate aerosol generator for use with any given patient interface in use during any of the clinical interventions.

354


Drug Delivery to the Lungs 27, 2016 -Can Aerosol Therapy Keep Pace with Innovations in Patient Care? A Review. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

Hickin S, Mac Loughlin R, Sweeney L, Tatham A, and Gidwani S: Comparison of mesh nebuliser versus jet nebuliser in simulated adults with chronic obstructive pulmonary disease. Poster at the College of Emergency Medicine Clinical Excellence Conference. 2014. Newman S, Steed K, Reader S, Hooper G, and Zierenberg B: Efficient delivery to the lungs of flunisolide aerosol from a new portable hand�held multidose nebulizer. Journal of pharmaceutical sciences. 1996;85:960-964. Alcoforado L, de Melo Barcelar J, Castor Galindo V, Cristina S. Brandão S, Fink J, B. , and Dornelas de Andrade A: Analysis of Deposition Radioaerosol Nebulizers Membrane in Healthy Subjects. ISAM poster presentation 2015. Dugernier J, Reychler G, Depoortere V, Roeseler J, Michotte JB, Laterre PF, Jamar F, and Hesse M: Tomoscintigraphiccomparison of lung deposition with a vibrating-mesh and a jet nebulizer. ERS conference poster 2015. Alhamad BR, Fink JB, Harwood RJ, Sheard MM, and Ari A: Effect of Aerosol Devices and Administration Techniques on Drug Delivery in a Simulated Spontaneously Breathing Pediatric Tracheostomy Model. Respiratory care. 2015. Ari A, Harwood R, Sheard M, Alquaimi MM, Alhamad B, and Fink JB: Quantifying Aerosol Delivery in Simulated Spontaneously Breathing Patients With Tracheostomy Using Different Humidification Systems With or Without Exhaled Humidity. Respiratory care. 2016. Reminiac F, Vecellio L, Loughlin RM, Le Pennec D, Cabrera M, Vourc'h NH, Fink JB, and Ehrmann S: Nasal high flow nebulization in infants and toddlers: An in vitro and in vivo scintigraphic study. Pediatric pulmonology. 2016. Dubus JC, Vecellio L, De Monte M, Fink JB, Grimbert D, Montharu J, Valat C, Behan N, and Diot P: Aerosol deposition in neonatal ventilation. Pediatric research. 2005;58:10-14. Dugernier J, Reychler G, Wittebole X, Roeseler J, Depoortere V, Sottiaux T, Michotte J-B, Vanbever R, Dugernier T, and Goffette P: Aerosol delivery with two ventilation modes during mechanical ventilation: a randomized study. Annals of Intensive Care. 2016;6:1. Ari A, Areabi H, and Fink JB: Evaluation of aerosol generator devices at 3 locations in humidified and non-humidified circuits during adult mechanical ventilation. Respiratory care. 2010;55:837-844. Ari A and Fink JB: Aerosol Drug Delivery During Mechanical Ventilation: Devices, Selection, Delivery Technique, and Evaluation of Clinical Response to Therapy. Clinical Pulmonary Medicine. 2015;22:7986. Berlinski A and Willis JR: Albuterol Delivery by 4 Different Nebulizers Placed in 4 Different Positions in a Pediatric Ventilator In-Vitro Model. Respiratory care. 2012. Berlinski A and Kumaran S: Particle Size Characterization of Nebulized Albuterol Delivered by a Vibrating Mesh Nebulizer Through Pediatric Endotracheal Tubes. Poster Presentation at ATS. 2016. Kelly P, O'Sullivan A, McKenna C, Sweeney L, and MacLoughlin R. Effect of Nebulizer Type and Position on Aerosol Drug Delivery During Support Mechanical Ventilation and Spontaneously Breathing for Tracheostomized Adult Patients. Poster presented at DDL27: Edinburgh, 2016. Galindo-Filho VC, Ramos ME, Rattes CS, Barbosa AK, Brandao DC, Brandao SC, Fink JB, and de Andrade AD: Radioaerosol Pulmonary Deposition Using Mesh and Jet Nebulizers During Noninvasive Ventilation in Healthy Subjects. Respiratory care. 2015;60:1238-1246. Abdelrahim ME, Plant P, and Chrystyn H: In-vitro characterisation of the nebulised dose during noninvasive ventilation. The Journal of pharmacy and pharmacology. 2010;62:966-972. McPeck M: Improved Aerosol Drug Delivery with an Electronic Mesh Nebulizer during Non-invasive Ventilation AARC poster. 2012. White CC, Crotwell DN, Shen S, Salyer J, Yung D, Zheng J, and DiBlasi RM: Bronchodilator delivery during simulated pediatric noninvasive ventilation. Respiratory care. 2013;58:1459-1466. Cortez FV, Niland WF, Boyd P, McGarrity G, and Buyer C. Heated nebulizer devices, nebulizer systems, and methods for inhalation therapy. Google Patents, 2013. Hogan B and XU H. METHOD FOR PRODUCING AN APERTURE PLATE. US Patent 20,150,336,115, 2015. Paykel F. https://www.fphcare.co.nz/products/airvo/. 2016. Accessed 22 July, 2016, 2016.

355


Drug Delivery to the Lungs 27, 2016 – Patrick M. Kelly et al. Effect of Nebulizer Type and Position on Aerosol Drug Delivery during Support Mechanical Ventilation and Spontaneously Breathing for Tracheostomized Adult Patients Patrick M. Kelly, Andrew O’Sullivan, Cathy McKenna, Louise Sweeney & Ronan MacLoughlin Aerogen, Galway Business Park, Dangan, Galway, Ireland Summary In tracheostomized patients the effect of nebulizer type/platform has not been studied in great detail. This study investigated tracheal dose delivery for a vibrating mesh nebulizer (VMN) and jet nebulizer (JN) in support (single limb) mechanical ventilation (SMV) and spontaneously breathing (SB) tracheostomized patients. Selection of nebulizer type can have a substantial influence when treating a patient, earlier studies have shown that the VMN [1] provides more aerosol than the conventional JN . The nebulizer was placed at two positions during the SMV, 1) Dry side of the humidifier pot and 2) at tracheostomy tube (no spacers were used between the nebulizer t-piece and tracheostomy tube) and at the tracheostomy tube for SB. Results: the VMN delivered significantly greater % aerosol delivery when compared to the JN across all positions evaluated. The VMN proximal to tracheostomy tube during SMV facilitated the highest tracheal dose (50.78 ± 2.65%) compared with JN (15.89 ± 1.41%) at this position. Conclusion: the VMN delivered a significantly larger fraction of aerosol (P-values <0.05) when compared to the JN for the tracheal dose for all positions evaluated for SMV and SB in tracheostomy patients. Introduction The administration of nebulized therapeutic agents for SMV and SB tracheostomized patients are common in reducing pulmonary complications. Nebulizer type has been shown in previous studies to have large effects on [2] the efficiency for the nebulized drug to be deposited in the lung, during mechanically invasive ventilation and [3] spontaneously breathing patients . However, in tracheostomized patients the effect of the type of nebulizer platform has not been studied in great detail. As part of this study, we investigated aerosol delivery from a VMN and JN in a humidified adult SMV circuit with a breathing simulator generating the adult SB. The nebulizer was placed at two positions during the SMV, 1) Dry side of the humidifier pot and 2) at tracheostomy tube (no spacers were used between the nebulizer t-piece and tracheostomy tube) and at the tracheostomy tube for SB. The objective of this study was to establish which nebulizer facilitates the highest drug delivery to the lung for a tracheostomized adult patient. Materials and methods Aerosol delivery performance was evaluated by characterising the Tracheal Dose (%) (drug delivered beyond the trachea). A 2.0 mL dose of Albuterol sulphate (1 mg/mL) was nebulised as a tracer aerosol using a 1) vibrating mesh nebulizer (VMN) (Aerogen Solo, Aerogen, Ireland), with an average volume mean diameter (VMD) of 4.73 µm and aerosol flow rate of 0.38 mL/min (measured using the Malvern Spraytec), 2) JN (Cirrus 2, Intersurgical, United Kingdom) with a driving gas flow rate of 8 LPM, with an average volume mean diameter (VMD) of 4.14 µm and aerosol flow rate of 0.27 mL/min (measured using the Malvern Spraytec). At the end of each dose the drug was extracted and quantified using UV spectrophotometry (at 276 nm). The mass of drug eluted from the filters was determined using spectrophotometry and interpolation on a standard curve of Albuterol sulphate concentrations (200 µg/mL down to 3.125 µg/mL). Results were expressed as the percentage of the nominal dose placed in the nebulizer’s medication cup that was delivered Support (single limb) mechanical ventilation (Figure 1, A-E) In all the experiments the model included a ventilator (Airox Supportair, Covidien, Ireland), a heated humidifier (Fisher & Paykel, Auckland, New Zealand), a heated-wire ventilator smoothbore circuit (Intersurgical, UK) and a tracheostomy tube (Shiley, Covidien, Ireland). The following mechanical ventilator breathing pattern was used (BPM 15, Vt 500 mL, I:E 1:2). The end of the tracheostomy tube cannula was inserted into a 3D printed fixture, a 22 male adaptor with a bore where the cannula was glued into position, which was then inserted into the housing of an absolute bacterial/viral filter (Respirgard II 303, Baxter, Ireland). The pass-over humidifier and ventilator circuit were heated during ventilation for approximately 20 – 30 min, until the temperature at the airway was stable at 35 ± 1°C. The tracheostomy tube was placed in an upright position. Spontaneously breathing (Figure 1, F-H) The nebulizer was connected to a breathing simulator (ASL 5000, Ingmar Medical, PA, USA) via an absolute filter. A simulated adult breath was used (BPM 15, Vt 500 mL, I:E 1:2). For the VMN, the Heat and Moisture exchanger (Hydro-Trach II, Intersurgical, UK) was connected to the 15 mm male end of the Aerogen T-piece (Paediatric T-piece, Aerogen, Ireland). The tracheostomy tube was placed in an upright position.

356


Drug Delivery to the Lungs 27, 2016 - Effect of Nebulizer Type and Position on Aerosol Drug Delivery during Support Mechanical Ventilation and Spontaneously Breathing

Figure 1 - Test Setup for SMV (A-E) and SB (F-H) with VMN and JN; A) Generic test setup for SMV, B) VMN place at the dry side of the humidifier, C) JN place at the dry side of the humidifier, D) VMN place at tracheostomy tube, E) JN place at tracheostomy tube; F) JN connected to ASL for SB, G) VMN connected to ASL for SB, H) VMN with HME connected to ASL for SB

Results and discussion Based on the results presented in Figure 2, the VMN was seen to have a significantly greater % aerosol delivery when compared to the JN across all positions evaluated. For the SMV setup, the largest aerosol delivery in this study was observed for the VMN at the tracheostomy tube at 50.78 ± 2.65% (D), this compared to the JN at 15.88 ± 1.41% (E). The % aerosol delivery for the dry side of the humidifier was substantially higher for the VMN at 28.28 ± 3.13% (B) when compared to the JN at 23.57 ± 1.36% (C). For the SB setup, the VMN with HME facilitated the highest drug delivery at 39.70 ± 0.84% (H), however a direct comparison for the JN with a HME was not completed as part of this study, as it was not possible to directly connect the HME to the T-piece of the JN without a series of connectors/adaptors. The only direct comparison completed as part of the tracheostomy SB study showed the VMN was also superior to the JN at 33.87 ± 2.609% (G) and 19.03 ± 1.03% (F). The lower delivered dose associated with JN was due to larger losses to the circuit as a result of the high gas flow and the residual drug remaining in the reservoir upon ceasing aerosol generation (Note: was not quantified as part of this [4],[5] ). study, but has been reported up to 50% remaining

357


Drug Delivery to the Lungs 27, 2016 – Patrick M. Kelly et al.

Aerosol delivery for Adult Tracheostomy Patients 60

P=0.006

P=0.007

AEROSOL DELIVERY (%)

50

P=0.023

40 30

Solo II Jet Nebulizer

20 10 0 Single Limb @ tracheostomy tube

Single Limb @ Dryside of Humidifier

ASL - Adult Spontaneous Breathing

ASL with HME Adult Spontaneous Breathing

Figure 2 - Aerosol delivery to for Adult Tracheostomy Patients for SMV and SB

Conclusion Results demonstrate that the VMN delivered significantly larger fractions of aerosol (P-values <0.05) for the tracheal dose for all positions evaluated for SMV and SB in tracheostomy patients when compared to the JN. The delivered aerosol dose for the VMN and JN was seen to range from 33.87-50.784 % and 15.88-23.57 % respectively, depending on position. The positioning of both nebulizers on the spontaneous breathing patients may require additional review as the position may not be clinically applicable due to the close proximity to the patient. Further evaluation should be undertaken to understand the influence of spacers placed between the nebulizers and tracheostomy tube on aerosol delivery. References 1. Ari A, Atalay OT, Harwood R, Sheard M, Aljamhan EA, Fink JB, Influence of Nebulizer Type, Position, and Bias Flow on Aerosol Drug Delivery in Simulated Pediatric and Adult Lung Models During Mechanical Ventilation, Respir Care 2010; 55 (7): pp845– 851. 2. Berlinski A, Willis JR, Albuterol Delivery by 4 Different Nebulizers Placed in 4 Different Positions in a Pediatric Ventilator In Vitro Model, Respir Care, 2013; 58 (7): pp1124–1133. 3. Ari A, Dornelas de Andrade A, Sheard M, AlHamad B, Fink JB: Performance Comparisons of Jet and Mesh Nebulizers Using Different Interfaces in Simulated Spontaneously Breathing Adults and Children, Journal of Aerosol Medicine and Pulmonary Drug Delivery, 2015; 28 (4): pp281–289 4. Fink JB. Humidity and aerosol therapy, Mosby’s Respiratory Care Equipment. St Louis: Mosby-Elsevier; 2010:91-140. 27. 5. Fink JB. Aerosol drug therapy, Egan’s Fundamentals of Respiratory Care, 9th edition. St Louis: Mosby Elsevier; 2009:801-842

358


Drug Delivery to the Lungs 27, 2016 – R F Oliveira et al. Numerical Simulation of Salbutamol Deposition in VHC Devices 1

2

3

R F Oliveira , S F C F Teixeira , H M Cabral-Marques & J C F Teixeira

1

1

MEtRiCS R&D Centre, University of Minho, 4800-058 Guimarães, Portugal ALGORITMI R&D Centre, University of Minho, 4800-058 Guimarães, Portugal 3 iMed.ULisboa R&D Centre, Universidade de Lisboa, 1649-003 Lisboa, Portugal 2

Summary Valved holding chamber (VHC) devices are an essential piece of technology for pressurized Metered-Dose Inhaler (pMDI) users who cannot properly coordinate their inhalation with the actuation of the spray. The unique design characteristics of the VHC devices, particularly the one-way inhalation valve, make this device performance evaluation complex to execute. Another VHC role is to reduce the throat deposition, caused by the pMDI high velocity spray. When it is fired it into the device, the coarser particle fraction of the emitted plume is reduced by directly impacting onto the walls. A computational model was developed for a better understanding of the geometric features that affect the reduction of the spray plume by the VHC. This model predicted the transport, evaporation of propellant and wall deposition of the pMDI spray inside the VHC device. A constant flowrate of 60 L/min, passing through the device, was assumed. This model was applied for six VHCs: Aerochamber, A2A Spacer, Compact SpaceChamber, SpaceChamber, Nebuchamber and Volumatic. The results show that the use of a VHC against a solitary pMDI, reduced the throat deposition by 57.9% (Compact SpaceChamber) to 79.7% (A2A Spacer). It was verified that droplets with diameter ≥ 17 µm (mainly composed by propellant) deposit in the further portion of the VHC body by direct impaction. Particles between 3 µm and 17 µm tend to sediment by gravity force. Drug particle mass deposited in the valve of the Volumatic (8.0 µg – Coin valve design) was highest and lowest for the Nebuchamber (1.8 µg – Duck valve design). The deposition in the VHC body was highest for the AeroChamber (31.9 µg) and lowest for the Nebuchamber (14.5 µg). VHC devices design could be improved by modifying the dimensions of the body to target a specific coarse particle diameter range with the help of numerical tools. Introduction Inhalation therapy success depends, in part, on the devices used in the delivery of the Active Pharmaceutical Ingredient (API) particles into the patient’s lungs. Additionally, a correctly executed inhalation manoeuvre with the delivery device is essential to an effective treatment. In the specific case of the pressurized metered dose inhaler (pMDI), the elderly and younger patients usually cannot correctly coordinate the inhalation with the actuation of [1] the device, resulting in substantial drug loss by impaction onto the patient’s throat . The spacers, particularly the valved holding chamber (VHC), are add-on devices developed to overcome this difficulty by allowing the patient to [2] freely tidal breathe the drug treatment. More so, they reduce the throat deposition in more than 80% , which minimizes the risk of health hazard by the systemic absorption of corticosteroids. The VHC device has a unique operating mechanism triggered by the one-way inhalation valve. This feature, present in all commercial devices, show different designs that can be grouped in (see Figure 1): Coin, Duck, Leaflets and Annulus Flap. Each design will differently influence the throat deposition and emitted plume by the VHC, the Coin design (in the Volumatic) is a higher obstacle to the coarse particle transport than a Leaflets valve design (the Compact SpaceChamber). The aim of this work is to better understand the deposition caused by different valve designs and VHC geometries available in the market, by applying a Computational Fluid Dynamics (CFD) model to predict the drug transport and surface deposition. Methodology Geometry Three-dimensional representations of commercial pMDI VHC devices were used in the numerical simulations, to predict salbutamol pMDI HFA-134a spray deposition in the device surfaces, throat and the delivered dose. The ® six commercial devices geometrically represented in this study were (see Figure 1): the Space Chamber Plus ® ® SC (from Medical Developments International ); Compact Space Chamber Plus - CSC (from Medical ® ® ® Developments International ); the Volumatic - VOL (from GlaxoSmithKline ); the AeroChamber Plus - AC (from ® ® Trudell Medical International ); the A2A Spacer – A2A (from Clement Clarke International ) and the ® Nebuchamber – NC (currently off-market from AstraZeneca ). Their volumes are: SC – 230 mL; CSC - 160 mL; VOL – 750 mL; AC - 149 mL; A2A – 210 mL and NC – 250 mL. Since the simulations were performed assuming a constant flowrate of 60 L/min, the VHC valves were drawn open. For this task, dimensional information of the aperture was taken from photography of the valve operation at that flowrate. Some degree of uncertainty may arise from this technique, specifically for VOL and AC valves. All VHC geometries are connected to a USP Induction Port (IP) 90º bend, designed according to standardized dimensions. Additionally, the pMDI actuator alone into the IP was simulated. All geometries presented a longitudinal symmetry; hence only half domain was discretized to reduce the computational effort.

359


Drug Delivery to the Lungs 27, 2016 - Numerical Simulation of Salbutamol Deposition in VHC Devices

Figure 1 - Representation of the VHC devices used in this study.

Grid ®

The discretization of the geometries was carried out using the Meshing software from ANSYS , by applying an automated method with refinement upon proximity and curvature geometric characteristics. This resulted in unstructured grids with 10-15 layers near wall composed by prismatic elements, to keep the y+ of the first layer close to unity. With first elements layer height between 0.02 – 0.05 mm, the domain maximum element was size kept between 3.5 – 10 mm. The resulting grids have a number of elements ranging from 89k (for the pMDI solo) [3] was performed for the CSC geometry, using three grids to 883k (for the VOL). A grid uncertainty analysis (257k, 582k and 1.3M), which showed a difference of 3.14% in the FPM emitted between the medium and the fine grids. This means that the medium size grid (with the characteristics described above) is adequate to perform the simulations. Grid analysis was performed for the Element Quality ( ≈ 0.5), Orthogonal Quality ( ≈ 0.8), Aspect Ratio ( ≈ 30) and Skewness ( ≈ 0.3) metrics; showing similar values for all the grids, the values are within a good quality range. Solver Configuration ®

The simulations were performed in FLUENT v15 from ANSYS , a commercial finite volume CFD code. The drug droplet/particle transport in the fluid was modelled as a discrete phase, considering an Euler-Lagrange approach that accounts for: drag, turbulent dispersion, propellant evaporation, gravity sedimentation, surface collision and heat exchange. The fluid was considered a mixture of H2O, O2, HFA-134a and N2, at 20ºC that flows at 60 L/min. The turbulence was modelled using a k-ω SST model allowing the calculation of large stagnation areas near the VHC walls. The Ventolin HFA-134a salbutamol spray has a duration of 0.1 s, with a mass of 100 µg per actuation, an initial velocity of 120 m/s. Its cone angle is 40º, the nozzle diameter is 0.50 mm and its injection temperature is -58ºC. The droplets diameter distribution was measured experimentally using laser diffraction technique (particle [4] sizer model 2600 from Malvern, UK) . It was observed that this distribution is poly-dispersed, hence the data were not modelled using a statistical function (e.g. log-normal or Rosin-Rammler), but directly inputted into the numerical model. The droplets were considered to be a mixture of liquid propellant and solid salbutamol particles (m/m = 0.95%). Particle-surface collision was modelled using a two phase criteria: first entrapment of any droplet in which the propellant did not fully evaporated. Secondly, if the particle had no propellant a critical velocity criteria [5,6] . This model includes the surfaces material characteristics in model was applied, as described elsewhere [7] collision. The particles drag coefficient was modelled accordingly to the law proposed by Morsi and Alexander [8] with a slip Cunningham correction factor for particles with diameters bellow 1 µm . Results Figure 2 presents the salbutamol mass deposited in the VHC body, VHC valve surfaces, USP IP walls and the Sizable Mass (SM) emitted to the lungs. Results are given for an actuation of 100 µg.

360


Drug Delivery to the Lungs 27, 2016 – R F Oliveira et al.

Figure 2 - Salbutamol mass deposited in the domain surfaces and the emitted mass from the geometry.

Figure 3 provides the location of each particle deposition within the six VHC devices. The results are coloured by the particle/droplet diameter and the valve surfaces are coloured in dark grey.

Figure 3 -

Drug particles deposition locations in the domain surfaces, coloured by diameter, for the six VHC geometries under study.

Discussion Figure 2 predicts that half (54.6 Âľg) of the spray salbutamol mass is lost by direct impaction in the USP IP walls. The majority of the droplets had no time to fully evaporate and upon colliding with the throat surfaces (the model applied predicted entrapment). The addition of a VHC, provided a reduction of the throat deposition between 57.9% (for the Compact SpaceChamber) and 79.7% (for the A2A Spacer).

361


Drug Delivery to the Lungs 27, 2016 - Numerical Simulation of Salbutamol Deposition in VHC Devices Drug deposition in the valve surfaces is higher for the Volumatic (8.0 µg - Coin design), followed by the A2A Spacer (3.2 µg – Leaflets design), AeroChamber (3.1 µg – Annulus Flap design), Compact SpaceChamber (3.0 µg - Leaflets), SpaceChamber (2.3 µg - Leaflets) and Nebuchamber (1.8 µg - Duck). The valves designs with obstacles to the air jet transporting particles result in higher drug deposition, such as the case of Volumatic and A2A post-valve cross obstacle (that creates vortices leading to greater drug loss). The drug deposition in the VHC body surfaces ranges as AeroChamber (31.9 µg) > Compact SpaceChamber (31.6 µg) > SpaceChamber (30.4 µg) > Volumatic (22.6 µg) > A2A Spacer (22.6 µg) > Nebuchamber (14.5 µg). The data suggest that the main deposition mechanism in the VHC is the impaction due to the spray high velocity and therefore the deposition should be strongly related with the VHC body length and diameter. However, it was expected that the Volumatic had the lowest body deposition, since it has the largest dimensions. This was not verified, and might be related to: simplified mathematical modelling of the collision particle-wall collision (no electrostatic attraction was considered and full adhesion for liquid droplets), inaccurate spray inputs or simplifications in the solver spray modelling. The emitted sizable mass, induction port and valve deposition predictions are deeply related to the limitations of the VHC body drug deposition predictions, hence they inherit its influence. Figure 3 provides information on the spray droplets/particles preferable deposition locations for each diameter. It can be observed that the diameters ≥ 17 µm (mainly composed by propellant) are bound to deposit downstream (> 50% of its length) of the VHC body. Particles with 3 µm to 17 µm suffer mainly sedimentation by gravity in the bottom portion of the VHC body. The IP throat (not shown in Figure 3) presented a deposition of particles < 6 µm, being strongly influenced by turbulence generated by the valve air jet, which is dependent of the valve crosssection area and shape. Aerochamber valve design (Annulus Flap) shown a radially uniform distributed particle deposition in the upper throat wall, the same was verified for both SpaceChamber devices. Conclusions These observations suggest that the valve design has a strong influence in the throat deposition and drug targeting. The valve cross-section will dictate the throat deposition pattern. The regular VHC valve operation is at tidal breathing (and at lower flowrates) where it becomes mostly important to evaluate the valve leakage and inertia to opening/closing upon respiratory solicitation. The VHC body design could be improved to target the spray deposition, by impaction, within a desired particle diameter range. This can be made by changing the VHC length and/or diameter to meet the commercial pMDI products spray characteristics range. The numerical model gives a good insight of the drug particles transport inside of a VHC device, however the particle-wall collision model needs to be improved for better accuracy. Additionally, simulations at variable flowrate should be performed in the future. References 1.

Newman SP, Newhouse MT: Effect of add-on devices for aerosol drug delivery: deposition studies and clinical aspects. J Aerosol Med. 1996;9(1):55-70.

2.

Oliveira RF, Cabral Marques HM, Machado A V, Teixeira JC, Teixeira SF: VHC Performance Evaluation at Connstant Flow: 30 L/Min. In: Volume 3: Biomedical and Biotechnology Engineering. Houston, Texas, USA: ASME; 2015:V003T03A098. doi:10.1115/IMECE2015-52283.

3.

Celik IB, Ghia U, Roache PJ, Freitas CJ, Coleman H, Raad PE: Procedure for Estimation and Reporting of Uncertainty Due to Discretization in CFD Applications. J Fluids Eng. 2008;130(7):78001.

4.

Oliveira RF, Teixeira SFCF, Teixeira JC, Silva LF, Antunes H: pMDI Sprays: Theory , Experiment and Numerical Simulation. In: Liu C, ed. Advances in Modeling of Fluid Dynamics. Rijeka, Croatia: Intech; 2012:300. doi:10.5772/46099.

5.

Brach RM, Dunn PF: A Mathematical Model of the Impact and Adhesion of Microsphers. Aerosol Sci Technol. 1992;16(1):51-64.

6.

Milenkovic J, Alexopoulos AH, Kiparissides C: Flow and particle deposition in the Turbuhaler: A CFD simulation. Int J Pharm. 2013;448(1):205-213.

7.

Morsi SA, Alexander AJ: An investigation of particle trajectories in two-phase flow systems. J. Fluid Mech. 1972;55(2):193.

8.

Cunningham E: On the Velocity of Steady Fall of Spherical Particles through Fluid Medium. Proc. R. Soc. A Math. Phys. Eng. Sci. 1910;83(563):357-365.

362


Drug Delivery to the Lungs 27, 2016 - Daniel P. Jenkins et al. A Fresh Look at Designing Respiratory Health Devices 1

1

Daniel P. Jenkins , Chris Langley & Paul Draper 1

1

DCA Design International, 19 Church Street, Warwick, CV34 4AB, UK

Summary Many pharmaceutical companies see equivalence claims as the fastest and most cost effective route to market for new respiratory drug devices. In reality though, the process is often not as simple as first thought. By their very nature, equivalence claims mean adopting existing, often old, technologies and user interfaces. Furthermore, the task of demonstrating equivalence can be protracted and expensive. Thus, in many cases, manufacturers are left accepting sub-optimal devices, with known usability issues. Moreover, opportunities may be missed to explore different products, technologies and services that would improve usability, compliance, patient satisfaction and market share. This paper uses tools from the discipline of human factors to explore what a fresh approach to requirements capture and design might mean for devices designed from first principles. A structured approach is described for designing medical devices based on patient and stakeholder information requirements. The approach is based on the premise that better designs are informed by an explicit understanding of what information is required, where and when, along with an understanding of who needs it and how it should be presented. To develop more usable and engaging products these information requirements are explored at the earliest stages of the design process and revisited throughout the design. Introduction Ostensibly, decision making is at the heart of the safe and effective management of respiratory health. By exploring key decision points, it is possible to reveal rich insights into patient life. Opportunities can be identified for better ways of communicating with the patient and key stakeholders. Decisions relating the control of patients’ environments and the use of their medication will influence both their symptoms and their quality of life. Experimental methods There have been many attempts to model the decision making activity. Most approaches involve some form of observation of information, orientation to the current situation, a choice as to which action to adopt, and finally an [1] action. The decision ladder (see Figure 1) is a representation of this decision cycle commonly used within an [2] analysis approach used in human factors called Cognitive Work Analysis . GOALS

Evaluate performance

CHOSEN GOAL

OPTIONS

Predict consequences

SYSTEM STATE

TARGET STATE

Diagnose state

Definition of task

INFORMATION

TASK

Observe information and data, scanning for cues

Planning of procedure

ALERT

PROCEDURE

Activation

Execute

Figure 1 - the decision ladder model

363


Drug Delivery to the Lungs 27, 2016 - Daniel P. Jenkins et al.

Table 1 –

Process for eliciting information requirements for respiratory health (stage numbers relate to numbered circles in Figure 2)

Stage 1 – Determining the goal The process starts near the top of Figure 2. The goal at this stage of the process is simply ‘To safely and effectively manage respiratory health in order to maximise the patient’s quality of life.’ Stage 2 – Alert Moving to the base of the left leg of Figure 2, possible alerts that may indicate a change to the system state include physiological states such as shortness of breath, wheezing and coughing. Likewise, other alerts may be linked to situational aspects such as the time of day or a meal, or the relationship to exercise. Finally, other alerts might also be liked to environmental factors like the current weather. Stage 3 – Information The information elements are the ‘nuggets’ of information that can be brought together to understand the state of the system. In this case, they include information about the physical environment (e.g. pollen count, air quality), the individual patient (known irritants), the given situation (e.g. time of day, location), drug usage (e.g. when taken, how much) and breathing performance (e.g. FEV1, peak expiratory flow rate). Stage 4 – System state The system states represent a perceived understanding of the system based upon the interpretation of a number of different information elements. Questions such as ‘Is the pollen count to a level that is likely to impact breathing?’ can be assessed by considering a number of different information elements together. In this case, the current pollen level and the patient’s sensitivity to it. Stage 5 – Options At a high level, in this part of the decision making process there are four main options available to the operator: Is it possible to take drug as prescribed? Is it possible to increase dose? Is it possible to decrease dose? And Is it possible to seek medical assistance? Stage 6 – Chosen goal The chosen goal, at any one time, is determined by selecting which of the constraints receives the highest priority. For example is the priority at the given time, effectiveness, safety or quality of life. Stage 7 – Target state The target states mirror the option available (see Stage 5). Stage 8 – Task The tasks relate to the specific actions required to achieve the chosen goal. They are outside the scope of this paper. While each of the aspects of Figure 2 informs the design process, the left leg of the model is expected to be the most useful in understanding the information requirements for a medical device. There is an interlinked relationship between the alerts, information elements and system states. Information elements are expected to direct the identification of system states; likewise, as new system states are identified in the model, this often leads to the inclusion of additional information elements which, in turn, might identify new system states. The model is particularly useful for identifying information elements that users rely on but they often fail to articulate, such as changes in the weather. One useful approach for building the model and identify additional alerts, information elements and system states, is to question what the interviewee would you do if each of the identified element were missing. As with all theoretical models, there is challenge in identifying a suitable stopping rule – that is how many information elements should be included in the model, particularly when their relevance can be questioned. Ultimately, this is down to the analyst to determine. Once the model is considered to be complete, the list of alerts, information elements, system states and options provides a comprehensive description of ‘what’ information may be required to support decision making. This list is deliberately independent of context and thus represents the totality of what might be useful to know. The next stage of the process is to code this list to indicate when the information may be needed: (e.g. first thing in the morning, prior to drug taking, during drug taking, after drug taking); where it could be displayed (e.g. on the device, on an IFU, on packaging, on a companion app, via a healthcare professional, on a website); to whom (e.g. the patient, a carer, a healthcare professional, a pharmacist, a prescriber) and finally; ‘how’ (e.g. text, photos, graphics, video, audio files). Logistically, this is best done by converting the model to a tabular format, adding columns for each possibility and coding the matrix accordingly. The generated table can then be interrogated to establish if the constraints on the system are appropriate. Cells that are not coded may represent particular opportunities for product improvement.

365


Drug Delivery to the Lungs 27, 2016 - A Fresh Look at Designing Respiratory Health Devices Discussion and conclusions For complex systems, a structured approach is needed to ensure firstly, that all the required information elements are considered, and secondly that they are included in the optimal way to ensure an appropriate balance of system values (e.g. safety, efficacy, efficiency, usability and resilience). The approach described in this paper provides welcome structure to the process of eliciting and exploring information requirements that focus on end users and stakeholders decision making needs. One of the clear strengths of the approach is that it provides a very explicit link between the data collection, the analysis, and the resultant design that can be revisited and audited throughout the design process. Informed decision making is fundamental to safe management of respiratory conditions. There is a long established connection between the information available to decision makers and quality of decision making. This does not necessarily mean presenting more information; on the contrary, too much information can be as detrimental to performance as too little. Rather, to optimise system performance (e.g. safety, efficacy, efficiency, resilience), effective decision making must be supported by the right information, at the right time, in the right place, to the right actors, in a format that can be readily understood. The approach can be applied to existing designs as part of the equivalence process; however, it is contended that applying it at the earliest opportunity in the design process would allow better designs to be identified and explored. In the case of respiratory health, the approach highlights that the use of medication is impacted by a wide range of factors distributed across the environment, the individual, the given situation, the history of drug use and breathing performance. It is contended that better consideration and integration of this information could lead to better management of respiratory health. The approach does not directly leap to a solution (such as an app), rather it systematically questions the different option for, when, where, how and to whom the given information should be displayed. References 1

Rasmussen J: The human data processor as a system component: Bits and pieces of a model (Report No. Risø -M-1722). Roskilde, Denmark: Danish Atomic Energy Commission 1974.

2

Jenkins D P, Stanton N A, Salmon P M, Walker G H: Cognitive Work Analysis: Coping With Complexity. Farnham: Ashgate. 2009

3

Jenkins D P, Boyd M, Langley C: Using the decision ladder to reach a better design. Presented at: The Chartered Institute of ergonomics and Human Factors annual conference, Daventry, UK, June 19-22 April, 2016.

4

Jenkins D P: Using Cognitive Work Analysis to describe the role of UAVs in Military Operations. Theoretical Issues in Ergonomics Science. 2012, 13(3)335-357

366


Drug Delivery to the Lungs 27, 2016 – Andrew O’Sullivan et al. Assessment of Exhaled Aerosol Emissions using Two Prevalent Nebuliser Technologies in Clinical Use Today Andrew O’Sullivan, James McGrath , Miriam Byrne, Patrick Power & Ronan Mac Loughlin, PhD Aerogen, IDA Business Park, Dangan, Galway, Ireland. Summary The design and operation features of two prevalent nebuliser technologies which are in clinical use today were assessed for their emission of exhaled aerosols. Vibrating mesh/jet nebuliser and facemask/mouthpiece are examples of aerosol generators and non-invasive ventilatory support. This study investigated the effect of these in terms of quantities of secondary aerosol mass concentrations emitted at varying distances (0.8 m & 2.2 m). The characterisation of exhaled aerosol emissions for two combinations of commercially available aerosol generators and non-invasive ventilation interfaces was investigated. Mass & number concentrations, and size distribution of the airborne particles were continuously measured in real time at different distances relative to a simulated patient. Results: The VMN/filtered mouthpiece was found to emit the lowest mass concentration over time -3 -3 (0.00441 mg m at 0.8 m & 0.00436 mg m at 2.2 m). The JN/open facemask emitted the highest mass -3 concentrations (mg m ) at both distances. For 0.8 m the JN/open facemask facilitated a near two fold higher -3 -3 emitted mass concentration (mg m ) compared with the VMN/valved facemask (0.048 vs 0.025 mg m ), near -3 fivefold higher compared with the JN/filtered mouthpiece (0.048 vs 0.00980 mg m ) & tenfold higher compared -3 with the VMN/filtered mouthpiece (0.048 vs 0.00441 mg m ). This study successfully demonstrated the obvious escape of aerosol to the environment and further established the risk to caregivers and other bystanders during the course of a standard nebuliser treatment. Introduction The use of handheld nebulisers is widespread and facilitates non-invasive drug delivery to patients ranging from infant to adult. Whilst most hand held nebulisers are lightweight and easy to use, there is often the need for [1] caregiver intervention , e.g. due to poor mobility or young age and the need for proper supervised drug dosing. Nebuliser-generated medical aerosols are delivered to the lungs via a variety of patient interfaces, including facemasks & mouthpieces. The combinations vary in terms of the facemask or mouthpiece of choice but also in how the aerosols are generated. Venturi Jet Nebulisers (JN) require a driving gas flow for aerosol generation, meaning that there is the potential for aerosol to be driven out of the facemask & mouthpiece during both [2] inhalation and exhalation . Vibrating mesh nebulisers (VMN) generate low velocity aerosols. Whilst the focus of drug delivery is on the patient, bystander caregivers are often exposed to aerosol as it escapes or is exhaled through the nebuliser system. The potential for inhalation of medications, not required by [3, 4] , and has not been comprehensively described to date in the literature. that person, is relatively high We set out to attempt to characterise the risk to caregivers and other bystanders during the course of a standard nebuliser treatment. Materials and methods -3

Exhaled aerosol emissions were evaluated by characterising the mass concentrations (mg m ) & mass medium diameters (MMD) (μm) emitted from a simulated patient. Inhaled dose (%) delivered was also recorded. All testing carried out n=3. This study incorporated two combinations of commercially available aerosol generators and noninvasive ventilation interfaces, see Figure 1; VMN/valved facemask/filtered mouthpiece (Aerogen Solo/Ultra, Aerogen, Ireland) vs JN/open facemask/filtered mouthpiece (Cirrus 2, Intersurgical, United Kingdom). A supplemental gas flow rate of 6 LPM was used with the VMN and a driving gas flow of 8 LPM for the JN, while a 2.5 mL dose of albuterol sulphate (1 mg/mL) was nebulised as a tracer aerosol. Each aerosol generator/non-invasive ventilation interface combination was connected to a breathing simulator (ASL 5000, Ingmar Medical, PA, USA) via an absolute filter (RespirGard II 303, Baxter, Ireland). A simulated adult breath was used (BPM 15, Vt 500 mL, I:E 1:1). As shown in Figure 1, the primary aerosol instrument used in this study was the Aerodynamic Particle Sizer (APS) (APS, model 3321 TSI Inc., St. Paul, MN) measuring airborne particle size distributions from 0.5 to 20 μm. Throughout the experiments; mass and number concentrations, and size distribution of the airborne particles were continuously measured in real time using two APS’s located at different distances (0.8 m & 2.2 m). A 5-minute baseline level of airborne particles was established in the room pre nebulisation. Nebulisation was then initiated with the APS’s recording data for a total of 30 minutes. Facemask and mouthpiece tests were performed on separate days (and had different airflows as a result).

367


Drug Delivery to the Lungs 27, 2016 - Assessment of Exhaled Aerosol Emissions using Two Prevalent Nebuliser Technologies in Clinical Use Today.

Figure 1 - Aerodynamic Particle Sizer (APS) at varying distances relative to the simulated patient (ASL 5000) with VMN & JN facemask/mouthpiece iterations.

Results and discussion -3

Interface

-3

Mass concentrations (mg m ) 0.8 m APS

Mass concentrations (mg m ) 2.2 m APS

Valved facemask (VMN)

0.025

0.022

Open facemask (JN)

0.048

0.044

Filtered mouthpiece (VMN)

0.00441

0.00436

Filtered mouthpiece (JN)

0.00980

0.00868

Table 1 –

Mass concentrations (mg m-3) results for APS at a distance of 0.8 m & 2.2 m away.

Interface

MMD (μm) 0.8 m APS

MMD (μm) 2.2 m APS

Average (μm)

SD

Average (μm)

SD

Valved facemask (VMN)

1.35

0.22

1.35

0.19

Open facemask (JN)

1.19

0.13

1.20

0.09

Filtered mouthpiece (VMN)

1.432

0.241

1.184

0.180

Filtered mouthpiece (JN)

1.173

0.079

0.982

0.085

Table 2 –

MMD (μm) results for APS at a distance of 0.8 m & 2.2 m away.

368


Drug Delivery to the Lungs 27, 2016 – Andrew O’Sullivan et al.

Figure 2 - Illustration of total mass concentrations emitted by two aerosol generators and non-invasive ventilation interface combinations 0.8 m away from simulated patient.

Figure 3 - Illustration of total mass concentrations emitted by two aerosol generators and non-invasive ventilation interface combinations 2.2 m away from simulated patient.

Based on the myriad of aerosol generators & non-invasive ventilation interface combinations tested, the JN/open -3 facemask emitted the highest mass concentrations (mg m ). At 0.8 m the JN/open facemask facilitated a near -3 two fold higher emitted mass concentration (mg m ) compared with the VMN/valved facemask (0.048 vs 0.025 -3 -3 mg m ), near fivefold higher compared with the JN/filtered mouthpiece (0.048 vs 0.00980 mg m ) & tenfold -3 higher compared with the VMN/filtered mouthpiece (0.048 vs 0.00441 mg m ).

369


Drug Delivery to the Lungs 27, 2016 - Assessment of Exhaled Aerosol Emissions using Two Prevalent Nebuliser Technologies in Clinical Use Today. A similar trend continued for the APS at a distance of 2.2 m away from simulated patient, with the JN/open facemask having again emitted the highest mass concentration over the 30 minute run time. The VMN/filtered mouthpiece was found to have emitted the lowest mass concentration over time -3 -3 (0.00441 mg m at 0.8 m & 0.00436 mg m at 2.2 m) while emitted particles had a larger MMD. Finally, the VMN/filtered mouthpiece combination allowed for the highest inhaled dose (50.319 ± 1.898%) compared with the VMN/valved facemask (46.686 ± 3.055%), JN/open facemask (21.952 ± 0.659%), and JN/filtered mouthpiece (35.942 ± 1.811%). Conclusion The initial findings described herein successfully demonstrated the obvious escape of aerosol to the environment and further established the risk to caregivers and other bystanders during the course of a standard nebuliser treatment. Considering the results above it is evident that of the two patient interfaces (facemask and mouthpiece) and aerosol generators (VMN and JN) tested, the mouthpiece combinations released lower fractions of exhaled mass concentrations across both distances (0.8 m & 2.2 m). Additionally the VMN/mouthpiece combination delivered the highest inhaled dose (%) to the patient. References (1)

American Thoracic Society. Skills of the health team involved in out-of-hospital care for patients with COPD. Am Rev Respir Dis 1986;133:948–949.

(2)

Somogyi R, Vesely AE, Azami T, Preiss D, Fisher J, Correia J, Fowler RA. Dispersal of respiratory droplets with open vs closed oxygen delivery masks: implications for the transmission of severe acute respiratory syndrome. Chest 2004;125(3):1155–1157.

(3)

Chughtai, A.A., Seale, H. and MacIntyre, C.R., 2013. Availability, consistency and evidence-base of policies and guidelines on the use of mask and respirator to protect hospital health care workers: a global analysis. BMC research notes, 6(1), p.216.

(4)

Ari A, Fink JB, Pilbeam S. Secondhand aerosol exposure during mechanical ventilation with and without expiratory filters: An in-vitro study. Ind J Resp Care 2016; 5(1): 677-82.

370


Drug Delivery to the Lungs 27, 2016 – Francesca Buttini et al. Accessorized DPI: A Shortcut Towards Flexibility and Patient Adaptability in Dry Powder Inhalation 1,2

3

3

1

1,a

Francesca Buttini , James Hannon , Kristi Saavedra , Irene Rossi , Anna Giulia Balducci , Hugh 3,4 3,5 1 Smyth , Andy Clark & Paolo Colombo 1

2

Department of Pharmacy, University of Parma, Viale delle Scienze 27/A, Parma, 43124, IT Institute of Parmaceutical Science, King’s College London, 150 Stamford Street, London, SE19NH, UK 3 Respira Therapeutics Inc., 5901 Indian School Rd NE #107, Albuquerque, NM, 87110, US 4 College of Pharmacy, the University of Texas at Austin, Austin, TX, 78712, US 5 Aerogen Pharma Corporation, 1660 S Amphlett Blvd, Suite 360, San Mateo, CA, 94402, US a Present address: Chiesi Limited, Bath Road Industrial Estate, Chippenham, Wiltshire, SN140AB, UK

Summary In this work, a novel powder dispersion add-on device, the AOS (Axial Oscillating Sphere), was studied in conjunction with commercially available DPI devices to investigate the improvement of powder dispersion and to minimize the dependence of the device performance on the inspiratory effort. An ordered mixture of formoterol fumarate and lactose was selected. We studied the emission and dispersion of the drug at different flow rates, paying particular attention to a number of metrics of Fine Particle Dose (FPD). Two novel findings emerged from the data collected: the aerosol quality, measured as fine particle dose, was increased by adding the accessory promoting the formulation deagglomeration and the flow dependence of the [1] aerosol formation was reduced . Increasing inhaler performance can be achieved using an add-on accessory that enhances aerosol dispersion and minimizes flow rate dependency. Introduction Drug delivery and intrinsic lung deposition from a dry powder inhaler (DPI) are influenced by the inspiratory flow produced by the patient, the resistance of the inhaler and formulation characteristics. [2, 3]

It has been reported that clinical efficacy is related to the inhalation flow rates achieved through a DPI . Delivery of drug to the lung has been shown to be lower in children than in Chronic Obstructive Pulmonary Disease (COPD) patients, whereas adults with asthma produce the greatest dependency on inhalation flow [4] rates .Therefore, the inspiratory profile of the patient is an issue that must be addressed for an efficient DPI performance. Lung deposition differences between patients using DPIs underline the need of technologic expedients that are capable of minimizing flow rate and interpatient variability. One novel concept translated into product form has been the design of the accessory named AOS (Axial Oscillating Sphere, Respira Therapeutics Inc.), which is to be used in conjunction with dry powder devices to optimize the aerosol formation. The objective of this study was to evaluate the AOS as an accessory to different DPI devices, assessing its impact on drug aerosol dispersion and its ability to promote the formation of extra-fine aerosol particles while taking into account the impact of device resistance. The concept of using the AOS as a simple "upgrading accessory" to enhance inhalation [5] performance has been reported previously . The aim of this research was to study the performance of the AOS accessory in connection with RS01 DPI device (Plastiape SpA), a capsule based reservoir device, using RS01 DPI versions with two different intrinsic airflow resistances. In order to conduct the study a commercial drug product consisting of an ordered mixture of formoterol fumarate and lactose for inhalation was selected. We studied the emission and dispersion of the drug at different flow rates, paying attention not only to the fine particle dose (< 5 µm), but in particular to the extra-fine particle dose (< 3µm). Materials and Methods The drug product employed was formoterol fumarate lactose blend (Foradil Aerolizer, Novartis – inhalation powder in hard gelatine capsules) purchased from a local pharmacy. Each capsule is filled with 12 µg of formoterol fumarate blended with 25 mg of lactose. All chemicals used were of analytical grade and water was purified by ElixEssential (Merck Millipore, MA, USA). The device used in the study was RS01 (Plastiape Spa, Osnago LC, Italy), available in two different versions: medium and low resistance, coded as RS01_MR and RS01_LR respectively. The AOS accessory uses a small [5] spherical bead oscillating in a cylindrical chamber to promote powder dispersion, as reported by Hannon et al .

371


Drug Delivery to the Lungs 27, 2016 - Accessorized DPI: A Shortcut Towards Flexibility and Patient Adaptability in Dry Powder Inhalation

Figure 1 - Schematic cross-sections of RS01+AOS with the modified mouthpiece and added AOS dispersion chamber. The small bead in the AOS chamber primarily oscillates along the axis of the cylindrical chamber. A parallel flow path was included in the RS01-AOS to reduce overall device resistance.

The aerodynamic assessment was performed using the Next Generation Impactor (NGI) (Copley Scientific, Nottingham, UK). The methodology followed the USP36 guidelines for dry powder inhalers (Apparatus 5, United States Pharmacopoeia, Chapter 601). The flow rate used during each test was adjusted with a Critical Flow Controller TPK (Copley Scientific, Nottingham, UK) in order to produce a pressure drop of 2 kPa or 4 kPa across the inhalers. The resistance of the tested devices is reported in Table 1.

Table 1 -

Resistance values for RS01 inhalers in the different configuration tested. R (kPa0.5/LPM)

Inhaler RS01-MR

0.033

RS01-MR AOS

0.038

RS01-LR

0.019

RS01-LR AOS

0.036

The metered dose (MD), the mass of drug recovered, was quantified by HPLC by summing the drug recovered from the device and capsule and the impactor (induction port, pre-separator, stages 1 to 7 and MOC). The Emitted Dose (ED) was calculated as the amount of drug leaving the device, i.e. reaching the impactor (induction port, pre-separator, stages 1 to 7 and MOC). The mass median aerodynamic diameter (MMAD) was determined as indicated by USP 36. The Fine Particle Dose (FPD) was defined in two different ways: (i) mass of drug < 5 µm (calculated from log-probability plots), (ii) mass of drug < 3 µm (calculated from log-probability plots). Results and Discussion The emitted dose for all the cases studied was higher than 75% with no significant difference observed between the standard RS01 devices and those with the AOS attached. Figure 2 shows the two categories of fine particle dose, i.e., less than 5 µm and less than 3 µm. A similar trend is presented for both. The mass of drug < 5 µm, with or without the AOS accessory, depended on the flow rate (pressure drop). An increase of FPD to approx. 3.5 µg for air flow rates > 60 L/min was observed for the RS01 without accessory. When the AOS was connected to the device, the FPD values improved compared to the values without AOS at the same flow rate. The oscillating sphere of AOS enhanced the detachment of the drug particles from the lactose carriers, allowing the use of the DPI at lower flow rates while maintaining delivery performance. For the mass of drug < 3 µm, the values with AOS accessory were also consistently higher compared to those generated without AOS.

372



Drug Delivery to the Lungs 27, 2016 – Francesca Buttini et al.

Figure 2 - Fine Particle Dose of formoterol fumarate < 5 µm and < 3µm (circle: RS01-LR; square: RS01-MR. Empty symbols: without AOS; full symbols: with AOS).

Another value examined was the Mass Median Aerodynamic Diameter for both the two types of RS01 with or without AOS. An inverse relationship between MMAD and the airflow rate was observed with and without the AOS device with the MMAD of the aerosol generated by the AOS equipped devices showing and asymptote to that of the micronized drug powder (Figure 3).

Figure 3 - MMAD and flow rate for the RS01 devices with and without AOS (circle: RS01-LR; square: RS01-MR. Empty symbols: without AOS; full symbols: with AOS).

The fraction of an inhaled aerosol reaching, and depositing in, the lungs is dependent upon the filtering that takes place in the mouth and oropharynx. Traditionally, FPD is defined as the fraction of particles under 5 µm aerodynamic diameter. It is known that the best way to compare the performance of dry powder inhalers is to compare them at equal pressure drops. Therefore, the low resistance devices have to be tested at high flow rates, while, on the other hand, high resistance devices at low flow rates. Since the deposition within the airways is controlled by flow rate and particle size, not pressure, this underlines that ideally the size fraction considered to be “respirable” has to change with the device resistance, i.e. test flow rate. Over the past few decades, numerous definitions of “Fine Particle Dose” or “Respirable Fraction” have been employed. The principle behind all of these definitions is to estimate through in vitro tests the dose likely to deposit in the lungs of patients in vivo. In general, a respirable fraction of < 5 µm cut-off has been used; this is derived from the size distribution data obtained from multistage impaction instruments by plotting the size distribution and interpolating. More recently, studies using a compilation of published scintigraphy studies have shown that a more appropriate cut-off size is < 3 µm. This definition of FPD produces a 1:1 correspondence with [6] lung deposition, albeit with a rather large scatter about the mean . For this report we chose not to try and correct for the influence of flow rate on the definition of FPD, but rather to concentrate on two different fixed cut-off diameters (< 5 µm and < 3 µm) to compare the respirable dose emitted from the devices we investigated.

373


Drug Delivery to the Lungs 27, 2016 - Accessorized DPI: A Shortcut Towards Flexibility and Patient Adaptability in Dry Powder Inhalation Conclusion The presence of the AOS increased the deposition of the drug on the lower stages of the impactor (i.e. with smaller cut-offs diameters) at all flow rate and showed better dispersion, compared to the standard RS01 devices using both definitions of FPD. The oscillating sphere of AOS contributes to the detachment of drug particles from [1] the carrier and their subsequent dispersion . This was particularly evident at the lower test flow rates. Thus, the use of an AOS accessorized DPI enhances the dispersion and deagglomeration of the tested ordered blend and, as a consequence of its better performance at lower flow rates, makes the DPI performance less dependent on the inspiratory effort of the patient (pressure drop).

References 1

Buttini F, Hannon J, Saavedra K, Rossi I, Balducci AG, Smyth H, Clark A, Colombo P: Accessorized DPI: a Shortcut towards Flexibility and Patient Adaptability in Dry Powder Inhalation, Pharm Res 2016 In press , DOI 10.1007/s11095-016-2023-0

2

Buttini F, Brambilla G, Copelli D, Sisti V, Balducci AG, Bettini R &Pasquali I : Effect of Flow Rate on In Vitro Aerodynamic Performance of NEXThaler® in Comparison with Diskus® and Turbohaler® Dry Powder Inhalers, J Aerosol Med Pulm Drug Deliv 2015, pp167-178

3

Buttini F, Pasquali I, Brambilla G, Copelli D, Alberi MD, Balducci AG, Bettini R &Sisti V : Multivariate Analysis of Effects of Asthmatic Patient Respiratory Profiles on the In Vitro Performance of a Reservoir Multidose and a Capsule-Based Dry Powder Inhaler, Pharm Res 2015 1–15

4

Azouz W.,Chetcuti P., Hosker HSR., Saralaya D., Stephenson J., Chrystyn H. : The inhalation characteristics of patients when they use different dry powder inhalers , J. Aerosol Med Pulm Drug Deliv. 2014:27:1-8

5

Hannon, J., Donovan, M., Gibbons, A., Xu, Z., Curtis, R., Smyth, H., Clark, A. : Novel high efficiency inhalaer for PDE5i lung delivery, Respiratory Drug Delivery 2016, pp519-522

6

Newman SP, Chan H-K. In vitro/ In vivo comparisons in pulmonary drug delivery, J Aerosol Med Pulm Drug Deliv. 2008;21:77–84

374


Drug Delivery to the Lungs 27, 2016 –E. Hoffman et el Characterisation of Interspecies Differences between Human and Rat Airway Macrophage Responses in vitro 1

2

1

2

3

3

E. Hoffman , A. Kumar , R. Mahendran , A Patel , V. Millar , M. Clements , 2 2,4 1 B. Forbes , L. Dailey , & V. Hutter 1

2

University of Hertfordshire, College Lane, Hatfield, Herts, AL10 9AB, UK Kings College London, Waterloo Campus, Franklin-Wilkins Building, 150 Stamford Street, London, SE1 9NH, UK 3 GE Healthcare Life Sciences, Maynard Centre, Forest Farm, Whitchurch, Cardiff CF14 7YT, UK. 4 Institute of Pharmacy, Martin-Luther University Halle-Wittenberg, Wolfgang-Langenbeck Str. 4, 06120 Halle/Saale, Germany

Summary Few new inhaled medicines for the treatment of asthma and chronic obstructive pulmonary disease (COPD) have made it to market in the past decade despite the increasing global health burden of airways diseases. One of the reasons there has been limited success, despite considerable investment in this area to develop new inhaled drugs, is due to the observation of foamy macrophage responses in pre-clinical rat studies, which questions their safety for use in humans. The aim of this work was to develop in vitro cell culture assays to better characterise alveolar macrophage responses and to ascertain interspecies differences between human and rat macrophage models. Baseline profiles for cell health, morphology and lipid content parameters were generated for U937 (human) and NR8383 (rat) macrophage models using high content assays and analysis platform. Cells were then exposed to drug challenges with established pathological responses for assay validation. The rat and human in vitro macrophage models investigated had identical baseline profiles for all parameters tested. Typical response profiles were observed for both cell lines exposed to amiodarone (inducer of phospholipidosis) and staurosporine (apoptotic agent). These results indicate that the high content assays developed are suitable for understanding macrophage responses at the individual cell level, which may allow the characterisation of specific types of foamy macrophage responses. Future work is ongoing to further validate the assays to ascertain its potential, as an early pre-clinical screening tool, to predict the safety of candidate inhaled medicines. Introduction Airway diseases such as asthma and chronic obstructive pulmonary disease (COPD) still remain an area of [1, 2] . Many new unmet clinical need despite considerable investment to develop new therapeutics in this area inhaled medicines that perform well in pre-clinical animal studies fail to demonstrate safety or efficacy in humans, [1-4] . Foamy macrophage responses are often which questions the relevance of current preclinical models observed in histological lung slices of rats from preclinical in vivo studies, which are typically characterised by a [2, 3] highly vacuolated appearance and larger cell size . These cell responses may be associated with or without [2, 3] . As the mechanism for induction of the foamy other immune cell infiltration or remodelling of lung tissue alveolar macrophage phenotype and its relation to lung pathophysiology are not well understood, safe exposure [2-4] . Due to these levels are set without knowing if these observations are truly an adverse response or not [3] concerns over safety, inhaled compounds often fail in human studies due to lack of efficacy . The main objective of this work was to develop in vitro cell culture assays to better characterise alveolar macrophage responses and to ascertain interspecies differences between human and rat macrophage models. By obtaining detailed information regarding cell health and morphological parameters for macrophages exposed to a range of compounds with established responses in vitro, we aim to gain an in depth understanding of the different types of foamy macrophage response to better predict if all foamy macrophage responses are truly adverse. The ultimate goal of the research will be to develop a high-throughput screening process that is suitable to test new drug candidates, and with the power to predict macrophage responses to inhaled candidate drugs. It is anticipated that these assays may also help reduce the numbers of animal experiments required to achieve an accurate prediction of the safety of new inhaled medicines in humans. Materials and Methods Cell culture: Rat macrophage (NR8383) and human monocyte (U937) cell lines were purchased from LCG Standards (Teddington, Middlesex, UK) and used between passage 2 and 20 from purchase. NR8383 cells were cultured in Kaighn’s modified Ham’s F12 (K-F12) medium with 15 %v/v fetal bovine serum (FBS) and supplemented with 100 IU/ml penicillin-100 µg/ml streptomycin solution and 2 mM L-glutamine. U937 cells were cultured in RPMI with 10 %v/v FBS and supplemented with 100 IU/ml penicillin-100 µg/ml streptomycin solution and 2 mM L-glutamine. Cells were cultured in a humidified atmosphere at 37 °C with 5 %v/v CO2 and cell number 5 6 was maintained between 1 x 10 to 2 x 10 cells/ml. For experiments, cells were seeded onto 96 well plates at a 4 density of 3 x 10 cells/well in 100 µl of complete cell culture medium. U937 cells were differentiated to a macrophage phenotype using 4 nM phorbol myristate acetate (PMA) in complete cell culture media for 96 h followed by a 24 h rest period in complete cell culture media.

375


Drug Delivery to the Lungs 27, 2016 - Characterisation of Interspecies Differences between Human and Rat Airway Macrophage Responses in vitro Induction and characterisation of macrophage responses: Cells were incubated with amiodarone (0.003 – 100 µM) or staurosporine (0.0003 – 10 µM) in complete cell culture medium with 1 %v/v DMSO 24 h after seeding (NR8383) or after differentiation (U937). For cell health and morphology assessment, cells were stained with Hoechst 33342 (10 µg/ml), MitoTracker Red (300 nM), Image-It Dead Green (25 nM) for 30 min followed by fixation with 3.7 %w/v paraformaldehyde for 20 min. Fixed cells were then stained overnight with Cell Mask Deep Red (diluted 1:1000). For the determination of lipid content, cells were incubated with HCS LipidTox Phospholipid Red (diluted 1:1000) for 24 h and fixed with 3.7 %w/v paraformaldehyde containing Hoechst 33342 (10 µg/ml) for 20 min. Cells were then incubated with HCS LipidTox Green (diluted 1:1000) for 30 min for detection of neutral lipids. Cells from both assays were imaged using the In Cell Analyser 6000 (GE Healthcare, Little Chalfont, ucks, UK) with a 40x objective. Image analysis was performed using In Cell Developer v 1.9.2 (GE Healthcare, Little Chalfont, Bucks, UK). Cells were categorised as “healthy” and included in further analysis when possessing mitochondrial activity greater than the mean minus 2 standard deviations from the untreated cell population. Cell viability was calculated as the percentage of healthy cells per well. Nuclear area was classified as “normal” for nuclei size being greater than the mean minus 2 standard deviations and less than the mean plus 2 standard deviations. All remaining cell profile data was calculated as the percentage of cells greater than the mean plus 2 standard deviations of the healthy cell population. Results and Discussion The baseline profiles for cell health, morphology and lipid profile parameters were assessed for untreated rat alveolar macrophage (NR83838) and human monocyte derived macrophage (U937) cell lines. No significant difference (p>0.05) was observed for any of the parameters tested (Figure 1) and reproducible cell profiles were obtained over three passages tested for both species. The consistency in profiles generated for cell health, morphology and lipid characterisation for both cells lines supports their relevance as a tool to investigate interspecies differences in vitro. Additionally, as less than 10% of the cell populations were characterised as possessing atypical morphology or lipid content, these cell models provide a good platform for investigating abnormal or foamy macrophage responses.

Figure 1 - Cell health, morphology and lipid profiling of rat and human in vitro macrophage cell models. Multi-parameter charts profiling cell health, morphology and lipid changes to NR8383 rat macrophage cells (black line) and U937 human monocyte-derived macrophage cells (purple line). Data is presented as the mean of n=6 wells per plate from three experiments with different cell passage numbers.

Amiodarone was expected to induce phospholipids accumulation in the cells. A typical profile for phospholipidosis was observed for both cell lines in a concentration dependent manner (Figure 2). On exposure to amiodarone, mitochondrial activity increased for both cell lines until 10 µM. Above this concentration, cell viability was reduced which was associated with an increase in cell permeability. The presence of phospholipids was significantly increased (p<0.05) in cells treated with amiodarone in comparison with untreated cells. After 24 h exposure to 10 µM amiodarone, the presence of phospholipids was 6.5 fold greater in NR8383 cells and elevated 3.5 fold in U937 cells in comparison with untreated control cells. The number and area of vacuoles within cells was also significantly raised (p<0.05) for cells incubated with 10 µM concentrations of amiodarone. Additionally, 10 µM amiodarone significantly increased (p<0.05) cell permeability in U937, whereas no impact on NR8383 was observed.

376


Drug Delivery to the Lungs 27, 2016 –E. Hoffman et el

Figure 2 - Cellular response of in vitro macrophages to amiodarone. Multi-parameter charts profiling cell health, morphology and lipid changes to NR8383 rat macrophage cells (A) and U937 human monocyte-derived macrophage cells (B) exposed to amiodarone (0.1-10 µM) for 24 h. Data is presented as the mean of n=1 wells per plate from three experiments with different passage number.

Staurosporine induced apoptotic cell death in cells exposed to concentration above 1 µM and was characterised by a reduction in cell viability, increased cell permeability and increased cell area (Figure 3). The rat macrophage cell line was associated with an increase in both phospholipid and neutral lipid content when exposed to 1 µM staurosporine whereas no significant change (p>0.05) in the lipid profile was observed for the human U937 cell line.

Figure 3 - Cellular response of in vitro macrophages to staurosporine. Multi-parameter charts profiling cell health, morphology and lipid changes to NR8383 rat macrophage cells (A) and U937 human monocyte-derived macrophage cells (B) exposed to staurosporine (0.01-1 µM) for 24 h. Data is presented as the mean of n=1 wells per plate from three experiments with different passage number.

377


Drug Delivery to the Lungs 27, 2016 - Characterisation of Interspecies Differences between Human and Rat Airway Macrophage Responses in vitro Conclusion This study suggests that whilst the baseline profile of untreated lung macrophages in vitro is similar between rat and human models, there are observed differences in macrophage responses to drug challenges with established pathophysiology. Initial results indicate that the high content analysis approach to profiling macrophage behaviour in vitro may allow characterisation of different types of foamy macrophage response. This approach has the potential to offer an early pre-clinical screening tool to predict the safety of candidate inhaled medicines. It is anticipated that characterising foamy macrophage responses in this way may provide a better understanding of the pathophysiology of airway immune responses to inhaled medicines and permit the development of new, safe inhaled therapeutics for airway disease to reach the market. References 1.

Nikula KJ, McCartney JE, McGovern T, Miller GK, Odin M, Pino MV, and Reed MD. STP position paper: interpreting the significance of increased alveolar macrophages in rodents following inhalation of pharmaceutical materials. Toxicologic pathology. 2014;42:472-486.

2.

Lewis DJ, Williams TC, and Beck SL. Foamy macrophage responses in the rat lung following exposure to inhaled pharmaceuticals: a simple, pragmatic approach for inhaled drug development. Journal of applied toxicology : JAT. 2014;34:319-331.

3.

Forbes B, O'Lone R, Allen PP, Cahn A, Clarke C, Collinge M, Dailey LA, Donnelly LE, Dybowski J, Hassall D, Hildebrand D, Jones R, Kilgour J, Klapwijk J, Maier CC, McGovern T, Nikula K, Parry JD, Reed MD, Robinson I, Tomlinson L, and Wolfreys A. Challenges for inhaled drug discovery and development: Induced alveolar macrophage responses. Advanced drug delivery reviews. 2014;71:15-33.

4.

Jones RM and Neef N. Interpretation and prediction of inhaled drug particle accumulation in the lung and its associated toxicity. Xenobiotica; the fate of foreign compounds in biological systems. 2012;42:86-93.

378


Drug Delivery to the Lungs 27, 2016 – Melissa P. Manice et al. Connected drug delivery systems drive improved clinical outcomes 1

2

Melissa P. Manice , Benjamin Jung & Dan Weinstein 1

1

Cohero Health Inc., 335 Madison Avenue, New York, NY 10017, USA 2 H&T Presspart, Am Meilenstein 8-19, Marsberg 34431, Germany

Summary Compliant asthmatic patients are significantly less likely to undergo an exacerbation than their less-compliant [1] counterparts . Studies have shown that patient adherence can already be significantly increased by text [2] message reminders . Different connected drug delivery systems have been introduced in the respiratory arena in the last years. For example Cohero’s platform in connection with their Herotracker and H&T Presspart’s eMDI or the solutions of Propeller Health or Adherium. These systems use a sensor to detect usage pattern of inhalers, e.g. metered dose inhalers, dry powder inhalers or nebulizers, and transmit these patterns to hosts, for example mobile phones. Based on the collected information patients can be engaged and their adherence increases. These systems therefore have the ability to improve patient outcomes significantly. The resulting value creates opportunities for the different stakeholders in the health system – pharmaceutical industry, regulators, physicians, insurance companies, patients. The use of these and other electronic monitoring devices (EMDs) is becoming increasingly prevalent in clinical practice to harness cutting edge technology to achieve improved disease control and medication adherence. The systems described above offer extensive potential in both the clinical trials and clinical care environments. Especially in connection with lung function measurement devices, for example Cohero’s Spirometer, electronic data capture can furthermore drive improved clinical decision making and overall care management for the pulmonary community. Introduction [3]

Asthma is a major global disease, more than 300 million people suffer from it worldwide . Non-adherence to controller medication is a common problem in patients with an asthma diagnosis leading to avoidable [4] exacerbations . This work therefore describes potential solutions to increase adherence and their short to longterm areas of application and market penetration. Main Body of Text Asthma affects 334 million people worldwide, including approximately 14% of all children, and pharmacotherapy [5, 6] . Currently, $56 billion dollars is essential to prevent symptoms and relieving asthma attacks in this population [7] are spent in the US annually on direct and indirect costs of asthma . In order to provide therapeutic benefit, preventer (i.e. controller) medications need to be taken daily. [8]

However, non-adherence to controller medication is a common problem in patients with an asthma diagnosis . [9, 10] . Optimal Current rates of non-adherence to asthma controller medication are estimated at 45-50% adherence to inhaled corticosteroids requires patients to take their controller medication at least 80% of the time [11] in order to prevent exacerbations . Thus, substantial numbers of patients do not get the maximum benefit of medical treatment; this results in the overuse of reliever medication, increased asthma symptoms, more frequent [12, 13, 14] . The poor adherence with preventive treatments is asthma attacks, and increased hospital admissions [15] associated with reduced quality of life and leads to mortality . Based on refill data, the risk of an asthma [16] exacerbation is 21% to 68% lower for children who are adherent to controller medications . In General, the use of electronic monitoring devices (EMDs) is becoming increasingly prevalent in clinical practice to harness cutting edge technology to achieve improved disease control and medication adherence. As the cost of such devices is reduced and improvements in technology are made, there is increasing evidence that such [17] interventions can be a cost-effective tool for improving patient outcomes and decreasing healthcare costs . Examples can for example be found in diabetes care. EMDs in the form of connected drug delivery systems have been introduced in the respiratory arena in the last years, for example Cohero’s platform in connection with their Herotracker and H&T Presspart’s eMDI or the solutions of Propeller Health or Adherium. These systems use a sensor to detect usage pattern of inhalers, e.g. metered dose inhalers, dry powder inhalers or nebulizers, and transmit these patterns to hosts, for example mobile phones. They can be used as a tool – for example in form of an App – to generate discussion of adherence and provide reminders to patients; therewith patients can be engaged. Such technology has the potential to increase adherence significantly and therewith to prevent exacerbations and improve symptom [18] control . We expect that this technology will address the need for improved understanding of patient medication use behaviour and that patients using such technology will be better equipped for disease self-management leading to improved asthma control.

379


Drug Delivery to the Lungs 27, 2016 - Connected drug delivery systems drive improved clinical outcomes The systems described above offer extensive potential in both the clinical trials and clinical care environments. Especially in connection with lung function measurement devices, for example Cohero’s Spirometer, electronic data capture can furthermore drive improved clinical decision making and overall care management for the pulmonary community. In the future, as mHealth and telehealth continue to grow and become more cost-effective, we expect that use of such technology will become standard of care and become integrated into the current clinical workflow. Particularly for those patients who are most at-risk for exacerbations and subsequent hospitalizations, we expect that such technology will be used for improved coordination of care leading to decreased healthcare costs to both patients and to the system at large. The resulting value creates opportunities for all the different stakeholders in the health system – pharmaceutical industry, regulators, physicians, insurance companies, patients.

Aerosol Logo 1

Drug Delivery to the Lungs 27

DDL27 Logo

Sterrn L, Berman, J, Lumry W, Katz L, Wang L, Rosenblatt L, Doyle J J: Medication compliance and disease exacerbation in patients with asthma: a retrospective study of managed care data, in: Ann Allergy Asthma Immunol, 2006, Vol. 97, p. 402-408.

Wednesday 7th - Friday 9th December 2016

2

Miloh T, Annunziato R, Arnon R, Warshaw J, Parkar S, Suchy F J, Iyer K, Kerkar N: Improved Adherence and

3

Global Asthma Network: The Global Report, 2014, Drug Delivery toAsthma the Lungs 27 Auckland, New Zealand.

4

Sterrn L, Berman, J, Lumry W, Katz L, Wang L, Rosenblatt L, Doyle J J: Medication compliance and disease th exacerbation in patients asthma: a9retrospective study2016 of managed care data, in: Ann Allergy Asthma Immunol, - Friday December Wednesday 7thwith 2006, Vol. 97, p. 402-408.

Outcomes for Pediatric Liver Transplant Recipients by Using Test Messages, in: Pediatrics,, 2009, Vol. 124, p. 844Edinburgh 840. International Conference Centre, Scotland, UK

5

Thank you to our Platinum Sponsors

Global Asthma Network: The Conference Global Asthma Report, 2014,Scotland, Auckland, New Zealand. Edinburgh International Centre, UK

6

Sterrn L, Berman, J, Lumry W, Katz L, Wang L, Rosenblatt L, Doyle J J: Medication compliance and disease exacerbation in patients with asthma: a retrospective study of managed care data, in: Ann Allergy Asthma Immunol, 2006, Vol. 97, p. 402-408.

7

Barnett S B, Nurmagambetov T A: Platinum Costs of asthma in the United States: 2002-2007, in: J Allergy Clin Immunol, 2011, Thank you to our Sponsors Vol. 127, p. 145-152.

8

Sterrn L, Berman, J, Lumry W, Katz L, Wang L, Rosenblatt L, Doyle J J: Medication compliance and disease exacerbation in patients with asthma: a retrospective study of managed care data, in: Ann Allergy Asthma Immunol, 2006, Vol. 97, p. 402-408.

9

Wiliams L K, Pladevall M, Xi H, Peterson E L, Joseph C, Lafata, J E, Ownby D R, Johnson, C C: Relationship Addcorticosteroids Logos between adherence to inhaled and poor outcomes among adults with asthma. In: J Allergy Clin Immunol, 2004, Vol. 114, p.1288-1293.

10

Claxton A J, Cramer J, Pierce C.: A systematic review of the associations between dose regimens and medication compliance, in: Clin Ther, 2001, Vol. 23, p. 1296-1310.

11

Sterrn L, CHIESI Berman, J, Lumry W, Katz L, Wang L, Rosenblatt L, Doyle J J: Medication compliance and disease exacerbation in patients with asthma: a retrospective study of managed care data, in: Ann Allergy Asthma Immunol, 2006, Vol.Intertek 97, p. 402-408.

12

Wiliams L K, Pladevall M, Xi H, Peterson E L, Joseph C, Lafata, J E, Ownby D R, Johnson, C C: Relationship between adherence to inhaled corticosteroids and poor outcomes among adults with asthma. In: J Allergy Clin Oechsler Immunol, 2004, Vol. 114, p.1288-1293.

13

Miloh T, Annunziato R, Arnon R, Warshaw J, Parkar S, Suchy F J, Iyer K, Kerkar N: Improved Adherence and Outcomes for Pediatric Liver Transplant Recipients by Using Test Messages, in: Pediatrics, 2009, Vol. 124, p. 844840.

14

Sterrn L, Berman, J, Lumry W, Katz L, Wang L, Rosenblatt L, Doyle J J: Medication compliance and disease exacerbation in patients with asthma: a retrospective study of managed care data, in: Ann Allergy Asthma Immunol, 2006, Vol. 97, p.Visit 402-408. the DDL27 Event App

15

Wiliams L K, Pladevall M, Xi H, Peterson E L, Joseph C, Lafata, J E, Ownby D R, Johnson, C C: Relationship between adherence to inhaled corticosteroids and poor outcomes among adults with asthma. In: J Allergy Clin Immunol, 2004, Vol. 114, p.1288-1293.

16

Anderson W C, Szefler S J: New and Future Strategies to Improve Asthma Control in Children. National Center for Biotechnology Information. U.S. National Library of Medicine, 28 Aug. 2015. Web. 29 July 2016.

17

Chan A H, Harrison J, Black P N, Mitchell E A, Foster J M: Using electronic monitoring devices to measure inhaler adherence: a practical guide for clinicians, in: J Allergy Clin Immunol Pract, 2015, Vol. 3, p. 335-349.

18

Anderson W C, Szefler S J: New and Future Strategies to Improve Asthma Control in Children. National Center for Biotechnology Information. U.S. National Library of Medicine, 28 Aug. 2015. Web. 29 July 2016.

3M

Molins

Team Consulting

Visit the DDL27 Event App

Copyright 'The Aerosol Society' 380


Thank you to our Silver Sponsors


2016

Follow us on:

www.ddl-conference.com

DRUG DELIVERY TO THE LUNGS 27

t

www.aerosol-soc.com


Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.