Acta Orthopaedica Vol. 89 Issue 3 June 2018 by Acta

Acta

C\ Taylor & Francis ï¿½ Taylor&Francis Group

3/18

Volume 89, Number 3, June 20"18

Acta Orthopaedica is owned by the Nordic Orthopaedic Federation and is the official publication of the Nordic Orthopaedic Federation

E DITORIAL O FFICE

Acta Orthopaedica Department of Orthopedics Lund University Hospital SE–221 85 Lund, Sweden E-mail: acta.ort@med.lu.se Homepage: http://www.actaorthop.org

EDITOR

THE FOUNDATION BOARD OF

Anders Rydholm Lund, Sweden

THE NORDIC O RTHOPAEDIC F EDERATION AND A CTA O RTHOPAEDICA

DEPUTY EDITOR

Peter A Frandsen Odense, Denmark CO-EDITORS

Per Aspenberg Linköping, Sweden Nils Hailer Uppsala, Sweden Ivan Hvid Oslo, Norway Urban Rydholm Lund, Sweden Bart A Swierstra Nijmegen, The Netherlands Eivind Witsø Trondheim, Norway Rolf Önnerfält Lund, Sweden

Li Felländer-Tsai Sweden Peter Frandsen Denmark Ragnar Jonsson Iceland Jan van Mourik The Netherlands Anders Rydholm Sweden

WEB EDITOR

Magnus Tägil Lund, Sweden S TATISTICAL EDITOR

Jonas Ranstam Lund, Sweden P RODUCTION MANAGER

Kaj Knutson Lund, Sweden

Vol. 89, No. 3, 2018

SUBSCRIPTION INFORMATION Acta Orthopaedica [print 1745-3674, online 1745-3682] is a peerreviewed journal, published six times a year plus supplements by Taylor & Francis on behalf of Nordic Orthopaedic Federation. Annual Institutional Subscription, Volume 89, 2018 $1,173 £725 €939 The subscription fee purchases an online subscription. The price includes access to current content and back issues to January 1997 (if available). Printed copies of the journal are provided on request as a free supplementary service accompanying an online subscription. Supplements to the journal are also included in the subscription price. For more information, visit the journal’s website: http://www.tandfonline.com/IORT Manuscripts should be uploaded at http://www.manuscriptmanager.com/ao/ for further handling at: Acta Orthopaedica Editorial Office, Department of Orthopaedics, Lund University Hospital, SE-221 85 Lund, Sweden Correspondance concerning copyright and permissions should be sent to: Håkan Pårup P.O. Box 3255, SE-103 65 Stockholm, Sweden, Tel: +46 (0)8 440 80 40. Fax: +46 (0)8 440 80 50. E-mail: hakan.parup@informa.com. Ordering information: Please contact your local Customer Service Department to take out a subscription to the Journal: USA, Canada: Taylor & Francis, Inc., 530 Walnut Street, Suite 850, Philadelphia, PA 19106, USA. Tel: +1 800 354 1420; Fax: +1 215 207 0050. UK/ Europe/Rest of World: T&F Customer Services, Informa UK Ltd, Sheepen Place, Colchester, Essex, CO3 3LP, United Kingdom. Tel: +44 (0) 20 7017 5544; Fax: +44 (0) 20 7017 5198; Email: subscriptions@tandf.co.uk Dollar rates apply to all subscribers outside of Europe. Euro rates apply to all subscribers in Europe except the UK and Republic of Ireland. If you are unsure which applies, contact Customer Services. All subscriptions are payable in advance and all rates include postage. Journals are sent by air to the USA, Canada, Mexico, India, Japan and Australasia. Subscriptions are entered on an annual basis, i.e., January to December. Payment may be made by sterling check, US dollar check, euro check, international money order, National Giro, or credit card (Amex, Visa and Mastercard). Back issues: Taylor & Francis retains a two-year back issue stock of journals. Older volumes are held by our official stockists to whom all orders and enquiries should be addressed: Periodicals Service Company, 351 Fairview Ave., Suite 300, Hudson, New York 12534, USA. Tel: +1 518 537 4700; fax: +1 518 537 5899; e-mail: psc@ periodicals.com.

Airfreight and mailing in the USA by agent named Air Business Ltd, c/o Worldnet Shipping Inc., 156-15, 146th Avenue, 2nd Floor, Jamaica, NY 11434, USA. Periodicals postage paid at Jamaica NY 11431. US Postmaster: Send address changes to Acta Orthopaedica, Air Business Ltd, c/o Worldnet Shipping Inc., 156-15, 146th Avenue, 2nd Floor, Jamaica, NY 11434, USA. Subscription records are maintained at Taylor & Francis Group, 4 Park Square, Milton Park, Abingdon, OX14 4RN, United Kingdom.

Copyright © 2018 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by-nc/3.0 . Informa UK Limited, trading as Taylor & Francis Group makes every effort to ensure the accuracy of all the information (the “Content”) contained in its publications. However, Informa UK Limited, trading as Taylor & Francis Group, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Informa UK Limited, trading as Taylor & Francis Group. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Informa UK Limited, trading as Taylor & Francis Group shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. Terms & Conditions of access and use can be found at http://www.tandfonline. com/page/terms-and-conditions Indexed/abstracted in: Allied and Complementary Medicine Library (Amed); ASCA (Automatic Subject Citation Alert); Biological Abstracts; Chemical Abstracts; Cumulative Index to Nursing and Allied Health Literature(CINAHL); Current Advances in Ecological and Environmental Sciences; Current Contents/Clinical Medicine; Current Contents/Life Sciences; Developmental Medicine and Child Neurology; Energy Research Abstracts; EMBASE/ Excerpta Medica; Faxon Finder; Focus On: Sports Science & Medicine; Health Planning and Administration; Index Medicus/MEDLINE; Index to Dental Literature; Index Veterinarius; INIS Atomindex; Medical Documentation Service; Nuclear Science Abstracts (Ceased); Periodicals Scanned and Abstracted. Life Sciences Collection; Research Alert; Science Citation Index; SciSearch; SportSearch; Uncover Veterinary Bulletin. Printed in England by Henry Ling

Acta Orthopaedica

ISSN 1745-3674

Vol. 89, No. 3, June 2018 Editorials Are competing risks models appropriate to describe implant failure? 20 years of porous tantalum in primary and revision hip arthroplasty—time for a critical appraisal Competing risks models Are competing risks models appropriate to describe implant failure? Hip Trabecular metal acetabular components in primary total hip arthroplasty. Higher risk for revision compared with other uncemented cup designs in a c ollaborative register study including 93,709 hips Does surgeon experience affect patient-reported outcomes 1 year after primary total hip arthroplasty? A register-based study of 6,713 cases in western Sweden No effect of double nerve block of the lateral cutaneous nerve and subcostal nerves in total hip arthroplasty: A randomized controlled trial Revision surgery of metal-on-metal hip arthroplasties for adverse reactions to metal debris: A clinical update Influence of surgical approach on complication risk in primary total hip arthroplasty: Systematic review and meta-analysis A randomized controlled trial on maximal strength training in 60 patients undergoing total hip arthroplasty: Implementing maximal strength training into clinical practice Bone mineral density changes in the graft after acetabular impaction bone grafting in primary and revision hip surgery: A 2-year prospective follow-up Projections of primary hip arthroplasty in Germany until 2040 Obesity-related metabolic and endocrine disorders diagnosed during postoperative follow-up of slipped capital femoral epiphysis Knee RSA migration of total knee replacements. A systematic review and meta-analysis The effect of surgeon’s preference for hybrid or cemented fixation on the long-term survivorship of total knee replacement: An analysis of 39,623 prostheses from the Australian Orthopaedic Association National Joint Replacement Registry Patient-reported symptoms and changes up to 1 year after meniscal surgery: An observational cohort study of 641 adult patients with a meniscal tear Shoulder and children High risk for revision after shoulder arthroplasty for failed osteosynthesis of proximal humeral fractures: A matched pair analysis of 285 cases from the Danish Shoulder Arthroplasty Registry 9 years’ follow-up of 168 pin-fixed supracondylar humerus fractures in children Infection and tendon experiment Low sensitivity of α-defensin (Synovasure) test for intraoperative exclusion of prosthetic joint infection Negative effect of zoledronic acid on tendon-to-bone healing: In vivo study of biomechanics and bone remodeling in a rat model Information to authors

253 254

J Ranstam N Hailer

256

A Sayers, J T Evans, M R Whitehouse, and A W Blom

259

I Laaksonen, M Lorimer, K Gromov, A Eskelinen, O Rolfson, S E Graves, H Malchau, and M Mohaddes

265

P Jolbäck, O Rolfson, M Mohaddes, S Nemes, J Kärrholm, G Garellick, and H Lindahl

272

J L Bron, J Verhart, I N Sierevelt, D De Vries, H J Kingma, and M V Rademakers

278

G S Matharu, A Eskelinen, A Judge, H G Pandit, and D W Murray L E Miller, J S Gondusky, A F Kamath, F Boettner, J Wright, and S Bhattacharyya S B Winther, O A Foss, O S Husby, T S Wik, J Klaksvik, and V S Husby

289 295 302

D M J M Gerhardt, E De Visser, B W Hendrickx, B W Schreurs, and J L C van Susante

308 314

V Pilz, T Hanstein, and R Skripitz H Ucpunar, I Y Camurcu, S Duman, E Ucpunar, H Sofu, and A I Bayhan

320

B G Pijls, J W M Plevier, and R G H H Nelissen

329

C J Vertullo, S E Graves, Y Peng, and P L Lewis

336

S T Skou, K Pihl, N Nissen, U Jørgensen, and J B Thorlund

345

M R Kristensen, J V Rasmussen, B Elmengaard, S L Jensen, B S Olsen, and S Brorson

351

N Tuomilehto, A Sommarhem, and A Y Nietosvaara

357

R Scholten, J Visser, J L C van Susante, and C J M van Loon

360

G A Hjorthaug, E Søreide, L Nordsletten, J E Madsen, F P Reinholt, S Niratisairak, and S Dimmen

367

PER ASPENBERG BENBOKEN

tan ben i kroppen skulle vi inte klara oss! Då skulle vi bara vara en formlös klump. Ben är e klara oss! Då skulle vi bara vara tommenfastän i vårade kroppar, men i våra skroppar, flesta fastän de flesnde förhållande en sällan tanke. ägnar detta avgörande ta av oss sto mer på ben är läkaren och förhållande en tanke. berg. I BENBOKEN delar han med Någon som däremot tänkt desto yrkesliv i ”benets tjänst” och av mer påsom beninte är läkaren stiska material, bara göroch professorn den unikai ortopedi förmåganPer att Aspenberg. återskapa I Benboken delar han med sig av sina erfarenheter de reder författaren ut många från ett yrkesliv i ”benets tjänst” och professionella och lekmän: av sin fascination inför detta fantastiska material, som inte bara gör oss rer? raka i ryggen utan också har den unika och vilka ska inte göra det? förmågan att återskapa sig självt om det ? skadas. viktigt? Pedagogiskt och underhållande ka metoder för skapa pålitlig rederatt författaren ut många frågor, som och, inte minst, om den nya forskkan intressera både professionella och a på frakturläkning. lekmän: Aspenberg är läkare senior – Vad och är ben? essor i ortopedi vid Linköpings – Varför läker inte vissa frakturer? ersitet. Han forskar med stöd av – Vilka frakturer ska opereras och nskapsrådet, är författare till över vilkaartiklar, ska inte göra det? vetenskapliga vetenskap– Vad innebär daktör och mottagare av fleraen stressfraktur? nationella utmärkelser. – Varför är inte benskörhet så viktigt? Benboken handlar också om olika metoder för att skapa pålitlig kunskap, om briljanta forskare och, inte minst, om den nya forskning som kan komma att snabba på frakturläkning.

Per Aspenberg

BENBOKEN Om frakturer och forskning

K ARNE VAL FÖRL AG

»Boken är elegant skriven, illustrerad och redigerad ... en imponerande översikt som, ovanligt nog, med stor behållning kan läsas av flera olika läsekretsar. Den kan varmt rekommenderas till så skilda grupper som gymnasieelever som vill förstå ben, en populärvetenskapligt intresserad allmänhet, studenter på grundutbildningar, intresserade läkare som vill hålla sig uppdaterade om ben, men också till ortopeder som vill uppdatera sina kunskaper om benvävnad och frakturläkning.« olle svensson i läkartidningen

PER ASPENBERG: BENBOKEN . Inbunden. 159 sidor. Karneval förlag 2018. ISBN 978-91-88729-05-7

Beställ i bokhandeln eller direkt från förlaget: www.karnevalforlag.se

Acta Orthopaedica 2018; 89 (3): 253

253

Editorial

Are competing risks models appropriate to describe implant failure?

In this issue of Acta Orthopaedica, Sayers et al. raise the issue of whether competing risks models are appropriate to describe implant failure. The question is relevant because the number of publications presenting competing risks models is increasing rapidly. A PubMed search with the criteria “arthroplasty” and “competing risk” reveals that while no such paper was published before 2000, all but 2 of 46 published papers were published in 2010 or later, and of these not less than 17 were published during 2017 alone. It should thus not come as a surprise if many more papers using competing risks models are published during the coming years. What is a competing risks model, and why is this suddenly so interesting? First, implant survival has traditionally been analyzed using the Kaplan–Meier method. This allows inclusion also of observations with incomplete follow-up in the analysis; these are known as censored observations. The analysis is, however, problematic in the presence of competing events (such as death) that preclude the studied event (implant failure). This changes the interpretation of the failure-rate estimate. An alternative technique that accounts for the competing event, a competing risks model, is then often recommended. Sayers et al. have recently noticed several publications in which the differences in results between the Kaplan–Meier method and competing risks models have been misinterpreted. To explain the problem they describe two different measures, net and crude failure: the former is calculated using the Kaplan–Meier method and the latter with a competing risks model.

Both estimates can be useful in arthroplasty studies and both provide, in the absence of confounding and selection bias, unbiased estimates. The main difference between the two measures is that the net failure refers to a hypothetical existence in which competing risks are assumed to be eliminated. In real life, the (crude) failure rate is lower for elderly patients as they are more likely to be excluded from failure because of the competing risk of dying. Which estimate should be presented in scientific publications on implant failure and annual reports from arthroplasty registers? This depends on the application of interest. Net failure is the relevant measure when comparing the failure rates of different implants. It would clearly not be reasonable to include effects of patient survival in this comparison. Crude failure, on the other hand, is the relevant measure if patient survival is part of the problem, as for example when studying health economics and planning resources. An important aspect of the paper by Sayers et al. is that it draws attention to the general importance of recognizing why a study is performed and to the limitations in using the results for other purposes. Knowledge of a study’s aim, design, and analysis is usually necessary for a correct interpretation of its findings.

Jonas Ranstam Statistics Editor email: jonas.ranstam@med.lu.se Sayers A, Evans J T, Whitehouse M R, Blom A W. Are competing risk models appropriate to describe implant failure? Acta Orthop 2018; 89 [Ahead of print] DOI: 10.1080/17453674.2018.1444876

© 2018 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https://creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2018.1452470

254

Acta Orthopaedica 2018; 89 (3): 254–255

Editorial

20 years of porous tantalum in primary and revision hip arthroplasty—time for a critical appraisal After almost 2 decades of clinical use, porous tantalum has become an important component in the orthopedic surgeon’s toolbox, and its most common use today is as acetabular cups, cages, or augments in revision arthroplasty. For many revision surgeons, porous tantalum has significantly changed their practice when addressing more complex acetabular bone loss. However, now porous tantalum is also increasingly used as a coating for cups used in primary total hip arthroplasty (THA). It is time for a critical appraisal of porous tantalum in hip arthroplasty, and this issue of Acta Orthopaedica contains an article that investigates the use of porous tantalum cups in primary THA (Laaksonen et al. 2018). Porous tantalum in revision THA Tantalum was first described in 1802 by the Swedish chemist Anders Ekeberg, who isolated it from the rare mineral tantalite. In modern times, pure tantalum has been manufactured into 3-dimensional, macroporous structures that closely resemble cancellous bone, so-called “trabecular metal” (TM). Its mechanical and biological properties make porous tantalum well suited for orthopedic applications: • Mechanically, a fairly unique combination of high elasticity and a high coefficient of friction (Levine et al. 2006) gives porous tantalum implants their much appreciated “grip” in situations where primary stability is critical but difficult to achieve, such as in acetabular revision surgery with a high degree of bone loss and sclerotic acetabular rims. • Biologically, tantalum and titanium share a very pronounced thrombogenic potential, a property that may be important since blood clot formation is one of the first and essential steps in bone healing, and this distinguishes tantalum and titanium from other metals such as nickel (Hong et al. 2005). Furthermore, porous tantalum in the shape of “trabecular metal” has a very high porosity of 75–85%, and its pore size of more than 500 µm is also relatively large (Balla et al. 2010). These properties may at least in part explain the osteoconductivity that is observed in different in vivo models investigating the integration of porous tantalum into host bone (Levine et al. 2006). The first clinical reports on porous tantalum shells were on patients with technically demanding acetabular defects (Sporer and Paprosky 2006, Flecher et al. 2008). Further applications included the combination of porous tantalum shells with differently shaped augments or cages that are manufactured from

the same material, again for use in large acetabular defects or even in pelvic discontinuity (Abolghasemian et al. 2014). Comparative retrospective studies indicate that porous tantalum cups used in revision situations are at least as good as conventional uncemented titanium cups (Jafari et al. 2010, Mohaddes et al. 2015), and that they can be superior to Müller acetabular reinforcement rings in terms of a reduced risk of aseptic loosening (Brüggemann et al. 2017). On the other hand, dislocation seems to be a recurrent issue after the use of porous tantalum shells in revision THA (Skytta et al. 2011, Brüggemann et al. 2017), a problem that has not been investigated in depth. This instability could be related to difficulties in restoring a correct center of rotation‚ or in less than optimal abduction and anteversion angles, possibly related to the strategy of “going for bone” as opposed to reconstructing the acetabular bed, as in the old-fashioned techniques. The combination of dual-mobility cups with porous tantalum shells may reduce the problem of instability (Brüggemann et al. 2018), but this hypothesis needs further investigation. Porous tantalum cups in primary THA The next step of using porous tantalum as a cup material even in primary arthroplasty surgery was both logical and tempting, and initial medium-term reports showed good results (Malizos et al. 2008, Macheras et al. 2009). Subsequently, 2 randomized controlled trials compared porous tantalum with hemispherical titanium cups in primary THA. When measured by radiostereometry, porous tantalum cups had slightly better initial stability than titanium fiber-mesh cups, but they were accompanied by a similar degree of periprosthetic bone loss (Baad-Hansen et al. 2011). A randomized controlled trial comparing porous tantalum monoblock with porous-coated titanium monoblock cups found fewer radiolucencies around the porous tantalum cups (Wegrzyn et al. 2015), but of course none of these trials was designed or powered to detect differences in the risk of revision. Large-scale registry studies provide the opportunity to investigate survival data derived from large cohorts, and— unexpectedly—both the Australian (Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) 2017) and the Swedish arthroplasty registries (Swedish Hip Arthroplasty Register 2016) repeatedly reported a higher-than-expected revision risk for porous tantalum cups

© 2018 The Author(s). Published by Taylor & Francis on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License (https: //creativecommons.org/licenses/by/4.0) DOI 10.1080/17453674.2018.1463007

02-EDHailer18.indd 254

03-05-2018 11:20:23

Acta Orthopaedica 2018; 89 (3): 254–255

that were used in primary THA, in part due to an increased risk of dislocation. In contrast, a recently published registry study from the England and Wales National Joint Registry that compares porous tantalum-coated cups with titanium-coated cups from the same manufacturer finds a lower risk of revision due to aseptic loosening in the group of patients operated with porous tantalum cup (Matharu et al. 2018). Authors from the Australian and Swedish joint registries have now collaborated in a joint effort to compare the outcome after the use of porous tantalum cups in primary THA with that of other common uncemented cups. In the present issue of Acta Orthopaedica, Laaksonen et al. (2018) report a 1.5-fold higher risk of revision for the porous tantalum-coated cups. Porous tantalum—what needs to be done? Porous tantalum cups undoubtedly confer highly desirable effects in terms of excellent initial and long-term stability in demanding acetabular revision surgery, and they have certainly come to stay. Further studies are needed to investigate whether a combination of porous tantalum shells with dual-mobility systems reduces instability after complex revision procedures. However, what is good in revision surgery is not necessarily superior in standard primary THA surgery. Could it be that— in the context of primary THA—the terrific grip of porous tantalum cups makes them more likely to jam in suboptimal positions, maybe retroverted, or steeper than intended, but that the surgeon refrains from correcting their position simply because they are difficult to get out? This is entirely speculative, and a systematic analysis of the mechanisms underlying the potentially increased risk of dislocation after the use of porous tantalum cups has not yet been done. The orthopedic community will have to keep an eye on the evolving evidence in order to critically assess whether we should continue to be as tantalized by tantalum in primary THA as we are in complex revision situations. Nils Hailer Co-Editor Orthopaedics/Department of Surgical Sciences Uppsala University Hospital nils.hailer@surgsci.uu.se

Abolghasemian M, Tangsaraporn S, Drexler M, Barbuto R, Backstein D, Safir O, Kuzyk P, Gross A. The challenge of pelvic discontinuity: cup-cage reconstruction does better than conventional cages in mid-term. Bone Joint J 2014; 96-B: 195-200.

02-EDHailer18.indd 255

255

Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR). Annual report; 2017. Available from: http: //www.aoa.org.au Baad-Hansen T, Kold S, Nielsen P T, Laursen M B, Christensen P H, Soballe K. Comparison of trabecular metal cups and titanium fiber-mesh cups in primary hip arthroplasty: a randomized RSA and bone mineral densitometry study of 50 hips. Acta Orthop 2011; 82: 155-60. Balla V K, Bodhak S, Bose S, Bandyopadhyay A. Porous tantalum structures for bone implants: fabrication, mechanical and in vitro biological properties. Acta Biomater 2010; 6(8): 3349-59. Brüggemann A, Fredlund E, Mallmin H, Hailer N P. Are porous tantalum cups superior to conventional reinforcement rings? Acta Orthop 2017; 88: 35-40. Brüggemann A, Mallmin H, Hailer N P. Do dual-mobility cups cemented into porous tantalum shells reduce the risk of dislocation after revision surgery? Acta Orthop 2018; 89(2): 156–62. Flecher X, Sporer S, Paprosky W. Management of severe bone loss in acetabular revision using a trabecular metal shell. J Arthroplasty 2008; 23: 949-55. Hong J, Azens A, Ekdahl K N, Granqvist C G, Nilsson B. Material-specific thrombin generation following contact between metal surfaces and whole blood. Biomaterials 2005; 26(12): 1397-403. Jafari S M, Bender B, Coyle C, Parvizi J, Sharkey P F, Hozack W J. Do tantalum and titanium cups show similar results in revision hip arthroplasty? Clin Orthop Relat Res 2010; 468: 459-65. Laaksonen I, Lorimer M, Gromov K, Eskelinen A, Rolfson O, Graves S E, Malchau H, Mohaddes M. Trabecular metal acetabular components in primary total hip arthroplasty. Acta Orthop 2018; 89(3): 259-64. Levine B R, Sporer S, Poggie R A, Della Valle C J, Jacobs J J. Experimental and clinical performance of porous tantalum in orthopedic surgery. Biomaterials 2006; 27(27): 4671-81. Macheras G, Kateros K, Kostakos A, Koutsostathis S, Danomaras D, Papagelopoulos P J. Eight- to ten-year clinical and radiographic outcome of a porous tantalum monoblock acetabular component. J Arthroplasty 2009; 24: 705-9. Malizos K N, Bargiotas K, Papatheodorou L, Hantes M, Karachalios T. Survivorship of monoblock trabecular metal cups in primary THA: midterm results. Clin Orthop Relat Res 2008; 466: 159-66. Matharu G S, Judge A, Murray D W, Pandit H G. Trabecular metal acetabular components reduce the risk of revision following primary total hip arthroplasty: a propensity score matched study from the National Joint Registry for England and Wales. J Arthroplasty 2018; 33(2): 447-52. Mohaddes M, Rolfson O, Karrholm J. Short-term survival of the trabecular metal cup is similar to that of standard cups used in acetabular revision surgery. Acta Orthop 2015; 86: 26-31. Skytta E T, Eskelinen A, Paavolainen P O, Remes V M. Early results of 827 trabecular metal revision shells in acetabular revision. J Arthroplasty 2011; 26: 342-5. Sporer S M, Paprosky W G. Acetabular revision using a trabecular metal acetabular component for severe acetabular bone loss associated with a pelvic discontinuity. J Arthroplasty 2006; 21: 87-90. Swedish Hip Arthroplasty Register. Annual report 2016. Available from: https: //shpr.registercentrum.se/shar-in-english/annual-reports-from-theswedish-hip-arthroplasty-register/p/rkeyyeElz Wegrzyn J, Kaufman K R, Hanssen A D, Lewallen D G. Performance of porous tantalum vs. titanium cup in total hip arthroplasty: randomized trial with minimum 10-year follow-up. J Arthroplasty 2015; 30(6): 1008-13.

03-05-2018 11:20:24

256

Acta Orthopaedica 2018; 89 (3): 256–258

Are competing risks models appropriate to describe implant failure? Adrian SAYERS 1, 2, Jonathan T EVANS 1, Michael R WHITEHOUSE 1,3, and Ashley W BLOM 1,3

1 Musculoskeletal Research Unit, Bristol Medical School, Southmead Hospital, Bristol, 2 Population Health 3 National Institute for Health Research Bristol Biomedical Research Centre, University of Bristol, UK

Sciences, Bristol Medical School, Bristol,

Correspondence: adrian.sayers@bristol.ac.uk Submitted 2017-12-08. Accepted 2018-02-06.

Background and purpose — The use of competing risks models is widely advocated in the arthroplasty literature due to a perceived bias in comparison of simple Kaplan–Meier estimates. Proponents of competing risk models in the arthroplasty literature appear to be unaware of the subtle but important differences in interpretation of net and crude failure estimated by competing risk and Kaplan–Meier methods respectively. Methods — Using a simple simulation we illustrate the differences between competing risks and Kaplan–Meier methods. Results — Competing risk and Kaplan–Meier methods estimate different survival quantities, i.e., crude and net failure respectively. Estimates of crude failure estimated using competing risk methods will be less than net failure as estimated using Kaplan–Meier methods. Interpretation — Kaplan–Meier methods are appropriate for describing implant failure, whereas crude survival estimated using competing risk methods estimates the risk of surgical revision as it depends on both implant failure and mortality. Both competing risk models and Kaplan–Meier methods are useful in arthroplasty, and both provide unbiased estimates of crude and net failure in the absence of any confounding or selection respectively. Surgeons and researchers should carefully consider whether the use of competing risks is always justified. Lower estimates of failure from competing risk models may be misleading to surgeons who are attempting to select the best implants with the lowest failure rates for their patients. ■

We have recently noticed a number of incidences in the arthroplasty literature of authors espousing the benefits of using competing risk models in preference to Kaplan–Meier (KM) estimates to describe the failure of implants due to a perception that the observed high mortality rates in elderly patients may lead to biased estimates using the KM method (Biau et

al. 2007, Fennema and Lubsen 2010, Keurentjes et al. 2012, Lacny et al. 2015, Porcher 2015, Wongworawat et al. 2015, Martin et al. 2016, Lampropoulou-Adamidou et al. 2017). This recent trend is somewhat worrying as we believe there is a fundamental misinterpretation of what Kaplan–Meier (KM) (Kaplan and Meier 1958) or competing risks (CR) (Coviello and Boggess 2004) models estimate, and under which circumstances each method may be preferable. To correct this misunderstanding, we describe a simple simulation in a hypothetical situation with immortal patients, where no individuals are ever lost to follow-up. Figure 1 panel (a) illustrates this process using a line plot which illustrates when a patient becomes at risk and when a failure occurs and exits the study. In this situation, it is very easy to estimate implant survival at a time of interest, i.e., it is simply the proportion of those who fail. The numerator is the number of failures, and the denominator is the number of patients implanted. A simple proportion, KM estimates (Kaplan and Meier 1958), and the cumulative incidence function (CIF) (Coviello and Boggess 2004) from a CR model will give identical estimates. This scenario is the ideal scenario, as we need not concern ourselves with problems such as censoring (loss to follow-up or mortality), and we describe these estimates of failure as net failure, using the terminology of Lambert et al. (2010). However, some researchers are under the misguided belief that this hypothetical situation is the only scenario in which the KM estimator is appropriate (Biau et al. 2007). The title of Kaplan and Meier’s (1958) seminal work, “NonparametricEstimation from Incomplete Observations,” gives us a clue to why this is incorrect. The KM method was specifically developed to allow incomplete observations due to non-informative right censoring, i.e., individuals cease to be at risk of failure, but have not failed where the reason that they cease to be at risk is completely independent of the cause of failure.

12396 Sayers D.indd 256

5/3/2018 11:31:38 AM

Acta Orthopaedica 2018; 89 (3): 256–258

Patient F = Failure F

1 2

1 7

Time

0.2

F D 0

0.3

Time

0.4

D D

1 5

0.5

0.6

2 F

0.7

F D

KM immortal cohort KM mortal cohort CIF mortal cohort

0.1

F 1

1 0.9 0.8

F F

4 F

D D

F = Failure D = Death

7 D

Survival D

10 9

D = Death

10 9

Patient

257

Time

Observed primary “at risk” Observed revision “at risk” Unobserved primary “at risk”

Time

Figure 1. Panel (a) is a line plot that illustrates the time at risk of 10 patients entering a study following arthroplasty (time 0) and exiting the study after failure where the only possible mechanism of exiting the study is failure, i.e., no other cause of censoring occurs. Panel (b) is a line plot that illustrates a non-informative mortality profile of the same 10 patients entering a study following arthroplasty (time 0).

Figure 2. A line plot that illustrates the time at risk of 10 patients entering a study following arthroplasty (time 0) and the combination of a failure and mortality mechanism, i.e., mortal patients.

In arthroplasty failure studies, mortality is one possible cause of being censored. Figure 1 panel (b) illustrates a noninformative mortality profile of patients in Figure 1 panel (a). In this more complex and alternate situation with mortal patients, the failure process is more difficult to estimate due to the presence of a mortality process. This additional process removes patients from the study and calculation of failure becomes more complex—see Figure 2 which overlays the failure and mortality processes. Due to the complexity of this alternate situation with mortal patients, we are confronted with a choice of what to estimate. We can attempt to recover an estimate of net failure, which gives us an estimate of the failure of the implant, i.e., the failure estimate from the immortal cohort. Or, we can estimate crude failure, which represents the likely number of failures we see in practice, i.e., it is a composite of both the failure of the implants and the mortality process. The terminology used in this field is somewhat heterogeneous, therefore we use the terminology described by Lambert et al. (2010). Standard methods of conducting survival analysis, i.e., KM or Cox regression focus on net failure, are based solely on the hazard profile of the cause of interest. Competing risk methods estimate crude failure and depend on both the hazard of the event of interest and the hazard of the competing event. The differences in the KM estimate with immortal patients and mortal patients and the CIF (competing risks estimate) with mortal patients is presented in Figure 3. Here, we simply create 2 independent random uniform failure profiles between 0 and 10 years for 2 processes, (1) implant failure, and (2) mortality for 1,000 patients. Analysis of implant failure of immortal patients, ignoring the mortality process, can be considered the “truth,” and removing patients from the risk set due to a mortality event creates a mortal cohort, i.e., the observed. We expect the failure to be 100% at 10 years, and a straight line from 0 years to 10 years, i.e., a 45-degree line.

This clearly illustrates the CIF (competing risks estimate) is not the same as that of KM. It is a biased estimate of net failure, but an unbiased estimate of crude failure. Whilst the simulation is extreme, i.e., everyone fails and everyone dies, the results will hold in all circumstances that the censoring is non-informative. The degree to which the CIF is different from the KM profile depends on the mortality process. Prior to the first mortality event, KM and CIF are equal, and only following the first mortality event do they become unequal. In arthroplasty research differences between KM and CIF are likely to be more evident in series with long-term follow up, where mortality is inevitably higher, or in series with elderly or frail patients. These differences are well known to those with a methodological interest in survival analysis. For example, Gooley et al. (1999) note that if one is interested in evaluating a causespecific failure, the CIF may be misleading and inferences should be made from functions which are based solely on the hazard of failure from the cause of interest, i.e., use the KM estimator. Putter et al. (2007) similarly state that the “naive Kaplan–Meier estimator describes what would happen if the competing event could be prevented to occur, creating an imaginary world in which an individual remains at risk of failure from the event of interest,” i.e., an immortal patient cohort. Ranstam et al. (2011) describe this in an arthroplasty setting as the “implicit assumption that the patient will be alive until the implant fails.” Recently, we have similarly illustrated this result using a simulation study in the context of prosthesis benchmarking: we illustrate that KM provides unbiased estimates of net failure and provide nominal coverage, i.e., the confidence interval includes the true value on 95% of occasions (Sayers et al. 2017). In as far as we currently know, the mortality process is independent of whether implants are revised or not, i.e., mortality satisfies the non-informative censoring assumption. Our

12396 Sayers D.indd 257

Figure 3. KM survival curves and the 1 minus the cumulative incidence function in mortal and immortal cohorts.

5/3/2018 11:31:39 AM

258

belief in this assumption is based on the observation that even when an implant or group of implants fail in a large number of patients, e.g., metal on metal, this is not associated with any increase in pathologies, in the short term, such as cancer that in turn may lead to an excess of mortality (Smith et al. 2012a, 2012b, 2012c). However, it is important these assumptions are checked periodically; an absence of evidence is not evidence of absence, and future information may require analyses to be modified to account for an informative censoring profile. Simply, competing risk methods and non-competing risk methods estimate different quantities, and which quantity you should use depends on your application of interest. If you are interested in describing the failure of an implant, comparing the failure rate of a group of implants, looking for outliers, i.e., from a regulatory perspective, or attempting to select an implant for use that has the greatest longevity, you need estimates of net failure (KM). If you are interested in resource planning, health economics, or communicating with patients their likely chance of experiencing a revision, estimates of crude failure (CR) are more likely to be desirable. Just because the estimate of net implant failure is higher than crude failure does not mean they are not correct or desirable in many circumstances in arthroplasty. However, it also important to remember that whilst KM and the CIF are statistically unbiased estimates for net and crude failure respectively, they are both equally likely to display bias in the presence of confounding factors and selection effects, and simply choosing the appropriate approach is not a panacea against this immutable problem. Funding and conflict of interest AS was supported by a MRC strategic skills fellowship: MRC Fellowship MR/L01226X/1. JTE was supported by the National Joint Registry of England, Wales, Northern Ireland and the Isle of Man and Royal College of Surgeons of England Fellowship. This study was supported by the NIHR Biomedical Research Centre at the University Hospitals Bristol NHS Foundation Trust and the University of Bristol. The views expressed in this publication are those of the authors and not necessarily those of the NHS, the National Institute for Health Research, or the Department of Health. We have no competing interests to declare. See also Editorial in the June 2018 issue of Acta Orthopaedica.

Acta thanks Nicole Pratt and other anonymous reviewers for help with peer review of this study.

AS, JTE, MRW, AWB conceived the manuscript, interpreted data from simulation, and approved the final version of the manuscript. AS wrote the first draft and performed the simulation. JTE and AS reviewed the literature.

12396 Sayers D.indd 258

Acta Orthopaedica 2018; 89 (3): 256–258

Biau D J, Latouche A, Porcher R. Competing events influence estimated survival probability: when is Kaplan–Meier analysis appropriate? Clin Orthop Relat Res 2007; 462: 229-33. doi: 10.1097/BLO.0b013e3180986753. Coviello V, Boggess M. Cumulative incidence estimation in the presence of competing risks. Stata J 2004; 4(2): 103-11. Fennema P, Lubsen J. Survival analysis in total joint replacement: an alternative method of accounting for the presence of competing risk. J Bone Joint Surg Br 2010; 92(5): 701-6. doi: 10.1302/0301-620X.92B5.23470. Gooley T A, Leisenring W, Crowley J, Storer B E. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med 1999; 18(6): 695-706. Kaplan E L, Meier P. Nonparametric-estimation from incomplete observations. J Am Stat Assoc 1958; 53(282): 457-81. doi: 10.2307/2281868. Keurentjes J C, Fiocco M, Schreurs B W, Pijls B G, Nouta K A, Nelissen R G. Revision surgery is overestimated in hip replacement. Bone Joint Res 2012; 1(10): 258-62. doi: 10.1302/2046-3758.110.2000104. Lacny S, Wilson T, Clement F, Roberts D J, Faris P D, Ghali W A, Marshall D A. Kaplan–Meier survival analysis overestimates the risk of revision arthroplasty: a meta-analysis. Clin Orthop Relat Res 2015; 473(11): 343142. doi: 10.1007/s11999-015-4235-8. Lambert P C, Dickman P W, Nelson C P, Royston P. Estimating the crude probability of death due to cancer and other causes using relative survival models. Statistics in Medicine 2010; 29(7-8): 885-95. doi: 10.1002/ sim.3762. Lampropoulou-Adamidou K, Karachalios TS, Hartofilakidis G. Overestimation of the risk of revision with Kaplan-Meier presenting the long-term outcome of total hip replacement in older patients. Hip Int 2017; [Epub ahead of print]. doi: 10.5301/hipint.5000575. Martin C T, Callaghan J J, Gao Y B, Pugely A J, Liu S S, Warth L C, Goetz D D. What can we learn from 20-year followup studies of hip replacement? Clin Orthop Relat Res 2016; 474(2): 402-7. doi: 10.1007/s11999015-4260-7. Porcher R. CORR Insights((R)): Kaplan–Meier survival analysis overestimates the risk of revision arthroplasty: a meta-analysis. Clin Orthop Relat Res 2015; 473(11): 3443-5. doi: 10.1007/s11999-015-4291-0. Putter H, Fiocco M, Geskus R B. Tutorial in biostatistics: competing risks and multi-state models. Stat Med 2007; 26(11): 2389-430. doi: 10.1002/ sim.2712. Ranstam J, Karrholm J, Pulkkinen P, Makela K, Espehaug B, Pedersen A B, Mehnert F, Furnes O, NARA study group. Statistical analysis of arthroplasty data, II: Guidelines. Acta Orthop 2011; 82(3): 258-67. doi: 10.3109/17453674.2011.588863. Sayers A, Crowther M J, Judge A, Whitehouse M R, Blom A W. Determining the sample size required to establish whether a medical device is noninferior to an external benchmark. BMJ Open 2017; 7(8): e015397. doi: 10.1136/bmjopen-2016-015397. Smith A J, Dieppe P, Howard P W, Blom A W, National Joint Registry for England and Wales. Failure rates of metal-on-metal hip resurfacings: analysis of data from the National Joint Registry for England and Wales. Lancet 2012a; 380(9855): 1759-66. doi: 10.1016/S0140-6736(12)60989-1. Smith A J, Dieppe P, Porter M, Blom A W, National Joint Registry of England and Wales. Risk of cancer in first seven years after metal-on-metal hip replacement compared with other bearings and general population: linkage study between the National Joint Registry of England and Wales and hospital episode statistics. BMJ 2012b; 344: e2383. doi: 10.1136/bmj.e2383. Smith A J, Dieppe P, Vernon K, Porter M, Blom A W, National Joint Registry of England and Wales. Failure rates of stemmed metal-on-metal hip replacements: analysis of data from the National Joint Registry of England and Wales. Lancet 2012c; 379(9822): 1199-204. doi: 10.1016/S01406736(12)60353-5. Wongworawat M D, Dobbs M B, Gebhardt M C, Gioe T J, Leopold S S, Manner P A, Rimnac C M, Porcher R. Editorial: Estimating survivorship in the face of competing risks. Clin Orthop Relat Res 2015; 473(4): 1173-6. doi: 10.1007/s11999-015-4182-4.

5/3/2018 11:31:42 AM

Acta Orthopaedica 2018; 89 (3): 259–264

259

Trabecular metal acetabular components in primary total hip arthroplasty Higher risk for revision compared with other uncemented cup designs in a collaborative register study including 93,709 hips Inari LAAKSONEN 1,2, Michelle LORIMER 3, Kirill GROMOV 4, Antti ESKELINEN 5, Ola ROLFSON 6, Stephen E GRAVES 3, Henrik MALCHAU 1,2,6, and Maziar MOHADDES 6

1 Harris

Orthopedic Laboratory, Massachusetts General Hospital, Boston, USA; 2 Harvard Medical School, Boston, USA; 3 Australian Orthopaedic Association National Joint Replacement Registry, Adelaide, Australia; 4 Department of Orthopaedic Surgery, Copenhagen University Hospital Hvidovre, Copenhagen, Denmark; Danish Hip Arthroplasty Register, Aarhus, Denmark; 5 Coxa Hospital for Joint Replacement, Tampere, Finland; Finnish Arthroplasty Register, Helsinki, Finland; 6 Swedish Hip Arthroplasty Register, Department of Orthopaedics, Institute of Surgical Sciences, Sahlgrenska University Hospital, Gothenburg, Sweden Correspondence: hmalchau@partners.org Submitted 2017-04-06. Accepted 2017-11-25.

Background and purpose — Trabecular metal (TM) cups have demonstrated favorable results in acetabular revision and their use in primary total hip arthroplasty (THA) is increasing. Some evidence show that TM cups might decrease periprosthetic infection (PPI) incidence. We compared the survivorship of TM cups with that of other uncemented cups in primary THA, and evaluated whether the use of TM cups is associated with a lower risk of PPI. Patients and methods — 10,113 primary THAs with TM cup and 85,596 THAs with other uncemented cups from 2 high-quality national arthroplasty registries were included. The mean follow-up times were 3.0 years for the TM cups and 3.8 years for the other uncemented cups. Results — The overall survivorship up to 8 years for TM cups and other uncemented cups was 94.4% and 96.2%, respectively (p = < 0.001). Adjusting for relevant covariates in a Cox regression model the TM cups had a persistently higher revision risk than other uncemented cups (HR = 1.5, 95% CI 1.4–1.7, p = < 0.001). There was a slightly higher, though not statistically significant, revision rate for PPI in the TM group (1.2, 95% CI 1.0–1.6, p = 0.09). Interpretation — Risk of revision for any reason was higher for the TM cup than for other uncemented cups in primary THA. In contrast to our hypothesis, there was no evidence that the revision rate for PPI was lower in the TM cup patients. Regardless of the promising early and mid-term results for TM cups in hip revision arthroplasty, we would like to sound a note of caution on the increasing use of the TM design, especially in uncomplicated primary THAs, where uncemented titanium cups are considered to provide a reliable outcome. ■

Trabecular metal (ZimmerBiomet, Warsaw, IN, USA) is made of porous tantalum and has been shown to provide higher porosity, increased initial stability, and good bone ingrowth qualities (Bobyn et al. 1999, Beckmann et al. 2014). These advantages make it an attractive option in both primary and revision THA. While TM acetabular components are most commonly used in revision THA to manage poor bone quality and acetabular bone defects (Jafari et al. 2010, Kremers et al. 2012, Mohaddes et al. 2015), TM cups have also shown excellent results in primary THA (Baad-Hansen et al. 2011, Howard et al. 2011). However, there is a lack of data on whether TM cups are indeed a more reliable option for primary THA compared to other uncemented cups. In addition a recent study suggested that the use of TM acetabular components in hip revision arthroplasty might be protective against subsequent failure due to periprosthetic infection (PPI) (Tokarski et al. 2015). PPI is a devastating complication following THA. Infection rates around 1% after primary THA have been reported by major national registries (Lindgren et al. 2014, Gundtoft et al. 2015, Huotari et al. 2015). The total number of primary THAs is increasing and, in addition to this increase, studies have shown that the risk for infection has been increasing as well over recent decades (Dale et al. 2009, 2012). There is a lack of data on whether TM cups are a more reliable option for primary THA compared with other uncemented cups. A recent study suggested that the use of TM acetabular components in hip revision arthroplasty might be protective against subsequent failure due to infection (Tokarski et al. 2015); however, the effect of using a TM cup on infection rates following primary THA also remains unknown. The purpose of this collaborative registry study was to: (a) determine the overall revision rate of TM acetabular

11545 Malchau D.indd 259

5/3/2018 11:33:05 AM

260

Acta Orthopaedica 2018; 89 (3): 259–264

Table 1. Trabecular metal cups and 5 other most commonly used uncemented cups in primary THA in AOANJRR and SHAR. Values are frequency and (%) Cup design

SHAR

AOANJRR

Trident Pinnacle Trilogy Reflection Allofit Exceed TM Continuum Shell

3,390 (21) 1,113 (7) 6,972 (44)

33,704 (43) 18,005 (23) 9,335 (12) 4,165 (5) 5,231 (7)

889 (6) 792 (5) 1,979 (12) 817 (5)

3,858 (5) 3,449 (4)

components used in primary THA and to compare it with that of other frequently used uncemented cups; (b) investigate whether the use of a TM cup in primary THA will decrease the risk of early revision due to infection compared with other uncemented cup designs.

Patients and methods Data for this collaborative registry study were collected from the Swedish Hip Arthroplasty Register (SHAR) and the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR). SHAR has been collecting data on total hip replacements since 1979 and currently has information on more than 300,000 primary hip replacements. In Sweden all orthopedic units performing hip arthroplasties report to the SHAR. The completeness of this register has been reported as 99% in primary THA. Descriptive surgical data are com-

AOANJRR Uncemented cups reported during years 2006–2014 n = 213,314 Excluded Not metal on X-poly n = 120,919 Metal on X-poly n = 92,395

Tantalum cups n = 7,317

pleted on standard forms by the responsible surgical team at each center. Several validation steps are performed on a regular basis. AOANJRR began data collection of total hip and knee arthroplasties in 1999, and includes data on more than 98% of arthroplasty procedures performed nationally since 2002 (AOANJRR 2016) AOANJRR data are validated against patient-level data provided by each state and territory health departments in Australia using a sequential, multilevel matching process reaching 94% validation on the initial pass of the validation process (AOANJRR n.d.). Data are also matched biannually with the National Death Index to obtain information on the date of death. In both registries revision is defined as a new surgical intervention when any part of the implant is removed or exchanged. This study includes data starting from January 1, 2006, which was the time point when reporting of the use of TM cups started in both registers (AOANJRR 2016). Study population Between January 2006 and December 2014, 25,451 and 213,314 operations performed with an uncemented acetabular component were reported to the SHAR and the AOANJRR, respectively. During this study period 10,113 primary THAs performed with a TM design (Trabecular Metal Tantalum or Continuum (ZimmerBiomet, Warsaw, IN, USA) were registered in SHAR (n = 2,796) and AOANJRR (n = 7,317). The 5 most commonly used uncemented acetabular components from each register (uncemented n = 83,596, SHAR n = 13,156, AOANJRR n = 70,440) were identified (Table 1). The patient selection is described in a flowchart (Figure 1). Characteristics of the study population The average age of the patients at the time of the primary operation was 68 (11–100) years in both groups. There were

SHAR Uncemented cups reported during years 2006–2014 n = 25,451 Excluded Not metal on X-poly n = 7,727 Metal on X-poly n = 17,724

Excluded Less frequently used cup designs n = 14,638

Excluded Less frequently used cup designs n = 1,772

Other uncemented cup designs (5 most frequently used) n = 70,440

Other uncemented cup designs (5 most frequently used) n = 13,156

Tantalum cups n = 2,796

Figure 1. Flowchart of patient selection.

11545 Malchau D.indd 260

5/3/2018 11:33:05 AM

Acta Orthopaedica 2018; 89 (3): 259–264

261

for age, sex, diagnosis, femoral head size, and fixation of the stem. In addition, all analyses were repeated following stratifiSHAR AOANJRR cation of the study population to include Data TM Uncemented TM Uncemented only patients with a primary diagnosis of Age, years (SD) 61 (11) 60 (11) 70 (11) 71 (11) osteoarthritis, who received a cemented Sex stem, and whose femoral head size was Male 1,448 (52) 6,982 (53) 3,039 (42) 29,311 (42) Female 1,348 (48) 6,174 (47) 4,278 (58) 41,129 (58) 32 mm (SHAR, n = 708 and AOANJRR, Side n = 16,824). This was done to remove the Left 1,457 (52) 6,921 (53) 3,330 (46) 32,211 (46) possibility of confounding by head size, Right 1,339 (48) 6,235 (47) 3,987 (54) 38,229 (54) Diagnosis stem fixation, and primary diagnosis. In Osteoarthritis 2,311 (83) 11,074 (84) 62,46 (85) 62,003 (88) this additional subgroup analysis the estiRheumatoid arthritis 61 (2) 307 (2) 66 (1) 626 (1) mated relative revision risks were adjusted Fracture or secondary to fracture 88 (3) 455 (3) 382 (5) 4,289 (6) for age and sex. Schoenfeld’s residuals Developmental dysplasia 212 (8) 788 (6) 119 (2) 477 (1) were used to control for proportional hazOsteonecrosis 106 (4) 507 (4) 267 (4) 2,164 (3) ards assumption. The primary outcome in Others 18 (1) 25 (0.2) 237 (3) 881 (1) Femoral stem fixation all analyses was revision for any reason Uncemented 2,487 (89) 11,347 (86) 4,109 (56) 32,124 (46) and the secondary outcome was revision Cemented 309 (11) 1,797 (14) 3,208 (44) 38,316 (54) for periprosthetic infection. Revision was Head size, mm < 32 401 (14) 3,268 (25) 621 (8) 17,525 (25) described as a change or removal of at 32 1,869 (67) 7,253 (55) 3,207 (44) 33,947 (48) least 1 component. It is possible that indi> 32 526 (19) 2,635 (20) 3,489 (48) 18,968 (27) vidual cup types might interact in different Revised No 2,712 (97) 12,804 (97) 7,041 (96) 68,710 (98) ways with the femoral stem component, Yes 84 (3) 352 (3) 276 (4) 1,730 (2) and therefore revisions where the stem Follow-up time, years (SD) 2.0 (1.7) 3.7 (2.5) 3.2 (2.2) 3.9 (2.5) only was exchanged were also included. Survival data are presented as percentages with 95% confidence interval (CI). Cox 44% and 43% males in the TM and the uncemented control regression analysis is presented with hazard ratio (HR) and groups, respectively (Table 2). Primary osteoarthritis was the 95% CI. most common reason for surgery both in the TM group (85% of all operations) and in the uncemented control group (87% Ethics, funding, and potential conflict of interest Ethical approval was obtained from the Local Ethical Review of all operations) (Table 2). Board in Gothenburg, Sweden (669-16 dd 29/09/2016). Date of Operative data issue September 29, 2016. The AOANJRR has been approved In the TM group, uncemented stems were more frequently to use its own data as Federal Quality Control Activity and used (65% of all TM cups) compared with the uncemented therefore an individual IRB approval is not needed for publications on de-identified analysis. No funding directly related control group (52% of all uncemented cups). In the TM group, large femoral heads (> 32 mm) were to this study was received. No competing interests declared. implanted in 40% (n = 4,015) of the cases. The corresponding proportion in the uncemented control group was 26% (n = 21,603). The average follow-up was 3.0 (0–9) years in the TM Results group and 3.8 (0–9) years in the uncemented control group. Implant survival Statistics In Kaplan–Meier analysis, the up to 8-year survivorship of the The time to first revision was defined using Kaplan–Meier TM group was 94.4% (CI 92.8–96.0) and that of the unceestimates of survivorship. Survival curves were excluded mented control group 96.2% (CI 96.0–96.4) (p = < 0.001) when numbers at risk in any of the groups were below 100 (Figure 2). After adjustment for age, sex, indication for pricases. All analyses were performed using R statistical soft- mary THA, femoral head size, and stem fixation, the TM ware (RStudio: Integrated Development for R. RStudio, Inc., group had a 1.5 (CI 1.4–1.7, p = < 0.001) times higher risk for Boston, MA, USA). The log rank test was used to compare revision compared with the uncemented control group. In the survival at 8 years. To reduce the risk of possible selection subgroup analysis, which included only patients with a pribias towards more difficult cases being treated with TM cups mary diagnosis of osteoarthritis, cemented stem, and femoral we adjusted the estimated relative revision risks in the Cox head size 32 mm (total n = 17,532; TM n = 1,222; other unceregression models performed in the whole study population mented cups n = 16,310), TM cups were also revised more Table 2. Demographics by registry. Values are frequency and (%) unless otherwise stated

11545 Malchau D.indd 261

5/3/2018 11:33:06 AM

262

Acta Orthopaedica 2018; 89 (3): 259–264

Table 3. Reasons for revision and type of revision by registry. Values are frequency and (%) unless otherwise stated

Data

Figure 2. Kaplan–Meier survival for TM cups and other uncemented cups in primary THA with revision for any reason as the end-point. 95% CI levels presented around the curves in light blue and light grey.

Reason for revision Infection Fracture Dislocation Loosening Others Type of revision Cup + stem exchange Stem exchange Cup exchange Liner +/- head exchange Femoral head exchange Extraction Others

Figure 3. Kaplan–Meier survival for TM cups and other uncemented cups in primary THA with revision for infection as the end-point. 95% CI levels presented around the curves in light blue and light grey.

often than other uncemented cups (HR = 2.3, p = < 0.001, CI 1.6–3.2). Infection In the TM group, including all TM cups, 79 of all 360 revisions were performed due to infection, whereas, in the uncemented control group 27% (n = 561) of all revisions were performed due to infection (Table 3). There was a slightly higher, though not statistically significant, revision rate due to infection in the TM group compared with the uncemented control group (HR = 1.2, CI 1.0-–1.6, p = 0.09) (Figure 3). In the TM group, in 33% of cup revisions the stem was revised simultaneously (n = 120), whereas in groups of other uncemented cups 41% of cup revisions had simultaneous stem revision (n = 848).

Discussion Since the introduction of trabecular metal in 1997, TM cups have shown promising results in acetabular revisions (Siegmeth et al. 2009, Jafari et al. 2010, Davies et al. 2011, Skyttä et al. 2011, Mohaddes et al. 2015). Mid- to long-term survivorship of TM cups has been promising in primary THAs as

11545 Malchau D.indd 262

SHAR Uncemented

AOANJRR Uncemented

36 (43) 7 (8) 30 (36) 1 (1) 10 (12)

125 (36) 55 (16) 76 (22) 40 (11) 56 (16)

43 (16) 53 (19) 108 (39) 41 (15) 31 (11)

436 (25) 387 (22) 431 (25) 350 (20) 126 (7)

7 (8) 16 (19) 7 (8) 38 (45) 11 (13) 2 (2) 3 (4)

46 (13) 95 (27) 60 (17) 78 (22) 24 (7) 8 (2) 42 (12)

16 (6) 81 (29) 50 (18) 80 (29) 14 (5) 15 (5) 20 (7)

157 (9) 550 (32) 293 (17) 511 (30) 72 (4) 106 (6) 41 (2)

well, where the use of a TM cup has been increasing lately (Macheras et al. 2009, Baad-Hansen et al. 2011, Howard et al. 2011, Wegrzyn et al. 2015, De Martino et al. 2016). In this collaborative register-based study we found that the risk for revision for the TM cups was 50% higher in the unrestricted analysis and 128% higher in the restricted analysis than that of the 5 most frequently used uncemented cup designs after primary THA revision for any reason as the end-point. In addition, there was weak evidence of association between the use of a TM cup and higher risk for revision for PPI compared with other uncemented designs. Our findings that TM cups have a higher revision rate compared with the control group of other uncemented cups is in contrast to some earlier studies. When adjusted for confounding factors in the Cox model, the TM cups still had a 1.5 times higher risk for revision for any reason when compared with the 5 most commonly used uncemented cup designs. Previous studies have reported excellent results with tantalum cups used in primary THA (Jafari et al. 2010, Noiseux et al. 2014, De Martino et al. 2016), or have not found any statistically or clinically significant difference in the survival results with TM cups compared with uncemented titanium cups in primary THA (Baad-Hansen et al. 2011, Mohaddes et al. 2015). Wegrzyn et al. (2015) presented 100% survivorship for the TM cup in a randomized controlled trial at 10 years, and they also observed statistically significantly fewer radiolucencies around the TM cups compared with other uncemented titanium cups. We were not able to assess radiographs and therefore could not study radiolucencies. The use of TM cups has been recommended for more demanding cases with larger acetabular bone defects because of the reliable bone ingrowth observed with this material (Bobyn et al. 1999). Although increased use in more complicated primary procedures might be one reason explaining

5/3/2018 11:33:06 AM

Acta Orthopaedica 2018; 89 (3): 259–264

the higher overall and infection revision rate of the TM cups found in this study, the evidence we have indicates that the use of these devices in Sweden and Australia has largely been in routine primary procedures. To verify this, we looked at the distribution of use by hospital in Sweden and by surgeon in Australia. In Sweden 28% of hospitals performing 40 or more uncemented primary cups per year used TM in at least 30% of all cases. In Australia 22% of surgeons have used a TM cup. For surgeons doing more than 25 procedures a year almost 14% use the TM cup in over 50% of their procedures and 11% use it in over 75%. Despite its reputation of being a specialty cup, in practice it is being used in these countries as a routine cup. Our results may indicate that this is not a beneficial trend from either the cost point of view or more importantly the outcome. However, TM cups used with large bone deficiencies in revision surgery have been reported to have excellent results (Weeden and Schmidt 2007, Sternheim et al. 2012). We do not recommend that our results be extrapolated to either complex primary or revision surgery. Reducing the rate of infection remains an important imperative in joint replacement surgery. In a recent study by Tokarski et al. (2015), TM cups used for first-time cup revisions showed lower second revision rate for infection than titanium cups. The risk of second revision due to infection in cases revised due to infection was substantially lower with a tantalum cup than with a titanium cup (3% vs. 18%, respectively). The authors commented that this might be due to better osteointegration or directly associated with tantalum’s biological properties. The data from our study indicate that if there is a beneficial effect of TM with respect to infection risk it is not evident when used for routine primary procedures. There was, in fact, a slightly higher revision rate due to PPI in the TM group compared with other uncemented cups, but this was not statistically significant (p = 0.09). As previously stated, TM cups have been recommended for use in more complex primary procedures. It is possible to speculate that the trend to increased infection may indicate that patients operated on using TM cups in this study may have been at a greater risk for infection; however, we have no evidence for this. At the very least it can be stated that in this large population-based study we have not been able to confirm that there is a beneficial effect on infection risk when TM cups are used in primary procedures. We acknowledge a number of limitations in this study, the most important being the potential for confounding. Our major concern was that TM cups might be selectively used in more demanding cases. A number of approaches were taken to address this, including the adjustments used in the regression analyses as well as in the subgroup analysis. There are, however, a number of other reasons why we believe that selection bias is not a major factor in this study. During the study period 2006–2014, over 10% of all uncemented cups used in primary THA were TM cups. The use of the TM cup has also rapidly increased with a 5-fold increase during the study period. The TM is clearly a commonly used acetabular component with

11545 Malchau D.indd 263

263

use comparable to other commonly used cementless acetabular components, indicating frequent use in routine procedures. The lack of radiological data and patient-reported outcome measures, usual for large registry-based studies, is a further limitation. It is likely that there are unrevised patients in both groups with pain and/or poor function. Although revision is an excellent primary outcome measure for large population-based studies, the number of patients with unsatisfactory results will be larger than the revised population. There is no reason, however, to suspect that the indications for revision vary between the different populations. Another possible problem in registry collaboration is aggregating data from individual registries. Nonetheless, due to the similar data structure in both SHAR and AOANJRR we are not concerned about combining data in this study. Although the maximum follow-up of this study is 9 years, the mean follow-up time is 3 years. This in part reflects the increased use of these devices in both Sweden and Australia in recent years. Risk of revision due to dislocation or infection is highest during the first 2 years after primary THA (Pulido et al. 2008, Jameson et al. 2011). This is well covered by the mean follow-up of this study. Determining the long-term outcomes of TM cups requires studies with longer follow-up. In summary, we found the early and mid-term revision rate of TM cups is significantly higher than that of other uncemented cups in primary THA. Further, there was no statistically significant difference in revision rate for infection between these 2 groups. Although TM cups may be a good option in complex primary or revision hip arthroplasty, there does not appear to be any reason to use TM cups in routine primary THAs, where uncemented titanium cups provide patients with a good and reliable outcome.

All authors participated in the study design. MM and ML performed the statistical analyses. Interpretation of the results was done by IL, SEG, AE, HM, and MM. All authors contributed to the writing of manuscript.

Acta thanks Adrian Sayers and Willem Schreurs for help with peer review of this study. AOANJRR. AOANJRR validation process. Available from: https://aoanjrr. sahmri.com/en/data [last accessed December 12, 2017]. AOANJRR. Annual Report 2016. Available from: https://aoanjrr.sahmri.com/ annual-reports-2016 [last accessed December 12, 2017]. Baad-Hansen T, Kold S, Nielsen P T, Laursen M B, Christensen P H, Soballe K. Comparison of trabecular metal cups and titanium fiber-mesh cups in primary hip arthroplasty: a randomized RSA and bone mineral densitometry study of 50 hips. Acta Orthop 2011; 82(2): 155-60. Beckmann N A, Weiss S, Klotz M C M, Gondan M, Jaeger S, Bitsch R G. Loosening after acetabular revision: comparison of trabecular metal and reinforcement rings. A systematic review. J Arthroplasty 2014; 29(1): 22935. Bobyn J D, Stackpool G J, Hacking S A, Tanzer M, Krygier J J. Characteristics of bone ingrowth and interface mechanics of a new porous tantalum biomaterial. J Bone Joint Surg 1999; 81(5): 907-14.

5/3/2018 11:33:06 AM

264

Dale H, Fenstad A M, Hallan G, Havelin L I, Furnes O, Overgaard S, et al. Increasing risk of prosthetic joint infection after total hip arthroplasty. Acta Orthop 2012; 83(5): 449-58. Dale H, Hallan G, Hallan G, Espehaug B, Havelin L I, Engesaeter L B. Increasing risk of revision due to deep infection after hip arthroplasty. Acta Orthop 2009; 80(6): 639-45. Davies J H, Laflamme G Y, Delisle J, Fernandes J. Trabecular metal used for major bone loss in acetabular hip revision. J Arthroplasty 2011; 26(8): 1245-50. De Martino I, De Santis V, Sculco P K, D’Apolito R, Poultsides L A, Gasparini G. Long-term clinical and radiographic outcomes of porous tantalum monoblock acetabular component in primary hip arthroplasty: a minimum of 15-year follow-up. J Arthroplasty 2016; 31(9): 110-14. Gundtoft P H, Overgaard S, Schønheyder H C, Møller J K, KjærsgaardAndersen P, Pedersen A B. The incidence of surgically treated deep prosthetic joint infection after 32,896 primary total hip arthroplasties: a prospective cohort study. Acta Orthop 2015; 86(3): 326-34. Howard J L, Kremers H M, Loechler Y A, Schleck C D, Harmsen WS , Berry D J, et al. Comparative survival of uncemented acetabular components following primary total hip arthroplasty. J Bone Joint Surg (Am) 2011; 93(17)1597-604. Huotari K, Peltola M, Jämsen E. The incidence of late prosthetic joint infections: a registry-based study of 112,708 primary hip and knee replacements. Acta Orthop 2015; 86(3):3 21-5. Jafari S M, Bender B, Coyle C, Parvizi J, Sharkey P F, Hozack W J. Do tantalum and titanium cups show similar results in revision hip arthroplasty? Clin Orthop Relat Res 2010; 468(2): 459-65. Jameson S S, Lees D, James P, Serrano-Pedraza I, Partington P F, Muller S D, et al. Lower rates of dislocation with increased femoral head size after primary total hip replacement: a five-year analysis of NHS patients in England. J Bone Joint Surg (Br) 2011; 93: 876-80. Kremers H M, Howard J L, Loechler Y, Schleck C D, Harmsen W S, Berry D J, et al. Comparative long-term survivorship of uncemented acetabular components in revision total hip arthroplasty. J Bone Joint Surg (Am) 2012; 94(12): e82.

11545 Malchau D.indd 264

Acta Orthopaedica 2018; 89 (3): 259–264

Lindgren J V, Gordon M, Wretenberg P, Kärrholm J, Garellick G. Validation of reoperations due to infection in the Swedish Hip Arthroplasty Register. BMC Musculoskelet Disord 2014; 15: 384. Macheras G, Kateros K, Kostakos A, Koutsostathis S, Danomaras D, Papagelopoulos P J. Eight- to ten-year clinical and radiographic outcome of a porous tantalum monoblock acetabular component. J Arthroplasty 2009; 24(5): 705-9. Mohaddes M, Rolfson O, Kärrholm J. Short-term survival of the trabecular metal cup is similar to that of standard cups used in acetabular revision surgery. Acta Orthop 2015; 86(1): 26-31. Noiseux N O, Long W J, Mabry T M, Hanssen A D, Lewallen D G. Uncemented porous tantalum acetabular components: early follow-up and failures in 613 primary total hip arthroplasties. J Arthroplasty 2014; 29(3): 617-20. Pulido L, Ghanem E, Joshi A, Purtill J J, Parvizi J. Periprosthetic joint infection: the incidence, timing, and predisposing factors. Clin Orthop Relat Res 2008; 466(7): 1710-15. Siegmeth A, Duncan C P, Masri B A, Kim W Y, Garbuz D S. Modular tantalum augments for acetabular defects in revision hip arthroplasty. Clin Orthop Relat Res 2009; 467(1): 199-205. Skyttä E T, Eskelinen A, Paavolainen P O, Remes V M. Early results of 827 trabecular metal revision shells in acetabular revision. J Arthroplasty 2011; 26(3): 342-5. Sternheim A, Backstein D, Kuzyk P R, Goshua G, Berkovich Y, Safir O, Gross A E. Porous metal revision shells for management of contained acetabular bone defects at a mean follow-up of six years: a comparison between up to 50% bleeding host bone contact and more than 50% contact. J Bone Joint Surg Br 2012; 94(2): 158-62. Tokarski A T, Novack T A, Parvizi J. Is tantalum protective against infection in revision total hip arthroplasty? Bone Joint J 2015; 97-B(1): 45-9. Weeden S H, Schmidt R H. The use of tantalum porous metal implants for Paprosky 3A and 3B defects. J Arthroplasty 2007; 22(6 Suppl): 151-5. Wegrzyn J, Kaufman K R, Hanssen A D, Lewallen D G. Performance of porous tantalum vs. titanium cup in total hip arthroplasty: randomized trial with minimum 10-year follow-up. J Arthroplasty 2015; 30(6): 1008-13.

5/3/2018 11:33:06 AM

Acta Orthopaedica 2018; 89 (3): 265–271

265

Does surgeon experience affect patient-reported outcomes 1 year after primary total hip arthroplasty? A register-based study of 6,713 cases in western Sweden Per JOLBÄCK 1–3, Ola ROLFSON 1,2, Maziar MOHADDES 1,2, Szilárd NEMES 1, Johan KÄRRHOLM 1,2, Göran GARELLICK 1,2, and Hans LINDAHL 1–3

1 Swedish Hip Arthroplasty Register, Gothenburg; 2 Department of Orthopaedics, Institute of Clinical Sciences, The Sahlgrenska Academy, University of Gothenburg, Gothenburg; 3 Department of Orthopaedics, Skaraborgs Hospital, Lidköping, Sweden Correspondence: per.jolback@vgregion.se Submitted 2017-05-18. Accepted 2018-01-16.

Background and purpose — Several studies have reported on the influence of various factors on patient-reported outcomes (PROs) after total hip arthroplasty (THA), but very few have focused on the experience of the surgeon. We investigated any association between surgeons’ experience and PROs 1 year after primary THA. Patients and methods — Patient characteristics and surgical data at 10 hospitals in western Sweden were linked with PROs (EQ-5D-3L, Satisfaction Visual Analogue Scale (VAS), Pain VAS). These data were retrieved from the Swedish Hip Arthroplasty Register (SHAR). The surgeon’s level of experience was divided into 4 subgroups related to experience: < 8 years, 8–15 years, and > 15 years of clinical practice after specialist certificate. If no specialist certificate was obtained the surgery was classified as a trainee surgery. Surgeons with > 15 years’ experience as an orthopedic specialist were used as reference group in the analyses. Results — 8,158 primary THAs due to osteoarthritis were identified. We identified the surgeons’ level of experience in 8,116 THAs. Data from SHAR on pre- and postoperative PROs and satisfaction at 1 year were available for 6,713 THAs. We observed a statistically significant difference among the 4 groups of surgeons regarding mean patient age, ASA classification, Charnley classification, diagnosis, and fixation technique. At 1-year followup, there were no statistically significant differences in Pain VAS, EQ-5D index, or EQ VAS among the subgroups of orthopedic specialists. Patients operated on by orthopedic trainees reported less satisfaction with the result of the surgery compared with the reference group. Interpretation — These findings indicate that patients can expect similar health improvements, pain reduction, and satisfaction 1 year after a primary THA operation irrespective of years in practice after specialty certification as an orthopedic surgeon. ■

During 2014, 16,565 total hip arthroplasties (THA) were carried out in Sweden, which corresponds to 331 procedures per 100,000 inhabitants aged 40 years or older (Garellick et al. 2014). The forecast increase in primary THAs in Sweden during the next 10–15 years is an annual volume of 18,000 in 2020 and 20,000 in 2030 (Nemes et al. 2014). During the last decade, there has been an increased focus on patient-reported outcomes (PROs) following primary THAs. One explanation of the increased interest in PROs might be a non-negligible group of patients reporting dissatisfaction after THA. Some studies report that 7–10 % (Anakwe et al. 2011, Rolfson et al. 2011) of patients are uncertain about or dissatisfied with the result after THA. A number of studies have investigated different factors influencing PROs after THA. In addition to demographic factors and socioeconomics, patient expectations and surgical factors have also been shown to influence PROs (Rolfson et al. 2011, Neuburger et al. 2013, Gordon et al. 2014, Greene et al. 2014, Jameson et al. 2014, Lindgren et al. 2014, Palazzo et al. 1999, Graves et al. 2016). Traditionally in Sweden, training surgeons begin their education by assisting an experienced surgeon. After having performed a number of “more or less” supervised procedures followed by a postoperative discussion with their mentor, they perform the entire operation independently. After minimum of 5.5 years of training and achievement of the aims outlined for becoming an orthopedic specialist, trainees can apply for specialist certification at the Swedish National Board of Health and Welfare. The overall objective of the orthopedic education is that this learning procedure will not interfere with the outcome. In other surgical specialties, no statistically significant difference in outcomes after surgery have been shown between surgeons with different levels of experience

11732 Jolba¦êck D.indd 265

5/3/2018 11:34:13 AM

266

(Bradbury et al. 1997, Naylor et al. 1998, Paisley et al. 1999, Goodwin et al. 2001, Asimakopoulos et al. 2006, Borowski et al. 2008). Nonetheless, surgical training programs may raise ethical issues. Patients might have a desire to be operated by a surgeon with long experience and this understandable desire might be in conflict with the education of surgeons. Only 3 studies have reported an association between surgeons’ experience and PROs after THAs. In a study based on data from the New Zealand Joint Registry (Inglis et al. 2013), surgeons had better results than trainees, whereas the other 2 found no difference (Palan et al. 2009, Reidy et al. 2016). In these 3 papers, there was a discrepancy in the classification of experience (i.e., consultant vs. non-consultant, consultant vs. supervised trainees, and unsupervised trainees and consultant vs. junior trainees and senior trainees) but also in which PROM (patient-reported outcome measures) instruments were used, which varied between Harris Hip Score (HHS) and Oxford Hip Score (OHS). We were unable to find any study in which the experience of the surgeon was measured as the time after obtaining a specialist certificate in orthopedics and its relationship to PROs after THA. Our main objective was to study any association between surgical experience and PROs 1 year after primary THAs. We hypothesized that patients operated on by more experienced orthopedic surgeons report better PROs when compared with patients operated on by less experienced surgeons.

Patients and methods The inclusion criteria were: a patient operated with a primary THA with either cemented, uncemented, hybrid, or reversed hybrid fixation and diagnostic indication of osteoarthritis (OA) as defined by International Statistical Classification of Diseases and Related Health Problems 10th Revision codes (ICD-10 codes) M16.0–M16.7 and M16.9. The surgery should also have been performed in a public hospital managed by the county council in the region of western Sweden between years 2007 and 2012. There should also be a complete preoperative and 1-year follow-up PROMs questionnaire, plus complete information on surgeon’s year for license to practice and/or specialist certification in orthopedics (Figure). Western Sweden In 2012, the population in western Sweden amounted to around 1.6 million (17% of the total Swedish population). During the study period, 10 hospitals managed by the county council were performing primary THAs. In Sweden, the vast majority of hospitals are managed by the relevant county council. The Swedish Hip Arthroplasty Register (SHAR) PROMs program The SHAR aims (Kärrholm 2010) to register all primary

11732 Jolba¦êck D.indd 266

Acta Orthopaedica 2018; 89 (3): 265–271

Primary THAs due to OA extracted from local hospital medical records n = 8,301 Excluded (n = 1,588): – inaccurate ICD or fixation method, 143 – missing data related to surgeon, 31 – missing data on year of orthopedic specialist certificate, 11 – missing preoperative PROs, 842 – missing 1-year follow-up PROs, 561 Primary THAs due to OA with preoperative and 1-year follow-up PROs and data on surgeons’ experience n = 6,713

Flow chart.

THAs and reoperations performed in Sweden. Participation is voluntary for both clinics and patients. The coverage is 100% (Garellick et al. 2012) for clinics and the completeness of individual patients is 98% (Garellick et al. 2014). Individual patient data, such as age, sex, height, weight, diagnosis, fixation technique, surgical approach, and type of implant used, as well as PRO data, were registered for every THA included in this study. We used data from the register’s PROM program which begun in 2002. The PROMs data were collected at the outpatient clinic visit shortly before surgery and at the 1-year follow-up by a mailed questionnaire. Non-respondents receive the first and only reminder after 8 weeks. The respondent rate for individual registration at national level in 2008 was 86% preoperatively and for both surveys 79% (Rolfson et al. 2011). All the hospitals in western Sweden have been participating in collecting PROs since the start of the PROMs program. The SHAR has used the 3-level version of EQ5D (EQ5D-3L), which was developed by the EuroQol Group and introduced in 1990 (EuroQol 2015, Rabin and de Charro 2001). The EQ-5D index extends from minimum values of –0.594 (worse than death) to the maximum of 1.0 (full health) using a UK value set (Dolan 1997), as there is no specific Swedish value for the investigated period. The EQ-5D-3L also contains a health state visual analog scale (VAS) (EQ VAS) ranging from 0 (worst imaginable) to 100 (best imaginable). The PROMs program in the SHAR also includes questions on hip pain and satisfaction with the outcome of the operation. For the registration of pain, a VAS with a range of 0 (no pain) to 100 (worst pain imaginable) is used (Pain VAS). The question addresses the average pain experienced from the current hip during the last month. VAS is also used to address satisfaction with the outcomes of the surgery (hereafter only named “satisfaction”) with a range from 0 (satisfied) to 100 (dissatisfied) (Rolfson et al. 2011, Garellick et al. 2012, Kärrholm et al. 2015). In the SHARs PROMs program, there is also a question about musculoskeletal comorbidity based on the Charnley classification (Charnley 1972). This patient-reported Charnley

5/3/2018 11:34:13 AM

Acta Orthopaedica 2018; 89 (3): 265–271

classification is assigned by using 2 self-administered questions: “Do you have any symptoms from the other hip” and “Do you have problems walking because of other reasons” (e.g., pain from other joints, back pain, angina, or any other medical condition). The difference between preoperative and the 1-year followup value was calculated for EQ-5D index, EQ VAS, and Pain VAS (Change Score). The Change Score could be interpreted as the amount of gain in quality of life and reduction of pain experienced by the patient 1 year after the operation. Sources of data Local hospital medical records, the SHAR, and the Swedish National Board of Health and Welfare register of licensed health-care professionals were used as data sources. Data from local hospital medical records containing a 10-digit personal identity number, the name of the hospital, the name of the surgeon, the date of surgery, diagnosis, the fixation technique, and the ASA classification were extracted. This data were then linked with EQ-5D-3L, Charnley classification, Pain VAS, and Satisfaction VAS at 1-year follow-up from the SHAR. Patient characteristics, BMI, ASA classification, sex and age, diagnosis, and fixation technique were also available from the SHAR. Linking was performed using the 10-digit personal identity number, the name of the hospital, and the date of surgery. For each surgeon involved, data on the year for license to practice and/or specialist certificate in orthopedics were obtained from publicly available data from the Swedish National Board of Health and Welfare register of licensed health-care professionals. Surgeons’ experience Our classification of surgical experience was based on years between time for certification as an orthopedic specialist (both license to practice and orthopedic specialist certification) and the time of the THA surgery. If the surgeon has only a license to practice but no orthopedic specialist certification he was classified as an orthopedic trainee. Surgeons who did not appear with an orthopedic specialist’s certificate in the Swedish National Board of Health and Welfare register of licensed health-care professionals were classified as orthopedic trainees. The surgeons were divided into 4 groups: orthopedic trainees, orthopedic specialists with less than 8 years’ experience, 8–15 years’ experience, and more than 15 years’ experience after specialist certification. The categorization of experience was decided in advance in the research group during the planning of the study. Statistics SPSS version 21 (IBM Corp, Armonk, NY, USA) and R version 3.2.3 (R Foundation for Statistical Computing, Vienna, Austria) were used for the statistical analysis. We used the

11732 Jolba¦êck D.indd 267

267

Kruskal–Wallis H test (variables: age, BMI, ASA classification, Charnley classification, fixation technique, diagnosis, EQ-5D index, EQ VAS, Pain VAS, Satisfaction VAS, EQ-5D index Change Score, EQ VAS Change Score, and Pain VAS Change Score) and Pearson’s chi-square test (gender). A linear regression analysis for EQ-5D index, EQ VAS, Pain Vas, and Satisfaction VAS (simple and multivariable) was performed. The regression coefficients (β) denote the change in the outcome for one-unit change in the exposure, or in case of categorical variables the mean difference between the reference and the different categories. Data were adjusted for age, sex, BMI, ASA classification, diagnosis, and Charnley classification at 1-year postoperatively. Surgeons with more than 15 years’ experience were used as a reference in the linear regression. The p-value for statistical significance was set at p < 0.05. Data in the linear regression are presented with a 95% confidence interval (CI). Sensitivity analysis We performed 3 different sensitivity analyses with use of the same statistical methods as presented above except that other time limits for categorizing surgical experience were used. In first analysis surgical experience was categorized into 3 groups: trainees, 9 years or less, and 10 years or more as orthopedic specialist. In the second analysis there were 5 groups: trainees, 5 years or less, 6–9, 10–15, and 16 years or more as specialist. In the third analysis there were 4 groups now separated into trainees, 6 years or less, 7–18, and 19 years or more. Analysis of respondents versus non-respondents A respondents versus non-respondents analysis was performed to evaluate whether non-responders differed in characteristic variables (age, sex, ASA classification, and BMI) and surgical data (fixation technique and diagnosis). Ethics, funding, and potential conflicts of interests The study was approved by the Regional Ethical Review Board in Gothenburg, DNR 205-16. A research grant for the project was received from Skaraborgs Hospital research foundation. There is no conflict of interest.

Results The analyzed cohort comprised 6,713 primary THAs, of which 16% were performed at the university hospital, 29% at central hospitals and 56% at rural hospitals. A total of 219 surgeons performed the 6,713 cases. Of these cases, 8% were performed by trainees and 92% by orthopedic specialists. Patient characteristic The sex distribution was even between the groups of surgeons. Less experienced surgeons operated on older patients than

5/3/2018 11:34:14 AM

268

Acta Orthopaedica 2018; 89 (3): 265–271

Table 1. Patient characteristics and surgical data divided into subgroups based on surgeons’ experience at the time of surgery Trainees (n = 538)

< 8 years (n = 2,181)

Age, years, mean All (SD) 73 (8) 71 (9) Missing, n (%) 0 (0) 0 (0) Gender, n (%) Male 215 (40) 935 (43) Missing, n (%) 0 (0) 0 (0) BMI, mean All (SD) 27 (4) 28 (5) Missing, n (%) 150 (28) 527 (24) ASA classification, n (%) I 91 (17) 580 (27) II 320 (60) 1,197 (55) III/IV 85 (16) 279 (13) Missing 42 (8) 125 (6) Diagnosis, n (%) Primary OA 536 (100) 2,136 (98) Secondary OA 2 (0) 45 (2) Missing 0 (0) 0 (0) Fixation technique, n (%) Cemented 481 (90) 1,749 (80) Uncemented 29 (5) 237 (11) Hybrids 5 (1) 62 (3) Reverse hybrids 23 (4) 133 (6) Missing 0 (0) 0 (0) Charnley classification at 1 year, n (%) A 229 (43) 925 (42) B 52 (10) 179 (8) C 257 (48) 1,077 (49) Missing 0 (0) 0 (0)

8–15 years > 15 years (n = 984) (n = 3,010)

All (n = 6,713)

69 (10) 0 (0)

67 (11) 0 (0)

69 (10) 0 (0)

424 (43) 0 (0)

1,264 (42) 0 (0)

2,838 (42) 0 (0)

27 (4) 287 (29)

27 (5) 854 (28)

27 (5) 1,818 (27)

326 (33) 510 (52) 91 (9) 57 (6)

882 (29) 1,523 (51) 258 (9) 347 (12)

1,879 (28) 3,550 (53) 713 (11) 571 (9)

964 (98) 20 (2) 0 (0)

2,893 (96) 116 (4) 1 (0.0)

6,529 (97) 183 (3) 1 (0)

764 (78) 2,034 (68) 126 (13) 567 (19) 21 (2) 82 (3) 72 (7.3) 322 (11) 1 (0) 5 (0)

5,028 (75) 959 (14) 170 (3) 550 (8) 6 (0)

p-value < 0.001 0.6 0.2 < 0.001

< 0.001

SHAR’s PROMs program. The nonrespondents group were older (mean age 71) compared with the respondents (mean age 69) (p = 0.001). The sex distribution among respondents and non-respondents also differed, with a lower quota of males (39%) in the non-respondents group and 42% males among respondents. There was a higher proportion of ASA classification III/IV among the non-respondents (18% versus 11%, p < 0.001). Cemented fixation was also more common among non-respondents, 75% versus 62% among respondents (p = 0.01), but there were no differences regarding the diagnosis for the operation, with 96%/97% hips with primary OA in the non-respondents/respondents’ groups (p = 0.2). BMI was similar between responders (mean 27.5) and non-responders (mean 27.4).

< 0.001

Outcomes The preoperative EQ-5D index was lowest in the patient group operated on by the surgeons with the shortest time Kruskal–Wallis H test for age, BMI, ASA classification, diagnosis, fixation technique, and Charnin practice and it increased in groups ley classification at 1-year follow-up, and Pearson’s chi-square test for sex. SD = standard deviawith increasing experience but with tion. BMI = body mass index. OA = osteoarthritis. ASA = American Society of Anesthesiologists. no statistically significant difference. At 1 year the values for EQ-5D index, those who had longer experience (p < 0.001). There was a EQ VAS, Pain VAS, and Satisfaction VAS were unevenly statistically significant difference in the distribution of ASA distributed among the groups with a tendency toward better classification (p < 0.001) and Charnley classification (p < results for those surgeons with longer experience but with no 0.001) between the groups of surgeons with different levels statistically significant difference (Table 2). The change scores of experience. There was no statistically significant difference from the preoperative evaluation to the follow up at 1 year, in BMI (p = 0.6). Patients with primary OA were more com- however, did not differ. Patients operated on by surgeons with the shortest time in monly operated on by less experienced surgeons (p < 0.001) practice (both trainees and surgeons with less than 8 years’ (Table 1). experience) reported a higher value on the VAS for satisfacSurgical data tion (i.e., they were less satisfied) than those patients operated Cemented fixation had been used in 75%, uncemented in 14%, on by the most experienced surgeons (p < 0.001) (Table 2). Both the simple and multivariable linear regression showed reverse hybrid in 8%, and hybrid in 3% of the hips. Trainees had a higher percentage of cemented THAs compared with that longer time in practice as an orthopedic surgeon was not more experienced surgeons. This percentage decreased as the associated with better PROs 1 year after THAs measured experience of the surgeon increased in favor of an increasing by the EQ-5D index, EQ VAS. In terms of satisfaction and pain, the simple analysis showed that both trainees and the proportion of hybrids and uncemented fixations (p < 0.001). group with less than 8 years’ experience had a lower level Respondents versus non-respondents of satisfaction and higher pain level compared with the most Non-respondents rate in the PRO survey in our data totaled experienced surgeon with more than 15 years in orthopedics. around 11% preoperatively and, pooled at 1-year follow-up, However, after adjusting for differences in age, sex, BMI, about 18%. There were some differences in patient demo- diagnosis, ASA classification, and Charnley classification graphics between respondents and non-respondents in the in the multivariable linear regression, there were no statisti-

11732 Jolba¦êck D.indd 268

464 (47) 87 (9) 433 (44) 0 (0)

1,406 (47) 307 (10) 1,297 (43) 0 (0)

3,024 (45) 625 (9) 3,064 (46) 0 (0)

5/3/2018 11:34:14 AM

Acta Orthopaedica 2018; 89 (3): 265–271

269

Table 2. PROs divided into subgroups based on surgeons’ experience at the time of surgery

Discussion

Our main objective was to study any association between surgical experiEQ-5D index ence and PROs 1 year after primary preoperative 0.41 (0.31) 0.41 (0.31) 0.41 (0.31) 0.43 (0.31) 0.42 (0.31) 0.05 THAs. We did not find any associa1-year postop. 0.75 (0.24) 0.76 (0.25) 0.77 (0.25) 0.77 (0.24) 0.77 (0.24) 0.005 change 0.34 (0.35) 0.35 (0.34) 0.36 (0.34) 0.34 (0.34) 0.35 (0.34) 0.7 tion between surgeons’ experience EQ VAS and EQ-5D index, EQ VAS, and preoperative 56 (22) 56 (21) 57 (21) 57 (21) 57 (21) 0.2 Pain VAS 1 year after primary THA. 1-year postop. 74 (20) 74 (21) 75 (20) 76 (20) 75 (20) 0.02 change 18 (23) 18 (23) 18 (22) 19 (23) 18 (23) 0.9 However, patients operated on by Pain VAS orthopedic trainees reported lower preoperative 61 (16) 61 (16) 61 (16) 60 (17) 61 (16) 0.6 satisfaction VAS at 1-year follow-up 1-year postop. 17 (19) 15 (19) 14 (18) 15 (18) 15 (18) 0.002 change 44 (23) 46 (23) 47 (22) 45 (23) 46 (23) 0.3 than those operated by surgeons with Satisfaction VAS more than 15 years’ experience. 1-year postop. 21 (23) 19 (22) 18 (21) 18 (22) 19 (22) < 0.001 Our findings are in accordance with Kruskal–Wallis H test used for all variables. All values are presented as mean value and stanthose reported by Palan et al. (2009) dard deviation (SD). VAS = visual analogue scale. and Reidy et al. (2016), with similar PROs except for less satisfaction with the outcome of surgery in patients Table 3. Linear regression analysis of PROs operated on by trainees. Inglis et al. (2013) recorded the OHS in the New Zealand Joint Registry at 6 months and related this outcome to surgeons’ experience. They found a higher Simple Multivariable a ß-coefficient (95% CI) ß-coefficient (95% CI) mean OHS for patients operated on by surgeons compared with operations performed by supervised and unsupervised EQ-5D index trainees. Their study included only 6-month follow-up data > 15 years Reference Reference 8–15 years 0.00 (–0.02 to 0.02) –0.01 (–0.03 to 0.01) on 20% of randomly selected patients and no adjustment was < 8 years –0.01 (–0.03 to 0.00) –0.01 (–0.03 to 0.01) made for preoperative PROs as these data not were collected. trainee –0.02 (–0.05 to 0.00) –0.02 (–0.05 to 0.00) Thus, any potential difference in baseline values between the EQ VAS > 15 years Reference Reference subgroups could not be accounted for. 8–15 years –0.05 (–1.50 to 1.41) –0.82 (–2.47 to 0.83) In our study, there were significant differences in patient < 8 years –1.29 (–2.40 to –0.17) –0.45 (–1.71 to 0.80) characteristics between those operated by more or less expetrainee –1.79 (–3.65 to 0.07) –0.91 (–3.03 to 1.21) Pain VAS rienced surgeons and there was also a statistically significant > 15 years Reference Reference difference between the groups of surgeons as regards choice 8–15 years –0.14 (–1.46 to 1.18) 0.49 (–1.06 to 2.05) of fixation technique. This bias in patient selection and fixa< 8 years 0.84 (–0.17 to 1.85) 0.78 (–0.41 to 1.96) trainee 2.37 (0.69 to 4.05) 1.88 (–0.12 to 3.88) tion technique could be expected because, in Sweden, trainees Satisfaction VAS usually start by learning cemented fixation, which is still the > 15 years Reference Reference gold standard (Troelsen et al. 2013). Uncemented components 8–15 years –0.22 (–1.81 to 1.37) 0.55 (–1.32 to 2.42) < 8 years 1.58 (0.37 to 2.80) 0.72 (–0.70 to 2.13) are regarded as being more difficult to handle because of the trainee 3.61 (1.58 to 5.63) 2.64 (0.24 to 5.03) risk for complications arises intraoperatively (Kärrholm et al. 2015) and this method is commonly used by surgeons later in a Adjusted for age, sex, BMI, ASA classification, diagnosis, and Charnley classification at 1 year. CI = confidence intervals, BMI = their career. The reason for the difference in mean age between body mass index, ASA = American Society of Anesthesiologists, the groups could be an effect of selection of type of fixation. VAS = visual analogue scale. Reference is surgeon with > 15 years’ Uncemented, hybrid, and reverse hybrid are more suitable in experience as orthopedic specialists. younger patients. The difference in mean age between the subgroups may also affect the observed difference in ASA clascally significant differences in satisfaction at 1-year follow- sification due to the fact that younger patients are expected to ups for surgeons with less than 8 years’ experience. However, be healthier. We did note the same differences in postoperative PROs as patients operated on by trainees reported lower satisfaction those reported by Inglis et al. (2013). However, after adjusting scores (Table 3). All 3 sensitivity analyses showed similar results with respect for patient characteristics and surgical data, these differences to the distribution of patient characteristics, surgical data, and became statistically non-significant. Non-respondent rate in the PRO survey in our data totaled PROs (data not shown). around 11% preoperatively and, pooled at 1-year follow-up, about 18%. This is better than in the SHARs PROM for all Trainees

11732 Jolba¦êck D.indd 269

< 8 years 8–15 years

> 15 years

All

p-value

5/3/2018 11:34:14 AM

270

clinics in Sweden where the preoperative respondent rate was 14% and in both surveys 21%. Females at a greater age, with higher ASA classification and treated with cemented fixation, were underrepresented among the respondents in our study. The reason for this is not known, but high age and presence of more concomitant diseases could have had an influence. This patient group could be expected to report poorer health status and quality of life, but whether excess dropout of these patients actually had any influence on the results remains unknown. One could speculate that the majority of these patients were operated by less experienced surgeons and thus, rather, would make the differences observed more obvious than vice versa. Overall the pooled response rate of 82% is probably difficult to improve in a study like this. Despite the skewed distribution in dropouts we think that the conclusions made in our study are reasonably valid. Our study has limitations. The extent to which the experience of the surgeon restricted to the annual number of hip arthroplasties is reflected in patient-reported outcomes is unclear. The surgeon might have operated on an unknown number of fractures and undertaken other types of orthopedic surgery, for example. The comparatively wide variation among surgeons regarding the volume of primary THAs might partly be an effect of the performance of other types of arthroplasty, including hip and knee revisions. Some surgeons might therefore have a much larger total annual volume of operations, including hip revisions and hemiarthroplasties, which could be a source of bias when compared with surgeons with a low overall surgical volume. However, we decided to subgroup surgeons based on years after specialist degree rather than number of THR surgeries. In this this way experience from other type of operations than primary THR and clinical practice in general, which includes a continuing learning process, will be included. We lack information on socioeconomic factors and educational level, which are known to influence PROs (Borowski et al. 2008, Neuburger et al. 2013). These factors are not available in SHAR or local medical records. In this study, we did not link our data with Statistics Sweden (Greene et al. 2014, Krupic et al. 2014). This lack of information may affect satisfaction if there is a skewed distribution of patients with a good economic situation and a high educational level in any of the groups. We have no information on the quality of the preoperative information given to the patient, or whether this information was given only verbally, only in writing, or both. Nor do we know whether the preoperative information was given by the same surgeon who performed the operation, or whether the information was repeated. However, there is little evidence in the literature to support the value of preoperative patient education (McDonald et al. 2004, Aydin et al. 2015), which might suggest that its presence may have no or a minor influence on the outcome after surgery. Another limitation is that we do not have any knowledge of the supervision of the trainee. The level of supervision may

11732 Jolba¦êck D.indd 270

Acta Orthopaedica 2018; 89 (3): 265–271

vary between completely unsupervised to fully supervised individuals. Earlier studies have, however, not shown any substantial difference in PROs after up to 10 years of followup between unsupervised and supervised trainees (Reidy et al. 2016). It is therefore also possible that this potential confounder had a limited influence in our study. Of the possible confounders mentioned here, we feel that socioeconomic factors and continued patient treatment by one and the same surgeon may have the greatest influence on PROs and patient satisfaction. These hypotheses require further studies. As highlighted above there are a number of confounding factors that we could not consider, and likely there is a degree of residual confounding that we could not control for. As with other observational studies, the results of this study should be considered in terms of association rather than causality. In summary, to our knowledge, this is the largest report analyzing the influence of surgeon experience on postoperative PRO adjusted for Charnley classification and comorbidity. We found considerable differences in patient characteristics and surgical data depending on surgeons’ experience. After adjusting for these covariates, we found no associations between orthopedic trainees and/or orthopedic specialists with variable length of practice after certification and EQ-5D index, EQ VAS, and Pain VAS 1 year after primary THA. Patients operated on by orthopedic trainees did, however, report less satisfaction VAS with the outcome of surgery than those patients operated on by surgeons with more than 15 years’ practice. No such difference in level of satisfaction was observed between any of the orthopedic specialist subgroups. These findings indicate that patients can expect similar health improvements, pain reduction, and satisfaction 1 year after a primary THA operation irrespective of years in practice after specialty certification as orthopedic surgeon.

PJ had the original idea for the study, processed the data, undertook the writing of the manuscript, and performed the statistical analyses but not the linear regression analysis. SN performed the linear regression analysis. OR, MM, JK, GG, and HL took part in the interpretation of the data and in writing the manuscript. GG contributed significantly to the prolonged ethical approval process. All the authors have read and approved the final manuscript.

Acta thanks Yvette Pronk and other anonymous reviewers for help with peer review of this study.

Anakwe R E, Jenkins P J, Moran M. Predicting dissatisfaction after total hip arthroplasty: a study of 850 patients. J Arthroplasty 2011; 26(2): 209-13. Asimakopoulos G, Karagounis A P, Valencia O, Rose D, Niranjan G, Chandrasekaran V. How safe is it to train residents to perform off-pump coronary artery bypass surgery? Ann Thorac Surg 2006; 81(2): 568-72. Aydin D, Klit J, Jacobsen S, Troelsen A, Husted H. No major effects of preoperative education in patients undergoing hip or knee replacement: a systematic review. Dan Med J 2015; 62(7): A5106.

5/3/2018 11:34:14 AM

Acta Orthopaedica 2018; 89 (3): 265–271

Borowski D W, Ratcliffe A A, Bharathan B, Gunn A, Bradburn D M, Mills S J, Wilson R G, Kelly S B. Involvement of surgical trainees in surgery for colorectal cancer and their effect on outcome. Colorectal Dis 2008; 10(8): 837-45. Bradbury A W, Brittenden J, Murie J A, Jenkins A M, Ruckley C V. Supervised training in carotid endarterectomy is safe. Br J Surg 1997; 84(12): 1708-10. Charnley J. The long-term results of low-friction arthroplasty of the hip performed as a primary intervention. J Bone Joint Surg Br 1972; 54(1): 61-76. Dolan P. Modeling valuations for EuroQol health states. Med Care 1997; 35(11): 1095-108. EuroQol. EQ-5D-3L User GuideL: basic information on how to use the EQ5D-3L instrument. 2015. Garellick G, Rogmark C, Kärrholm J, Rolfson O. Swedish Hip Arthroplasty Register Annual Report 2012. Garellick G, Kärrholm J, Lindahl H, Malchau H, Rogmark C, Rolfsson O. Swedish Hip Arthroplasty Register Annual Report 2014. Goodwin A T, Birdi I, Ramesh T P, Taylor G J, Nashef S A, Dunning J J, Large S R. Effect of surgical training on outcome and hospital costs in coronary surgery. Heart 2001; 85(4): 454-7. Gordon M, Frumento P, Skoldenberg O, Greene M, Garellick G, Rolfson O. Women in Charnley class C fail to improve in mobility to a higher degree after total hip replacement. Acta Orthop 2014; 85(4): 335-41. Graves S C, Dropkin B M, Keeney B J, Lurie J D, Tomek I M. Does surgical approach affect patient-reported function after primary THA? Clin Orthop Relat Res 2016; 474(4): 971-81. Greene ME, Rolfson O, Nemes S, Gordon M, Malchau H, Garellick G. Education attainment is associated with patient-reported outcomes: findings from the Swedish Hip Arthroplasty Register. Clin Orthop Relat Res 2014; 472(6): 1868-76. Inglis T, Dalzell K, Hooper G, Rothwell A, Frampton C. Does orthopedic training compromise the outcome in total hip arthroplasty? J Surg Educ 2013; 70(1): 76-80. Jameson S S, Mason J M, Baker P N, Elson D W, Deehan D J, Reed M R. The impact of body mass index on patient reported outcome measures (PROMs) and complications following primary hip arthroplasty. J Arthroplasty 2014; 29(10): 1889-98. Kärrholm J. The Swedish Hip Arthroplasty Register (www.shpr.se). Acta Orthop 2010; 81(1): 3-4.

11732 Jolba¦êck D.indd 271

271

Kärrholm J, Lindahl H, Malchau H, Mohaddes M, Rogmark C, Rolfsson O. The Swedish Hip Arthroplasty Register Annual Report 2015. Krupic F, Garellick G, Gordon M, Kärrholm J. Different patient-reported outcomes in immigrants and patients born in Sweden: 18,791 patients with 1-yearof follow-up in the Swedish Hip Arthroplasty Register. Acta Orthop 2014; 85(3): 221-8. Lindgren J V, Wretenberg P, Kärrholm J, Garellick G, Rolfson O. Patientreported outcome is influenced by surgical approach in total hip replacement: a study of the Swedish Hip Arthroplasty Register including 42,233 patients. Bone Joint J 2014; 96-B(5): 590-6. McDonald S, Hetrick S, Green S. Pre-operative education for hip or knee replacement. Cochrane Database Syst Rev 2004; (1): CD003526. Naylor A R, Thompson M M, Varty K, Sayers R D, London N J, Bell P R. Provision of training in carotid surgery does not compromise patient safety. Br J Surg 1998; 85: 939-42. Nemes S, Gordon M, Rogmark C, Rolfson O. Projections of total hip replacement in Sweden from 2013 to 2030. Acta Orthop 2014; 85(3): 238-43. Neuburger J, Hutchings A, Black N, van der Meulen J H. Socioeconomic differences in patient-reported outcomes after a hip or knee replacement in the English National Health Service. J Public Health (Oxf) 2013; 35(1): 115-24. Palan J, Gulati A, Andrew JG, Murray DW, Beard DJ. The trainer, the trainee and the surgeons’ assistant: clinical outcomes following total hip replacement. J Bone Joint Surg Br 2009; 91(7): 928-34. Palazzo C, Jourdan C, Descamps S, Nizard R, Hamadouche M, Anract P, Boisgard S, Paisley A M, Madhavan K K, Paterson-Brown S, Praseedom R K, Garden O J. Role of the surgical trainee in upper gastrointestinal resectional surgery. Ann R Coll Surg Engl 1999; 81(1): 40-5. Rabin R, de Charro F. EQ-5D: a measure of health status from the EuroQol Group. Ann Med 2001; 33(3): 337-43. Reidy M J, Faulkner A, Shitole B, Clift B. Do trainee surgeons have an adverse effect on the outcome after total hip arthroplasty?: a ten-year review. Bone Joint J 2016; 98-B(3): 301-6. Rolfson O, Kärrholm J, Dahlberg LE, Garellick G. Patient-reported outcomes in the Swedish Hip Arthroplasty Register: results of a nationwide prospective observational study. J Bone Joint Surg Br 2011; 93(7): 867-75. Troelsen A, Malchau E, Sillesen N, Malchau H. A review of current fixation use and registry outcomes in total hip arthroplasty: the uncemented paradox. Clin Orthop Relat Res 2013; 471(7): 2052-9.

5/3/2018 11:34:14 AM

272

Acta Orthopaedica 2018; 89 (3): 272–277

No effect of double nerve block of the lateral cutaneous nerve and subcostal nerves in total hip arthroplasty A randomized controlled trial Johannes L BRON 1,5, Jeanette VERHART 2, Inger N SIEREVELT 2, Dirk DE VRIES 3, Hylke J KINGMA 4 and Maarten V RADEMAKERS 1

1 Department

of Orthopaedic Surgery, Spaarne Gasthuis Hoofdorp, 2 Spaarne Gasthuis Academy (formerly: Lineaus Institute), Hoofddorp, 3 Department of Anaesthesiology, Spaarne Gasthuis Hoofddorp, 4 Pharmacy Foundation of the Haarlem Hospitals (SAHZ), Haarlem, 5 Currently: Department of Orthopaedic Surgery, Antonius Hospital, Sneek, The Netherlands Correspondence: MRademakers@spaarnegasthuis.nl Submitted 2017-09-18. Accepted 2018-01-16.

Background and purpose — The use of local infiltration anesthesia (LIA) has become one of the cornerstones of rapid recovery protocols in total knee arthroplasty patients during the past decade. In total hip arthroplasty (THR), however, the study results are more variable and LIA has therefore not yet been generally accepted. There is no consensus on which structure should be infiltrated and the cutaneous nerves are generally neglected. Hence, we hypothesized a pain-reducing effect of specifically blocking these nerves. Patients and methods — We performed a single-center randomized placebo-controlled trial in 162 subjects to evaluate the infiltration of the lateral cutaneous femoral and subcostal nerve with ropivacaine in patients undergoing total hip arthroplasty via a straight lateral approach. The primary endpoint was pain at rest after 24 hours. Patients were followed up to 6 weeks postoperatively. Results — After correction for multiple testing, no statistically significant differences in pain scores were found between the ropivacaine compared with the placebo group after surgery. In addition, no differences were observed in the use of escape pain medication, complications, and the length of hospital stay. Interpretation — We found no clinically meaningful differences in pain scores between placebo and ropivacaine patients in the postoperative period after THA performed via a straight lateral approach under spinal anesthesia and a multimodal pain regimen. Moreover, our primary endpoint, pain reduction after 24 hours, was not met. Further research should focus on the composition and volume of the LIA suspension, the optimal localization of the infiltration, and should be evaluated for every surgical approach separately. ■

While the use of local infiltration anesthesia (LIA) in total knee arthroplasty (TKA) has been embraced by orthopedic surgeons worldwide, the use in total hip arthroplasty (THA) patients has remained a matter of debate (Andersen and Kehlet 2014). LIA in TKA has unequivocally been shown to reduce early postoperative pain with positive effects on mobilization and hospitalization (Andersen et al. 2008a and b, Andersen et al. 2009, Dillon et al. 2012, Gibbs et al. 2012, Tran and Schwarzkopf 2015). Studies with the use of LIA in THA patients show mixed results (Andersen et al. 2007a, Busch et al. 2010, Andersen et al. 2011, Banerjee and McLean 2011, Lunn et al. 2011, Dobie et al. 2012, Murphy et al. 2012, Rikalainen-Salmi et al. 2012, Kuchalik et al. 2013, Pandazi et al. 2013, Zoric et al. 2014, den Hartog et al. 2015). Although most research consists of well-conducted randomized controlled trials (RCTs), study designs and set-up vary considerably (Wang et al. 2017). Differences are found in the treatment of the control groups, the composition of the LIA suspension, and the endpoints investigated. Moreover, surgical factors like surgical approach and the structures that are actually being infiltrated differ remarkably (Table 1). Our prior observations in THA patients via a straight lateral approach showed that most patients suffer from superficial wound pain instead of deeper groin pain. We therefore hypothesized that blocking the nerves innervating the wound area might result in reduced postoperative pain. This study is a single-center, randomized placebo-controlled trial in patients undergoing THA via straight lateral approach by infiltrating the nerves innervating the area of the incision— the lateral cutaneous femoral nerve and the subcostal nerve— with ropivacaine compared with placebo (NaCl 0.9%). All

12095 Rademakers D.indd 272

5/3/2018 11:35:23 AM

Acta Orthopaedica 2018; 89 (3): 272–277

273

Table 1. Overview of the trials with LIA performed in THA patients. Note the remarkable variation in the LIA suspension, localization of infiltration, surgical approach, and endpoint studied Study Bianconi et al. 2003 Andersen et al. 2007a/b Andersen et al. 2007a/b Busch et al. 2010 Lunn et al. 2011 Andersen et al. 2011 Banerjee&McLean 2011 Murphy et al. 2012 Dobie et al. 2012 Rikalaine et al. 2012

Set up Control Approach a Suspension b Infiltration c

? 80 40 64 120 12 d 204 91 96 60

RCT RCT RCT RCT RCT RCT retro RCT RCT RCT

NaCl epidural NaCl – NaCl NaCl – NaCl – –

PL PL PL SL PL PL ? PL PL ?

R R/K/A R/K/A R/K/A R/A R/A R/K/A bupi Bupi/A B/K/A

SC WI + IA WI C + SC C + M + SC WI WI C + M + SC LC+C+M+SC C + M + SC

Kuchalik et al. 2013

RCT

epidural

R/K/A

C + M + SC

Pandazi et al. 2013 Zoric et al. 2014 den Hartog et al. 2015

63 60 75

RCT RCT RCT

PCA NaCl NaCl

? PL A

R/K/A R R/A

C+M C + M + SC C + M + SC

Results Lower VAS first 72h 2 days earlier discharge Less pain 2 weeks, better function after 1 week Less pain/medication use first 24 h No differences in pain score No differences in pain score and length of stay Faster mobilisation and discharge Les opioid consumption first 12 h No differences pain/ mobilisation No difference VAS, less urinary retention, faster mobilisation Less pain during mobilisation less escape medication Less pain and opioid consumption first 24h No differences opioid consumption/ mobilisation No differences

a Approach: PL = posterolateral, SL = straight b Suspension: R = ropivacaine, K = ketorolac,

lateral, A = anterior A = adrenalin c Infiltration: SC = subcutaneous, WI = wound infiltration, IA = intraarticular, C = capsule, M = muscle, LC = lateral cutaneous nerve d bilateral

patient received the ropivacaine or placebo in addition to multimodal pain management in a fast-track recovery protocol. The primary endpoint was pain at rest 24 hours after surgery.

Patients and methods Patients After written informed consent had been obtained, 162 patients with primary osteoarthritis undergoing elective THA at the Department of Orthopedic Surgery, Spaarne Gasthuis Hospital location Hoofddorp, from August 2014 to September 2016 were enrolled in this randomized trial. Exclusion criteria included known allergy to ropivacaine, general anesthesia, opioid dependency, malignancies, co-morbidities compromising pain perception (e.g., neurologic of psychiatric disorders), ASA grades 3–4. In addition, patients suffering from peroperative complications interfering with postoperative pain and perception or mobilization were excluded. Preoperatively, patient demographics, visual analogue pain scores (VAS) in rest and during mobilization, as well as the Hip disability and Osteoarthritis Outcome Score (HOOS), Short form Survey (SF-12) and Oxford Hip Score (OHS) questionnaires were documented. Methods All patients were admitted on the day of surgery. A standardized pain protocol was used. 2 hours preoperatively patients received 1000 mg acetaminophen, 15 mg meloxicam, and 600 mg gabapentin. All surgeries were performed under spinal anesthesia, using low-dose bupivacaine. Postoperative

12095 Rademakers D.indd 273

patients received 1000 mg acetaminophen every 6 hours, 15 mg meloxicam once daily both at least until discharge, and 600 mg gabapentin 3 times a day for a period of 5 days. When the protocol medication was insufficient to control pain, escape medication consisting of Oramorph oral solution (morphine sulfate) or Dipidolor (piritramide) intramuscular (both 5–10 mg) was available and its usage documented. On the day of surgery, blinded numbered 50 cc syringes containing either ropivacaine (ropivacaine HCl 7.5 mg/mL injection fluid; Fresenius Kabi, Zeist, The Netherlands) or placebo (0.9% NaCl) were provided by the department of pharmacy, which used a block randomization scheme to control group volume. Surgeons, patients, nurses, and investigators were all blinded for treatment. Standard hip arthroplasty surgeries (uncemented Zweymuller steel, Durasul cup (Zimmer Orthopaedics)) were performed by 6 different orthopedic surgeons or coupled residents via a straight lateral approach according to Hardinge. Prior to the surgeries an instruction video was shown to all surgeons on multiple occasions. The video shows exactly how to infiltrate with the highest chance of blocking the targeted nerves. In addition, a single physician assistant was instructed to control the infiltration and was present at the theatre at all surgeries. At the end of the surgeries before wound closure, 0.5 mL/kg body weight of the syringes was infiltrated at 2 different locations: (1) half of the volume was infiltrated anterior to the superior iliac spine, where the LCFN passes below the inguinal ligament; (2) the other half of the volume was infiltrated in the tissues of the upper part of the wound extending from the anterior iliac spine to the iliac crest just cranial of the wound to anaesthetize all sensible end nerves of the subcostal nerves.

5/3/2018 11:35:23 AM

274

Acta Orthopaedica 2018; 89 (3): 272–277

Enrollment

Table 2. Overview of patient demographics and baseline characteristics in both groups

Assessed for eligibility n = 166 Excluded (n = 4): – ASA grade 3

Allocation

Randomized n = 162

Allocated to treatment group (n = 82): – received allocated intervention, 82 – did not receive allocated intervention, 0 Lost to follow-up (n = 0) Analyzed (n = 82)

Allocated to control group (n = 80): – received allocated intervention, 80 – did not receive allocated intervention, 0

Follow-up Analysis

Lost to follow-up (n = 0) Analyzed (n = 80)

Figure 1. Flow chart of the study.

Postoperatively, VAS at rest, escape medication, and complications every 6 hours on the first day and every 8 hours on the second day were registered. Nausea, vomiting, and urinary retention were also documented. The primary endpoint of the study was the VAS at rest 24 hours postoperatively. Discharge criteria included stable vital signs, the ability to walk with crutches, to walk the stairs, and VAS < 4. After discharge, patients were seen at the outpatient clinic 6 weeks postoperatively and the HOOS, SF-12, and OHS questionnaires as well as complications were documented. After the last patient was seen in the outpatient clinic the randomization key was provided to the investigators by the pharmacy department. Statistics Intention to treat analyses were performed by use of IBM SPSS Statistics for Windows version 24 (IBM Corp, Armonk, NY, USA). All variables are summarized according to their distribution by use of means with standard deviations (SD) or frequencies with accompanying percentages. Patient characteristics and baseline variables were compared between treatment groups using Student’s t-tests and chi-squared tests (or Fisher’s exact tests) for continuous and categorical variables respectively. The effect of the intervention on patientreported outcome measures (NRS pain, HOOS, OHS, and SF-12) was analyzed by use of univariate and multivariate regression analysis to correct for potential confounders (age, gender, BMI) at primary (24 hours) and secondary endpoints, and mean between group difference with 95% confidence intervals were calculated at each follow-up moment. Additionally, escape medication usage and complications were compared by use of chi-squared tests (or Fisher’s exact tests). For the main analysis (NRS pain at 24 hours) a p-value of ≤ 0.05 was considered statistically significant. Holm’s procedure was used to correct for multiple testing for secondary outcomes. Sensitivity analysis, using multiple imputation, was performed for all cases to assess presence of bias. 10 fully observed datasets were imputed with complete data and

12095 Rademakers D.indd 274

Sex, n (%) Male Female Age, mean (SD) BMI, mean (SD) ASA, n (%) ASA 1 ASA 2 HOOS, mean (SD) Symptoms Pain ADL Sport/rec QOL OHS, mean (SD) SF12, mean (SD) PCS MCS

Placebo (n=80)

Ropivacaine (n=82)

32 (40) 48 (60) 66 (9.4) 28 (4.9)

36 (44) 46 (56) 67 (8.7) 28 (5.0)

22 (28) 58 (72)

24 (29) 58 (71)

45 (16) 46 (17) 46 (16) 25 (18) 30 (12) 23 (7.4)

40 (16) 45 (18) 44 (15) 31 (24) 27 (14) 24 (7.6)

42 (5.4) 45 (5.7)

40 (6.3) 46 (5.2)

No statistically significant differences are observed.

were analyzed to conform with the previously stated protocol, and results were aggregated. Ethics, registration, funding, and potential conflicts of interest The randomized double-blind placebo-controlled study was approved by the local ethics committee (protocol no. M013034). Written informed consent was obtained from all participants. The study was registered at EudraCT registry (2013004031-71) and Dutch CCMO registry (NL45740.094.13). No competing interest declared. No funding was received for this study.

Results 166 patients were randomized into 2 groups (Figure 1). 4 of the subjects were excluded due to their ASA (ASA III) classification, which was a pre-defined exclusion criterion. Of the remainder 80 were enrolled in the placebo group and 82 in the ropivacaine group. No subjects were lost to follow-up. Baseline characteristics and patient demographics are given in Table 2. The mean age of the subjects was 66 years in both groups with a female predominance. The results of the VAS measurements are shown in Table 3 and Figure 2. After 6 hours there was a statistically lower VAS at rest in the study group compared with the control (p = 0.04). At later time points, this difference ceased (Figure 2). When using Holm’s procedure to correct for multiple comparisons, however, the significance after 6 hours was lost (significance level according to Holm’s procedure < 0.008). There is no statistically signifi-

5/3/2018 11:35:23 AM

Acta Orthopaedica 2018; 89 (3): 272–277

275

Table 3. Overview of results of pain scores and questionnaires in both groups. Values are mean (95% CI)

NRS pain at rest 6 hours 12 hours 18 hours 24 hours 32 hours 40 hours HOOS (at 6 weeks) Symptoms Pain ADL Sports/Rec QOL OHS (at 6 weeks) SF12 (at 6 weeks) PCS MCS

Between group difference Unadjusted Adjusted a

Placebo

Ropivacaine

p-value

4.2 (3.6–4.9) 3.5 (2.9–4.0) 3.6 (3.2–4.1) 2.9 (2.4–3.3) 2.6 (2.2–2.9) 2.0 (1.6–2.5) (n=53) 74 (70–79) 83 (79–87) 77 (73–82) 55 (48–62) 57 (51–62) 13 (11–15)

3.3 (2.6–3.9) 3.6 (2.9–4.4) 3.3 (2.9–3.7) 2.5 (2.1–2.9) 2.2 (1.8–2.6) 2.6 (2.0–3.2) (n=57) 74 (70–79) 82 (78–86) 78 (74–82) 58 (50–65) 61 (55–67) 12 (10–14)

0.9 (0.1 to 1.7) -0.1 (-1.1 to 0.8) 0.3 (-0.3 to 0.9) 0.4 (-0.2 to 0.9) 0.4 (-0.1 to 0.8) -0.5 (-1.3 to 0.2)

0.9 (0.04 to 1.7) -0.2 (-1.1 to 0.8) 0.3 (-0.3 to 0.9) 0.4 (-0.2 to 1.0) 0.4 (-0.1 to 0.8) -0.6 (-1.3 to 0.2)

0.04 0.8 0.3 0.2 0.2 0.1

-0.2 (-6.4 to 6.1) 0.7 (-5.2 to 6.6) -0.7 (-6.8 to 5.4) -2.5 (-12.6 to 7.7) -4.3 (-12.4 to 3.8) 1.1 (-1.7 to3.8)

-0.2 (-6.0 to 6.5) 0.5 (-5.5 to 6.5) -0.8 (-7.0 to 5.4) -2.3 (-12.7 to 8.0) -4.3 (-12.5 to 3.8) 1.1 (-1.7–3.9)

0.9 0.9 0.8 0.7 0.3 0.4

45 (43–47) 45 (43–46)

45 (43–48) 44 (42–46)

-0.6 (-3.7 to 2.6) 0.6 (-2.3 to 3.5)

-0.7 (-3.9 to 2.5) 0.6 (-2.4 to 3.5)

0.7 0.7

a Between-group

differences adjusted for age, gender, and BMI. Only at T = 6 hours is the pain score in the LIA group significantly lower compared with the placebo group with a p-value of 0.04. However, note that the p-value required for significance is < p = 0.008 after Holm’s procedure.

thetic joint infection (placebo group). Finally, the length of stay was not statistically different between the groups (data not shown). Sensitivity analysis revealed similar outcomes, except after 6 hours at which point the difference was not statistically significant anymore.

VAS score 6

Discussion

Placebo Ropivacaine

0 6

Hours after surgey

Figure 2. Pain scores of both groups decline in time postoperatively. a At the first time point the LIA group has a statistically significant lower pain score compared with the placebo group.

cant difference in the use of escape pain medication between the two groups. In addition, no differences are observed in the HOOS, OHS, and SF-12 questionnaires. The occurrence of nausea, vomiting, and urinary retention was equally distributed between the two groups (data not shown). The presence of adverse events was also similar between the two groups, with none of the events related to the infiltration of ropivacaine. The 3 serious adverse events included one subject with postoperative hemorrhage requiring surgical evacuation (ropivacaine group), cardiac ischemia requiring coronary bypass surgery (ropivacaine group), and a subject with early peripros-

12095 Rademakers D.indd 275

We found that a double nerve block of the lateral cutaneous nerve and subcostal nerves does not result in significant pain relief in the postoperative period. No differences in complications or other adverse events were observed. Moreover, the pre-determined primary endpoint of the current trial, a lower pain score after 24 hours, was not met. Although initially a significant reduction of pain scores after 6 hours was found, the difference was lost after using a multiple comparisons procedure to control for the inflation of the type 1 error rate. And even when significance had remained, it was not clinically relevant, since the VAS was only 0.9 point in favor of the LIA group; it has been suggested that differences in VAS become clinically meaningful when exceeding 1.3 points (Gallagher et al. 2001). A well-known limitation of the VAS measurements is the ceiling effect that becomes more significant at later time points due to the low values that approach the values of healthy subjects. Postoperative pain is a complex clinical condition that is still not fully understood. One of the main causes is the local tissue damage due to the surgical approach. This may be the main reason why LIA has reliable outcomes in TKA patients in whom the surgical approach hardly differs between trials.

5/3/2018 11:35:27 AM

276

In THA on the other hand, there is a highly variable surgical preference to approaching the hip, which is reflected by the different approaches used in LIA studies (see Table 1). It is important to mention that local infiltration is performed as part of an effective multimodal pain regimen consisting of acetaminophen, meloxicam and gabapentin. The latter especially is not always well tolerated and may result in early postoperative dizziness and more effective local infiltration might therefore allow a reduction of these medications. We used a straight lateral approach, which was only used in 2 prior studies (Bianconi et al. 2003, Busch et al. 2010). Both studies showed a reduction in postoperative pain by the use of LIA. However, further comparison with our trial is controversial since the groups were much smaller compared with our group and the tissues that were infiltrated differed. Both RCTs used LIA to infiltrate primarily the subcutaneous tissues and Busch et al. also targeted the capsule tissue. Our aim was to show whether the infiltration of the nerves supplying the skin around the incision area is of additional value as a part of traditional LIA. The reason for this type of infiltration is that in our prior observations most patient suffer from superficial wound pain after a THA performed by a straight lateral approach; the groin pain after surgery is mostly very mild. Only a single study infiltrated the lateral cutaneous femoral nerve, but the surgeries in that study were actually performed via a posterolateral approach (Dobie et al. 2012). In that study also, no differences were found in pain and mobilization in the LIA compared with a not further specified control group. The study, however, additionally infiltrated the surgical tissues and had a much smaller sample size than our study. One of the possible limitations of our study is that the suspension contained only ropivacaine. Earlier studies in TKA showed superior results with the addition of an anti-inflammatory agent such as ketorolac (Andersen et al. 2013). Also, studies that used ketorolac in the suspension in THA generally show more favorable results (Andersen et al. 2007a, Andersen et al. 2007b, Busch et al. 2010, Banerjee and McLean, 2011, Kuchalik et al. 2013). Our aim, however, was to block the nerves supplying the surgical area in order to evaluate specifically the pain contribution from these 2 nerves and we did not directly infiltrate the damaged tissues in the surgical field. It might therefore be questioned whether the addition of ketorolac would have been more effective. Moreover, the addition of ketorolac to the LIA suspension can be debated by the absence of NSAID receptors in the tissues around the hip and the agents may act primarily by a general instead of a local effect. We did administer an NSAID to all patients orally postoperatively. The only trial studying the infiltration of the cutaneous femoral nerve as part of LIA used bupivacaine (and adrenalin (epinephrine)) (Dobie et al. 2012). Our primary endpoint was pain reduction after 24 hours. It might be questioned whether ropivacaine is still working at that time point, in view of its half-life of 2–6 hours. However, we surmised if pain was reduced early postoperatively

12095 Rademakers D.indd 276

Acta Orthopaedica 2018; 89 (3): 272–277

this would have resulted in earlier effective mobilization and thereby a longer reduction of pain and earlier discharge. Obviously, choosing an earlier endpoint in the current study would not have resulted in another conclusion since no differences were observed at all. The maximum follow-up was 6 weeks, since we did not expect any differences in the long term. Few studies have looked at the long-term effects of LIA in arthroplasty patients. Interestingly, a recent study showed that LIA in hip arthroplasty patients reduced postsurgical pain after 1 year (Wylde et al. 2015). Because of these findings, they also showed that the LIA administration is more cost effective in hip compared with knee arthroplasty patients (Marques et al. 2015). Since in our study all differences have ceased after 6 hours postoperatively, the interesting findings in these trials seem incompatible with our results. In summary, we show that a double nerve block of the lateral cutaneous nerve and subcostal nerves does not result in a clinically relevant reduction of postoperative pain in THA patients with surgery performed by a straight lateral approach under spinal anesthesia. Further research should focus on the composition and volume of the LIA suspension, the optimal localization of the infiltration, and should be evaluated for every surgical approach separately. The authors would like to thank all the patients for their participation in the study. In addition, the authors are grateful to Paul Spruijt, José de Droog and the orthopedic surgeons performing the surgeries as well as the people involved in the outpatient clinic, operating theater, and clinical department.

JLB and MV designed and coordinated the study, JV was involved in data collection, INS performed the statistical analysis, DdV wrote the pain protocol, HJK performed the randomization, JLB drafted the manuscript and MV revised the manuscript.

Acta thanks Henrik Husted and Johan Rader for help with peer review of this study.

Andersen L O, Kehlet H. Analgesic efficacy of local infiltration analgesia in hip and knee arthroplasty: a systematic review. Br J Anaesth 2014; 113: 360-74. Andersen K, Pfeiffer-Jensen M, Haraldsted V, Søballe, Kjeld. Reduced hospital stay and narcotic consumption, and improved mobilization with local and intraarticular infiltration after hip arthroplasty. Acta Orthop 2007a; 78: 180. Andersen L J, Poulsen T, Krogh B, Nielsen T. Postoperative analgesia in total hip arthroplasty: a randomized double-blinded, placebo-controlled study on peroperative and postoperative ropivacaine, ketorolac, and adrenaline wound infiltration. Acta Orthop 2007b; 78: 187-92. Andersen L Ø, Husted H, Otte K S, Kristensen B B, Kehlet H. High-volume infiltration analgesia in total knee arthroplasty: a randomized, doubleblind, placebo-controlled trial. Acta Anaesthesiol Scand 2008a; 52: 1331-5. Andersen L Ø, Husted H, Otte K S, Kristensen B B, Kehlet H. A compression bandage improves local infiltration analgesia in total knee arthroplasty. Acta Orthop 2008b; 79: 806-11. Andersen L Ø, Gaarn-Larsen L, Kristensen B B, Husted H, Otte K S, Kehlet H. Subacute pain and function after fast-track hip and knee arthroplasty. Anaesthesia 2009; 64: 508-13.

5/3/2018 11:35:28 AM

Acta Orthopaedica 2018; 89 (3): 272–277

Andersen L Ø, Otte K S, Husted H, Gaarn-Larsen L, Kristensen B, Kehlet H. High-volume infiltration analgesia in bilateral hip arthroplasty: a randomized, double-blind placebo-controlled trial. Acta Orthop 2011; 82: 423-6. Andersen K V, Nikolajsen L, Haraldsted V, Odgaard A, Søballe K. Local infiltration analgesia for total knee arthroplasty: should ketorolac be added? Br J Anaesth 2013; 111: 242-8. Banerjee P, McLean C. The efficacy of multimodal high-volume wound infiltration in primary total hip replacement. Orthopedics 2011; 34: e522-9. Bianconi M, Ferraro L, Traina G C, Zanoli G, Antonelli T, Guberti A, et al. Pharmacokinetics and efficacy of ropivacaine continuous wound instillation after joint replacement surgery. Br J Anaesth 2003; 91: 830-5. Busch C A, Whitehouse M R, Shore B J, MacDonald S J, McCalden R W, Bourne R B. The efficacy of periarticular multimodal drug infiltration in total hip arthroplasty. Clin Orthop Relat Res 2010; 468: 2152-9. Dillon J P, Brennan L, Mitchell D. Local infiltration analgesia in hip and knee arthroplasty: an emerging technique. Acta Orthop Belg 2012; 78: 158-63. Dobie I, Bennett D, Spence D J, Murray J M, Beverland D E. Periarticular local anesthesia does not improve pain or mobility after THA. Clin Orthop Relat Res 2012; 470: 1958-65. Gallagher E J, Liebman M, Bijur P E. Prospective validation of clinically important changes in pain severity measured on a visual analog scale. Ann Emerg Med 2001; 38: 633-8. Gibbs D M R, Green T P, Esler C N. The local infiltration of analgesia following total knee replacement: a review of current literature. J Bone Joint Surg Br 2012; 94: 1154-9. den Hartog Y M, Mathijssen N M C, van Dasselaar N T, Langendijk P N J, Vehmeijer S B W. No effect of the infiltration of local anaesthetic for total hip arthroplasty using an anterior approach: a randomised placebo controlled trial. Bone Joint J 2015; 97-B: 734-40. Kuchalik J, Granath B, Ljunggren A, Magnuson A, Lundin A, Gupta A. Postoperative pain relief after total hip arthroplasty: a randomized, double-blind comparison between intrathecal morphine and local infiltration analgesia. Br J Anaesth 2013; 111: 793-9.

12095 Rademakers D.indd 277

277

Lunn T H, Husted H, Solgaard S, Kristensen B B, Otte K S, Kjersgaard A G, et al. Intraoperative local infiltration analgesia for early analgesia after total hip arthroplasty: a randomized, double-blind, placebo-controlled trial. Reg Anesth Pain Med 2011; 36: 424-9. Marques E M, Blom A W, Lenguerrand E, Wylde V, Noble S M. Local anaesthetic wound infiltration in addition to standard anaesthetic regimen in total hip and knee replacement: long-term cost-effectiveness analyses alongside the APEX randomised controlled trials. BMC Med 2015; 13: 151. Murphy T P, Byrne D P, Curtin P, Baker J F, Mulhall K J. Can a periarticular levobupivacaine injection reduce postoperative opiate consumption during primary hip arthroplasty? Clin Orthop Relat Res 2012; 470: 1151-7. Pandazi A, Kanellopoulos I, Kalimeris K, Batistaki C, Nikolakopoulos N, Matsota P, et al. Periarticular infiltration for pain relief after total hip arthroplasty: a comparison with epidural and PCA analgesia. Arch Orthop Trauma Surg 2013; 133: 1607-12. Rikalainen-Salmi R, Förster J G, Mäkelä K, Virolainen P, Leino K A, Pitkänen M T, et al. Local infiltration analgesia with levobupivacaine compared with intrathecal morphine in total hip arthroplasty patients. Acta Anaesthesiol Scand 2012; 56: 695-705. Tran J, Schwarzkopf R. Local infiltration anesthesia with steroids in total knee arthroplasty: a systematic review of randomized control trials. J Orthop 2015; 12: S44-50. Wang Y, Gao F, Sun W, Wang B, Guo W, Li Z. The efficacy of periarticular drug infiltration for postoperative pain after total hip arthroplasty: a systematic review and meta-analysis. Medicine (Baltimore) 2017; 96(12): e6401. Wylde V, Lenguerrand E, Gooberman-Hill R, Beswick A D, Marques E, Noble S, et al. Effect of local anaesthetic infiltration on chronic postsurgical pain after total hip and knee replacement: the APEX randomised controlled trials. Pain 2015; 156: 1161-70. Zoric L, Cuvillon P, Alonso S, Demattei C, Vialles N, Asencio G, et al. Singleshot intraoperative local anaesthetic infiltration does not reduce morphine consumption after total hip arthroplasty: a double-blinded placebo-controlled randomized study. Br J Anaesth 2014; 112: 722-8.

5/3/2018 11:35:28 AM

278

Acta Orthopaedica 2018; 89 (3): 278–288

Revision surgery of metal-on-metal hip arthroplasties for adverse reactions to metal debris A clinical update Gulraj S MATHARU 1, Antti ESKELINEN 2, Andrew JUDGE 1, Hemant G PANDIT 1, and David W MURRAY 1

1 Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Nuffield Orthopaedic Centre, Oxford, United Kingdom; 2 Coxa Hospital for Joint Replacement, Tampere, Finland Correspondence: gsm@doctors.org.uk Submitted 2017-07-11. Accepted 2018-01-16.

Background and purpose — The initial outcomes following metalon-metal hip arthroplasty (MoMHA) revision surgery performed for adverse reactions to metal debris (ARMD) were poor. Furthermore, robust thresholds for performing ARMD revision are lacking. This article is the second of 2. The first article considered the various investigative modalities used during MoMHA patient surveillance (Matharu et al. 2018a). The present article aims to provide a clinical update regarding ARMD revision surgery in MoMHA patients (hip resurfacing and large-diameter MoM total hip arthroplasty), with specific focus on the threshold for performing ARMD revision, the surgical strategy, and the outcomes following revision. Results and interpretation — The outcomes following ARMD revision surgery appear to have improved with time for several reasons, among them the introduction of regular patient surveillance and lowering of the threshold for performing revision. Furthermore, registry data suggest that outcomes following ARMD revision are influenced by modifiable factors (type of revision procedure and bearing surface implanted), meaning surgeons could potentially reduce failure rates. However, additional large multi-center studies are needed to develop robust thresholds for performing ARMD revision surgery, which will guide surgeons’ treatment of MoMHA patients. The long-term systemic effects of metal ion exposure in patients with these implants must also be investigated, which will help establish whether there are any systemic reasons to recommend revision of MoMHAs ■

Many metal-on-metal hip arthroplasties (MoMHAs) have been implanted worldwide in the form of hip resurfacing arthroplasty (HRA) and total hip arthroplasty (THA) (Bozic et al.

2009, NJR 2016). Despite the high failure rates of MoMHAs (Smith et al. 2012a, 2012c), it is estimated that at least 80% of these implants remain in situ worldwide (AOANJRR 2016, NJR 2016). Adverse reactions to metal debris (ARMD) represent an almost unique mode of arthroplasty failure associated with MoMHAs. ARMD is very different from conventional modes of arthroplasty failure, such as dislocation, loosening, and infection, given the associated soft-tissue problems, the potentially progressive and destructive disease nature, as well as its development in patients with asymptomatic and seemingly well-functioning MoMHAs (Grammatopoulos et al. 2009, Liddle et al. 2013, Matharu et al. 2016c). Given the prevalence of ARMD revision surgery is increasing (Liddle et al. 2013, AOANJRR 2016, Matharu et al. 2016b, NJR 2016) it is expected that many more MoMHA patients will undergo future revision. However a systematic review suggested that there was little good quality evidence available regarding the outcomes following MoMHA revision surgery performed for ARMD (Matharu et al. 2014a), which represents the commonest cause of revision (Matharu et al. 2016b). This therefore makes it difficult for surgeons to counsel patients about the risks of undergoing further procedures. Furthermore, robust thresholds for performing revision for ARMD in MoMHA patients have not been established due to a lack of evidence (Matharu et al. 2015). This is reflected in the variable recommendations proposed by worldwide regulatory authorities for considering MoMHA revision surgery (Canada 2012, MHRA 2012, Therapeutic Goods Administration 2012, FDA 2013, Hannemann et al. 2013). Knowledge of any prognostic factors of outcome following ARMD revision would assist surgeons when making decisions about the threshold (when to recommend revision) and type of revision

11904 Matharu D.indd 278

5/3/2018 11:36:54 AM

Acta Orthopaedica 2018; 89 (3): 278–288

surgery (which components require revision, and which bearing surface/fixation methods to use) required in MoMHA patients with ARMD. This article is the second of 2, which aims to provide a clinical update on the investigation and management of MoMHA patients. The first article considered the various investigative modalities used during MoMHA surveillance, with specific focus on blood metal ion sampling and imaging (Matharu et al. 2018a). The present article considers the threshold for performing revision, the surgical strategy, and the outcomes following ARMD revision surgery when performed in patients with MoM HRAs and MoM THAs with femoral head sizes of 36 mm or greater.

Threshold for ARMD revision surgery The poor outcomes initially reported following ARMD revision surgery (Grammatopoulos et al. 2009, de Steiger et al. 2010) led regulatory authorities and surgeons to widely recommend performing early revision in MoMHAs with ARMD (Grammatopoulos et al. 2009, De Smet et al. 2011, Haddad et al. 2011, MHRA 2012, FDA 2013). However, there are currently no robust thresholds for performing ARMD revision surgery, because of lack of evidence. Therefore surgeons have difficulty when managing patients with ARMD: which patients to revise and which to keep under surveillance? The limited evidence base and problems clinicians experience is highlighted by expert opinion (level 5 evidence) (Oxford 2011) commonly being used to manage these patients (Hannemann et al. 2013, Berber et al. 2015). In 2012, one UK center introduced an internet-based multidisciplinary team to assist other centers in managing their MoMHA patients with cases referred electronically (Berber et al. 2015). The multidisciplinary team consists of experts (specialist surgeons and radiologists), who meet weekly to recommend management decisions based on their experience, regulatory guidance (MHRA 2012, FDA 2013), and the latest evidence. The proposed management recommendations have been implemented by referring institutions in 92% of cases (Berber et al. 2015). For some cases management is clear. Timely revision surgery is needed for symptomatic patients with pseudotumors that are solid, large, invasive, and destructive to the soft tissues and bone, sometimes with neurovascular damage (Pandit et al. 2008, Grammatopoulos et al. 2009, Langton et al. 2010, Langton et al. 2013, Liddle et al. 2013). Similarly, asymptomatic patients with normal investigations (no imaging abnormalities and blood metal ion levels below 2 µg/L) are also straightforward and do not require revision. This leaves a lot of patients where management remains uncertain. Such patients include those with moderate to no symptoms with abnormal investigations, which can include non-destructive cystic pseudotumors and/or moderately raised blood metal ion concentrations (Almousa et al. 2013, Hasegawa et al. 2014, Goldstein et al.

11904 Matharu D.indd 279

279

2016, Kwon et al. 2016, Matharu et al. 2016c). This uncertainty is highlighted by a recent study where 10 MoMHA clinical scenarios were used to assess the management decisions made by international experts from 6 centers (Berber et al. 2016). Agreement was inconsistent between centers specifically when managing patients with raised or rising blood metal ion concentrations, cystic pseudotumors, and peri-acetabular osteolysis (Berber et al. 2016). The issue of systemic disease in MoMHA patients continues to generate interest. The long-term effects of high concentrations of cobalt and chromium in the body remain unknown. Patients with ARMD may develop systemic symptoms due to exposure to high cobalt and chromium concentrations. A review reported that MoMHA patients with systemic features had median serum cobalt concentrations of 35 (14–288) µg/L, with symptoms often resolving after revision to a non-MoM articulation (Zywiel et al. 2016). Systemic features appear to be extremely rare but can be divided into neurological (hearing and visual impairment/loss, peripheral neuropathy, and cognitive impairment), cardiovascular (cardiomyopathy, breathlessness), and endocrine (hypothyroidism, fatigue, malaise, depression) (Tower 2010, Bradberry et al. 2014, Cheung et al. 2016, Zywiel et al. 2016). Deaths due to cardiac failure secondary to cobalt toxicity have been reported in MoMHA patients, even following revision (Gilbert et al. 2013, Martin et al. 2015). Analysis of Australian Veterans data reported an association between ASR XL MoM THAs and hospital admission for heart failure in elderly male patients (Gillam et al. 2017). However, a larger study using National Joint Registry (NJR) data from England and Wales demonstrated that MoMHA patients were not at increased risk of heart failure compared with non-MoMHA patients (Sabah et al. 2018), with the same authors reporting that MoMHA patients with high blood metal ion concentrations undergoing comprehensive cardiac investigations had no detectable heart pathology (Berber et al. 2017). Although an established relationship exists between high metal ion exposure and the development of certain cancers in the occupational setting (Keegan et al. 2007), population data from the NJR and Finnish Arthroplasty Register have yet to demonstrate any increased risk of cancer or mortality in MoMHA patients compared with conventional THA patients at shortterm follow-up (Makela et al. 2012, McMinn et al. 2012, Smith et al. 2012b, Kendal et al. 2013, Makela et al. 2014). Currently regulatory authorities do not make recommendations for considering revision surgery in MoMHA patients presenting with systemic symptoms (Matharu et al. 2015).

Intraoperative findings and surgical strategy The surgical management of ARMD has evolved over time, with the heterogeneity of ARMD not being appreciated initially (Grammatopoulos et al. 2009, De Smet et al. 2011,

5/3/2018 11:36:54 AM

280

Liddle et al. 2013). Therefore MoMHA revisions performed for ARMD can range from simple to more complex cases (Liddle et al. 2013). Simple cases to manage surgically include synovitis with minimal tissue damage and metallosis, with a satisfactory reconstruction usually achieved with primary conventional THA implants. Furthermore, when revising MoM THAs some surgeons have advised retaining acetabular and/or femoral components when these are well fixed and well positioned, with taper adapters recommended if femoral tapers are not severely damaged (Munro et al. 2014, Lainiala et al. 2015, Plummer et al. 2016). In such instances procedures may be performed where either 1 MoMHA component is revised (acetabular or femoral), or both components are retained with exchange of the modular components only (femoral head and acetabular liner, with or without the use of a taper adapter). Complex cases include large pseudotumors that are destructive to the peri-prosthetic bone and soft tissues, which pass through tissue planes and also involve vital neurovascular structures (such as the femoral vessels or the sciatic nerve). Such complex cases are likely to require extensive reconstruction with revision THA implants with or without augments. Occasionally staged revision procedures for ARMD may be considered for severe cases or when infection is also a concern. In addition, surgical expertise from other specialties (plastics, vascular, orthopedic oncology, or pelvic surgeons) may also be required depending on the lesion anatomy, nature of the destruction, and the neurovascular involvement. This highlights the importance of performing detailed preoperative investigations, and planning for surgeons with other expertise to be available at revision as well as the necessary reconstructive implants (Liddle et al. 2013). Dealing with large soft-tissue lesions around MoMHAs represents a new challenge for arthroplasty surgeons. Given the potential risk of ARMD recurrence (Grammatopoulos et al. 2009, Matharu et al. 2014a), the aim of revision surgery is to excise all metal debris within the soft tissues, including pseudotumors, similar to an oncological resection. However, in reality this aim can become compromised depending on the lesion anatomy, namely if it involves vital neurovascular structures and/or the primary soft tissues contributing to hip stability. There is little guidance available regarding the specific type of reconstruction to manage ARMD revisions in MoMHAs, namely which components to revise, and which component fixation methods and bearing surfaces to implant. On the basis of the limited current available evidence we have made some recommendations about these particular aspects given the authors have significant experience with revision of MoMHAs for ARMD (Matharu et al. 2014b, Lainiala et al. 2015, Matharu et al. 2016b). Selective component revisions are simpler surgically compared with revising the entire construct, especially in MoM THAs. Theoretically single component and modular compo-

11904 Matharu D.indd 280

Acta Orthopaedica 2018; 89 (3): 278â&#x20AC;&#x201C;288

nent only revisions reduce the risks associated with removing the acetabular and/or femoral components. However even well-fixed components, especially the acetabulum, should be removed if they are clearly malpositioned (De Smet et al. 2011). On the femoral side special attention should be paid to assess the trunnion in MoM THAs; if it is severely worn the use of a taper adapter is not appropriate. Instead the stem should be removed, with extended trochanteric osteotomies sometimes required to facilitate the removal of a well-fixed femoral component. Therefore the use of single component or modular component only revisions should not be overused, with recent NJR data reporting that employing the latter strategy when revising MoM THAs with ARMD was associated with twice the risk of re-revision compared with all component revisions and compared with hips undergoing acetabular component only revisions (Matharu et al. 2017a). Instability and infection were the commonest reasons for re-revision following these modular component only exchanges (Matharu et al. 2017a). No studies have yet shown the superiority of one type of fixation method over another following ARMD revision. Uncemented implants have been favored at ARMD revision, presumably because patients who received MoMHAs were young and active (De Smet et al. 2011, Matharu et al. 2017a). If concerns exist about osseointegration of the revision prosthesis on the femoral side, which may be due to associated bone loss or elderly patients with osteoporotic bone, we suggest using cemented primary implants or revision implants (often uncemented), with proximal femoral replacements sometimes needed in severe cases. Uncemented revision implants with or without augments, pelvic plating, and/or cages are often used for managing bone loss on the acetabular side (Liddle et al. 2013, Munro et al. 2014, Lainiala et al. 2015). Recent NJR data in MoMHAs revised for ARMD suggested acetabular bone grafting was associated with twice the risk of re-revision compared with not grafting, with the authors suggesting that the higher failures associated with bone grafting may be either due to the need for more complex reconstructions thus necessitating the graft, or because of problems with the graft itself (such as infection) (Matharu et al. 2017a). It is universally accepted that ARMD revisions should receive a non-MoM bearing surface, given the potential for high wear from MoM bearings (Kwon et al. 2010, Langton et al. 2010) and the poor outcomes subsequently reported when MoM bearings were used at revision (Liddle et al. 2013, Matharu et al. 2014b, Pritchett 2014). Recent NJR data observed that, in ARMD revisions, implanting ceramic-on-ceramic bearing surfaces was associated with an 86% increased risk of re-revision compared with ceramic-on-polyethylene bearings; the outcomes for metal-on-polyethylene were not significantly different when compared with either of these ceramic bearing couples (Matharu et al. 2017a). However, this study was not able to perform further analyses on the type of polyethylene (standard vs. highly cross-linked) as this particular information was not available for the authors to assess.

5/3/2018 11:36:54 AM

Acta Orthopaedica 2018; 89 (3): 278â&#x20AC;&#x201C;288

In cases with well-functioning abductors and good soft-tissue balance, we currently recommend using ceramic femoral heads combined with highly cross-linked polyethylene liners. This strategy also minimizes the risk of further metal wear debris and/or corrosion that can be generated when using metal-on-polyethylene revision bearings, which can be a source of ARMD recurrence (Munro et al. 2014, Whitehouse et al. 2015, Plummer et al. 2016). When stability is compromised because of soft-tissue problems, options include largehead ceramic-on-ceramic bearing surfaces, large ceramic heads with thin cross-linked polyethylene liners, dual mobility cups, or constrained acetabular liners (De Smet et al. 2011, Liddle et al. 2013, Pritchett 2014, Lainiala et al. 2015). In these complex cases with poor soft tissues there is no compelling evidence to recommend any of these options over the others (Jones 2016). However, retaining the original 1-piece MoM acetabular component (with a worn internal surface) and combining it with a Dual Mobility head could potentially increase the risk of polyethylene wear.

Outcomes following ARMD revision surgery Evidence available and patient demographics A systematic review undertaken in 2013 identified 6 studies reporting the outcomes (complications, re-revision surgery, and post-revision functional outcomes) following 216 MoMHA revisions performed for ARMD (Matharu et al. 2014a). This review was updated to include eligible studies up to the end of 2016 using the methods described (see Supplementary data). Including the 6 initial studies a total of 15 unique studies were eligible for inclusion with 803 MoMHA revisions for ARMD (Grammatopoulos et al. 2009, De Smet et al. 2011, Ebreo et al. 2011, Rajpura et al. 2011, Liddle et al. 2013, Su and Su 2013, Matharu et al. 2014b, Munro et al. 2014, Norris et al. 2014, Pritchett 2014, Cip et al. 2015, Lainiala et al. 2015, Stryker et al. 2015, van Lingen et al. 2015, Liow et al. 2016). 1 further study reported the outcomes following 16 ARMD revisions at extended follow-up (median of 10 years) (Matharu et al. 2017b). All included studies were either case-control or cohort studies graded as level 4 evidence (Oxford 2011), with the majority being retrospective. We acknowledge that a number of registry reports of outcomes following MoMHA revision surgery exist (de Steiger et al. 2010, Wong et al. 2015, Penrose et al. 2016). However, these studies were not considered formally, as they did not stratify outcomes following MoMHA revision according to the reason for revision of the primary MoMHA implant. Included studies had certain limitations such as small sample size, and data quality problems including missing data, which can all make synthesis of the available evidence problematic. 10 of the 15 unique studies involved small cohorts of less than 50 ARMD revisions. Apart from the extended follow-up report (Matharu et al. 2017b) all studies had a mean follow-up

11904 Matharu D.indd 281

281

after ARMD revision of 5 years or shorter, with most having a mean follow-up of 3 years or less. A number of studies did not specifically provide patient demographics and outcomes for the ARMD revision procedures. Some studies reported on complications and re-revision surgery but not on post-revision functional outcomes, or vice versa (Norris et al. 2014, Stryker et al. 2015). The patient demographics and outcomes following ARMD revision surgery for all included studies are summarized in Table 1. Most revisions were performed in females, in young patients (under 60 years), and within 5 years of primary MoMHA. Revisions were almost equally split between HRAs and MoM THAs. The initial systematic review (Matharu et al. 2014a) concluded that there was limited evidence on outcomes following MoM THA revisions performed for ARMD (Munro et al. 2014). Therefore a number of studies have subsequently reported on the outcomes following MoM THA revisions. Defining ARMD The term ARMD was introduced in 2010 as an umbrella term for painful MoMHA failures with one or more of the following features in the absence of another plausible diagnosis (Langton et al. 2010, Langton et al. 2011): large sterile effusions (including pseudotumor), macroscopic tissue necrosis, metallosis and ALVAL (or aseptic lymphocytic vasculitis associated lesions, which refers to a specific lymphocyte-dominated histopathological appearance). The term ARMD has generally been accepted as the most inclusive for describing this heterogeneous condition, which can have varying degrees of disease severity (discussed above in â&#x20AC;&#x153;Intra-operative findings and surgical strategyâ&#x20AC;?). Continued research in this particular field has resulted in the definition for ARMD evolving with time. We would make a diagnosis of ARMD in MoMHAs if there was cross-sectional imaging and intra-operative evidence of a pseudotumor (a cystic, solid, or mixed mass communicating with the hip joint), or if there was no pseudotumor but abnormalities including pathological effusions, significant metallosis, synovitis, tissue damage and/or necrosis (Pandit et al. 2008, Langton et al. 2010, Langton et al. 2011, Hart et al. 2012, Nishii et al. 2012, Lainiala et al. 2015). Ideally this would be supported by appropriate histopathological findings, namely evidence of lymphocytic infiltrates (including ALVAL) and a phagocytic macrophage response to metal wear debris, with or without tissue necrosis (Willert et al. 2005, Campbell et al. 2010, Grammatopoulos et al. 2013). However, we recognize that surgery is planned based on preoperative investigations with any histopathological diagnosis of ARMD typically available only after revision has been performed. ARMD can also coexist with other abnormalities (such as loosening, osteolysis, instability) in the same way, for example, infection can coexist with other abnormalities (De Smet et al. 2011, Matharu et al. 2014b). Of the 15 unique studies reviewed, which reported outcomes following ARMD revision, all provided some defini-

5/3/2018 11:36:54 AM

282

tion for ARMD. However, only 11 studies provided enough details about both the clinical and histopathological features to satisfy our working diagnosis of ARMD (Grammatopoulos et al. 2009, De Smet et al. 2011, Rajpura et al. 2011, Liddle et al. 2013, Matharu et al. 2014b, Munro et al. 2014, Norris et al. 2014, Pritchett 2014, Cip et al. 2015, Lainiala et al. 2015, Liow et al. 2016). No study formally stated how it dealt with multiple revision indications (e.g., by using a hierarchy to establish the primary diagnosis for revision). However, it was clear from the data presented that most studies had reported cases of ARMD alone, as well as ARMD coexisting with other abnormalities (such as loosening, osteolysis, instability). In addition to this variability between studies when diagnosing ARMD it is important to note that the diagnosis can be subjective even when the same criteria are used, with various surgeons potentially interpreting any abnormal features differently. Although reviews of the literature like the present one are inherently limited by the quality and reporting of the published studies reviewed, the variability in how ARMD has been diagnosed in the different studies has important implications when making comparisons between studies and must be considered when interpreting our review of the literature. Complications and re-revision surgery The frequency of complications (up to 68%) and re-revision surgery (up to 38%) were variable amongst the short-term studies but generally high (Table 1). In the first 16 ARMD revisions performed at 1 center there was an increase in the frequency of complications (50% to 69%) and re-revision (38% to 44%) from the initial study at a mean of 3 years follow-up (Grammatopoulos et al. 2009) to the report at a median of 10 years (Matharu et al. 2017b). Some studies did report implant survival rates following ARMD revision, which were 88% at both 3 years (Liow et al. 2016) and 5 years (Matharu et al. 2014b), and 56% at 10 years (Matharu et al. 2017b). Recent data from the NJR for England and Wales reported that following 2,535 ARMD revision procedures in MoMHA patients the 5-year implant survival rate was 90% (Matharu et al. 2017a). This was similar to the 5-year implant survival reported in the NJR following all-cause non-MoMHA revision surgery (88%â&#x20AC;&#x201C;89% depending on bearing surface and component fixation) (NJR 2016). Another recent study comparing outcomes following all-cause acetabular only revisions in MoMHAs with all-cause acetabular only revisions in metalon-polyethylene THAs observed similar incidences of local complications, dislocation, infection, and re-revision between the groups at up to 2-year follow-up (Penrose et al. 2016). Early observations suggested short-term outcomes following ARMD revision were poor, with half of patients sustaining major complications and over one-third requiring further operations (Grammatopoulos et al. 2009). Similar observations were reported in subsequent small cohorts reporting their initial experience following ARMD revision surgery (Rajpura et al. 2011, Munro et al. 2014). Furthermore the frequency of

11904 Matharu D.indd 282

Acta Orthopaedica 2018; 89 (3): 278â&#x20AC;&#x201C;288

complications and re-revision following the first ARMD revision cases were higher compared with MoMHA revisions for non-ARMD indications (fracture, loosening, infection), and compared with matched patients undergoing primary conventional THA (Grammatopoulos et al. 2009). The poor outcomes following ARMD revision were thought to relate to the invasive and destructive nature of these lesions (Pandit et al. 2008, Grammatopoulos et al. 2009, Langton et al. 2010, Haddad et al. 2011). In addition to regular MoMHA patient surveillance these initial observations led regulatory authorities and orthopedic surgeons to widely recommend performing early revision in MoMHAs with ARMD (Grammatopoulos et al. 2009, De Smet et al. 2011, Haddad et al. 2011, MHRA 2012, FDA 2013). It was thought that this strategy would improve outcomes following ARMD revision. Therefore over time surgeons subsequently adopted a lower threshold for performing revision for ARMD. There is evidence to suggest that outcomes following ARMD revision in MoMHAs may have improved with time. The reason for this is likely multifactorial and may include regular patient surveillance, lowering of the threshold for performing revision, increasing surgical experience with ARMD revisions (including employing different strategies to manage soft-tissue and bone damage, and reduce ARMD recurrence), and patients now undergoing revision at longer intervals from primary surgery rather than early after MoMHA. One single surgeon series observed that the frequency of complications and re-revisions were significantly reduced in their latest 31 ARMD revision cases compared with their first 17 revisions (De Smet et al. 2011). In addition to lowering the threshold for performing ARMD revisions in the later cases the surgeon made numerous other changes to their practice including more intensive surveillance with routine blood metal ion sampling, revising both HRA components, implanting larger ceramic femoral head sizes, and postoperatively using anti-dislocation braces. Furthermore outcomes following ARMD revision were not significantly inferior compared with the 65 non-ARMD HRA revisions performed by the same surgeon (De Smet et al. 2011). Although this was contrary to earlier observations (Grammatopoulos et al. 2009), we acknowledge that when this initial cohort were reviewed at extended follow-up there was no longer a difference in the frequency of complications and re-revision following ARMD revisions compared with non-ARMD revisions (Matharu et al. 2017b). 2 of the largest and more recent studies (Table 1) have reported the lowest frequency of complications and re-revision (Pritchett 2014, Lainiala et al. 2015). 1 study was a single surgeon series of 90 revisions (Pritchett 2014), and the other was the largest published study involving 215 ARMD revisions performed by six surgeons (Lainiala et al. 2015). Therefore ARMD revisions performed in specialist centers by experienced and high-volume MoMHA revision surgeons, not surprisingly, seem to also lead to better outcomes,

5/3/2018 11:36:54 AM

Acta Orthopaedica 2018; 89 (3): 278–288

283

Table 1. Studies reporting the outcomes following metal-on-metal hip arthroplasty revision surgery performed for ARMD A

Grammatopoulos et al. (2009)

16 R

100% (16)

51.3 (20–71)

1.6 a (0.01–6.7)

3.0 a (0.8–7.2)

50% (8)

38% (6)

Rajpura et al. (2011)

11 R

De Smet et al. (2011)

48 R

36% (4) 61% (NS) a

53.5 (22–67) 52.5 (18–71)

3.8 (1.3–7.3) 2.7 (0.3–8.4)

1.8 (1.0–3.3) 3.3 a (0.3–10.1)

18% (2) < 23% (11) a

18% (2) < 13% (6) *

Ebreo et al. (2011) Liddle et al. (2013)

42 R+T 32 R

55% (23) 81% (26)

Median 61 (NS) 57.7 (25–74)

4.7 a (1.3–7.8) 4.3 (0.9–10.9)

2.2 a (1.2–4.0) Median 2.5 (1.0–4.5) a

≤ 10% (4) a ≤ 6%

≤ 2% (1) a ≤ 6%

Su and Su (2013)

13 R

Munro et al. (2014)

19 T

85% (11) 37% (NS) a

57.5 (46–76)

2.8 a (0.6–4.9)

2.3 a (0.7–6.7) 2.1 a (0.8–4.0)

≤ 15% (2) a 68% (13)

≤ 15% (2) a 21% (4)

Pritchett (2014)

90 R

48% (43)

49.8 (32–71)

2.8 (1.3–4.9)

5.1 (3.0–9.8)

4% (4)

3% (3)

Matharu et al. (2014b) Norris et al. (2014)

46 R 18 T 35 R

13% (8) NS

Stryker et al. (2015)

58 T

5.5 (1.1–13.8) 4.3 (1.5–9.6) 4.6 a (2.7–6.7) 3.9 a (0.1–9.5)

20% (13) NS

20 T

57.8 (31–79) 58.0 (30–76) 49.6 a (21–61) 60.0 a (17–84)

4.5 (1.0–14.6) NS

Cip et al. (2015)

72% (46) 71% (25) 47% (NS) a 65% (NS) a

2.3 (1.5–3.1) 1.2 a (0–10.2)

10% (2) 20% (23)

5% (1) 16% (18)

60% (130) 69% (NS) a 36% (35)

62.1 (SD 10.1) 63.0 a (44–75) 62.0 (41–85)

4.7 2.3 (SD 1.3) (1.0–NS) Median 3.7 3.1 a (1.0–6.5) a (2.1–4.7) 5.1 2.5 (1.4–18.3) (2.2–4.3)

5% (11) 24% (9) 14% (14)

3% (6) 8% (3) 7% (7)

100% (16)

51.3 (20–71)

1.6 a (0.01–6.7)

Lainiala et al. (2015)

49 R 166 T van Lingen et al. (2015) 38 T Liow et al. (2016)

25 R 77 T

Matharu et al. (2017b) b

16 R

Median 10.3 69% (7–15) a (11)

44% (7)

I Dislocation ± ARMD recurrence (4) Loose cup (2) ARMD recurrence (2) Loose cup or stem (2) Infection (2) ARMD recurrence (1) Infection (1) Dislocation (1) ARMD recurrence with loose cup (1) Infection (2) Dislocation and/or loose cup (3) ARMD recurrence (1) ARMD recurrence (1) Infection (1) Loose cup (1) Dislocation (2) ARMD recurrence (2) NS Infection (1) Infection (7) Loose cup or stem (6) Dislocation (4) Dislocation (4) Infection (1) Dislocation (3) ARMD recurrence (3) Dislocation (2) Loose cup (2) Dislocation ± ARMD recurrence (5) Loose cup (2)

J Mean OHS 20.9 Mean OHS 35.3 Mean HHS 93.1 a Mean OHS 23.7 a Median OHS 36.5 a Mean HHS 96.4 Mean WOMAC (pain) 78 (function) 83 Mean HHS 93.2 Median OHS 39 Mean OHS 33 Mean HHS 85.1 NS Median OHS 40 Mean HOOS 61.9 a Mean HSS 75.6 Median OHS 21

NS = not stated, SD = standard deviation A. Study author and year B. Hips revised for ARMD (adverse reactions to metal debris) R: Resurfacing arthroplasty T: Total hip arthroplasty C. Female hips, % (n) D. Mean age (range) at revision in years E. Mean time to revision (range) in years F. Mean follow-up time after revision (range) in years G. Frequency of complications, % (n) H. Frequency of re-revision, % (n) I. Main reasons for re-revision surgery, % (n) j. Functional outcome: Functional outcome scoring systems: OHS (Oxford Hip Score) = 0–48 (48 best outcome) (Dawson et al. 1996, Murray et al. 2007); HHS (Harris Hip Score) = 0–100 (100 best outcome) (Harris 1969); HOOS (Hip disability and osteoarthritis outcome score) = 0–100 (100 best outcome) (Klassbo et al. 2003); WOMAC (Western Ontario and McMaster Universities Arthritis Index) = 0–100 (0 best outcome) (Bellamy et al. 1988). a Studies did not provide the relevant data specifically for the cohort of patients undergoing revision for adverse reactions to metal debris (but rather for the whole cohort of metal-on-metal hip arthroplasty revisions that they reported on). b Updated report on Grammatopoulos et al. (2009)

which has been postulated previously (De Smet et al. 2011, Liddle et al. 2013). Recent registry data suggest a weak effect between the time interval between primary and revision for ARMD and re-revision rates, with revisions performed early

11904 Matharu D.indd 283

after MoMHA more likely to require re-revision (Matharu et al. 2017a). This is plausible given MoMHAs are scarcely implanted now, thus the time interval between primary and revision for patients with MoMHAs now developing ARMD

5/3/2018 11:36:54 AM

284

is increasing. It is therefore suspected that those presenting with ARMD nowadays will tend to have more benign disease and hence better outcomes following revision compared with the aggressive and destructive lesions observed in MoMHAs which required early revision for ARMD (Grammatopoulos et al. 2009, Munro et al. 2014). A recent NJR analysis of MoMHA revisions compared the subsequent outcomes in 1,288 ARMD revisions matched to 1,288 non-ARMD revision indications (Matharu et al. 2018b). That study observed that patients revised for ARMD had approximately half the risk of re-revision compared with patients undergoing non-ARMD revision surgery. This was contrary to the initial observations (Grammatopoulos et al. 2009, Rajpura et al. 2011, Munro et al. 2014). The authors concluded that surgeons and worldwide regulatory authorities have positively influenced outcomes following ARMD revision by promoting regular patient surveillance and widely recommending a lower threshold for performing revision surgery (Matharu et al. 2018b). Interestingly the highest risk of rerevision surgery was when MoMHAs were revised for infection and dislocation/subluxation (5-year implant survival rates of 81% and 82% respectively) (Matharu et al. 2018b). Reasons for re-revision surgery In line with previous evidence (Matharu et al. 2014a), the commonest indications for re-revision surgery following ARMD revision in the included studies were dislocation, ARMD recurrence, aseptic component loosening, and infection (Table 1). Dislocation can be related to ARMD lesions that are destructive and/or require extensive soft-tissue debridement. Both can compromise the hip abductors and/or short external rotators and therefore lead to hip instability (De Smet et al. 2011). Other risk factors for dislocation include reducing the large-diameter MoM bearing when exchanging to a nonMoM articulation, and not correcting any residual component malposition at ARMD revision (for example when performing modular revisions in MoM THAs) (Stryker et al. 2015). Using large femoral head sizes (>36 mm) at ARMD revision has not prevented dislocations (Munro et al. 2014). This suggests that instability following ARMD revision is a complex problem, with some surgeons using anti-dislocation braces, dual-mobility acetabular components, and an overall lower threshold for performing ARMD revision procedures in an attempt to reduce the subsequent dislocation risk (De Smet et al. 2011, Pritchett 2014). Most studies have reported at least 1 case of ARMD recurrence following ARMD revision (see Table 1). Recurrence can occur due to incomplete excision of metal debris at revision. Reasons for incomplete excision include surgical error and being unable to excise all the metal debris because of its proximity to vital neurovascular structures (Grammatopoulos et al. 2009, Liddle et al. 2013). In such cases serial post-revision cross-sectional imaging can be useful for monitoring progres-

11904 Matharu D.indd 284

Acta Orthopaedica 2018; 89 (3): 278â&#x20AC;&#x201C;288

sion of any residual ARMD (Grammatopoulos et al. 2009, Munro et al. 2014). Recurrence of ARMD can also occur if there is another potential source of metal wear debris and/ or corrosion, including implantation of another MoM bearing (Liddle et al. 2013, Matharu et al. 2014b, Pritchett 2014) or using a cobalt-chrome femoral head on a titanium alloy or cobalt-chrome femoral stem (Munro et al. 2014, Whitehouse et al. 2015, Plummer et al. 2016). In light of these observations most surgeons now use ceramic femoral heads when performing ARMD revisions (Grammatopoulos et al. 2009, De Smet et al. 2011, Liddle et al. 2013), with a taper adapter used if the stemmed femoral component is retained given the theoretical risk of ceramic head fracture when impacted onto retained tapers (Plummer et al. 2016). Aseptic loosening, usually of the acetabular component, may be due to invasive and destructive ARMD causing significant osteolysis and/or iatrogenic bone loss when removing a well-fixed component, which increases the risk of failed implant osseointegration. 1 study experienced a number of failures due to aseptic acetabular loosening despite the absence of bone defects, but this was thought to be due to the fiber metal-backed acetabular components implanted at revision, with better results obtained when porous tantalum components were used (Munro et al. 2014). Risk factors for infection may include multiple operations, incomplete excision of metal debris or necrotic tissue at revision, and the retention of primary MoMHA components (Liddle et al. 2013, Munro et al. 2014). Functional outcomes Patient-reported outcomes after the initial ARMD revisions were poor (Grammatopoulos et al. 2009, Ebreo et al. 2011, Munro et al. 2014), with these also being inferior compared with MoMHA revisions for non-ARMD indications, and compared with primary conventional THAs (Grammatopoulos et al. 2009). The poor functional outcomes reported in an early study of 16 ARMD revisions (Grammatopoulos et al. 2009) persisted at extended follow-up, and remained inferior compared with MoMHA revisions for non-ARMD indications (Matharu et al. 2017b). These poor functional outcomes were also considered a consequence of the invasive and destructive nature of ARMD lesions, with a lower threshold for performing such revision procedures widely adopted (Grammatopoulos et al. 2009, De Smet et al. 2011, Haddad et al. 2011, MHRA 2012, FDA 2013). The majority of subsequent studies have reported good or excellent functional outcomes following ARMD revision (De Smet et al. 2011, Rajpura et al. 2011, Liddle et al. 2013, Su and Su 2013, Matharu et al. 2014b, Pritchett 2014, Cip et al. 2015, Lainiala et al. 2015), defined as a mean Oxford Hip Score (OHS) of 34 or above or a Harris Hip Score (HHS) of 80 or above (Harris 1969, Kalairajah et al. 2005, Murray et al. 2007). However, not all of the more recent studies achieved good functional outcomes following ARMD revision (Norris

5/3/2018 11:36:54 AM

Acta Orthopaedica 2018; 89 (3): 278â&#x20AC;&#x201C;288

et al. 2014, van Lingen et al. 2015, Liow et al. 2016), which suggests other factors may influence outcomes after these procedures. Predicting re-revision and guiding surgical decisions Identifying prognostic factors of outcome following ARMD revision would provide surgeons with thresholds for performing revision and help guide decisions regarding the type of reconstruction to perform. However, most of the evidence regarding factors predictive of poor outcomes following ARMD revision surgery is limited to small studies that were not adequately powered to identify prognostic factors (De Smet et al. 2011, Liddle et al. 2013, Matharu et al. 2014b, Liow et al. 2016). In some of the earliest MoMHAs revised for ARMD a significantly higher risk of re-revision and inferior functional outcomes were reported when another MoM bearing was implanted at revision (Liddle et al. 2013, Matharu et al. 2014b, Pritchett 2014). This is suspected to be due to the continued source of metal wear debris and/or corrosion, and has led to using non-MoM bearing surfaces at ARMD revision. Solid pseudotumors have been a risk factor for re-revision surgery following ARMD revision (Liddle et al. 2013). Similarly a recent study of 102 ARMD revisions identified MARS-MRI evidence of solid lesions with abductor deficiency to be predictive of post-revision complications, with pre-revision radiographic implant loosening also predicting poor outcomes (Liow et al. 2016). However, this study did not find other factors to be predictive of post-revision complications, such as pre-revision blood metal ions, type of revision articulation, and femoral head size (Liow et al. 2016). Recent analysis of NJR data identified 4 predictors of rerevision surgery following ARMD revision in MoMHAs: high BMI at revision, modular component only THA revisions (exchange of the head and liner only with/without taper adapter), ceramic-on-ceramic revision bearings, and acetabular bone grafting (Matharu et al. 2017a). The authors concluded that as these predictors included modifiable factors (type of THA revision procedure, revision bearing surface, and possibly the use of acetabular bone graft) surgeons could potentially reduce failure rates further following ARMD revision. By contrast, smaller studies from the Australian Joint Replacement Registry have reported that re-revision following aseptic MoM HRA revision surgery was not influenced by the type of revision procedure performed or the revision bearing surface implanted (de Steiger et al. 2010, Wong et al. 2015).

Conclusions Numerous studies have now reported on the short- to mediumterm risks associated with ARMD revision surgery in MoMHA patients, which can be used to counsel patients informatively pre-revision. Evidence suggests that outcomes following

11904 Matharu D.indd 285

285

ARMD revision may have improved with time. The reason for this is multifactorial and may include regular patient surveillance, lowering of the threshold for performing revision, surgical experience with ARMD revisions, and patients now undergoing revision at longer intervals from primary surgery with such cases potentially being less severe compared with revisions performed early after MoMHA. However, we consider that the threshold for performing ARMD revision surgery need not be lowered much further as this introduces the potential for surgical risk to outweigh any benefits. By contrast, robust thresholds for performing ARMD revision surgery are lacking. Although non-registry studies have attempted to identify predictors of cross-sectional imaging progression (Reito et al. 2014, Briant-Evans et al. 2015, Kwon et al. 2016, Matharu et al. 2016a), and predictors of poor outcomes following ARMD revision (De Smet et al. 2011, Liddle et al. 2013, Matharu et al. 2014b, Dimitriou et al. 2016), these have been short-term studies that were underpowered for identifying predictors. Recent registry data suggest that modifiable factors exist which may allow surgeons to reduce failure rates further following ARMD revision (Matharu et al. 2017a), although these require validation. Future studies should focus on obtaining good quality evidence to develop robust thresholds for performing ARMD revision surgery and to inform the surgical strategy at ARMD revision, which will both assist surgeons when managing MoMHA patients. Most of these questions require large welldesigned multi-center studies with extended follow-up that can assess potentially important predictors of complications, re-revision, and functional outcomes following ARMD revision surgery (such as pre-revision blood metal ion levels and cross-sectional imaging findings). These issues simply cannot be addressed using current arthroplasty registries, which do not collect such data. Multi-center studies would provide important information on the role of blood metal ions and cross-sectional imaging in predicting disease severity, tissue destruction, and outcomes following ARMD revision. This information can then be used to guide thresholds for performing ARMD revision. Both multi-center studies and retrospective registry cohorts are required to inform the surgical strategy at ARMD revision, which will include establishing the optimal implant fixation methods, the optimal bearing surface (comparing highly cross-linked polyethylene articulating with ceramic vs. metal heads), and the management of instability in association with soft-tissue problems (dual mobility cups vs. constrained acetabular liners). Finally, the long-term systemic effects of metal ion exposure in MoMHA patients must also be investigated, particularly the potential oncological and cardiac consequences. This will also help establish whether there are any systemic reasons to recommend revision of MoMHAs in the absence of local ARMD. Addressing the outlined research questions would allow a more evidence-based approach for the management of MoMHA patients who may require revision for ARMD.

5/3/2018 11:36:54 AM

286

Supplementary data The Appendix is available in the online version of this article, http://dx.doi.org/ 10.1080/17453674.2018.1440455.

GSM: literature review, manuscript draft, and revision. AJ, AE, HGP, DWM: contributed to the literature review, and manuscript revision.

The authors would like to thank the following organizations for funding this research: Arthritis Research UK (grant reference number 21006), the Royal College of Surgeons of England, the Orthopaedics Trust and the Royal Orthopaedic Hospital Hip Research and Education Charitable Fund.

GM has undertaken medico-legal work for Leigh Day, which includes work relating to metal-on-metal hip replacements. AJ has undertaken medico-legal work for Freshfields Bruckhaus Deringer, which includes work relating to metal-on-metal hip replacements. HGP provides expert testimony to Kennedys Law, which includes work relating to metal-on-metal hip replacements. None of the other authors have any conflicts of interest relating specifically to this work. No commercial companies were involved in the planning of this review, analysis and interpretation of data, or writing of the manuscript. DWM has undertaken medico-legal work for Herbert Smith Freehills, which includes work relating to metal-on-metal hip replacements

Acta thanks Johan Kärrholm and other anonymous reviewers for help with peer review of this study.

Almousa S A, Greidanus N V, Masri B A, Duncan C P, Garbuz D S. The natural history of inflammatory pseudotumors in asymptomatic patients after metal-on-metal hip arthroplasty. Clin Orthop Relat Res 2013; 471(12): 3814-21. AOANJRR. Australian Orthopaedic Association National Joint Replacement Registry(AOANJRR) Hip, Knee & Shoulder Arthroplasty Annual Report. 2016: https://aoanjrr.sahmri.com/annual-reports-2016. Bellamy N, Buchanan W W, Goldsmith C H, Campbell J, Stitt L W. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol 1988; 15(12): 1833-40. Berber R, Pappas Y, Khoo M, Miles J, Carrington R, Skinner J, Hart A. A new approach to managing patients with problematic metal hip implants: the use of an internet-enhanced multidisciplinary team meeting: AAOS exhibit selection. J Bone Joint Surg Am 2015; 97(4): e20. Berber R, Skinner J, Board T, Kendoff D, Eskelinen A, Kwon Y M, Padgett D E, Hart A, IsccoMh. International metal-on-metal multidisciplinary teams: do we manage patients with metal-on-metal hip arthroplasty in the same way? An analysis from the International Specialist Centre Collaboration on MOM Hips(ISCCoMH). Bone Joint J 2016; 98-B(2): 179-86. Berber R, Abdel-Gadir A, Rosmini S, Captur G, Nordin S, Culotta V, Palla L, Kellman P, Lloyd G W, Skinner J A, Moon J C, Manisty C, Hart A J. Assessing for cardiotoxicity from metal-on-metal hip implants with advanced multimodality imaging techniques. J Bone Joint Surg Am 2017; 99(21): 1827-35. Bozic K J, Kurtz S, Lau E, Ong K, Chiu V, Vail T P, Rubash H E, Berry D J. The epidemiology of bearing surface usage in total hip arthroplasty in the United States. J Bone Joint Surg Am 2009; 91(7): 1614-20. Bradberry S M, Wilkinson J M, Ferner R E. Systemic toxicity related to metal hip prostheses. Clin Toxicol (Phila) 2014; 52(8): 837-47.

11904 Matharu D.indd 286

Acta Orthopaedica 2018; 89 (3): 278–288

Briant-Evans T W, Lyle N, Barbur S, Hauptfleisch J, Amess R, Pearce A R, Conn K S, Stranks G J, Britton J M. A longitudinal study of MARS MRI scanning of soft-tissue lesions around metal-on-metal total hip arthroplasties and disease progression. Bone Joint J 2015; 97-B(10): 1328-37. Campbell P, Ebramzadeh E, Nelson S, Takamura K, De Smet K, Amstutz H C. Histological features of pseudotumor-like tissues from metal-on-metal hips. Clin Orthop Relat Res 2010; 468(9): 2321-7. Canada. Health Canada. Metal-on-metal hip implants—information for orthopaedic surgeons regarding patient management following surgery— for health professionals. 2012: http://healthycanadians.gc.ca/recall-alertrappel-avis/hc-sc/2012/15835a-eng.php. Cheung A C, Banerjee S, Cherian J J, Wong F, Butany J, Gilbert C, Overgaard C, Syed K, Zywiel M G, Jacobs J J, Mont M A. Systemic cobalt toxicity from total hip arthroplasties: review of a rare condition, Part 1: history, mechanism, measurements, and pathophysiology. Bone Joint J 2016; 98-B(1): 6-13. Cip J, Bach C, Widemschek M, Luegmair M, Martin A. Revision of articular surface replacement (ASR) total hip arthroplasty: correlation of perioperative data and early post-revision outcome results. J Arthroplasty 2015; 30(9): 1607-17. Dawson J, Fitzpatrick R, Carr A, Murray D. Questionnaire on the perceptions of patients about total hip replacement. J Bone Joint Surg Br 1996; 78(2): 185-90. De Smet K A, Van Der Straeten C, Van Orsouw M, Doubi R, Backers K, Grammatopoulos G. Revisions of metal-on-metal hip resurfacing: lessons learned and improved outcome. Orthop Clin North Am 2011; 42(2): 25969, ix. de Steiger R N, Miller L N, Prosser G H, Graves S E, Davidson D C, Stanford T E. Poor outcome of revised resurfacing hip arthroplasty. Acta Orthop 2010; 81(1): 72-6. Dimitriou D, Liow M H, Tsai T Y, Leone W A, Li G, Kwon Y M. Early outcomes of revision surgery for taper corrosion of dual taper total hip arthroplasty in 187 patients. J Arthroplasty 2016; 31(7): 1549-54. Ebreo D, Khan A, El-Meligy M, Armstrong C, Peter V. Metal ion levels decrease after revision for metallosis arising from large-diameter metal-onmetal hip arthroplasty. Acta Orthop Belg 2011; 77(6): 777-81. FDA. U.S. Food and Drug Administration (FDA). Medical Devices. Metalon-metal hip implants. information for orthopedic surgeons. 2013: http:// www.fda.gov/MedicalDevices/ProductsandMedicalProcedures/ImplantsandProsthetics/MetalonMetalHipImplants/ucm241667.htm. Gilbert C J, Cheung A, Butany J, Zywiel M G, Syed K, McDonald M, Wong F, Overgaard C. Hip pain and heart failure: the missing link. Can J Cardiol 2013; 29(5): 639 e1-2. Gillam M H, Pratt N L, Inacio M C, Roughead E E, Shakib S, Nicholls S J, Graves S E. Heart failure after conventional metal-on-metal hip replacements. Acta Orthop 2017; 88(1): 2-9. Goldstein J M, Fehring T K, Fehring K A. Cystic adverse local tissue reactions in asymptomatic modular metal-on-metal total hips may decrease over time. J Arthroplasty 2016; 31(7): 1589-94. Grammatopoulos G, Pandit H, Kwon Y M, Gundle R, McLardy-Smith P, Beard D J, Murray D W, Gill H S. Hip resurfacings revised for inflammatory pseudotumour have a poor outcome. J Bone Joint Surg Br 2009; 91(8): 1019-24. Grammatopoulos G, Pandit H, Kamali A, Maggiani F, Glyn-Jones S, Gill H S, Murray D W, Athanasou N. The correlation of wear with histological features after failed hip resurfacing arthroplasty. J Bone Joint Surg Am 2013; 95(12): e81. Haddad F S, Thakrar R R, Hart A J, Skinner J A, Nargol A V, Nolan J F, Gill H S, Murray D W, Blom A W, Case C P. Metal-on-metal bearings: the evidence so far. J Bone Joint Surg Br 2011; 93(5): 572-9. Hannemann F, Hartmann A, Schmitt J, Lutzner J, Seidler A, Campbell P, Delaunay C P, Drexler H, Ettema H B, Garcia-Cimbrelo E, Huberti H, Knahr K, Kunze J, Langton D J, Lauer W, Learmonth I, Lohmann C H, Morlock M, Wimmer M A, Zagra L, Gunther K P. European multidisciplinary consensus statement on the use and monitoring of metal-on-metal bearings for total hip replacement and hip resurfacing. Orthop Traumatol Surg Res 2013; 99(3): 263-71.

5/3/2018 11:36:54 AM

Acta Orthopaedica 2018; 89 (3): 278–288

Harris W H. Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty. An end-result study using a new method of result evaluation. J Bone Joint Surg Am 1969; 51(4): 737-55. Hart A J, Satchithananda K, Liddle A D, Sabah S A, McRobbie D, Henckel J, Cobb J P, Skinner J A, Mitchell A W. Pseudotumors in association with well-functioning metal-on-metal hip prostheses: a case-control study using three-dimensional computed tomography and magnetic resonance imaging. J Bone Joint Surg Am 2012; 94(4): 317-25. Hasegawa M, Miyamoto N, Miyazaki S, Wakabayashi H, Sudo A. Longitudinal magnetic resonance imaging of pseudotumors following metal-onmetal total hip arthroplasty. J Arthroplasty 2014; 29(12): 2236-8. Jones S A. CORR Insights ((R)): Complications are not increased with acetabular revision of metal-on-metal total hip arthroplasty. Clin Orthop Relat Res 2016; 474(10): 2143-4. Kalairajah Y, Azurza K, Hulme C, Molloy S, Drabu K J. Health outcome measures in the evaluation of total hip arthroplasties: a comparison between the Harris hip score and the Oxford hip score. J Arthroplasty 2005; 20(8): 1037-41. Keegan G M, Learmonth I D, Case C P. Orthopaedic metals and their potential toxicity in the arthroplasty patient: a review of current knowledge and future strategies. J Bone Joint Surg Br 2007; 89(5): 567-73. Kendal A R, Prieto-Alhambra D, Arden N K, Carr A, Judge A. Mortality rates at 10 years after metal-on-metal hip resurfacing compared with total hip replacement in England: retrospective cohort analysis of hospital episode statistics. BMJ 2013; 347: f6549. Klassbo M, Larsson E, Mannevik E. Hip disability and osteoarthritis outcome score: an extension of the Western Ontario and McMaster Universities Osteoarthritis Index. Scand J Rheumatol 2003; 32(1): 46-51. Kwon Y M, Glyn-Jones S, Simpson D J, Kamali A, McLardy-Smith P, Gill H S, Murray D W. Analysis of wear of retrieved metal-on-metal hip resurfacing implants revised due to pseudotumours. J Bone Joint Surg Br 2010; 92(3): 356-61. Kwon Y M, Liow M H, Dimitriou D, Tsai T Y, Freiberg A A, Rubash H E. What is the natural history of “asymptomatic” pseudotumours in metal-onmetal hip arthroplasty? Minimum 4-year metal artifact reduction sequence magnetic resonance imaging longitudinal study. J Arthroplasty 2016; 31(9 Suppl): 121-6. Lainiala O, Reito A, Elo P, Pajamaki J, Puolakka T, Eskelinen A. Revision of metal-on-metal hip prostheses results in marked reduction of blood cobalt and chromium ion concentrations. Clin Orthop Relat Res 2015; 473(7): 2305-13. Langton D J, Jameson S S, Joyce T J, Hallab N J, Natu S, Nargol A V. Early failure of metal-on-metal bearings in hip resurfacing and large-diameter total hip replacement: a consequence of excess wear. J Bone Joint Surg Br 2010; 92(1): 38-46. Langton D J, Jameson S S, Joyce T J, Gandhi J N, Sidaginamale R, Mereddy P, Lord J, Nargol A V. Accelerating failure rate of the ASR total hip replacement. J Bone Joint Surg Br 2011; 93(8): 1011-16. Langton D J, Sidaginamale R P, Joyce T J, Natu S, Blain P, Jefferson R D, Rushton S, Nargol A V. The clinical implications of elevated blood metal ion concentrations in asymptomatic patients with MoM hip resurfacings: a cohort study. BMJ Open 2013; 3(3): e001541. Liddle A D, Satchithananda K, Henckel J, Sabah S A, Vipulendran K V, Lewis A, Skinner J A, Mitchell A W, Hart A J. Revision of metal-on-metal hip arthroplasty in a tertiary center: a prospective study of 39 hips with between 1 and 4 years of follow-up. Acta Orthop 2013; 84(3): 237-45. Liow M H, Dimitriou D, Tsai T Y, Kwon Y M. Preoperative risk factors associated with poor outcomes of revision surgery for “pseudotumors” in patients with metal-on-metal hip arthroplasty. J Arthroplasty 2016; 31(12): 2835-42. Makela K T, Visuri T, Pulkkinen P, Eskelinen A, Remes V, Virolainen P, Junnila M, Pukkala E. Risk of cancer with metal-on-metal hip replacements: population based study. BMJ 2012; 345: e4646. Makela K T, Visuri T, Pulkkinen P, Eskelinen A, Remes V, Virolainen P, Junnila M, Pukkala E. Cancer incidence and cause-specific mortality in patients with metal-on-metal hip replacements in Finland. Acta Orthop 2014; 85(1): 32-8.

11904 Matharu D.indd 287

287

Martin J R, Spencer-Gardner L, Camp C L, Stulak J M, Sierra R J. Cardiac cobaltism: a rare complication after bilateral metal-on-metal total hip arthroplasty. Arthroplast Today 2015; 1(4): 99-102. Matharu G S, Pynsent P B, Dunlop D J. Revision of metal-on-metal hip replacements and resurfacings for adverse reaction to metal debris: a systematic review of outcomes. Hip Int 2014a; 24(4): 311-20. Matharu G S, Pynsent P B, Sumathi V P, Mittal S, Buckley C D, Dunlop D J, Revell P A, Revell M P. Predictors of time to revision and clinical outcomes following revision of metal-on-metal hip replacements for adverse reaction to metal debris. Bone Joint J 2014b; 96-B(12): 1600-9. Matharu G S, Mellon S J, Murray D W, Pandit H G. Follow-up of metalon-metal hip arthroplasty patients is currently not evidence based or cost effective. J Arthroplasty 2015; 30(8): 1317-23. Matharu G S, Janardhan S, Brash L, Pynsent P B, Dunlop D J, James S L. The utility of repeat ultrasound imaging in the follow-up of metal-on-metal hip arthroplasty patients. Ann R Coll Surg Engl 2016a; 98(2): 143-9. Matharu G S, Judge A, Murray D W, Pandit H G. Prevalence of and risk factors for hip resurfacing revision: a cohort study into the second decade after the operation. J Bone Joint Surg Am 2016b; 98(17): 1444-52. Matharu G S, Ostlere S J, Pandit H G, Murray D W. What is the natural history of asymptomatic pseudotumours in metal-on-metal hip resurfacing patients? Hip Int 2016c; 26(6): 522-30. Matharu G S, Judge A, Pandit H G, Murray D W. Which factors influence the rate of failure following metal-on-metal hip arthroplasty revision surgery performed for adverse reactions to metal debris? An analysis from the National Joint Registry for England and Wales. Bone Joint J 2017a; 99-B(8): 1020-27. Matharu G S, Pandit H G, Murray D W. Poor survivorship and frequent complications at a median of 10 years after metal-on-metal hip resurfacing revision. Clin Orthop Relat Res 2017b; 475(2): 304-14. Matharu G S, Judge A, Eskelinen A, Murray D W, Pandit H G. What is appropriate surveillance for metal-on-metal hip arthroplasty patients? A clinical update. Acta Orthop 2018a; 89 (1): 29–39. Matharu G S, Judge A, Murray D W, Pandit H G. Outcomes after metal-onmetal hip revision surgery depend on the reason for failure: a propensity score matched study. Clin Orthop Relat Res 2018b; 476 (2): 245–58. McMinn D J, Snell K I, Daniel J, Treacy R B, Pynsent P B, Riley RD. Mortality and implant revision rates of hip arthroplasty in patients with osteoarthritis: registry based cohort study. BMJ 2012; 344: e3319. MHRA. Medicines & Healthcare products Regulatory Agency (MHRA). Medical Device Alert: all metal-on-metal (MoM) hip replacements. MDA/2012/036. 2012: http://www.mhra.gov.uk/. Munro J T, Masri B A, Duncan C P, Garbuz D S. High complication rate after revision of large-head metal-on-metal total hip arthroplasty. Clin Orthop Relat Res 2014; 472(2): 523-8. Murray D W, Fitzpatrick R, Rogers K, Pandit H, Beard D J, Carr A J, Dawson J. The use of the Oxford hip and knee scores. J Bone Joint Surg Br 2007; 89(8): 1010-14. Nishii T, Sakai T, Takao M, Yoshikawa H, Sugano N. Ultrasound screening of periarticular soft tissue abnormality around metal-on-metal bearings. J Arthroplasty 2012; 27(6): 895-900. NJR. National Joint Registry (NJR) for England, Wales, Northern Ireland and the Isle of Man 13th Annual Report. 2016: http://www.njrcentre.org.uk/njrcentre/Portals/0/Documents/England/Reports/13th Annual Report/07950 NJR Annual Report 2016 ONLINE REPORT.pdf. Norris R J, McArthur J, Parsons H, Smith N A, Sprowson A P, Foguet P. A case series of 35 hip revisions for adverse reactions to metal debris following Cormet hip resurfacing. Hip Int 2014; 24(6): 575-81. Oxford. Oxford Centre for Evidence-Based Medicine Levels of Evidence Working Group. Oxford Centre for Evidence-Based Medicine Levels of Evidence. 2011: http://www.cebm.net/wp-content/uploads/2014/06/ CEBM-Levels-of-Evidence-2.1.pdf. Pandit H, Glyn-Jones S, McLardy-Smith P, Gundle R, Whitwell D, Gibbons C L, Ostlere S, Athanasou N, Gill H S, Murray D W. Pseudotumours associated with metal-on-metal hip resurfacings. J Bone Joint Surg Br 2008; 90(7): 847-51.

5/3/2018 11:36:54 AM

288

Penrose C T, Seyler T M, Wellman S S, Bolognesi M P, Lachiewicz P F. Complications are not increased with acetabular revision of metal-on-metal total hip arthroplasty. Clin Orthop Relat Res 2016; 474(10): 2134-42. Plummer D R, Berger R A, Paprosky W G, Sporer S M, Jacobs J J, Della Valle C J. Diagnosis and management of adverse local tissue reactions secondary to corrosion at the head-neck junction in patients with metal on polyethylene bearings. J Arthroplasty 2016; 31(1): 264-8. Pritchett J W. One-component revision of failed hip resurfacing from adverse reaction to metal wear debris. J Arthroplasty 2014; 29(1): 219-24. Rajpura A, Porter M L, Gambhir A K, Freemont A J, Board T N. Clinical experience of revision of metal on metal hip arthroplasty for aseptic lymphocyte dominated vasculitis associated lesions (ALVAL). Hip Int 2011; 21(1): 43-51. Reito A, Elo P, Puolakka T, Pajamaki J, Nieminen J, Eskelinen A. Repeated magnetic resonance imaging in 154 hips with large-diameter metal-onmetal hip replacement. Acta Orthop 2014; 85(6): 570-6. Sabah S A, Moon J C, Jenkins-Jones S, Morgan C L, Currie C J, Wilkinson J M, Porter M, Captur G, Henckel J, Chaturvedi N, Kay P, Skinner J A, Hart A H, Manisty C. The risk of cardiac failure following metal-on-metal hip arthroplasty. Bone Joint J 2018; 100-B(1): 20-7. Smith A J, Dieppe P, Howard P W, Blom A W, National Joint Registry for England and Wales. Failure rates of metal-on-metal hip resurfacings: analysis of data from the National Joint Registry for England and Wales. Lancet 2012a; 380(9855): 1759-66. Smith A J, Dieppe P, Porter M, Blom A W, National Joint Registry of England and Wales. Risk of cancer in first seven years after metal-on-metal hip replacement compared with other bearings and general population: linkage study between the National Joint Registry of England and Wales and hospital episode statistics. BMJ 2012b; 344: e2383. Smith A J, Dieppe P, Vernon K, Porter M, Blom A W, National Joint Registry of England and Wales. Failure rates of stemmed metal-on-metal hip replacements: analysis of data from the National Joint Registry of England and Wales. Lancet 2012c; 379(9822): 1199-204.

11904 Matharu D.indd 288

Acta Orthopaedica 2018; 89 (3): 278â&#x20AC;&#x201C;288

Stryker L S, Odum S M, Fehring T K, Springer B D. Revisions of monoblock metal-on-metal THAs have high early complication rates. Clin Orthop Relat Res 2015; 473(2): 469-74. Su E P, Su S L. Surface replacement conversion: results depend upon reason for revision. Bone Joint J 2013; 95-B(11 Suppl A): 88-91. Therapeutic Goods Administration DoH, Australian Government. Metal-onmetal hip replacement implantsâ&#x20AC;&#x201D;information for general practitioners, orthopaedic surgeons and other health professionals. 2012: http://www. tga.gov.au/hp/information-devices-mom-hip-implants.htm. Tower S S. Arthroprosthetic cobaltism: neurological and cardiac manifestations in two patients with metal-on-metal arthroplasty: a case report. J Bone Joint Surg Am 2010; 92(17): 2847-51. van Lingen C P, Ettema H B, Bosker B H, Verheyen C C. Revision of a single type of large metal head metal-on-metal hip prosthesis. Hip Int 2015; 25(3): 221-6. Whitehouse M R, Endo M, Zachara S, Nielsen T O, Greidanus N V, Masri B A, Garbuz D S, Duncan C P. Adverse local tissue reactions in metal-onpolyethylene total hip arthroplasty due to trunnion corrosion: the risk of misdiagnosis. Bone Joint J 2015; 97-B(8): 1024-30. Willert H G, Buchhorn G H, Fayyazi A, Flury R, Windler M, Koster G, Lohmann C H. Metal-on-metal bearings and hypersensitivity in patients with artificial hip joints: a clinical and histomorphological study. J Bone Joint Surg Am 2005; 87(1): 28-36. Wong J M, Liu Y L, Graves S, de Steiger R. What is the rerevision rate after revising a hip resurfacing arthroplasty? Analysis From the AOANJRR. Clin Orthop Relat Res 2015; 473(11): 3458-64. Zywiel M G, Cherian J J, Banerjee S, Cheung A C, Wong F, Butany J, Gilbert C, Overgaard C, Syed K, Jacobs J J, Mont M A. Systemic cobalt toxicity from total hip arthroplasties: review of a rare condition, Part 2: measurement, risk factors, and step-wise approach to treatment. Bone Joint J 2016; 98-B(1): 14-20.

5/3/2018 11:36:54 AM

Acta Orthopaedica 2018; 89 (3): 289–294

289

Influence of surgical approach on complication risk in primary total hip arthroplasty Systematic review and meta-analysis Larry E MILLER 1, Joseph S GONDUSKY 2, Atul F KAMATH 3, Friedrich BOETTNER 4, John WRIGHT 5, and Samir BHATTACHARYYA 5

1 Miller Scientific Consulting, Inc., Asheville, 2 Jordan-Young Institute, Virginia Beach, 3 Penn Medicine, Department of Orthopedic Surgery, Leonard Davis Institute of Health Economics, Philadelphia, 4 Hospital for Special Surgery, New York, 5 DePuy Synthes, Raynham, United States Correspondence: larry@millerscientific.com Submitted 2017-10-14. Accepted 2018-01-18.

Background and purpose — Systematic comparisons of anterior approach (A) versus posterior approach (P) in primary total hip arthroplasty (THA) have largely focused on perioperative outcomes. In this systematic review with meta-analysis, we compared complication risk of A versus P in studies of primary THA with at least 1-year mean follow-up. Patients and methods — We performed a systematic review of prospective and retrospective studies with at least 1-year mean follow-up that reported complications of A and P primary THA. Complications included infection, dislocation, reoperation, thromboembolic event, heterotopic ossification, wound complication, fracture, and nerve injury. Random effects meta-analysis was used for all outcomes. Complication risk was reported as rate ratio (RR) to account for differential follow-up durations; values > 1 indicated higher complication risk with A and values < 1 indicated lower risk with A. Results — 19 studies were included; 15 single-center comparative studies with 6,620 patients (2,278 A; 4,342 P) and 4 multicenter registries with 157,687 patients (18,735 A; 138,952 P). Median follow-up was 16 (12–64) months) with A and 18 (12–110) months with P. Anterior approach was associated with lower rate of infection (RR = 0.55, p = 0.002), dislocation (RR = 0.65, p = 0.03), and reoperation (RR = 0.84, p < 0.001). No statistically significant differences were observed in rate of thromboembolic event (RR = 0.59, p = 0.5), heterotopic ossification (RR = 0.63, p = 0.1), wound complication (RR = 0.93, p = 0.8), or fracture (RR = 1.0, p = 0.9). There was a higher rate of patient-reported nerve injury with A (RR = 2.3, p = 0.01). Interpretation — Comparing A with P in primary THA, A was associated with lower risk of reoperation, dislocation, and infection, but higher risk of patient-reported nerve injury. ■

The durability of total hip arthroplasty (THA) is excellent with 10-year survivorship exceeding 90% (Hailer et al. 2015, Makela et al. 2014). All standard approaches to the hip have been shown to be safe and effective, with certain advantages and disadvantages of each approach (Mjaaland et al. 2017). While the anterior approach (A) has been increasingly used in the United States, little is known about the safety of the A relative to other common surgical approaches. Several groups (Higgins et al. 2015, Meermans et al. 2017, Putananon et al. 2018) have performed systematic reviews comparing the A with the posterior approach (P) in primary THA. However, follow-up durations of the included studies varied widely, with most studies having less than 1-year follow-up. Comparative safety evaluation of these surgical techniques over a longer period is warranted. The purpose of this systematic review with meta-analysis was to compare the complication risk of A versus P in studies with at least 1-year mean follow-up.

Methods Literature search and data extraction In accordance with the PRISMA guidelines, we searched MEDLINE and EMBASE for comparative studies of primary THA performed using the A or P. Therapeutic search terms consisting of THA and total hip arthroplasty were combined with the following surgical approach-specific search terms: anterior, direct, posterior, posterolateral, and Smith-Petersen. We also manually searched the Directory of Open Access Journals (DOAJ), Google Scholar, and the reference lists of included papers and relevant systematic reviews. No language or date restrictions were applied to the searches. The final search was conducted on June 30, 2017.

12203 Miller D.indd 289

03-05-2018 18:23:11

290

Study eligibility was determined by 2 independent researchers (LM, DF). Disagreements were resolved by discussion. Main inclusion criteria included comparison of A versus P in primary THA, predominant diagnosis of osteoarthritis, mean follow-up duration at least 1 year, and extractable complication data. Titles and abstracts were initially screened to exclude review articles, commentaries, letters, case reports, and obviously irrelevant studies. Full-texts of the remaining articles were retrieved and reviewed. Studies were excluded if patients received revision or bilateral THA. When multiple studies included overlapping series of patients, only the study with the largest sample size was included. Data were independently extracted from eligible peer-reviewed articles by the same 2 researchers. Data discrepancies were resolved by discussion.

Acta Orthopaedica 2018; 89 (3): 289–294

Identification Records identified through database searching n = 337 Screening Records screened n = 340 Eligibility Full-text articles assessed for eligibility n = 114

Inclusion Included articles n = 19

Records excluded n = 226 Full-text articles excluded (n = 95): – mean follow-up < 1 year, 27 – no relevant outcomes reported, 25 – no anterior vs. posterior groups, 20 – review paper, 12 – single arm study, 4 – letter/commentary, 3 – cross-sectional study, 1 – duplicate publication, 1 – revision THA, 1 – bilateral THA, 1

PRISMA study flow diagram.

Definitions and outcomes When data were reported at multiple intervals during followup, the final value was extracted for analysis. Complications included infection, dislocation, reoperation (for any reason), thromboembolic event, heterotopic ossification, wound complication, fracture, and nerve injury. To account for differential follow-up durations, complication data were extracted by determining the number of events and then calculating the number of person-years in each group to determine incidence rates. Risk of bias in each study was assessed with the Cochrane Collaboration tool, which included evaluations of sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, and other sources of bias (Higgins et al. 2011). Data analysis We assumed heterogeneous effects among studies a priori and conservatively applied a random effects model for all outcomes. Denominators were adjusted to include the number of patients or hips, as appropriate. The rate ratio (RR) was the effect size statistic of interest, which indicates the ratio of incidence rates (events per person-year) between A and P. A RR value > 1 indicates higher complication incidence rate with A and a value < 1 indicates lower complication incidence rate with A. For each complication, the RR and 95% confidence interval (CI) were calculated in each study and pooled among all studies. Inconsistency in complication risk among studies was quantified with the I2 statistic; values of ≤ 25%, 50%, and ≥ 75% represented low, moderate, and high inconsistency, respectively (Higgins et al. 2003). Publication bias was visually assessed with funnel plots (not shown) and quantitatively assessed using Egger’s regression test. Post hoc random effects meta-regression using the Knapp–Hartung method (Knapp and Hartung 2003) was performed to assess the possible influence of study design, median surgery year,

12203 Miller D.indd 290

Records identified through other sources n=3

inclusion of learning cases, and follow-up duration on complication risk. P-values were 2-sided with a significance level < 0.05. Analyses were performed using Comprehensive Metaanalysis (version 3.3, Biostat, Englewood, NJ, USA). Funding and potential conflicts of interest This work was supported by DePuy Synthes (Raynham, MA, USA). LM received a research grant from DePuy Synthes for data analysis. JW and SB are employees of DePuy Synthes. JG, AK, and FB declare no conflict of interest in this work.

Results Study selection After screening 340 records for eligibility, 19 studies were included in this review, including 15 single-center comparative studies with 6,620 patients (2,278 A; 4,342 P) and 4 multicenter registries with 157,687 patients (18,735 A; 138,952 P). Primary reasons for study exclusion included mean follow-up less than 1 year (27 studies), complications not reported (25 studies), and no comparison of A with P (20 studies) (Figure). Study and patient characteristics This review included 4 randomized controlled trials, 1 prospective nonrandomized study, 10 retrospective studies, and 4 multicenter registries. Surgeries in each group occurred during the same period in 11 studies. In 7 studies, learning curve cases comprised some or all of the A group. Median follow-up duration was 16 months (range: 12–64 months) with A and 18 months (range: 12–110 months) with P. Comparing patients treated with A versus P, baseline patient characteristics were well matched for age (median 63 years per group), female sex (median 60% versus 58%), and BMI (median 28 per group)

03-05-2018 18:23:11

Acta Orthopaedica 2018; 89 (3): 289–294

291

Table 1. Study and patient characteristics

Study Comparative studies: Balasubramaniam et al. 2016 Barrett et al. 2013 Batailler et al. 2017 Fransen et al. 2016 Luo et al. 2016 Malek et al. 2016 Newman et al. 2016 Rathod et al. 2014 Rodriguez et al. 2014 Sugano et al. 2009 Taunton et al. 2014 Tripuraneni et al. 2016 Tsukada and Wakui 2015 Watts et al. 2015 Zhang et al. 2006 Registries Amlie et al. 2014 Mjaaland et al. 2017 Sheth et al. 2015 Zijlstra et al. 2017

Study Treatment design a period

Mean Parallel Learning follow-up, treatment cases months period included A P

RN RCT RN RN RCT RN RN RN PN RN RCT RN RN RN RCT

2006–2011 2010–2011 2013–2015 2012 2014 2010–2014 – 2007–2011 2010 2005–2007 2012 2012–2015 2000–2009 2010–2014 2002–2004

No Yes Yes Yes Yes Yes NR No Yes No Yes Yes No NR Yes

Yes No Yes Yes No No NR No No NR No Yes NR NR NR

12 12 12 12 14 14 12 12 14 14 18 18 24 24 16 30 12 12 24 24 12 12 14 13 64 110 12 12 20 20

RN RN RN RN

2008–2010 2008–2013 2001–2011 2007–2015

Yes Yes No No

No Yes Yes Yes

24 52 36 40

Sample size b A P 50 43 201 45 52 265 235 286 60 33 27 66 139 716 60

Mean age, years A P

Female, % A P

Mean BMI A P

42 44 101 38 52 183 120 293 60 39 27 66 177 3,040 60

63 61 72 64 62 71 63 62 60 56 62 60 67 64 61

57 63 74 63 64 70 59 61 59 57 66 60 62 62 63

50 33 65 67 67 56 54 55 53 88 56 61 90 51 58

67 57 65 63 58 53 57 57 57 92 52 61 83 51 53

31 30 31 29 26 28 25 28 23 24 29 29 29 34 26 26 27 28 23 23 28 29 28 28 23 24 29 30 – –c

30 421 421 52 2,017 5,961 36 1,851 31,747 40 14,446 100,823

67 67 65 –

66 65 66 –

69 67 60 68

64 65 58 68

– – 28 –

– – 29 –

A = anterior approach; P = posterior approach; NR = not reported RCT = randomized controlled trial; RN = retrospective nonrandomized.

a Study design: PN = prospective nonrandomized; b Reported as number of patients or hips. c All patients with BMI ≤ 27 kg/m2.

Table 2. Cochrane risk of bias assessment Study Comparative studies: Balasubramaniam et al. 2016 Barrett et al. 2013 Batailler et al. 2017 Fransen et al. 2016 Luo et al. 2016 Malek et al. 2016 Newman et al. 2016 Rathod et al. 2014 Rodriguez et al. 2014 Sugano et al. 2009 Taunton et al. 2014 Tripuraneni et al. 2016 Tsukada and Wakui 2015 Watts et al. 2015 Zhang et al. 2006 Registries Amlie et al. 2014 Mjaaland et al. 2017 Sheth et al. 2015 Zijlstra et al. 2017

● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

● ● ● ●

Notes: ● low bias risk; ● uncertain bias risk; ● high bias risk. A. Random sequence generation B. Allocation concealment C. Blinding of participants D. Blinding of personnel E. Blinding of outcome assessment F. Incomplete outcome data G. Selective outcome reporting

12203 Miller D.indd 291

(Table 1). The primary risks of bias were attributable to inclusion of retrospective nonrandomized studies (Table 2). Complications The A was associated with lower rates of infection (RR = 0.55, p = 0.002 from 7 studies), dislocation (RR = 0.65, p = 0.03 from 11 studies), and reoperation (RR = 0.84, p < 0.001 from 16 studies). In a subgroup analysis of infection, the rate of superficial (RR = 0.47, p = 0.5) and deep infection (RR = 0.23, p = 0.1) remained low with A, but neither was statistically significant. When explicitly reported, the most common reasons for reoperation were aseptic loosening, dislocation, fracture, and infection in the A group and dislocation, aseptic loosening, infection, and fracture in the P group. No statistically significant differences were observed in the rate of thromboembolic event (RR = 0.59, p = 0.5 from 4 studies), heterotopic ossification (RR = 0.63, p = 0.1 from 4 studies), wound complication (RR = 0.93, p = 0.8 from 5 studies), or fracture (RR = 1.0, p = 0.9 from 10 studies). Most fracture reports were of intraoperative periprosthetic fractures; however, type and time to fracture was not consistently reported. There was a higher rate of patient-reported nerve injury with A vs. P (RR = 2.3, p = 0.01 from 2 studies). Nerve injuries were described as patient-reported sensory deficit (Luo et al. 2016) or patient-reported nerve injury with no distinction between

03-05-2018 18:23:11

292

Acta Orthopaedica 2018; 89 (3): 289–294

Table 3. Complication rates with anterior versus posterior approach in primary total hip arthroplasty

Outcome

Studies

Infection Thromboembolic event Heterotopic ossification Dislocation Reoperation Wound Fracture Patient-reported nerve injury

7 4 4 11 16 5 10 2

Event rate per 100 person-years A P 0.2 0.5 1.5 0.2 0.6 1.7 0.3 3.0

Effect size Rate ratio (95% CI) a

p-value

0.55 (0.38–0.80) 0.59 (0.14–2.43) 0.63 (0.35–1.13) 0.65 (0.44–0.95) 0.84 (0.75–0.93) 0.93 (0.54–1.63) 1.02 (0.75–1.38) 2.31 (1.22–4.39)

0.002 0.5 0.1 0.03 < 0.001 0.8 0.9 0.01

0.4 1.1 2.3 0.2 0.7 1.9 0.1 1.3

Heterogeneity Publication bias (I2), % (Egger’s p-value) 0 0 0 17 0 0 0 0

0.5 0.2 0.3 0.5 1.0 0.4 0.2 b

Notes: A = anterior approach; P = posterior approach. a Rate ratio >1 indicates higher complication incidence rate with anterior approach; rate ratio < 1 indicates lower complication incidence rate with anterior approach. b Inadequate number of studies to calculate value.

Table 4. Subgroup analysis of study design on complication rates with anterior versus posterior approach in primary total hip arthroplasty

Outcome

Comparative studies Rate ratio Studies (95% CI) a

Infection 6 Thromboembolic event 4 Heterotopic ossification 3 Dislocation 8 Reoperation 12 Wound 5 Fracture 9 Patient-reported nerve injury 1

0.66 (0.16–2.7) 0.59 (0.14–2.4) 0.58 (0.30–1.2) 0.55 (0.17–1.8) 1.03 (0.60–1.8) 0.93 (0.54–1.6) 1.7 (0.79–3.7) 5.0 (0.24–104)

Registries Rate ratio Studies (95% CI) a 1 0 1 3 4 0 1 1

0.55 (0.37–0.80) – 0.81 (0.24–2.7) 0.74 (0.39–1.4) 0.83 (0.72–0.95) – 0.93 (0.66–1.3) 2.2 (1.2–4.3)

p-value b 0.8 – 0.6 0.7 0.5 – 0.2 0.6

a Rate

ratio > 1 indicates higher complication incidence rate with anterior approach; RR < 1 indicates lower complication incidence rate with anterior approach. b Comparison of rate ratio in comparative studies versus registries, derived from Knapp– Hartung random effects meta-regression model.

sensory and motor involvement (Amlie et al. 2014). For each complication, heterogeneity among studies was low and publication bias was not evident (Table 3). Post hoc meta-regression Post hoc meta-regression was performed to assess the possible influence of study design, median surgery year, inclusion of learning cases, and follow-up duration on complication risk. No covariate was statistically significantly associated with the risk of any complication. In comparative studies, there was no statistically significant difference between A vs. P in the rate of any complication. In registries, the rate of patient-reported nerve injury was higher with A while the rates of infection and reoperation were lower with A (Table 4).

12203 Miller D.indd 292

Discussion We conducted a systematic review and meta-analysis of comparative studies of A versus P primary THA with at least 1-year mean follow-up. An anterior approach was associated with a lower risk of reoperation, dislocation, and infection, but higher risk of patient-reported nerve injury. No difference was seen in the rate of thromboembolic event, heterotopic ossification, wound complication, or fracture. While heterogeneity or publication bias was not evident for any outcome, the possibility of such influences cannot be ruled out given the small number of studies reporting each complication. A criticism of the A in primary THA is the presence of a learning curve, during which complication rates may be elevated. In an analysis of over 5,000 THA procedures, 50 or more A procedures were required to overcome the learning

03-05-2018 18:23:11

Acta Orthopaedica 2018; 89 (3): 289–294

curve (de Steiger et al. 2015). In a single-surgeon experience with the first 500 A cases, the most dramatic reduction in complication rates occurred after the first 100 cases (Hartford and Bellino 2017). We identified no substantial influence of learning case inclusion on complication rates in meta-regression although this analysis was limited since it was not possible to determine the percentage of the entire sample comprising learning cases. We identified a higher rate of patient-reported nerve injury with A. In the study of Amlie et al. (2014), nerve injury was self-reported in 5.9% of A patients at 24 months follow-up and 3.3% of P patients at 30 months follow-up; however, there was no distinction between sensory or motor involvement. In another comparative study (Luo et al. 2016), sensory deficit was 3.8% with A and 0% with P at 14 months’ follow-up. While comparative nerve injury data were limited to these 2 studies, a high incidence of sensory deficit with A has been reported in other studies (Bhargava et al. 2010, Goulding et al. 2010). This is primarily attributable to likely iatrogenic injury of the lateral cutaneous femoral nerve. Despite the higher patient-reported nerve injury rate with A, long-term functional limitations or higher reoperation rates are unlikely with these events based on the findings from other studies (Bhargava et al. 2010, Goulding et al. 2010). In a meta-analysis comparing A and P (Higgins et al. 2015), there were no group differences in risk of intraoperative fracture and lower risk of dislocation with A. More recently, a systematic review compared anterior, posterior, and lateral approaches in primary THA (Meermans et al. 2017). In that review, complications were not systematically evaluated although the authors concluded that there were similar rates of complications between surgical approaches. In a network meta-analysis of randomized controlled trials (Putananon et al. 2018), complication risk was reported to be lower with P vs. A (1.0% vs. 1.4%); however, specific complications were not described. Among these reviews, follow-up duration varied considerably and was generally less than 1 year. Key differences in our meta-analysis are inclusion of only those studies with mean follow-up of at least 1 year, reporting of multiple specific complications, and statistical adjustment to account for differential follow-up periods among studies. Several aspects of our meta-analysis are novel including the longest duration follow-up of any A versus P review and a comprehensive assessment of complication rates. There are also several limitations. First, despite the longest mean follow-up of any review on this topic, it must be acknowledged that data derived from 16 (A) to 18 (P) months median follow-up must be considered preliminary. Further, while the RR statistic allows for group comparison of event rates on a common scale (per person-year), event rates that are nonconstant with respect to time may complicate interpretation of these results. Second, while osteoarthritis was the predominant diagnosis in each study, reporting of THA indications was inconsistent and may have confounded outcomes.

12203 Miller D.indd 293

293

Third, due to the small number of studies reporting certain complications, some complication estimates reported in this review may change with the addition of data by future studies. Further, the influence of study design on complication rates should be interpreted cautiously given the small number of studies for subgroup comparisons. Fourth, complication reporting was generally inconsistent among studies. Adherence to standardized complication reporting guidelines would greatly improve data transparency and consistency in the THA literature. Fifth, no conclusions regarding complication risk with anterolateral or lateral approaches in THA may be derived from this review. Finally, 14 of 19 included studies were retrospective in nature, which are inherently prone to bias. In summary, comparing A with P in primary THA, A was associated with a lower rate of reoperation, dislocation, and infection, but a higher rate of patient-reported nerve injury.

Conception and design: LM, SB. Data collection: LM. Data analysis: LM. Writing the article: LM. Critical revision of the article: LM, JG, AK, FB, JW, SB

The authors would like to thank David Fay, PhD for assistance with literature review.

Acta thanks Johan Kärrholm and other anonymous reviewers for help with peer review of this study

Amlie E, Havelin L I, Furnes O, Baste V, Nordsletten L, Hovik O, Dimmen S. Worse patient-reported outcome after lateral approach than after anterior and posterolateral approach in primary hip arthroplasty. A cross-sectional questionnaire study of 1,476 patients 1–3 years after surgery. Acta Orthop 2014; 85(5): 463-9. Balasubramaniam U, Dowsey M, Ma F, Dunin A, Choong P. Functional and clinical outcomes 244 following anterior hip replacement: a 5-year comparative study versus posterior approach. ANZ J Surg 2016; 86 (7-8): 58993. Barrett W P, Turner S E, Leopold J P. Prospective randomized study of direct anterior vs postero-lateral approach for total hip arthroplasty. J Arthroplasty 2013; 28(9): 1634-8. Batailler C, Fary C, Batailler P, Servien E, Neyret P, Lustig S. Total hip arthroplasty using direct anterior approach and dual mobility cup: safe and efficient strategy against post-operative dislocation. Int Orthop 2017; 41 (3): 499-506. Bhargava T, Goytia R N, Jones L C, Hungerford M W. Lateral femoral cutaneous nerve impairment after direct anterior approach for total hip arthroplasty. Orthopedics 2010; 33(7): 472. de Steiger R N, Lorimer M, Solomon M. What is the learning curve for the anterior approach for total hip arthroplasty? Clin Orthop Relat Res 2015; 473(12): 3860-6. Fransen B, Hoozemans M, Vos S. Direct anterior approach versus posterolateral approach in total hip arthroplasty : one surgeon, two approaches. Acta Orthop Belg 2016; 82 (2): 240-8. Goulding K, Beaule P E, Kim P R, Fazekas A. Incidence of lateral femoral cutaneous nerve neuropraxia after anterior approach hip arthroplasty. Clin Orthop Relat Res 2010; 468(9): 2397-404.

03-05-2018 18:23:11

294

Hailer N P, Lazarinis S, Makela K T, Eskelinen A, Fenstad A M, Hallan G, Havelin L, Overgaard S, Pedersen A B, Mehnert F, Karrholm J. Hydroxyapatite coating does not improve uncemented stem survival after total hip arthroplasty! Acta Orthop 2015; 86(1): 18-25. Hartford J M, Bellino M J. The learning curve for the direct anterior approach for total hip arthroplasty: a single surgeon’s first 500 cases. Hip Int 2017; 27(5): 483-8. Higgins B T, Barlow D R, Heagerty N E, Lin T J. Anterior vs. posterior approach for total hip arthroplasty, a systematic review and meta-analysis. J Arthroplasty 2015; 30(3): 419-34. Higgins J P, Altman D G, Gotzsche P C, Juni P, Moher D, Oxman A D, Savovic J, Schulz K F, Weeks L, Sterne J A, Cochrane Bias Methods G, Cochrane Statistical Methods G. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ 2011; 343: d5928. Higgins JP , Thompson S G, Deeks J J, Altman D G. Measuring inconsistency in meta-analyses. BMJ 2003; 327(7414): 557-60. Knapp G, Hartung J. Improved tests for a random effects meta-regression with a single covariate. Stat Med 2003; 22(17): 2693-710. Luo Z, Chen M, Shangxi F, Hu F, Ni Z, Cheng P, Ji X, Wu K, Zhang X. [Comparison of clinical efficacy of total hip arthroplasty and lateral posterior approach for total hip arthroplasty]. Chinese J Med Sci 2016; 96: 2807-12. Makela K T, Matilainen M, Pulkkinen P, Fenstad A M, Havelin L I, Engesaeter L, Furnes O, Overgaard S, Pedersen A B, Karrholm J, Malchau H, Garellick G, Ranstam J, Eskelinen A. Countrywise results of total hip replacement: an analysis of 438,733 hips based on the Nordic Arthroplasty Register Association database. Acta Orthop 2014; 85(2): 107-16. Malek I A, Royce G, Bhatti S U, Whittaker J P, Phillips S P, Wilson I R, et al. A comparison 287 between the direct anterior and posterior approaches for total hip arthroplasty: the role of an ‘Enhanced Recovery’ pathway. Bone Joint J 2016; 98-B (6): 754-60. Meermans G, Konan S, Das R, Volpin A, Haddad F S. The direct anterior approach in total hip arthroplasty: a systematic review of the literature. Bone Joint J 2017; 99-B(6): 732-40. Mjaaland K E, Svenningsen S, Fenstad A M, Havelin L I, Furnes O, Nordsletten L. Implant survival after minimally invasive anterior or anterolateral vs. conventional posterior or direct lateral approach: an analysis of 21,860 total hip arthroplasties from the Norwegian Arthroplasty Register (2008 to 2013). J Bone Joint Surg Am 2017; 99(10): 840-7. Newman E A, Holst D C, Bracey D N, Russell G B, Lang J E. Incidence of heterotopic ossification in direct anterior vs posterior approach to total hip arthroplasty: a 298 retrospective radiographic review. Int Orthop 2016; 40 (9): 1967-73.

12203 Miller D.indd 294

Acta Orthopaedica 2018; 89 (3): 289–294

Putananon C, Tuchinda H, Arirachakaran A, Wongsak S, Narinsorasak T, Kongtharvonskul J. Comparison of direct anterior, lateral, posterior and posterior-2 approaches in total hip arthroplasty: network meta-analysis. Eur J Orthop Surg Traumatol 2018; 28(2): 255-67. Rathod P A, Bhalla S, Deshmukh A J, Rodriguez J A. Does fluoroscopy with anterior hip arthroplasty decrease acetabular cup variability compared with a nonguided posterior approach? Clin Orthop Relat Res 2014; 472 (6): 1877-85. Rodriguez J A, Deshmukh A J, Rathod P A, Greiz M L, Deshmane P P, Hepinstall M S, et al. Does the direct anterior approach in THA offer faster rehabilitation and comparable safety to the posterior approach? Clin Orthop Relat Res 2014; 472 (2): 455-63. Sheth D, Cafri G, Inacio M C, Paxton E W, Namba R S. Anterior and anterolateral approaches for THA are associated with lower dislocation risk without higher revision risk. Clin Orthop Relat Res 2015; 473 (11): 3401-8. Sugano N, Takao M, Sakai T, Nishii T, Miki H, Nakamura N. Comparison of mini-incision total hip arthroplasty through an anterior approach and a posterior approach using navigation. Orthop Clin North Am 2009; 40 (3): 365-70. Taunton M J, Mason J B, Odum S M, Springer B D. Direct anterior total hip arthroplasty yields more rapid voluntary cessation of all walking aids: a prospective, randomized clinical trial. J Arthroplasty 2014; 29 (9 Suppl): 169-72. Tripuraneni K R, Munson N R, Archibeck M J, Carothers J T. Acetabular abduction and dislocations in direct anterior vs posterior total hip arthroplasty: a retrospective, matched cohort study. J Arthroplasty 2016; 31(10): 2299-302. Tsukada S, Wakui M. Lower dislocation rate following total hip arthroplasty via direct anterior approach than via posterior approach: five-year-average follow-up results. Open Orthop J 2015; 9: 157-62. Watts C D, Houdek M T, Wagner E R, Sculco P K, Chalmers B P, Taunton M J. High risk of wound complications following direct anterior total hip arthroplasty in obese patients. J Arthroplasty 2015; 30 (12): 2296-8. Zhang X L, Wang Q, Jiang Y, Zeng B F. [Minimally invasive total hip arthroplasty with anterior incision]. Zhonghua Wai Ke Za Zhi 2006; 44 (8): 512-5. Zijlstra W P, De Hartog B, Van Steenbergen L N, Scheurs B W, Nelissen R. Effect of femoral head size and surgical approach on risk of revision for dislocation after total hip arthroplasty. Acta Orthop 2017; 88(4): 395-401.

03-05-2018 18:23:11

Acta Orthopaedica 2018; 89 (3): 295–301

295

A randomized controlled trial on maximal strength training in 60 patients undergoing total hip arthroplasty Implementing maximal strength training into clinical practice Siri B WINTHER 1,2, Olav A FOSS 1,2, Otto S HUSBY 1,2, Tina S WIK 1,2, Jomar KLAKSVIK 1, and Vigdis S HUSBY 3

1 Orthopaedic Research Centre, Department of Orthopaedic Surgery, Clinic of Orthopaedics, Rheumatology and Dermatology, St Olavs Hospital HF, Trondheim; 2 Department of Neuromedicine and Movement Science, Faculty of Medicine and Health Science, Norwegian University of Science and Technology NTNU, Trondheim; 3 Department of Mental Health, Faculty of Medicine and Health Science, Norwegian University of Science and Technology NTNU, Postbox 8905 MTFS, NO-7491, Trondheim, Norway Correspondence: Siri.bjorgen@ntnu.no Submitted 2017-10-10. Accepted 2018-01-29.

Background and purpose — Total hip arthroplasty (THA) patients have reduced muscle strength after rehabilitation. In a previous efficacy trial, 4 weeks’ early supervised maximal strength training (MST) increased muscle strength in unilateral THA patients < 65 years. We have now evaluated muscle strength in an MST and in a conventional physiotherapy (CP) group after rehabilitation in regular clinical practice. Patients and methods — 60 primary THA patients were randomized to MST or CP between August 2015 and February 2016. The MST group trained at 85–90% of their maximal capacity in leg press and abduction of the operated leg (4×5 repetitions), 3 times a week at a municipal physiotherapy institute up to 3 months postoperatively. The CP group followed a training program designed by their respective physiotherapist, mainly exercises performed with low or no external loads. Patients were tested pre- 3, 6, and 12 months postoperatively. Primary outcomes were abduction and leg press strength at 3 months. Other parameters evaluated were pain, 6-min walk test, Harris Hip Score (HHS) and Hip disability and Osteoarthritis Outcome Score (HOOS) Physical Function Short-form score. Results — 27 patients in each group completed the intervention. MST patients were substantially stronger in leg press and abduction than CP patients 3 (43 kg and 3 kg respectively) and 6 months (30 kg and 3 kg respectively) postoperatively (p < 0.002). 1 year postoperatively, no intergroup differences were found. No other statistically significant intergroup differences were found. Interpretation — MST increases muscle strength more than CP in THA patients up to 6 months postoperatively, after 3 months’ rehabilitation in clinical practice. It was well tolerated by the THA patients and seems feasible to conduct within regular clinical practice. ■

Patients with osteoarthritis have reduced muscle strength in the affected limb preoperatively (Rasch et al. 2007), and muscle strength decreases further in the first postoperative week after total hip arthroplasty (THA) (Winther et al. 2016). The muscle strength is still reduced following the initial few weeks post THA (Holm et al. 2013, Winther et al. 2016). Reduced muscle mass, strength, and functionality seem to persist after completed rehabilitation (Reardon et al. 2001, Bertocci et al. 2004, Judd et al. 2014), in some studies up to years after surgery (Sicard-Rosenbaum et al. 2002, Rasch et al. 2010). Leg muscle strength affects ambulatory status following THA (Nankaku et al. 2014), and patients with the best muscle strength have higher physical function, quality of life scores, and lower pain levels (Rosenlund et al. 2016). Muscle strength is related to beneficial effects on functional performance such as chair raising, stair climbing, and gait performance (Samuel et al. 2012, Buirs et al. 2016, Unhjem et al. 2017), and seems important for minimizing postoperative limping (Horstmann et al. 2013). The prolonged deficits in muscle strength of the operated leg after THA indicate a potential for improvements in postoperative care (Judd et al. 2014). It is recommended that persistent asymmetries in hip flexor muscles should receive focused attention during rehabilitation (Friesenbichler et al. 2017) and that muscle-strengthening exercises should be continued for at least 1 year after THA (Shih et al. 1994). Although muscle strength of the operated leg is considered an important postoperative outcome, what specifically constitutes the optimal rehabilitation program after THA remains unclear (Westby et al. 2014). Conventional physiotherapy (CP) consisting of exercises with low or no external load is frequently used in rehabilitation after THA; however, the effect has been questioned

12187 Winther D.indd 295

5/3/2018 11:39:31 AM

296

Acta Orthopaedica 2018; 89 (3): 295–301

(Minns Lowe et al. 2009, Husted Assessed for eligibility Enrollment n = 80 2012). Physiotherapist-directed rehabilitation appears similarly effective Excluded (n = 20): performed unsupervised at home com– declined to participate, 11 – did not meet inclusion criteria, 5 pared with supervised by a physiother– other reasons, 4 apist in an outpatient setting (Coulter et al. 2013). Reasonably there is great Randomized potential for improvement as weightn = 60 bearing exercises and progressive Allocation strength training have shown favorable outcomes (Di Monaco and Castiglioni Allocated to MST (n = 31): Allocated to CP (n = 29): Did not receive allocated intervention (n = 1): Did not receive allocated intervention (n = 1): 2013, Skoffer et al. 2015). – personal reason – personal reason In a Fast-track perspective, it is suggested that studies should focus on Intervention period of 3 months early, intense postoperative strengthAnalysis ening (Rasch et al. 2010, Husted 2012, Holm et al. 2013, Judd et al. 2014). Analyzed at 3 months postoperatively (n = 27): Analyzed at 3 months postoperatively (n = 27): Immediate full weight bearing after Lost to follow-up, discontinued training (n = 3): Lost to follow-up, discontinued training (n = 1): – knee problem 1 week postoperatively, 1 – reoperated THA is safe (Wolf et al. 2010), and – pain, not related to training, 1 strength exercises do not exacerbate – personal reason, 1 postoperative pain (Mikkelsen et al. 2017). This enables patients to start Analyzed at 6 months postoperatively (n = 26): Analyzed at 6 months postoperatively (n = 26): their rehabilitation within the first Lost to follow-up (n = 1): deep infection Lost to follow-up (n = 1): contralateral THA week after surgery. In a previous effiAnalyzed at 12 months postoperatively (n = 25): Analyzed at 12 months postoperatively (n = 25): cacy study, patients < 65 years who Lost to follow-up (n = 1): personal reason Lost to follow-up (n = 1): personal reason underwent unilateral THA following a 4-week early maximal strength training (MST) program under supervision Figure 1. Patient inclusion and follow-up in the maximal strength training (MST) and conventional physiotherapy (CP) groups. from an exercise physiologist increased leg muscle strength more than CP after THA (Husby et al. 2009). In the present study, we aimed to evaluate the effect from 2 rehabilitation Patients programs, in regular clinical practice. The primary outcome Patients diagnosed with primary osteoarthritis, scheduled for was muscle strength at 3 months’ follow-up. The null hypoth- elective THA surgery at St Olavs University Hospital, Norway, esis is that there will be no intergroup differences. living within short travel distance to the hospital, were asked to participate in the study by a nurse at the admission office, and assessed by an orthopedic surgeon. Exclusion criteria were severe osteoarthritis of the contralateral hip, not fully Patients and methods recovered from previous THA surgery, communication diffiThe study was a prospective randomized controlled superi- culties, discharged to a rehabilitation institute, or any illness or ority trial with THA patients operated on using the posterior disorder that could influence the training and/or physical testapproach. All patients followed the standardized Fast-track ing performance. Between August 2015 and February 2016, course (Winther et al. 2015). Patients were randomly assigned 80 patients were assessed for eligibility. Finally, 60 patients receiving different postoperative rehabilitation at municipal gave written consent to participate (Figure 1). The randomizaphysiotherapy institutes—either MST or CP—and logged tion was stratified by sex and concealed by using a web-based their physical activity in a training diary up to the 6-monthly service provided by the research department at the university. follow-up. The primary outcome was 1RM in abduction and leg press strength, a surrogate outcome measure related Testing to functional performance. Secondary outcomes were pain Patients were tested preoperatively, and at 3, 6, and 12 months assessed by the numeric rating scale (NRS), distance covered postoperatively. 2 physiologists and 1 physiotherapist, all during the 6-minute walk test (6MWT), Harris Hip Score highly experienced with patient testing, conducted the physi(HHS) and Hip disability and Osteoarthritis Outcome Score cal tests starting with the 6MWT where the patient walked Physical function Shortform score (HOOS-PS). back and forth in a 50-meter hallway for 6 minutes. 1RM leg press was tested in an ergometer device (Steens Physical, Ring

12187 Winther D.indd 296

5/3/2018 11:39:32 AM

Acta Orthopaedica 2018; 89 (3): 295–301

297

Figure 2. Set-up for the (a) leg press ergometer and (b) abduction pulling apparatus.

Mekanikk AS, Norway) (Figure 2a). 1RM abduction strength was tested in a pulling apparatus (Pivot 820, Sports Master, Norway) (Figure 2b). Pain was assessed by the NRS: “On a scale from 0 to 10, where 0 is no pain and 10 is the worst pain imaginable, can you define the pain you have right now?” The disease-specific questionnaires HHS and HOOS-PS were obtained preoperatively and at 3 and 12 months postoperatively. Maximum score for the HHS is 100 points. HOOS-PS is scored from 0 to 100 with zero being optimal. Intervention In accordance with current practice, the patients contacted the preferred physiotherapy institute the day after discharge and scheduled the first appointment. The physiotherapy institutes prioritized these patients. Patients in the MST group could attend 1 out of 5 municipal physiotherapy institutes that had consented to participate in the study. The institutes received the training protocol and were instructed in what manner to supervise the patients performing the specific exercises: leg press and abduction strength of the operated leg. 1-week postoperatively, 1RM in leg press and abduction of the operated leg were tested at the hospital to decide the initial weight load for the patients during the first training session. Each training was provided individually and started with a preferable warm-up exercise followed by the 2 strength-training exercises. Patients performed 5 repetitions × 4 series, starting with a load equal to 85–90% of 1RM, with emphasis on maximal mobilization of force in the concentric part of the movement. The series were separated by 1- to 2-minute resting periods, and the load was increased when the patient could perform 6RM. MST patients were prescribed 3 weekly physiotherapy visits for 3 months in accordance with clinical practice, and the adherence was assessed by means of self-reported visits collected from the training diaries. If required, the patients could receive stretching guidance and treatment by the physiotherapist, but no additional strength training of the operated leg was offered. After the intervention period, training with the physiotherapist was optional up to 6 months.

12187 Winther D.indd 297

Patients in the CP group were instructed to follow the conventional rehabilitation regimen advised by the hospital—outpatient physical therapy for 3–6 months—and to follow their physiotherapist’s guidance. They could choose any municipal institute within the municipality in which they resided. CP consisted of different type of exercises performed with low or no external loads (10–20 repetitions in each series). Warm-up exercises were mainly cycling, step and treadmill walking. Other exercises used were aquatic exercises, balance training, range-of-motion exercises, massage, and sling exercises.

Group-size calculation To reveal an effect size of approximately 1 based on the primary outcome—1RM in leg press strength—a between-group difference of 20 kg (SD 21) was considered the minimally clinically important difference and used for sample size calculation. With a significance level of 5% and a power of 90%, 24 participants were required for each group. 60 patients were included to account for dropouts. Statistics A General Linear Mixed Model (GLMM) was used to analyze all outcome variables. Strength measures were expressed as percentage of the preoperative score from the non-operated leg. The measured values were used when analyzing the other outcome variables. The preoperative value of the tested variable was included as a covariate representing a baseline control as well as correcting for eventual initial imbalance between groups. The preoperative pain score was additionally included as a covariate in all analyses. The 2 groups and time points served as fixed factors in the analyses. Interaction terms were used to acquire detailed comparisons between groups and between time points. Robust estimation was chosen to handle violations of model assumptions. P-values were Bonferroni corrected to adjust for multiple comparisons. Normality of residuals was verified by histograms. The covariates included in the models were a priori selected based on clinical evaluation. The analysis was based on an intention-to-treat principle, but there were no patient crossovers between the two treatment groups. In total, 3.2% of all data points were missing from follow-ups. As the GLMM is robust in handling missing data, there were no data set imputations made. Statistical significant intergroup differences of both primary endpoints were required to reach a confirmatory conclusion. Figures 3–5 represent model estimates, adjusted for covariates. Statistical analyses were performed using the software package IBM SPSS Statistics for Windows (version 21.0; IBM Corp, Armonk, NY, USA).

5/3/2018 11:39:35 AM

298

Acta Orthopaedica 2018; 89 (3): 295–301

Table 1. Preoperative values from patients in the maximal strength training (MST) and conventional physiotherapy (CP) group. Values are mean (SD) or (range) MST (n=31) Sex (F/M) Age BMI Leg press: operated leg (kg) Abduction: operated leg (kg) Pain mobilization (NRS) 6MWT (m) HHS HOOS-PS

CP (n=29)

17/14 61 (35–77) 28 (4) 76 (29) 9 (3) 5.3 (2.2) 499 (124) 62 (13) 36 (16)

15/14 66 (44–83) 27 (3) 84 (33) 10 (4) 5.8 (2.0) 498 (125) 62 (14) 35 (14)

NRS: Numeric rating scale (0–10; 0 is no pain, 10 is worst pain imaginable), 6MWT: 6-minute walk test, HHS: Harris Hip Score, HOOS-PS: Hip disability and Osteoarthritis Outcome Score Physical function Short-form score.

Ethics, registration, funding, and potential conflicts of interest The study was approved by the regional ethics committee (2010/3373) and conducted in accordance with the Declaration of Helsinki. The study was registered at CilnicalTrials. gov (NCT02498093) and supported by the Liaison Committee between the Central Norway Regional Health Authority (RHA) and the Norwegian University of Science and Technology [grant number 2010/708/MOCA]. The funding sources had no impact on the analyses, interpretation, or presentation of the data.

Results Demographic variables of the patients are presented in Table 1. Compliance MST started 15 (SD 4) days after surgery. The average number of physiotherapy visits during the 3-month intervention period was 24 (4), and each training session was completed within 30 minutes. 6 patients in the MST group continued training with the physiotherapist from 3 to 6 months postoperatively, on an average of 2 (1) times per week, but did not perform the 1-leg MST. CP started 19 (8) days after surgery. The average number of physiotherapy visits until 3 months was 19 (7), each lasting 55 (12) minutes. 12 patients in the CP group continued training with the physiotherapist from 3 to 6 months postoperatively, at an average of 2 (1) times per week. Maximal strength Patients in the MST group were stronger in leg press strength of the operated leg than the CP group at the 3 and 6 months’ follow-up (p < 0.001). 1-year postoperatively, no statistically significant intergroup differences were found (p = 0.5) (Figure 3). Patients in the MST group were stronger in abduction strength of the operated leg than the CP group at the 3 and 6 months’ follow-up (p ≤ 0.002). 1-year postoperatively, no statistically significant intergroup differences were found (p = 0.2) (Figure 4).

Leg press strength – percent of preoperative value

Abduction strength – percent of preoperative value

Pain score

150

140

130

120

110

100

140

maximal strength training conventional physiotherapy

Preop.

3 months 6 months

12 months

Figure 3. Leg press strength of the operated leg compared with preoperative values (100%) of the non-operated leg in the maximal strength training (MST) and conventional physiotherapy (CP) groups at 3, 6, and 12 months postoperatively. Model estimate with 95% confidence intervals.

12187 Winther D.indd 298

maximal strength training conventional physiotherapy

1 0

Preop.

3 months 6 months

12 months

Figure 4. Abduction strength of the operated leg compared with preoperative values (100%) of the non-operated leg in the maximal strength training (MST) and conventional physiotherapy (CP) groups at 3, 6, and 12 months postoperatively. Model estimate with 95% confidence intervals.

Preop.

3 months 6 months

12 months

Figure 5. Pain score during mobilization on the numeric rating scale (NRS: 0–10) in the maximal strength training (MST) and conventional physiotherapy (CP) groups preoperatively and at 3, 6, and 12 months postoperatively. Model estimate with 95% confidence intervals.

5/3/2018 11:39:36 AM

Acta Orthopaedica 2018; 89 (3): 295–301

299

Table 2. Postoperative values from patients in the maximal strength training (MST) and conventional physiotherapy (CP) groups. Values are mean (SD)

Month 3 6 12

Leg press (kg) MST CP 120 (31) 114 (23) 101 (27)

77 (30) 84 (22) 100 (26)

Abduction (kg) MST CP 13 (4) 13 (3) 12 (4)

10 (4) 10 (4) 12 (4)

6MWT (m) MST CP 583 (97) 578 (129) 607 (94) 596 (111) 627 (96) 628 (110)

HHS MST CP

HOOS-PS MST CP

91 (14) 87 (12)

14 (12) 14 (13)

95 (7)

8 (10)

93 (9)

8 (10)

6MWT; 6-minute walk test, HHS; Harris Hip Score, HOOS-PS; Hip disability and Osteoarthritis Outcome Score Physical function Short-form score.

Pain, 6MWT, HOOS-PS, and HHS No statistically significant intergroup differences in pain score (p > 0.1) (Figure 5), 6MWT (p > 0.7), HOOS-PS (p > 0.6) or HHS (p > 0.3) were found at any follow-up (Table 2).

Discussion Early postoperative MST increases muscle strength more than CP in THA patients within 3 months postoperatively. This strength difference persists up to 6 months postoperatively. No statistically significant intergroup differences were observed with respect to pain, 6MWT, HHS, or HOOS-PS. Husby et al. (2009) previously demonstrated that MST 5 times a week for 4 weeks, initiated 1 week postoperatively, increased strength more than CP in THA patients. Patients included in that study were restricted to unilateral osteoarthritis diagnosis and were relatively young, i.e., < 65 years. In contrast to our study, the training was located at a rehabilitation institute and closely supervised by 2 exercise physiologists during the entire intervention period. Our study was completed in accordance with current clinical practice as the patients had outpatient physiotherapy at municipal institutes, and few exclusion criteria. This demonstrates that MST can be implemented into clinical practice and the results generalized outside the experimental setting. Muscle strength is considered an important outcome after primary THA (Westby et al. 2014). Consequently, we see this as a clinically meaningful surrogate measure. In a systematic review by Skoffer et al. (2015), weak evidence of a beneficial effect of progressive resistance training pre/post THA on muscle strength and functional capacity was found. Suetta et al. (2004) found that only 2 exercises with supervised progressive strength training (8–20 repetitions) of the operated leg increased muscle strength more than home-based training 12 weeks after THA. Conversely, Mikkelsen et al. (2014) did not find supervised progressive strength training twice a week superior to home-based exercise in improving muscle strength, 10 weeks after THA surgery. The training sessions in their study lasted longer than in our study, and consisted of 4 exercises with 10–12 repetitions. In the early postoperative phase, patients are physically reduced. Training should therefore be

12187 Winther D.indd 299

simple, task specific and focused on targeting the affected muscles (Rasch et al. 2010). In our study, each training session lasted about twice as long for patients in the CP compared with the MST group (55 vs. 30 minutes). Shorter training sessions may increase the performance during training. Muscle strength in the MST group decreased when the training was terminated. Still, they were stronger than the CP group, and exceeded the preoperative values of the nonoperated leg at all follow-ups. The gradually reduced muscle strength in the MST group after the training intervention might be expected as muscle strength has an expiration date and the benefits from the training cannot be stored for later. In the CP group, muscle strength continued to increase from 3 to 6 months and further up to the 1-year follow-up. Patients in the CP group did not reach the preoperative leg press values of the non-operated leg until the 1-year follow-up. Our results are consistent with previously reported findings of only modest strength improvements after CP (Minns Lowe et al. 2009), and the fact that strength increased after completed rehabilitation (Beaupre et al. 2014). It has therefore been suggested that THA patients should continue strength training of the operated hip for at least 1 year postoperatively or should be given a more advanced exercise program later in their recovery (Shih et al. 1994, Trudelle-Jackson et al. 2002). The MST group had on average a few more training sessions than the CP group during the intervention period, but each session lasted half the time so that the total volume was higher in the CP group. This demonstrates that performing MST in clinical practice requires less effort and resources than CR. Furthermore, only half as many patients in the MST group as the CP group preferred to continue training with the physiotherapist after the 3 months’ intervention period, due to subjective perception of equal bilateral leg strength and independence, and being competent to continue training on their own as reasons for terminating rehabilitation. The early substantial increase in muscle strength in the MST group suggests that 3 months’ rehabilitation might be sufficient when done efficiently. Early postoperative rehabilitation should focus on preventing the great reduction in muscle strength demonstrated in these patients. Therefore, especially in the early phase when the total training volume ought not be too high, strength training should be a priority. As soon as leg muscle strength is

5/3/2018 11:39:41 AM

300

regained, the patient’s endurance capacity and physical activity should be increased to prevent various lifestyle-related diseases. The MST and CP groups approach each other in muscle strength 1-year postoperatively with a decreasing/increasing trend respectively, with no statistical significant intergroup difference (Figure 3). A reasonable question is whether the patients are completely rehabilitated when compared with healthy age-matched controls 1 year postoperatively. Previous studies have shown that THA patients have less strength in the lower extremities than healthy age-matched controls 1 year postoperatively (Sicard-Rosenbaum et al. 2002, Judd et al. 2014). In the present study, the MST group probably could have maintained superior muscle strength, for example, by having 1 MST session a week (Lexell et al. 1995). Both groups expressed very little pain at the 3-month followup with a mean NRS score of 0.8 in the MST and 1.4 in the CP group, which consistently decreased by the 1-year followup in both groups, with no statistically significant intergroup differences (Figure 5). This is an important finding, as heavy strength training early after THA is anticipated to induce more pain than exercising at lower intensities. Our results are in accordance with findings from Mikkelsen et al. (2017) who found that substantial load progression during strength training did not exacerbate postoperative pain after THA. We did not find any statistically significant intergroup differences in 6MWT, HHS, or HOOS-PS. The 6MWT is foremost designed to evaluate cardiorespiratory fitness and does not appreciably challenge the leg muscle strength. The stairclimbing test could have differentiated between the groups as it evaluates tasks closely related to daily living situations/activities (Unver et al. 2015). All patients were treated according to the fast-track clinical pathway, previously shown to assign high scores for HHS and HOOS-PS (Winther et al. 2015). The ceiling effect of these scores might limit their validity for use in clinical trials and a more differentiating score, such as the forgotten joint score, could have been used. The main limitation of our study is the lack of a healthy agematched control group to identify the muscle strength of the normal population. This could have been informative when considering the results from the THA patients. Also, we did not record the patients’ exercise status from 6 to 12 months. The study was not designed with the power to evaluate the secondary endpoints, but they are interesting seen in context with the primary finding, and as descriptive information. As muscle strength is important for physical function in both patients and healthy people (Nankaku et al. 2016, Unhjem et al. 2017), it is important to explore rehabilitation methods that can restore and improve muscle strength postoperatively (Westby et al. 2014). Further, the proportion of THA patients < 70 years is increasing, and many are still employees and engaged in physical activities, demanding effective rehabilitation (Hobbs et al. 2011, Smith et al. 2015). As a result, patients might be able to return faster to physical activities and work. For future studies, it might be beneficial to train

12187 Winther D.indd 300

Acta Orthopaedica 2018; 89 (3): 295–301

these patients both pre- and postoperatively, and in combination with endurance training, since most patients experience a period of inactivity before surgery due to activity-related pain and contracture of the hip. In summary, MST increases muscle strength more than CP in THA patients after rehabilitation in clinical practice. The muscle strength difference persists up to 6 months postoperatively; however, the groups are approaching each other. 1 year postoperatively, no intergroup differences were found. MST is feasible to conduct in regular clinical practice and the results can be generalized to a wide THA population. All authors contributed substantially to interpreting the results, and drafting and revising the article. SBW, VSH, OSH, TSW and OAF contributed to the study design; SBW and VSH collected data; OAF and JK conducted the statistical analysis; OSH and TSW screened the patients.

The authors would like to thank the physiotherapy institutes participating; Moholt physiotherapy, Tiller physiotherapy and manual therapy, Leangen physiotherapy, Frisk physiotherapy and Orkanger physiotherapy and osteopathy. They also thank physiotherapist Sølvi Liabakk-Selli for contribution in patient testing, and nurses Ann Kristin Green-Helgetun, Ann-Merete Kjøsnes, and Elin Ida Sletner for contribution in patient recruiting and coordination.

Acta thanks Thomas Bandholm and Henrik Husted for help with peer review of this study. Beaupre L A, Masson E C, Luckhurst B J, Arafah O, O’Connor G J. A randomized pilot study of a comprehensive postoperative exercise program compared with usual care following primary total hip arthroplasty in subjects less than 65 years of age: feasibility, selection of outcome measures and timing of assessment. BMC Musculoskelet Disord 2014; 15: 192. Bertocci G E, Munin M C, Frost K L, Burdett R, Wassinger C A, Fitzgerald S G. Isokinetic performance after total hip replacement. Am J Phys Med Rehabil 2004; 83(1): 1-9. Buirs L D, Van Beers L W, Scholtes V A, Pastoors T, Sprague S, Poolman R W. Predictors of physical functioning after total hip arthroplasty: a systematic review. BMJ Open 2016; 6(9): e010725. Coulter C L, Scarvell J M, Neeman T M, Smith P N. Physiotherapist-directed rehabilitation exercises in the outpatient or home setting improve strength, gait speed and cadence after elective total hip replacement: a systematic review. J Physiother 2013; 59(4): 219-26. Di Monaco M, Castiglioni C. Which type of exercise therapy is effective after hip arthroplasty? A systematic review of randomized controlled trials. Eur J Phys Rehabil Med 2013; 49(6): 893-907, quiz 21-3. Friesenbichler B, Casartelli N C, Wellauer V, Item-Glatthorn J F, Ferguson S J, Leunig M, Maffiuletti N A. Explosive and maximal strength before and 6 months after total hip arthroplasty. J Orthop Res 2017. doi: 10.1002/ jor.23626. [Epub ahead of print] Hobbs N, Dixon D, Rasmussen S, Judge A, Dreinhofer K E, Gunther K P, Dieppe P. Patient preoperative expectations of total hip replacement in European orthopedic centers. Arthritis Care Res (Hoboken) 2011; 63(11): 1521-7. Holm B, Thorborg K, Husted H, Kehlet H, Bandholm T. Surgery-induced changes and early recovery of hip-muscle strength, leg-press power, and functional performance after fast-track total hip arthroplasty: a prospective cohort study. PLoS One 2013; 8(4): e62109. Horstmann T, Listringhaus R, Brauner T, Grau S, Mundermann A. Minimizing preoperative and postoperative limping in patients after total hip arthroplasty: relevance of hip muscle strength and endurance. Am J Phys Med Rehabil 2013; 92(12): 1060-9.

5/3/2018 11:39:41 AM

Acta Orthopaedica 2018; 89 (3): 295â&#x20AC;&#x201C;301

Husby V S, Helgerud J, Bjorgen S, Husby O S, Benum P, Hoff J. Early maximal strength training is an efficient treatment for patients operated with total hip arthroplasty. Arch Phys Med Rehabil 2009; 90(10): 1658-67. Husted H. Fast-track hip and knee arthroplasty: clinical and organizational aspects. Acta Orthop 2012; 83 (Suppl 346): 1-39. Judd D L, Dennis D A, Thomas A C, Wolfe P, Dayton M R, Stevens-Lapsley J E. Muscle strength and functional recovery during the first year after THA. Clin Orthop Relat Res 2014; 472(2): 654-64. Lexell J, Downham D Y, Larsson Y, Bruhn E, Morsing B. Heavy-resistance training in older Scandinavian men and women: short- and long-term effects on arm and leg muscles. Scand J Med Sci Sports 1995; 5(6): 329-41. Mikkelsen L R, Mechlenburg I, Soballe K, Jorgensen L B, Mikkelsen S, Bandholm T, Petersen A K. Effect of early supervised progressive resistance training compared to unsupervised home-based exercise after fasttrack total hip replacement applied to patients with preoperative functional limitations. A single-blinded randomised controlled trial. Osteoarthritis Cartilage 2014; 22(12): 2051-8. Mikkelsen L R, Petersen A K, Mechlenburg I, Mikkelsen S, Soballe K, Bandholm T. Description of load progression and pain response during progressive resistance training early after total hip arthroplasty: secondary analyses from a randomized controlled trial. Clin Rehabil 2017; 31(1): 11-22. Minns Lowe C J, Barker K L, Dewey M E, Sackley C M. Effectiveness of physiotherapy exercise following hip arthroplasty for osteoarthritis: a systematic review of clinical trials. BMC Musculoskelet Disord 2009; 10: 98. Nankaku M, Akiyama H, Kakinoki R, Nishikawa T, Tanaka Y, Matsuda S. Factors associated with ambulatory status 6 months after total hip arthroplasty. Physiotherapy 2014; 100(3): 263-7. Nankaku M, Ikeguchi R, Goto K, So K, Kuroda Y, Matsuda S. Hip external rotator exercise contributes to improving physical functions in the early stage after total hip arthroplasty using an anterolateral approach: a randomized controlled trial. Disabil Rehabil 2016: 1-6. Rasch A, Bystrom A H, Dalen N, Berg H E. Reduced muscle radiological density, cross-sectional area, and strength of major hip and knee muscles in 22 patients with hip osteoarthritis. Acta Orthop 2007; 78(4): 505-10. Rasch A, Dalen N, Berg H E. Muscle strength, gait, and balance in 20 patients with hip osteoarthritis followed for 2 years after THA. Acta Orthop 2010; 81(2): 183-8. Reardon K, Galea M, Dennett X, Choong P, Byrne E. Quadriceps muscle wasting persists 5 months after total hip arthroplasty for osteoarthritis of the hip: a pilot study. Intern Med J 2001; 31(1): 7-14. Rosenlund S, Holsgaard-Larsen A, Overgaard S, Jensen C. The Gait Deviation Index is associated with hip muscle strength and patient-reported outcome in patients with severe hip osteoarthritis: a cross-sectional study. PLoS One 2016; 11(4): e0153177.

12187 Winther D.indd 301

301

Samuel D, Rowe P, Hood V, Nicol A. The relationships between muscle strength, biomechanical functional moments and health-related quality of life in non-elite older adults. Age Ageing 2012; 41(2): 224-30. Shih C H, Du Y K, Lin Y H, Wu C C. Muscular recovery around the hip joint after total hip arthroplasty. Clin Orthop Relat Res 1994; (302): 115-20. Sicard-Rosenbaum L, Light K E, Behrman A L. Gait, lower extremity strength, and self-assessed mobility after hip arthroplasty. J Gerontol A Biol Sci Med Sci 2002; 57(1): M47-51. Skoffer B, Dalgas U, Mechlenburg I. Progressive resistance training before and after total hip and knee arthroplasty: a systematic review. Clin Rehabil 2015; 29(1): 14-29. Smith T O, Latham S, Maskrey V, Blyth A. Patientsâ&#x20AC;&#x2122; perceptions of physical activity before and after joint replacement: a systematic review with metaethnographic analysis. Postgrad Med J 2015; 91(1079): 483-91. Suetta C, Magnusson S P, Rosted A, Aagaard P, Jakobsen A K, Larsen L H, Duus B, Kjaer M. Resistance training in the early postoperative phase reduces hospitalization and leads to muscle hypertrophy in elderly hip surgery patients: a controlled, randomized study. J Am Geriatr Soc 2004; 52(12): 2016-22. Trudelle-Jackson E, Emerson R, Smith S. Outcomes of total hip arthroplasty: a study of patients one year postsurgery. J Orthop Sports Phys Ther 2002; 32(6): 260-7. Unhjem R, van den Hoven L T, Nygard M, Hoff J, Wang E. Functional performance with age: the role of long-term strength training. J Geriatr Phys Ther 2017. doi: 10.1519/JPT.0000000000000141. [Epub ahead of print] Unver B, Kahraman T, Kalkan S, Yuksel E, Karatosun V, Gunal I. Test-retest reliability of the stair test in patients with total hip arthroplasty. Hip Int 2015; 25(2): 160-3. Westby M D, Brittain A, Backman C L. Expert consensus on best practices for post-acute rehabilitation after total hip and knee arthroplasty: a Canada and United States Delphi study. Arthritis Care Res (Hoboken) 2014; 66(3): 411-23. Winther S B, Foss O A, Wik T S, Davis SP , Engdal M, Jessen V, Husby O S. 1-year follow-up of 920 hip and knee arthroplasty patients after implementing fast-track. Acta Orthop 2015; 86(1): 78-85. Winther S B, Husby V S, Foss O A, Wik T S, Svenningsen S, Engdal M, Haugan K, Husby OS. Muscular strength after total hip arthroplasty: a prospective comparison of 3 surgical approaches. Acta Orthop 2016; 87(1): 22-8. Wolf O, Mattsson P, Milbrink J, Larsson S, Mallmin H. Periprosthetic bone mineral density and fixation of the uncemented CLS stem related to different weight bearing regimes: A randomized study using DXA and RSA in 38 patients followed for 5 years. Acta Orthop 2010; 81(3): 286-91.

5/3/2018 11:39:41 AM

302

Acta Orthopaedica 2018; 89 (3): 302–307

Bone mineral density changes in the graft after acetabular impaction bone grafting in primary and revision hip surgery A 2-year prospective follow-up Davey M J M GERHARDT 1, Enrico DE VISSER 1, Baudewijn W HENDRICKX 2, Berend W SCHREURS 3, and Job L C VAN SUSANTE 1

1 Department

of Orthopedics, Rijnstate Hospital, Arnhem; 2 Department of Radiology and Nuclear Medicine, Rijnstate Hospital, Arnhem; 3 Department of Orthopedics, Radboud University Medical Center, Nijmegen, The Netherlands Correspondence: jvansusante@rijnstate.nl Submitted 2017-06-26. Accepted 2018-02-20.

Background and purpose — Impaction bone grafting (IBG) is an established method in hip revision surgery to reconstruct loss of bone stock. There is limited knowledge concerning the actual bone remodelling process within the allograft. We investigated with repeated bone mineral density (BMD) measurements the biological process of bone remodelling in the allograft in vivo. We hypothesized that an initial decrease in BMD would be followed by an increase towards baseline values. Patients and methods — Dual-energy X-ray absorptiometry (DXA) was used to measure BMD values in 3 regions of interest (ROI) in 20 patients (average age at surgery 70 years, 11 males) after an acetabular reconstruction with IBG and a cemented cup. A postoperative DXA was used as baseline and DXA was repeated at 3 and 6 months and at 1 and 2 years. The Oxford Hip Score (OHS), the 12-Item Short Form Health Survey (SF12), and a 0 to 100 mm visual analogue scale (VAS) for pain and satisfaction were obtained simultaneously. Results — The overall mean BMD in the IBG regions increased significantly by 9% (95% CI 2–15) at 2 years’ follow-up. In the cranial ROI BMD increased 14% (CI 6–22), whereas the BMD in the medial and caudal ROI showed an increase of 10% (CI 1–18) and 4% (CI –6–16), respectively. The OHS, SF12-mental, and VAS for pain all improved statistically significantly 2 years after surgery, with a mean VAS for satisfaction of 77 (CI 63–90) out of 100 points. The SF12-physical showed non-significant improvement. Interpretation — The BMD in the allograft gradually increased after IBG for acetabular reconstruction arthroplasties, particularly in the cranial ROI. An initial decrease in the BMD was not encountered. These BMD changes, as proxy measurements for bone remodeling, may indicate progressive apposition of vital new host bone in the grafted area. ■

In revision hip arthroplasty (THA), acetabular bone loss can be managed by impacting allograft bone chips (Slooff et al. 1996). An adequately impacted bone graft (IBG) provides initial stability for the implanted prosthesis and facilitates bone remodelling (Bolland et al. 2007). With a stable IBG, revascularization and incorporation of the graft into the host skeleton is stimulated, a process known as “creeping substitution” or bone graft incorporation. The impacted allograft serves as a non-vital matrix facilitating ingrowth of vital host bone and as such the restoration of host bone stock. There is little knowledge or histological data available about the speed of this process (Oakes et al. 2006). In a goat model Schimmel et al. (1998) reported graft resorption, bone apposition, and remodelling into new trabecular bone without hardly any graft remnants after 24 weeks. A subsequent human case series by van der Donk et al. (2002 ) investigated 24 acetabular biopsy specimens collected at different time points up to 9 years. First the graft consisted of nonvascularized graft remnants; at 3 to 5 months postoperatively a transition from the avital graft towards newly incorporated host bone was visible through a revascularization front. By 6 months approximately 30% of the graft had been incorporated. In biopsies from 6 months’ up to 9 years’ followup 70% of the graft had been incorporated and trabecular bone had been formed. Besides these studies no in-vivo data are available on the early biological process of bone graft remodelling after acetabular IBG. In earlier studies repeated dual energy X-ray absorptiometry (DXA) proved successful to monitor periprosthetic BMD changes after different types of hip arthroplasties (Smolders et al. 2010, Huang et al. 2013, Smolders et al. 2013, Lazarinis et al. 2014) and was also used to monitor the incorporation of bone graft after spinal fusion (Hagenmaier et al. 2013).

11844 Gerhardt D.indd 302

5/3/2018 11:40:50 AM

Acta Orthopaedica 2018; 89 (3): 302–307

303

Table 1. Clinical details of the 20 patients who received an acetabular reconstruction using bone impaction grafting Factor

THA

Sex (women/men) Mean BMI (SD) Mean age at surgery in years (SD) Diagnosis primary osteoarthritis with acetabular bone defects revision for aseptic loosening revision THA secondary to infection revision of hemiarthroplasty Mean blood loss in mL (SD) Mean surgery time in minutes (SD) Mean cup inclination (SD) Median acetabular cup size in mm (range)

11/9 27 (4) 70 (9) 2 14 3 1 558 (229) 103 (25) 50 (9) 49 (44–52)

We performed a 2-year prospective study with repeated DXA measurements, specifically to monitor BMD changes in vivo within the graft, to gain further insight into the process of bone graft incorporation after acetabular IBG. We hypothezised an initial BMD decrease or demineralization to occur within the IBG up to 6 months postoperatively as result of graft resorption, followed by a steadily increase in BMD as new vital trabecular bone is established.

Patients and methods Study design This exploratory study was designed to evaluate the BMD changes in the impacted bone graft region after acetabular reconstruction surgery. From December 2013 to December 2014, 25 consecutive patients with a clinical suspicion of cup loosening in the presence of an expected contained acetabular bony defect on plain pelvic radiographs were engaged in this exploratory study. Patients mean age was 70 (9) years (Table 1). The indication for acetabular IBG was determined preoperatively based on standard pelvic radiographs and sometimes combined with pelvic computerized tomography (CT) scanning to assess the degree of acetabular bone deterioration. Patients were excluded in the case of a current infection of the hip joint or other sites, a hip fracture, if immunocompromised or taking immunosuppressive medication, with a history of medication for osteoporosis (i.e., bisphosphonates), or neoplasm. Besides plain radiographs additional CT scanning of the pelvis was available in 10 cases. 20 patients (9 men) of 25 patients were intraoperatively confirmed to have cavitary defects type 2 (according to AAOS Classification of Acetabular Bone Loss) (D’Antonio et al. 1989), which could be reconstructed with acetabular bone impaction grafting without the use of metal augments. In 5 patients adequate reconstruction of the defect required metal meshes to ensure containment of the acetabular defect; these were excluded from DXA follow-up since these meshes would interfere with BMD measurements.

11844 Gerhardt D.indd 303

Study population and follow-up 14 acetabular revisions with IBG were performed for cup loosening with osteolytic bony defects due to profound polyethylene (PE) wear and 3 cups were revised as part of a 2-stage revision procedure for infection, which was considered healed at the second stage. The remaining 3 cases consisted of 1 hemiarthroplasty patient with acetabular protrusion, which was converted towards a THA, and 2 primary THA: 1 with profound acetabular protrusio and a 1 with a large post-traumatic acetabular bony defect. Of the 17 cup revisions (11 uncemented and 6 cemented), in 8 cases the cup revision was combined with a cemented stem revision (Exeter Stem, Stryker, Newbury, UK). All patients reached 1-year followup. In 3 patients BMD could not be measured at 2 years. 2 patients refused the final follow-up because of other illness and 1 patient was re-revised after 15 months towards a dual mobility cup for recurrent dislocations; however, clinical outcome scores could be assessed in all patients at 2 years excluding the 1 patient who was revised. Surgical technique All patients received the standard surgical procedure for revision hip arthroplasty with acetabular IBG through a posterolateral approach, performed by 3 experienced hip surgeons. Preoperative digital templating for implant positioning (Easyvision, Philips Medical Systems, Eindhoven, The Netherlands) was carried out in all patients. The bone impaction grafting technique used is described in detail by Schreurs et al. (1998). Briefly, via a posterolateral approach the acetabulum was exposed and in revision cases the failed component was removed. All existing cement and fibrous tissue was removed. The sclerotic acetabular wall was penetrated by 5–10 superficial drill holes using a 3 mm burr. The contained defect was packed and filled layer by layer with handmade bone chips (7 to 10 mm) from fresh frozen femoral head allografts using a rongeur. Subsequently, the acetabular socket was restored using incremental metal impactors and a metal hammer. Bone cement (40 grams) with broad spectrum antibiotics (COPAL gentamycin + clindamycin, Heraeus Holding, Hanau, Germany) was placed on top of the impacted bone graft and pressurized with a seal. Subsequently a cemented PE cup was implanted in all cases, aiming for a 2-mm-thick cement layer (Exeter Contemporary Flanged Cup, Stryker, Newbury, UK). Cup positioning was measured on the direct postoperative standard anterior–posterior pelvic radiograph. A median outer cup size of 49 mm (range 44–52) was implanted with a mean cup inclination of 50 (8) degrees. Patients received 24 hours of systemic cefazolin (preoperative 1 x 1 g, postoperative 2 x 1 g), NSAID (3 x 50 mg diclofenac) to prevent heterotopic ossification was given for 1 week and anticoagulation therapy for 3 months (nadroparin 2,850 IE daily). Mobilization was started 24 hours after surgery with partial weight bearing on crutches for 6 weeks.

5/3/2018 11:40:50 AM

304

Acta Orthopaedica 2018; 89 (3): 302–307

Figure 1. A: Example of an anterior–posterior (AP) hip radiograph with preoperative osteolytic bone deterioration with excessive cup protrusion and loosening. B: Postoperative AP radiograph with reconstruction of the contained acetabular bony defect using the bone impaction grafting technique without metal meshes. C: With dual-energy X-ray absorptiometry, bone mineral density was measured in 3 separate regions of interest covering the postoperative acetabular impacted bone graft: cranial (green), medial (red), and caudal (blue) to the polyethylene cup. The same ROI template was used for each subsequent time interval.

Bone densitometry The impacted bone graft BMD was measured with DXA (Lunar Prodigy, GE Healthcare, Little Chalfont, UK, software package Encore 2007, version 11.30.062). JvS identified and selected the impacted bone graft area on the postoperative DXA with reference to the intraoperative findings and preoperative radiographs and/or CT scan. Typically, a hemisphere surrounding the medial surface of the cup was selected with varying thickness depending on the volume of the grafted defect (Figure 1). This hemisphere or IBG area was divided into 3 regions (cranial, medial, and caudal) each covering approximately one-third of the surface. To enhance reproducibility, the defined template of the bone grafted area on the first postoperative DXA (Figure 1) was incorporated in each subsequent measurement by software recognizing the bony contours of the pelvis. Baseline measurements were performed within 2 weeks after surgery, at 3 and 6 months, and at 1 and 2 years. BMD values of the impacted bone graft obtained during the 2 years’ follow-up were compared with baseline levels (Table 2, see Supplementary data). BMD values are also expressed as a percentage against the original baseline levels (100%) (Table 2 and Figures 2 and 3). BMD values were categorized in a cranial, medial, and caudal ROI in relation to the acetabular cup. The software used in our study was designed to recognize the prosthesis and to measure periprosthetic acetabular BMD only (Figure 1). DXA scans and patient positioning were standardized according to a strict protocol; patients were positioned supine with their feet attached to a positioning device to obtain a reproducible 20° of internal rotation. A range of 15° internal to 15° external rotation yields a precision of 1.7% according to Mortimer et al. (1996). This precision error was also confirmed by a daily calibration procedure of the DXA where repeated measurements were performed on a phantom. In previous studies, regarding acetabular BMD changes after resurfacing and total hip arthroplasty, repeated measurements were also perfomed in study patients resulting in a mean coefficient of variation of 2.6% (SD 0.9) (Smolders et al. 2013); this was not repeated for the current study. Quality controls for

11844 Gerhardt D.indd 304

the DXA equipment were undertaken daily according to the manufacturer’s guidelines to verify the stability of the system. No change was observed during the entire study period. Clinical outcome measurements Clinical outcome measurements were completed preoperatively, at 3 and 6 months, and 1 and 2 years after surgery. This included the Oxford Hip Score (OHS), the 12-item Short Form Health Survey (SF12), a 0 (no pain) to 100 mm (maximum pain) visual analogue scale (VAS) for pain and a 0 (minimum) to 100 mm (maximum) VAS for satisfaction. Statistics 20 patients were included for 2-year follow-up in this exploratory study. This number of patients was selected and considered adequate to detect statistically significant differences regarding acetabular BMD changes based on earlier studies using this DXA technique (Smolders et al. 2010, Lazarinis et al. 2014). 20 patients have also proven to have adequate power monitoring BMD changes after bone grafting in a different field of interest, i.e., spinal fusion (Hagenmaier et al. 2013). For this reason, we performed no sample size calculation for the current study. All data were checked for normal distribution by means of the Shapiro–Wilk test. Normally distributed data are presented as mean (SD). Not-normal distributed data are presented as median (range). The absolute (g/cm2) and relative (%) BMD changes of each ROI over the observed period and clinical scores were compared with baseline values using linear mixed models with random intercept (patient) and random slope (time). Time (categorical) and sex were treated as fixed factors. Results from the mixed model were reported with use of point estimates with corresponding 95% confidence interval (CI). The assumptions for this model were checked and found to be adequately met. No adjustments for multiple testing were performed. Missing data were assumed to be missing at random; residuals of the model were normally distributed. Data of patients lost to follow-up were included up to their last measurement. Differences were considered

5/3/2018 11:40:50 AM

Acta Orthopaedica 2018; 89 (3): 302–307

305

Figure 2. Point estimates of relative mean bone mineral density (BMD) changes within the impacted acetabular bone graft (g/cm2) compared with direct postoperative baseline values with 95% confidence intervals during 2 years’ follow-up. The total mean BMD (gray) is divided into 3 regions of interest: cranial (green), medial (red), and caudal (blue) to the acetabular cup.

Figure 3. Spaghetti plot of measured bone mineral density (BMD) changes of the impacted acetabular bone graft (g/cm2) as percentage of the direct postoperative baseline values (%) at an individual patient level. Note: The outlier with a decrease in BMD of 20% corresponds with the early revision case due to recurrent dislocations of the hip.

statistically significant with a p-value < 0.05. All statistical analyses were performed using SPSS software version 21.0 (IBM Corp, Armonk, NY, USA).

area at an individual patient level are summarized in Figure 3. Most patients reveal an increase in BMD over time, and some patients fluctuate around baseline levels. 1 outlier was seen with a decrease in BMD close to 20% at 1-year followup. This patient was the single revision case at 15 months for recurrent (4 times) dislocation. Figure 2 represents mean values (%) for the entire group and as such the overall trend in BMD change is clearly visible (Figure 4 and Table 3, see Supplementary data).

Ethics, registration, funding, and potential conflicts of interest Approval from the regional ethics committee from the Radboud University Nijmegen Medical Centre was obtained (NL 46305.091.13). Written informed consent was obtained from all patients. The study was registered in the Clinical Trials registry (NCT02061904). Research funding was obtained from Rijnstate Vriendenfonds. There are no conflicts of interest to be reported by any of the authors.

Clinical outcome All clinical outcome scores had improved at 2-year follow-up (Table 3, see Supplementary data).

Results

Discussion

Bone mineral density The point estimates of the mean absolute (g/cm2) and relative (%) BMD values of the IBG are summarized in Table 2 (see Supplementary data) and are visualized in Figure 2. In the cranial region a higher BMD was measured at baseline compared with the medial (p = 0.02) and caudal regions (p = 0.07). In the overall grafted area, a BMD of 2.4 g/cm2 was measured at baseline. From postoperative baseline (100%) a gradual BMD increase of 9% was measured at 2 years’ followup for the overall grafted area. An initial decrease in BMD early after surgery, as hypothesized, was not seen. As for each specific ROI, the cranial region revealed the most pronounced BMD increase of 14%. In the medial region BMD increased 10% at 2 years’ follow-up, whereas for the caudal region this was 4%. Trends in measured BMD change for the grafted

There is a knowledge gap regarding the actual bone remodelling process of impacted bone grafts used in acetabular reconstruction surgery. This prospective DXA study was designed to measure BMD changes after an acetabular reconstruction with impacted bone grafting, monitoring the process of acetabular bone graft incorporation in vivo. The expected BMD decrease in the first 6 months after surgery, as observed in studies regarding acetabular BMD changes after primary hip arthroplasty, was not seen (Digas et al. 2006, Smolders et al. 2013, Lazarinis et al. 2014). In contrast to our hypothesis, BMD gradually increased directly from postoperative baseline levels. An increasing trend was seen in all 3 separate ROIs with the strongest increase of 14% cranial to the cup. In addition, baseline BMD levels (g/cm2) were significantly higher in this cranial region immediately after surgery compared with

11844 Gerhardt D.indd 305

5/3/2018 11:40:51 AM

306

the medial and caudal regions. These higher direct postoperative BMD values correspond with the direction of impaction force at the time of implantation and may thus represent the presence of denser bone graft. The encountered BMD increase up to 2 years after surgery may be indicative of the gradual apposition of vital new bone whereas initial bone depletion and excessive loss of density of the graft does not seem to occur. Interestingly, as noted in Figure 2, 3 patients do show minor BMD changes or even a slight decrease in BMD within the IBG. This is an interesting result, on top of which 1 outlier is seen with a decrease in BMD close to 20% at 1 year-followup. This outlier experienced recurrent hip dislocations and the inability of full weight-bearing may have affected the process of graft incorporation in this specific case, resulting in a steep BMD decrease. The functional outcome scores all improved significantly after surgery and patients reported a high satisfaction rate in accordance with earlier literature (Arumugam et al. 2015, Te Stroet et al. 2015b). Bone remodelling after IBG The impaction bone grafting technique has already proven its value in hip arthroplasty since it was introduced (Slooff et al. 1984). With this biological approach in reconstructing the acetabulum, satisfying long-term results on implant survival and preservation of the bone stock have been published (Schreurs et al. 2004, Comba et al. 2006, Schreurs et al. 2009, Te Stroet et al. 2015a). The process of bone graft incorporation depends on the initial mechanical stability of the IBG and the biological interaction with the host bone (Giesen et al. 1999). The bone graft resorption and bone ingrowth should be well balanced in order to retain its mechanical stability and to prevent cup migration, as confirmed by our findings as we did not observe a BMD decrease. BMD measurements after IBG Limited data are available regarding the process of acetabular bone graft incorporation with special reference to its speed and correlation with BMD changes (Buma et al. 1996, Ullmark and Obrant 2002, van der Donk et al. 2002). The available literature so far concerns BMD changes after IBG on the femoral side (Nesse et al. 2003, Grochola et al. 2008). The groups of Ullmark et al. (2009) and Piert et al. (1999) have reported on the use of positron emission tomography (PET) in monitoring bone metabolism within the IBG. By measuring the uptake of [18F]fluoride in various regions of the IBG this active process of bone graft incorporation could be quantified. In a small cohort of 7 cup revisions with segmental and cavitary defects reconstructed with IBG, an increased uptake of [18F]fluoride was measured up to 4 months after surgery compared with the uptake measured in the healthy contralateral hip (Ullmark et al. 2009). Subsequently normalizing uptake levels were measured at 1 year postoperatively, indicating the IBG having been transformed to living bone stock, similar to our results.

11844 Gerhardt D.indd 306

Acta Orthopaedica 2018; 89 (3): 302â&#x20AC;&#x201C;307

Strength and limitations The strength of this study is that BMD has prospectively been monitored in a consecutive series of patients with simple cavitary defects reconstructed with IBG. There are some limitations. First, we included only 20 patients with varying size and volume of acetabular defects. On the other hand, this is the reality in clinical practice and the observed mean trend in BMD changes appeared to overlap globally with the curves for each individual patient (Figure 3). We do not believe that a larger number of patients or standardization of the defects would have given a different outcome. However, future studies with a larger number of patients may allow further insight into the differences in BMD changes in different ROI, as our study clearly indicates a more pronounced BMD increase in the cranial (weightbearing) area. Second, CT scanning could also have been considered to monitor BMD changes over time. CT scanning may be more accurate, evaluating true bone mineral density alterations specifically within the IBG in a 3-dimensional manner, thus excluding over-projection of cortical bone at the acetabular rim, which is inevitable with two-dimensional DXA measurements. On the other hand, CT scanning also generates significantly higher radiation doses and costs and the bias from concomitant BMD change in the cortical walls is expected to be minimal. In our opinion, at this exploraory stage of evaluating the process of bone remodeling in IBG, DXA is useful due to its proven reproducible nature, low radiation dose, and costs. Finally, template selection for the DXA measurements was performed by a single individual and could not be standardized. All defects were contained and, in each case, the inner surface of the acetabular component thus had to be surrounded by a layer of bone graft, which area appeared to be recognizable on the first postoperative DXA. Only the size and volume of the defect could differ between patients, which resulted in differing thickness of the bone graft layer. Defining the grafted area on the first DXA scan together with the subsequent division in a cranial, medial, and caudal graft area had to be individualized. Since we used a conservative approach where the template was chosen within the boundaries of the grafted area and the cup cement mantle this limitation is of little importance. Also, the same template was used in all subsequent DXA measurements. Summary A gradual increase in BMD within the grafted area, particularly in the cranial ROI, was encountered up to 2 years after surgery. In contrast to our hypothesis, there was no initial decrease in BMD. The profound BMD decrease in the one early revision case supports our belief that early weightbearing is important for adequate incorporation of the graft in at least the cavitary defects. Supplementary data Tables 2 and 3, and Figure 4 are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/17453674.2018.1460776

5/3/2018 11:40:57 AM

Acta Orthopaedica 2018; 89 (3): 302–307

DG: Local investigator and PhD student on bone mineral density changes after hip arthroplasty. EdV: manuscript review. BH: responsible for all the patients’ DXA follow-up. BS: manuscript review. JvS: Principle investigator, study design, writing manuscript.

Acta thanks Stergios Lazarinis and Gösta Ullmark for help with peer review of this study.

Arumugam G, Nanjayan S K, Quah C, Wraighte P, Howard P. Revision hip arthroplasty using impacted cancellous bone and cement: a long-term follow-up study. Eur J Orthop Surg Traumatol 2015; 25(8): 1279-84. Bolland B J, New A M, Madabhushi S P, Oreffo R O, Dunlop D G. Vibrationassisted bone-graft compaction in impaction bone grafting of the femur. J Bone Joint Surg Br 2007; 89(5): 686-92. Buma P, Lamerigts N, Schreurs BW, Gardeniers J, Versleyen D, Slooff T J. Impacted graft incorporation after cemented acetabular revision: histological evaluation in 8 patients. Acta Orthop Scand 1996; 67(6): 536-40. Comba F, Buttaro M, Pusso R, Piccaluga F. Acetabular reconstruction with impacted bone allografts and cemented acetabular components: a 2- to 13-year follow-up study of 142 aseptic revisions. J Bone Joint Surg Br 2006; 88(7): 865-9. D’Antonio J A, Capello W N, Borden L S, Bargar W L, Bierbaum B F, Boettcher W G, Steinberg M E, Stulberg S D, Wedge J H. Classification and management of acetabular abnormalities in total hip arthroplasty. Clin orthop Relat Res 1989; (243): 126-37. Digas G, Karrholm J, Thanner J. Different loss of BMD using uncemented press-fit and whole polyethylene cups fixed with cement: repeated DXA studies in 96 hips randomized to 3 types of fixation. Acta Orthop Scand 2006; 77(2): 218-26. Giesen E B, Lamerigts N M, Verdonschot N, Buma P, Schreurs B W, Huiskes R. Mechanical characteristics of impacted morsellised bone grafts used in revision of total hip arthroplasty. J Bone Joint Surg Br 1999; 81(6): 1052-7. Grochola L F, Habermann B, Mastrodomenico N, Kurth A. Comparison of periprosthetic bone remodelling after implantation of anatomic and straight stem prostheses in total hip arthroplasty. Arch Orthop Trauma Surg 2008; 128(4): 383-92. Hagenmaier F, Kok D, Hol A, Rijnders T, Oner F C, van Susante J L. Changes in bone mineral density in the intertransverse fusion mass after instrumented single-level lumbar fusion: a prospective 1-year follow-up. Spine 2013; 38(8): 696-702. Huang Q, Shen B, Yang J, Zhou Z K, Kang P D, Pei F X. Changes in bone mineral density of the acetabulum and proximal femur after total hip resurfacing arthroplasty. J Arthroplasty 2013; 28(10): 1811-15. Lazarinis S, Milbrink J, Mattsson P, Mallmin H, Hailer N P. Bone loss around a stable, partly threaded hydroxyapatite-coated cup: a prospective cohort study using RSA and DXA. Hip Int 2014; 24(2): 155-66. Mortimer E S, Rosenthall L, Paterson I, Bobyn J D. Effect of rotation on periprosthetic bone mineral measurements in a hip phantom. Clin Orthop Relat Res 1996; (324): 269-74. Nesse E, Nielsen E W, Bastian D. [Cemented versus cementless revision femoral stems using morselized allograft: a prospective, randomized study with 5 years follow-up]. Z Orthop Ihre Grenzgeb 2003; 141(6): 678-83.

11844 Gerhardt D.indd 307

307

Oakes D A, Cabanela M E. Impaction bone grafting for revision hip arthroplasty: biology and clinical applications. J Am Acad Orthop Surg 2006; 14(11): 620-8. Piert M, Winter E, Becker G A, Bilger K, Machulla H, Muller-Schauenburg W, Bares R, Becker H D. Allogenic bone graft viability after hip revision arthroplasty assessed by dynamic [18F]fluoride ion positron emission tomography. Eur J Nucl Med 1999; 26(6): 615-24. Schimmel J W, Buma P, Versleyen D, Huiskes R, Slooff T J. Acetabular reconstruction with impacted morselized cancellous allografts in cemented hip arthroplasty: a histological and biomechanical study on the goat. J Arthroplasty 1998; 13(4): 438-48. Schreurs B W, Slooff T J, Buma P, Gardeniers J W, Huiskes R. Acetabular reconstruction with impacted morsellised cancellous bone graft and cement: a 10- to 15-year follow-up of 60 revision arthroplasties. J Bone Joint Surg Br 1998; 80(3): 391-5. Schreurs B W, Bolder S B, Gardeniers J W, Verdonschot N, Slooff T J, Veth R P. Acetabular revision with impacted morsellised cancellous bone grafting and a cemented cup. A 15- to 20-year follow-up. J Bone Joint Surg Br 2004; 86(4): 492-7. Schreurs B W, Keurentjes J C, Gardeniers J W, Verdonschot N, Slooff T J, Veth R P. Acetabular revision with impacted morsellised cancellous bone grafting and a cemented acetabular component: a 20- to 25-year follow-up. J Bone Joint Surg Br 2009; 91(9): 1148-53. Slooff T J, Huiskes R, van Horn J, Lemmens A J. Bone grafting in total hip replacement for acetabular protrusion. Acta Orthop Scand 1984; 55(6): 593-6. Slooff T J, Buma P, Schreurs B W, Schimmel J W, Huiskes R, Gardeniers J. Acetabular and femoral reconstruction with impacted graft and cement. Clin Orthop Relat Res 1996; (324): 108-15. Smolders J M, Hol A, Rijnders T, van Susante J L. Changes in bone mineral density in the proximal femur after hip resurfacing and uncemented total hip replacement: a prospective randomised controlled study. Clin Orthop Relat Res. [Randomized Controlled Trial] 2010; 92(11): 1509-14. Smolders J M, Pakvis D F, Hendrickx B W, Verdonschot N, van Susante J L. Periacetabular bone mineral density changes after resurfacing hip arthroplasty versus conventional total hip arthroplasty: a randomized controlled DEXA study. J Arthroplasty [Randomized Controlled Trial] 2013; 28(7): 1177-84. Te Stroet M A, Keurentjes J C, Rijnen W H, Gardeniers J W, Verdonschot N, Slooff T J, Schreurs B W. Acetabular revision with impaction bone grafting and a cemented polyethylene acetabular component: comparison of the Kaplan-Meier analysis to the competing risk analysis in 62 revisions with 25 to 30 years follow-up. Bone Joint J 2015a; 97-B(10): 1338-44. Te Stroet M A, Rijnen W H, Gardeniers J W, van Kampen A, Schreurs B W. Satisfying outcomes scores and survivorship achieved with impaction grafting for revision THA in young patients. Clin Orthop Relat Res 2015b; 473(12): 3867-75. Ullmark G, Obrant K J. Histology of impacted bone-graft incorporation. J Arthroplasty 2002; 17(2): 150-7. Ullmark G, Sorensen J, Nilsson O. Bone healing of severe acetabular defects after revision arthroplasty. Acta Orthop 2009; 80(2): 179-83. van der Donk S, Buma P, Slooff T J, Gardeniers J W, Schreurs B W. Incorporation of morselized bone grafts: a study of 24 acetabular biopsy specimens. Clin Orthop Relat Res 2002; (396): 131-41.

5/3/2018 11:40:57 AM

308

Acta Orthopaedica 2018; 89 (3): 308–313

Projections of primary hip arthroplasty in Germany until 2040 Veronika PILZ 1, Tim HANSTEIN 1,2, and Ralf SKRIPITZ 2,3

1 Heraeus Medical GmbH, Wehrheim; 2 Universitätsmedizin Rostock, Orthopädische Klinik und Poliklinik, 3 Roland Klinik, Zentrum für Endoprothetik, Fußchirurgie, Kinder- und Allgemeine Orthopädie, Bremen;

Rostock, Germany;

Correspondence: veronika.pilz@heraeus.com Submitted 2017-11-21. Accepted 2018-02-13.

Background and purpose — The number of hip replacements in Germany has increased considerably during the last 2 decades but lately levelled off with no significant increase in operation rates. We analyzed the future trend of hip arthroplasty and projected the number of primary hip replacements that will be performed in Germany until 2040. Patients and methods — We used prevalence data of hip arthroplasty patients from 2010 to 2016 from the nationwide inpatient statistics and population forecasts from the German Federal Bureau of Statistics up to the year 2040. We used Poisson regression to estimate the expected annual number of arthroplasty surgeries with calendar year and patient age as covariates to account for differences among age groups and changes over time. Results — The number of primary hip replacements performed in Germany in 2040 was estimated to grow by 27% to 288 x 103 (95% CI 250 x 103–332 x 103) from 2010. Projected counts were highest for patients aged 60 to 70 years. The estimated incidence rate was projected to 360 (95% CI 312–414) per 100,000 residents. However, incidence rates for individual age classes were found to be constant with a slight decrease over time for individual age classes. Interpretation — Our findings suggest a growth in the total hip arthroplasty count whereas incidence rate remained constant over age classes. We consider the future demographic change to an older population as well as the increasing life expectancy to be the main reasons for the increasing patient numbers rather than a general increase in the operation frequency. ■

Germany is ranked second among the countries of the Organization for Economic Cooperation and Development (OECD) concerning the incidence of hip replacements (OECD 2015). 283 hip arthroplasties per 100,000 residents were performed

in 2014 and Germany is thereby at the head of many European countries like Austria, Belgium, and the Nordic countries. The US is a little above average with an incidence of 204 per 100,000 residents. However, incidence rates have increased rapidly in most OECD countries and they are suspected to grow further (OECD 2015). The future rate of hip arthroplasty in Germany might considerably increase with the aging of the baby boomer generation starting to reach 65 in the year 2021 (Nowossadeck 2012). The aging of the population will certainly affect incidences of chronic diseases that are more common in the elderly and in a similar way the inpatient care. At present, hip arthroplasty is 1 of the 10 most commonly performed inpatient procedures in Germany in 2016, which is contributing strongly to the rapidly increasing hospital costs (Statistisches Bundesamt 2017). On the other hand, hospitals face higher constraints on their budgets since DRG reimbursement rates for hip replacement procedures have been reduced. The high frequency of hip replacements in Germany is the subject of ongoing discussion. Therefore, a reliable projection of the demand for future hip replacements would be valuable for decision-makers but also for other stakeholders like insurers or hospitals. There is only a limited amount of data available on the incidence of primary hip replacements in Germany (Schäfer et al. 2013, Wengler et al. 2014) and no study dealing with estimated future demands. Some publications exist outside Germany such as in the US (Kurtz et al. 2007, 2009), the UK (Culliford et al. 2015), and in the Nordic countries (Pedersen et al. 2009, Otten et al. 2010, Nemes et al. 2014, 2015). In this study, we investigated the epidemiology of hip arthroplasty performed in Germany and projected patient numbers up to the year 2040. We hypothesized that we would observe increasing rates of hip replacements over time and that the rates would vary with respect to the age distribution among the population.

12345 Pilz D.indd 308

5/3/2018 11:42:19 AM

Acta Orthopaedica 2018; 89 (3): 308–313

Methods Data We analyzed data from the nationwide inpatient statistics (DRG statistics), which include treatment data on all inpatient cases processed according to the DRG system. Inpatient episode reporting is mandatory and therefore this study can be considered a nationwide survey (except military and psychiatry services) (Müller-Bergfort and Fritze 2007). In 2016, the DRG statistics included approximately 19 million hospital cases with on average 3.6 procedures (Statistisches Bundesamt 2017). These statistics incorporate anonymized data from the Federal Bureau of Statistics, which undergo plausibility checks and data validations and verifications on a medical and on an economic level (Spindler 2017). The data comprise DRG information including medical procedures but no comprehensive demographic or hospital admission information. Medical procedures are coded according to the national classification of operations and procedures (OPS). It should be mentioned that these hospitalization statistics count admissions, rather than patients. We studied data from the year 2010 up to 2016 and corresponding OPS versions have been used (DIMDI 2016). We identified all patients with a procedure code for hip arthroplasty including either total hip arthroplasty, hemiarthroplasties, or other implant types (Table 1, see Supplementary data). For the years 2010 through 2016, there were no coding changes for hip arthroplasty. In our analysis, all patients were included from birth, which incorporated around 232,000 in 2016. Age was categorized in groups of < 45, 45–54, 55–59, 60–64, 65–69, 70–74, 75–79, 80–84, 85–89, and 90 years and above. Population data were available from the German Federal Bureau of Statistics as well as official statistics of population projection until 2060 (Pötzsch and Rößger 2015) (Figure 1, see Supplementary data). These population projections take into account the future mortality, increased life expectancy for the oldest population groups, and immigration rate. We used the projection model assuming constant trend of birth and death rates with a higher rate of immigration (Figure 2, see Supplementary data). In general, population forecasts are based on hypothetical assumptions about future conditions and factors and are therefore uncertain. Internal validations showed good agreement between predicted and proven results when comparing former German population forecasts since 1998 in the short and medium term with the actual population size and structure (Pötzsch 2016, Statistisches Bundesamt 2016). Statistics We used historical data from 2010 to 2016 and population forecasts up to the year 2040 in order to project the annual incidence of hip replacement surgery in Germany. We used Poisson regression to estimate the expected annual number of arthroplasty surgeries with calendar year and patient age as

12345 Pilz D.indd 309

309

Table 2. Annual number of total population, of hip replacement procedures, and incidence per 100,000 German residents for the years 2010–2016 in Germany Year

Total population

Cases

Incidence

2010 2011 2012 2013 2014 2015 2016

81,757,471 81,779,210 80,425,879 80,645,665 80,982,525 81,686,633 82,345,000

213,697 213,935 212,304 210,384 219,325 227,293 232,746

261 262 264 261 271 278 283

covariates to account for differences among age groups as well as changes over time. We chose an offset variable for the size of the population to ensure that the operation rate did not rise above the total population number. The regression estimates were limited to grow to the amount of the total population to prevent overestimation. The incidence was calculated by dividing the estimated number of surgical procedures for the national total and for each age subgroup by the corresponding population forecast from the Federal Bureau of Statistics. In order to overcome overdispersion problems that could result in an underestimation of the variance, we used a robust sandwich covariance matrix estimator for variance calculation. We checked the robustness of our projection and calculated a conservative model for the forecast. As the Poisson regression is assuming continuous growth, the incidence could be modelled to an unreasonably high rate. In the conservative approach, we assumed the THR incidence rate to be constant and thereby only accounted for population changes over time. The incidence of THR procedure is held constant based on the average of historical incidence data from the years 2010 to 2016. All statistical analyses were performed with R Version 3.4.0 (R Development Core Team, The R Foundation for Statistical Computing, Vienna, Austria). Funding and potential conflicts of interest The study received no funding. There is no conflict of interest involving any of the authors.

Results Historical view on hip arthroplasty in Germany Within the DRG statistics data there were approximately 213,000 patients undergoing primary hip arthroplasties in 2010, which increased by 9% to 232,000 patients in the year 2016. Thus, there was a small but steady increase in patient numbers as well as in the incidence rate, which amounted to 283 per 100,000 in 2016 (Table 2). Since 2014, the incidence rate as well as the patient numbers increased by 3% approximately every year.

5/3/2018 11:42:19 AM

310

Acta Orthopaedica 2018; 89 (3): 308–313

Figure 3. Projected incidence of hip replacements in Germany from 2010 to 2040 with points indicating historical patient numbers. Shading in green indicates the 95% confidence interval.

Figure 4. Projected number of hip replacements from 2010 to 2040 by age group.

Population (million)

Projection of hip arthroplasty in Germany For modelling the forecasts of the numbers of hip replacements up to the year 2040, the Poisson regression model indicated evidence of overdispersion. The problem of overdispersion could result in underestimating the variance of the estimated parameter. Therefore, we decided to account for overdispersion from the theory of quasi-likelihood and applied a quasiPoisson regression on our data (Wedderburn 1974). Under the quasi-Poisson projection model the annual number of hip replacements is forecast to grow to 288 x 103 (95% CI 250 x 103–332 x 103) by 2040. The estimated incidence rate was projected to 360 (95% CI 312–414) per 100,000 German residents, which results in a growth of 38% from 2010 to 2040. Figure 3 shows the projected incidence until 2040 with points representing historical data that were used for setting up the model. Even though variability increases over time, the total incidence rate increases significantly from 2010 to 2040. Figure 4 shows the projections of hip replacement counts for the years 2010 to 2040 colored by age groups. The highest number of patients in 2040 was estimated in the age group 75–80 years, which doubles to up to 66,575 from the year 2010. The highest increase in patients was modelled in the group of patients older than 90, which increased by 160%. On the other hand, hip replacements in patients up to the age of 70 were found to follow a constant or even negative trend. Procedure counts of patients aged 45–55, for instance, decreased by 20% approximately. Overall, it can be concluded that there is a shift towards older patients undergoing hip replacement surgery. In comparison with the increase in the total number of hip replacements or the total incidence, respectively, the incidence per age group is modelled to be constant. In Figure 5 the population forecast is compared with the number of hip replacements and the incidence by age group for the years 2010, 2020, 2030, and 2040. Our model computes the increasing hip replacements count, but in relation to the simultaneously increasing population count the incidence rates per age group are constant over time. Sensitivity analysis Our sensitivity analysis assuming constant incidence rates per

12345 Pilz D.indd 310

THA patients, n Year 2010

60,000

Incidence per 100,000

Year 2010

1,500

30 40,000

1,000

20,000

500

Age

THA patients, n

Population (million) 40

Age

Year 2020

60,000

Incidence per 100,000

Year 2020

1,500

30 40,000

1,000

20,000

500

Age

Population (million) 40

THA patients, n Year 2030

60,000

Incidence per 100,000

Year 2030

1,500

30 40,000

1,000

20,000

500

Age

THA patients, n

Population (million) 40

Age

Year 2040

60,000

Incidence per 100,000

Year 2040

1,500

30 40,000

1,000

20,000

500

Age

Figure 5. Population forecasts compared with the number of hip replacements and the incidence by age group for the years 2010, 2020, 2030, and 2040. For colour codes, see Figure 4.

5/3/2018 11:42:19 AM

Acta Orthopaedica 2018; 89 (3): 308–313

age group held fixed by the average of the years 2010–2016 showed similar results to the quasi-Poisson regression. The average total incidence rate was 268 hip replacements per 100,000 German residents. The annual number of THRs was forecast to grow to 295,422 patients in 2040 compared with 287,955 patients calculated by the quasi-Poisson regression (Table 3, see Supplementary data).

Discussion Summary This epidemiologic study investigated the trend for primary hip arthroplasty in Germany from 2010 to 2040. Given the official population forecasts, the number of primary hip replacements performed in Germany in 2040 was estimated to grow by 27% to 288 x 103 (95% CI 250 x 103 to 332 x 103) from 2010. The estimated incidence rate was projected to rise from 283 in 2016 to 360 (95% CI 312–414) in 2040 per 100,000 German residents. However, incidence rates per age group were modelled to be constant with a slight decrease over time. Furthermore, the increase in hip replacements is mainly based on the trend within different age groups. Whereas a constant or even a negative trend was found for patients aged 70 or younger, patient numbers older than 70 years will increase considerably to 2040. Comparison with other German studies In the literature, only a few publications can be found regarding the trend in hip arthroplasty in Germany. Wengler et al. (2014) examined the current situation over the years 2005– 2011 with no projection of hip replacement counts with the use of the DRG statistics data. In their analysis, several inclusion criteria restricted the dataset so that the patient numbers for the years 2010 and 2011 differ slightly from ours. They reported a growth of 11% in elective primary hip replacement operations in Germany from 170 to 190 per 100,000 persons. However, after correction for demographic changes, a 3% increase remained. Thus, Wengler et al. (2014) concluded that the increase in the volume of hip arthroplasty in Germany was most likely attributable to the demographic changes, as we did. Comparison with studies from other countries It has to be mentioned that a direct comparison between countries is complicated by various factors, i.e., different modelling techniques or assumptions and different lifestyle factors as well as social and economic factors influencing hip replacement surgeries. Furthermore, it must be considered that the extent to which crude figures can be interpreted is limited by demographic differences as well as different patient inclusion criteria (elective vs. acute hip fracture surgeries). Therefore, we simply wish to sum up recent findings. In an international comparison the incidence of hip arthroplasty varies consider-

12345 Pilz D.indd 311

311

ably, with Germany being among the top countries with the second largest hip replacement incidence behind Switzerland (Pabinger and Geissler 2014, OECD 2015). Similarly, the projected future hip replacements counts vary dependent on the country, whereas all predict the incidence to grow. However, there are few studies describing how future hip arthroplasty incidence might change in future years. In recent years some publications have existed outside Germany, for example in the United States (Kurtz et al. 2007, 2009), the UK (Culliford et al. 2015) or in the Nordic countries (Pedersen et al. 2009, Otten et al. 2010, Nemes et al. 2014, 2015). Kurtz et al. (2007) forecasted the biggest growth of 174% between 2005 and 2030. On the other hand, Culliford et al. (2015) expected an increase of 32% from 2015 to 2035 and a Swedish publication reported only a 25% increase in total hip arthroplasty from 2013 to 2030 (Nemes et al. 2014, 2015). A 66% increase in the incidence rate of primary THAs between 2013 and 2046 is projected for Australia (Inacio et al. 2017). In our study, we modelled an increase of 38% for hip arthroplasties in Germany. Our steady increase in incidence rate appears to be consistent with the reports from other countries. However, the incidence rates per age group were more or less constant over time. Similarly, some publications found a slowing growth rate (Bini et al. 2011, Nemes et al. 2014). For instance, Bini et al. (2011) reported a slowing demand for total joint replacement, with growth rates decreasing from 18% in 2002 to 3% in 2009, and in the Swedish population also the growth of THR decreased after 2000 (Nemes et al. 2014). Rising incidences in older age The incidence of hip osteoarthritis tends to increase with age, which will most likely result in an increasing number of hip arthroplasties. Our study results support this hypothesis as we found the highest number of patients in 2040 in the age group of 75–80 years, which doubled compared with 2010. Pedersen et al. (2009) had similar findings since the incidence rate peaked for patients aged 70–79 during 1996–2002. On the other hand, a study by Pabinger et al. (2014) reported that patients aged 64 or younger have a 7-fold higher growth rate in hip replacement surgeries compared with older patients within the OECD countries. However, a smaller trend was seen for Germany and the categorization of older or younger than 65 might not be suitable for the German population, as it is on average older than other countries. Furthermore, they also identified GDP and health-care expenditures as significant determinants of surgery rates, which could indeed have a greater impact in Germany than in other countries. In other studies a similar trend was found for the United States and in Finland (Kurtz et al. 2009, Skyttä et al. 2011). Limitations Our projection might be limited as it is based on the historical growth of hip arthroplasty counts and the population projection of the Federal Bureau of Statistics. Thus, our projection

5/3/2018 11:42:27 AM

312

may be biased from earlier trends inferred from the historical data but these might change in the future. We could take into account the longer life expectancy of the future population by using the official population forecast. On the other hand, we could not adjust for a possible change in lifestyle such as becoming more active in the later phase of life, which would possibly increase the demand for hip replacements. Further factors that might influence the incidence rate are not taken into account such as the trend in available surgeons, new technologies, or techniques making a hip replacement obsolete, economic trends constraining hospital budgets, or new healthcare policies. However, there is a lack of data on future need and many of these factors are hard to predict reliably, i.e., when they can only be based on the opinions of a small group of experts. Furthermore, every forecast—even to the near future—bears uncertainties that are hard to predict. Unfortunately, we could only include 7 years of historical primary hip arthroplasty counts due to limited access to the data. It is clear that the regression model would be more robust when relying on a larger database. When interpreting our results, one major aspect needs to be considered: we could not separate the different kinds of hip replacement surgeries. Thus, our projection includes elective patients as well as patients undergoing acute hip fracture surgery, making our projection population heterogeneous. However, Wengler et al. (2014) examined the share of primary hip replacements for fracture in Germany, which was constant at around 21% during the study period from 2005 to 2011. Furthermore, we checked the OPS codes for hemiarthroplasty, which are predominantly used in acute hip fracture cases, and also saw a constant share over the last years. Consequently, we do not expect a strong rise in hip fractures in the coming years, which would bias our projection model. Nevertheless, when comparing our results with the OECD countries or other publications one has to bear in mind that different patient inclusion criteria may complicate or even hinder a direct comparison. In our projection model we could only include calendar year and patient age as covariates but did not account for body mass index (BMI), which could also influence the incidence rate of hip arthroplasty (Culliford et al. 2015). However, evidence for BMI as a risk factor for hip osteoarthritis is mixed (Reijman et al. 2007) and it is assumed that the association between BMI and knee osteoarthritis is stronger than for the hip (Jiang et al. 2011). Conclusion Our findings suggest a growth in the total hip arthroplasty count whereas incidence rate remained constant over all age classes. We consider the general future demographic change to an older population as well as the increasing life expectancy to be the main reasons for the increasing patient numbers rather than a general increase in the operation frequency of hip arthroplasty. From an organizational perspective this demographic change might be one of the biggest challenges

12345 Pilz D.indd 312

Acta Orthopaedica 2018; 89 (3): 308–313

for orthopedics hospitals over the coming decades to maintain the current standard of care and—to avoid increased waiting times—infrastructure, as well as operation theatre capacity and personnel resources, must follow future demand. For the health-care system the increase in procedures might have a significant financial impact. Furthermore, with the increase in primary replacements, the number of revision procedures is expected to grow as well. Supplementary data Tables 1 and 3 and Figures 1 and 2 are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/17453674.2018.1446463

VP: literature review, statistical analysis and interpretation, manuscript draft, revision, and approval. TH, RS: study conception, data collection and data preparation, critical manuscript revision and approval.

Acta thanks Ola Rolfson and Eerik T Skyttä for help with peer review of this study.

Bini S A, Sidney S, Sorel M. Slowing demand for total joint arthroplasty in a population of 3.2 million. J Arthroplasty 2011; (26(6)): 124-8. Culliford D, Maskell J, Judge A, Cooper C, Prieto-Alhambra D, Arden N K. Future projections of total hip and knee arthroplasty in the UK: Results from the UK Clinical Practice Research Datalink. Osteoarthritis Cartilage 2015; 23(4): 594-600. DIMDI. Systematisches Verzeichnis - Operationen- und Prozedurenschlüssel - Internationale Klassifikation der Prozeduren in der Medizin (OPS), Cologne: DIMDI unter Beteiligung der Arbeitsgruppe OPS des Kuratoriums für Fragen der Klassifikation im Gesundheitswesen (KKG); 2016. Inacio M C S, Graves S E, Pratt N L, Roughead E E, Nemes S. Increase in total joint arthroplasty projected from 2014 to 2046 in Australia: a conservative local model with international implications. Clin Orthop Relat Res 2017; 475(8): 2130-7. Jiang L, Rong J, Wang Y, Hu F, Bao C, Li X, Zhao Y. The relationship between body mass index and hip osteoarthritis: A systematic review and meta-analysis. Joint Bone Spine 2011; 78 (2): 150-5. Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am 2007; 89(4):7 80-5. Kurtz S M, Lau E, Ong K, Zhao K, Kelly M, Bozic KJ. Future young patient demand for primary and revision joint replacement: national projections from 2010 to 2030. Clin Orthop Relat Res 2009; 467(10): 2606-12. Müller-Bergfort S, Fritze J. Diagnose- und Prozedurendaten im deutschen DRG-System. Bundesgesundheitsblatt-Gesundheitsforschung-Gesundheitsschutz. 2007; 50(8): 1047-54. Nemes S, Gordon M, Rogmark C, Rolfson O. Projections of total hip replacement in Sweden from 2013 to 2030. Acta Orthop 2014; 85(3): 238-43. Nemes S, Rolfson O, W-Dahl A, Garellick G, Sundberg M, Kärrholm J, Robertsson O. Historical view and future demand for knee arthroplasty in Sweden. Acta Orthop 2015; 86(4): 426-31. Nowossadeck E. Population aging and hospitalization for chronic disease in Germany. Dtsch Arztebl Int 2012; 109(9): 151-7. OECD. Health at a glance. 2015. Paris: OECD Publishing; 2015. Otten R, van Roermund P M, Picavet H S J. Trends in aantallen knie- en heupartroplastieken: De vraag naar knie- en heupprotheses blijft voorlopig toenemen. Ned Tijdschr Geneeskd 2010; 154: A1534.

5/3/2018 11:42:27 AM

Acta Orthopaedica 2018; 89 (3): 308–313

Pabinger C, Geissler A. Utilization rates of hip arthroplasty in OECD countries. Osteoarthritis Cartilage 2014(22): 734-41. Pedersen A B, Johnsen S P, Overgaard S, Søballe K, Sørensen H T, Lucht U. Total hip arthroplasty in Denmark. Acta Orthop 2009; 76(2): 182-9. Pötzsch O, editor. (Un-)Sicherheiten der Bevölkerungsvorausberechnungen: Rückblick auf die koordinierten Bevölkerungsvorausberechnungen für Deutschland zwischen 1998 und 2015. Wiesbaden: Federal Statistical Office; 2016. Pötzsch O, Rößger F. Demographic analyses, methods and projections, births and deaths: Germany’s population by. 2060. Results of the 13th coordinated population projection. Wiesbaden: Federal Statistical Office; 2015. Reijman M, Pols H A, Bergink A P, Hazes J , Belo J N, Lievense A M, Bierma-Zeinstra S M. Body mass index associated with onset and progression of osteoarthritis of the knee but not of the hip: the Rotterdam Study. Ann Rheum Dis 2007; 66(2): 158-62. Schäfer T, Pritzkuleit R, Jeszenszky C, Malzahn J, Maier W, Günther K P, Niethard F. Trends and geographical variation of primary hip and knee joint replacement in Germany. Osteoarthritis Cartilage 2013; 21(2): 279-88. Skyttä ET, Jarkko L, Antti E, Huhtala H, Ville R. Increasing incidence of hip arthroplasty for primary osteoarthritis in 30- to 59-year-old patients: a population based study from the Finnish Arthroplasty Register. Acta Orthop 2011; 82(1): 1-5.

12345 Pilz D.indd 313

313

Spindler J. Fallpauschalenbezogene Krankenhausstatistik: Diagnosen und Prozeduren der Krankenhauspatienten auf Basis der Daten nach §. 21 Krankenhausentgeltgesetz. In: Klauber J, Geraedts M, Friedrich J, Wasem J, editors. Krankenhaus-Report 2017. Stuttgart: Schattauer Verlag; 2017, Pp. 321-53. Statistisches Bundesamt. Qualitätsbericht zur Berechnung der koordinierten Bevölkerungsvorausberechnungen für Deutschland und die Bundesländer: 2015-2017. In: Statistisches Bundesamt Bevölkerungsvorausberechnung, editor. Wiesbaden: Statistisches Bundesamt; 2016. Statistisches Bundesamt. Gesundheit: Fallpauschalenbezogene Krankenhausstatistik(DRG-Statistik) Operationen und Prozeduren der vollstationären Patientinnen und Patienten in Krankenhäusern 2016. Wiesbaden: Statistisches Bundesamt; 2017. Wedderburn R W. Quasi-likelihood functions, generalized linear models, and the Gauss-Newton method. Biometrika 1974; 61(3): 439. Wengler A, Nimptsch U, Mansky T. Hip and knee replacement in Germany and the USA: analysis of individual inpatient data from German and US hospitals for the years 2005 to 2011. Dtsch Arztebl Int 2014; (111): 407-16.

5/3/2018 11:42:27 AM

314

Acta Orthopaedica 2018; 89 (3): 314–319

Obesity-related metabolic and endocrine disorders diagnosed during postoperative follow-up of slipped capital femoral epiphysis Hanifi UCPUNAR 1, Ismet Yalkin CAMURCU 1, Serda DUMAN 2, Esra UCPUNAR 3, Hakan SOFU 1, and Avni Ilhan BAYHAN 4

1 Erzincan

University Faculty of Medicine, Department of Orthopaedics and Traumatology; 2 Diyarbakir Selahaddin Eyyubi State Hospital, Department of Orthopaedics and Traumatology; 3 Erzincan University Faculty of Health Sciences, Department of Public Health; 4 Baltalimani Bone and Joint Diseases Education and Research Hospital, Department of Pediatric Orthopaedics, Turkey Correspondence: hanifiucpunar@yahoo.com Submitted 2017-09-28. Accepted 2018-01-22.

Background and purpose — Patients with slipped capital femoral epiphysis (SCFE) are phenotypically overweight or obese and may therefore require clinical follow-up of obesity-related disorders. We evaluated obesity-related disorders such as dyslipidemia, type 2 diabetes mellitus (DM), and vitamin-D deficiency during the postoperative period in patients with SCFE. Patients and methods — 51 patients who were operated and followed-up for SCFE and 62 healthy adolescents without SCFE (control group) were included in this retrospective study. Patients’ BMI, serum lipid profile (total cholesterol, LDL-C, HDL-C, triglyceride), fasting blood glucose, HbA1c, and serum vitamin D levels were evaluated. Results — At the time of surgery, 45 patients in the SCFE group were overweight or obese (BMI > 25). At the latest follow-up, 42 patients in the SCFE group and 53 patients in the control group were overweight/obese. Abnormal serum lipid profile and ratio of total dyslipidemia were similar between the groups. 8 patients had abnormal HbA1c levels in the SCFE group and mean HbA1c levels were significantly higher in the SCFE group (p = 0.03). All patients and controls had low levels of vitamin D. Interpretation — Although serum lipid profile and vitamin D levels were detected as similar in SCFE and control groups, the potential risk of type 2 DM identified via abnormal HbA1c levels was significantly higher in patients with SCFE. We recommend that patients diagnosed with SCFE should be considered as potential candidates for type 2 DM; thus follow-up after surgical treatment should include not only orthopedic outcomes but also evaluation of future risk for DM.

The incidence of slipped capital femoral epiphysis (SCFE) varies with demographic factors such as age, sex, and ethnicity (Loder et al. 2008). However, the exact etiological factors and the leading pathologic mechanism has not yet been clarified. Constitutional femoral retroversion, high BMI, increased vertical slope of the proximal femoral physeal line, and endocrine pathologies have been reported as potential predisposing factors for the pathogenesis of the disorder (Weiner 1996, Witbreuk et al. 2013). Although the majority of the adolescents diagnosed with SCFE may not have hormonal, metabolic, or chronic diseases they have been reported as phenotypically obese and fast growing (Witbreuk et al. 2013). Novais and Millis (2012) reported that the incidence of SCFE was correlated with increased BMI. Due to this relationship of the disease with obesity, the patients may require clinical follow-up of obesityrelated systemic disorders (Murray and Wilson 2008). Additionally, clinical management of these patients may become more complicated because it is still unclear whether the hormonal abnormalities are the reason for or the result of SCFE. The purpose of this study was to evaluate obesity-related metabolic and endocrine disorders such as dyslipidemia frequency, the prevalence of type 2 diabetes mellitus, and serum vitamin D levels during the clinical follow-up in patients who underwent surgical treatment for SCFE in comparison with overweight/obese healthy adolescents without SCFE.

■

12117 Ucpunar D.indd 314

5/3/2018 11:43:35 AM

Acta Orthopaedica 2018; 89 (3): 314–319

315

Serum levels of total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDLC), and triglyceride (TG) were determined Excluded (n = 12): Excluded (n = 8): Established diagnosis of Established diagnosis of by enzymatic colorimetric method to evalusystemic metabolic or systemic metabolic or ate the lipid profile. The results were classiendocrine disorder endocrine disorder at the – hypothyroidism, 8 time of surgical treatment fied as acceptable, borderline high, and high – polycystic ovary syndrome, 4 – hypothyroidism, 4 – hyperthyroidism, 2 for TC, LDL-C, and TG whereas acceptable, – hypogonadism, 1 borderline low, and low for HDL-C (NCEP – renal osteodystrophy , 1 1992). The patients with high serum levels Obese patients without Surgically treated patients on any of the measured TC, LDL-C, or TG established diagnosis of without established diagnosis systemic metabolic or of systemic metabolic or as well as those with low or borderline low endocrine disorder endocrine disorder HDL-C were identified as dyslipidemia. Fast(n = 75) (n = 68) ing blood glucose (FBG) was determined by Excluded (n = 13): Excluded (n = 17): enzymatic colorimetric method and hemoglo– insufficient clinical records, 13 – lost to follow-up, 5 bin A1c (HbA1c) using the immunoturbidi– past medical history of steroid treatment, 4 metric method. Fasting blood glucose > 125 – insufficient clinical records, 4 – polycystic ovary syndrome, 4 mg/dL was defined as impaired fasting glucose (IFG). According to an HbA1c test, the Control group SCFE group patients with a result of ≤ 5.7% were classified analyzed analyzed as normal glucose tolerance (NGT), between (n = 62) (n = 51) 5.7% and 6.4% as at risk for diabetes melliFlow chart demonstrating excluded patients. tus (DM), and > 6.5% as obvious type 2 DM (Nowicka et al. 2011). The serum vitamin D (25-OH-cholecalciferol) level was determined using the elecPatients and methods trochemiluminescence immunoassay (ECLIA) method. The Study population results were classified as deficient (< 20 ng/mL), insufficient Clinical data of the patients were retrospectively evaluated. (21–29 ng/mL), and sufficient (> 30 ng/mL) (Holick 2009). The SCFE group consisted of patients who were surgically We also recorded serum levels of free tri-iodothyronine (F-T3), treated between 2008 and 2014 with the diagnosis of unilat- free thyroxine (F-T4), thyroid stimulating hormone (TSH, eral or bilateral idiopathic SCFE. Control individuals were creatinine (kinetic colorimetric test, Jaffé compensated), urea recruited from the adolescents admitted to our outpatient (Urease method), alanine transaminase (ALT)-aspartate aminoclinic. The control group consisted of mainly overweight transferase (AST) (kinetic method, IFFC without pyrphosp.), or obese (BMI > 25) otherwise healthy adolescents without which was determined using the electrochemiluminescence SCFE. Patients and controls with an established diagnosis of immunoassay (ECLIA) method. All biochemical tests were any specific systemic, metabolic, or endocrine disorder at the analyzed with an Auto-Analyzer (Roche/Hitachi Cobas c501; time of surgery or who were under medical treatment which Roche Diagnostics USA, Indianapolis, IN, USA). has potential negative effects on glucose, lipid, and vitamin D metabolism were excluded from the study. 51 patients with Statistics SCFE and 62 patients without SCFE were eligible for the Statistical analysis was performed using the MedCalc Statistical Software version 12.7.7 (MedCalc Software bvba, Ostend, study (Figure). Belgium; http://www.medcalc.org; 2013). The normality of Clinical data collection continuous variables was investigated by the Shapiro–Wilk The clinical data of the 2 groups were obtained from our clini- test. Descriptive statistics were presented using mean (SD) for cal database by analyzing laboratory test results and clinical continuous variables. Non-parametric statistical methods were consultation notes of the pediatrics department. The BMI of used for values with skewed distribution. For comparison of all patients at the first control as well as the latest follow-up 2 non-normally distributed groups the Mann–Whitney U-test visit were recorded. BMI values were classified using the was used. The chi-square test was used for categorical varipercentile chart as underweight when < 5%, normal weight ables and expressed as observation counts (and percentages). between 5% and 85%, overweight between 85% and 95%, and Fisher’s exact test was used instead of a chi-square test when obese > 95% (Ogden et al. 2002). there was at least 1 expected value less than 5. Multivariate Blood samples for evaluation of the serum levels of various binary logistic regression analysis was used to determine the parameters were obtained following a 12-hour fasting period relationship between HbA1c and independent variables such without any previous change in the routine diet of the patients. as SCFE, overweight or obesity, blood lipid profile, fasting Obese patients without SCFE (n = 87)

12117 Ucpunar D.indd 315

SCFE (n = 76)

5/3/2018 11:43:36 AM

316

Acta Orthopaedica 2018; 89 (3): 314–319

weight or obese at the time of surgery. At latest follow-up, the number of the overweight or obese patients in the SCFE group had decreased to 42. SCFE group (n = 51) Healthy group Data At time of surgery Last control (n = 62) p-value In the control group 53/62 individuals were overweight or obese. There was no statistically siga c Female/male 4 / 47 5 / 57 1 nificant difference between the 2 groups regardAge (years) b 13.4 (11–16) 16.6 (16–23) 16.9 (15–21) 0.9 d CI [13.0–13.7] [16.1–17.1] [16.1–16.9] ing TC, LDL-C, HDL-C, and TG levels (Table 2). Follow-up (years) b – 3.1 (2–7) – Total dyslipidemia frequency in the SCFE group b d BMI percentile 91 (65–99) 89 (59–99) 90 (65–98) 0.9 was 68% and in the control group it was 59% CI [89–94] [87–92] [87–92] BMI percentile a with no statistically significant difference between Obese 16 13 15 the groups. According to serum HbA1c levels, 8 e Overweight 29 29 38 0.6 patients were diagnosed as either at risk of DM Normal weight 6 9 9 or with obvious type 2 DM, and 2 of the SCFE a Number of patients. patients were diagnosed as IFG according to the b Values are mean, (range), and [95% confidence interval]. c p-values (2-sided) according chi-square/Fisher’s exact test and FBG test (Table 3). In the control group, 2 adolesd p-values (2-sided) according to Student’s t-test. p-values indicate comparison of cents were classified as at risk of DM according to last control values in SCFE group and values in healthy group. HbAc1 levels. The frequency of elevated HbA1c e p-value that expresses comparison of frequencies of obese or overweight levels and higher mean serum HbA1c were statisindividuals in the SCFE group and control group. tically significantly different in SCFE group when compared with controls (Table 3). Multivariate blood glucose, and vitamin-D levels. Statistical significance binary logistic regression analysis showed a significant relawas accepted when a 2-sided p-value was < 0.05. tionship of HbA1c only for the SCFE group (Table 4). Serum vitamin D measurements showed that none of the individuals Ethics, funding, and potential conflicts of interest in either group had a sufficient level of 25-OH-cholecalciferol, This study was performed after receiving approval from the and this was similar between the groups (Table 3). F-T3, F-T4, institutional ethical review board (33216249-604.01.02- TSH, creatinine, urea, ALT, and AST were measured within E.16361). Informed consent was obtained from all patients normal reference ranges in both the SCFE and control group. and from individuals in the healthy control group. No funding was received for this study. All authors declare that they have no conflict of interest regarding the submission and publicaDiscussion tion of this manuscript. The clinical management of patients with SCFE has evolved to a multidisciplinary treatment approach because of their accompanying endocrine and metabolic disorders as well as Results obesity-related health conditions. Impaired mobility status, Table 1 summarizes demographic data of the patients and con- secondary skeletal deformities like tibia vara, and SCFE during trols. 45 of the 51 patients in the SCFE group were either over- the pubertal age, as well as systemic disorders such as glucose Table 1. Demographic data

Table 2. Comparison of last-follow up serum lipid profile and serum 25-OH vitamin D levels of the patients versus control group SCFE group (n = 51) Lipid profile (mg/dL) Total cholesterol c LDL-C c TG c HDL-C d Vit-D (25-OH) (ng/mL) e

36 / 8 / 7 30 / 6 / 15 18 / 5 / 28 14 / 13 / 24 0 / 39 / 12

160 (33) [150–169] 122 (96) [95–149] 167 (142) [127–207] 45 (11) [42–48] 17 (8)

[15–19]

Control group (n = 62) 47 / 8 / 7 41 / 14 / 7 29 / 25 / 8 10 / 17 / 35 0 / 46 / 16

158 (30) [151–166] 112 (89) [89–135] 149 (134) [114–184] 46 (10) [43–49] 17 (8)

[15–19]

p-value a p-value b 0.9 0.2 0.3 0.4

0.5 0.7 0.2 0.7

0.7

0.6

a p-values

indicate the comparison of the sum of borderline-high and high frequencies of total cholesterol, LDL-C, TG, the sum of borderline-low and low frequencies of HDL-C and insufficient frequencies of Vit-D of the 2 groups. b Adjusted by sex and age. c Values are number of patients with acceptable / borderline high / high levels, mean serum level (SD), and [95% CI] d Values are number of patients with low / borderline low / acceptable levels, mean serum level (SD), and [95% CI] e Values are number of patients with sufficient / deficient / insufficient levels, mean serum level (SD), and [95% CI]

12117 Ucpunar D.indd 316

5/3/2018 11:43:39 AM

Acta Orthopaedica 2018; 89 (3): 314–319

317

ning of the contralateral hip in unilateral SCFE cases with a BMI-for-age > 95% according to percentile chart SCFE group Control group was also mentioned as one of the (n = 51) (n = 62) p-value mostly discussed controversies in the HbA1c SCFE literature (Manoff et al. 205). Serum level (mg/dL) a 5.5 (0.3) [5.4–5.6] 5.3 (0.2) [5.3–5.4] < 0.01 c 0.04 d In our patients treated surgically for Normal glucose tolerance b 43 60 0.04 e At risk for diabetes mellitus b 7 2 SCFE, four-fifths of the patients were b Type 2 diabetes mellitus 1 0 either overweight or obese preoperaFasting blood glucose tively and this ratio was reduced only Serum level (mg/dL) a 99 (14) [95–103] 99 (SD) [95–102] 0.7 c Normal fasting glucose b 49 62 NS slightly postoperatively. Obviously, b Impaired fasting glucose 2 0 persistently high BMI ratios demonstrated the underestimated clinical relNS = not studied a Values are mean (standard deviation), and [confidence interval] evance of weight-control programs as b Values are number of patients. well as the ignored importance of obec p-value indicates the comparison of the means of serum levels of HbA1c and fasting blood sity and obesity-related disorders in glucose of two groups according to Mann–Whitney U-test. d Adjusted by sex and age. patients surgically treated for SCFE. e p-value indicates the comparison of the sum of at risk for diabetes mellitus and type 2 diabetes Obesity-related metabolic and mellitus frequencies of two groups according to chi-square/Fisher’s exact (F) test. endocrine disorders have negative effects on physeal development during the fast growing phase of the pre-pubertal and puberTable 4. Multivariate logistic regression analyses of potential high HbA1c level related to confounding variables tal ages (Witbreuk et al. 2013). Dyslipidemia was reported in between 40% and 45% of obese children of school age in Germany and Turkey (Korsten-Reck et al. 2008, Cizmecioğlu OR (Exp(B)) 95% CI p-value et al. 2008, Elmaogullari et al. 2015). It was emphasized SCFE 6.0 1.1–32 0.03 a that atherosclerosis-related systemic disorders during adultBlood glucose 1.3 0.9–1.1 0.4 hood could begin in childhood and adolescence (Berenson et BMI percentile 1.2 0.9–1.5 0.7 Age 1.2 0.6–2.4 0.5 al. 2016). Elmaogullari et al. (2015) mentioned that the risk LDL-C 0.9 0.9–1.0 0.4 of dyslipidemia increased in correlation with age and BMI. Total cholesterol 1.0 0.9–1.0 0.7 The ratio of dyslipidemia was similar between the groups in Triglycerides 1.0 0.9–1.0 0.2 Vitamin D 0.9 0.8–1.0 0.3 our study, and additionally no significant differences were observed between groups regarding TC, LDL-C, HDL-C, and OR = odds ratio a HbA1c levels were found to be related only to the SCFE group. TG. Our results are consistent with previous studies evaluating Other variables, especially weight, which has the potential to raise the lipid profile of overweight or obese children in similar age HbA1c levels, were irrelevant. The confidence interval is rather wide groups (Korsten-Reck et al. 2008, Cizmecioğlu et al. 2008, because the numbers of abnormal HbA1c results, particularly for Elmaogullari et al. 2015, Berenson et al. 2016). However, no non-SCFE cases, are very small. specific study exists evaluating the lipid profile in children with SCFE and comparing results with obese individuals in intolerance, type 2 DM, hyperlipidemia, non-alcoholic fatty the literature. The data that we acquired in this study are insufliver, cholelithiasis, and primary hypertension, are reported ficient to explain whether dyslipidemia was an etiologic factor with increased prevalence in overweight or obese individu- for the slippage or simply the result of obesity-related metaals (Norman et al. 2005, Taylor et al. 2006, Newmark and bolic syndrome. It is well known that most obese individuals have accomAnhalt 2007, Torun et al. 2013). More than 80% of the patients with a diagnosis of SCFE were reported as obese individuals panying blood glucose abnormalities as well as type 2 DM (Manoff et al. 2005). Nasreddine et al. (2013) evaluated 173 (Pinhas-Hamiel et al. 1996). Bowen et al. (2009) observed 3 cases of SCFE and noted that the ratio of obese patients was separate obesity-related phenotypes in adolescents with no 80% preoperatively, which decreased to 78% during a 3-year overlap of disease at initial presentation among SCFE, adofollow-up period. Aversano et al. (2016) demonstrated an asso- lescent tibia vara, and type 2 DM. They questioned the coexisciation between BMI-for-age and risk for bilateral SCFE at tence of type 2 DM at the time of SCFE diagnosis before surpresentation as well as overall incidence of developing bilat- gical treatment. In their retrospective study evaluating patienteral SCFE in obese children. They concluded that by defining reported outcomes as well as comorbidities with an average the at-risk population through BMI-for-age, physicians could of 19-year follow-up period, Escott et al. (2015) reported DM develop early strategies for therapeutic weight loss, which may in 8% of cases. Moreover, they noted an increase in the mean reduce the incidence of SCFE. Furthermore, prophylactic pin- BMI of 10 during postoperative follow-up. The rate of obeTable 3. Comparison of last follow-up HbA1c and fasting blood glucose levels

12117 Ucpunar D.indd 317

5/3/2018 11:43:39 AM

318

sity and obesity-related comorbidities was detected as higher than in the general population. In our study, 7 patients were detected as at risk of DM and 1 patient was diagnosed as obvious type 2 DM according to HbA1c results. However, in the control group only 2 patients were detected as at risk of DM according to IFG levels. There was a statistically significant difference between the 2 groups regarding the mean of HbA1c levels and abnormal HbA1c frequency. Additionally, in a multivariate logistic regression with confounding factors that may affect HbA1C value, HbA1c levels were found to be related only to the SCFE group. Other variables, especially weight, which may have the potential to raise HbA1c levels were irrelevant. In the SCFE group, the frequency of patients with abnormal serum glucose level was higher than the ratio reported by Escott et al. (2015). They evaluated the coexistence of DM with SCFE preoperatively in patients who underwent surgical treatment. Additionally, we evaluated DM postoperatively at 3-year follow-up compared with obese patients without SCFE. Therefore, we recommend that the patients diagnosed with and treated for SCFE should be followed up for potential abnormalities of glucose metabolism, which may not be detected on the first admission. Vitamin D deficiency has been noted in young patients admitted to hospital with orthopedic problems, as well as in healthy adolescents (Davies et al. 2011). Buyukinan and Ozen (2012) reported vitamin D deficiency in obese adolescents as high as 96%. According to a study analyzing the results after in situ pinning, a negative correlation was found between vitamin D levels and the fusion time in the physis line (Judd et al. 2015). Furthermore, Judd et al. (2015) also reported that only 2 of the 27 patients in their series had a normal serum vitamin D level at the time of SCFE diagnosis. On the other hand, Arkader et al. (2015) reported that all patients diagnosed with SCFE between the ages of 9 and 14 years had normal vitamin D levels. In their series of 15 cases with SCFE, Madhuri et al. (2013) also reported that none of the patients had normal serum vitamin D according to the tested reference range. None of our patients and controls had normal levels of vitamin D. Our patients and healthy individuals belong to the same ethnicity and live in the same city, and blood-taking periods were evenly distributed in all seasons of the year. Maybe surgeons dealing with the treatment of such patients should consider routine follow-up of vitamin D levels at least during the fusion time of the epiphysis after surgical intervention. Although a possible relationship of SCFE with endocrinological disorders has been evaluated by several studies, routine clinical screening for endocrine pathologies has not been recommended due to low rate of reported coexistence (Ogden and Southwick 1977, Mann et al. 1988). We also noted normal serum levels of F-T3, F-T4, TSH, PTH, ALT, AST, creatinine, and urea in our patients. We note some limitations of our study. First, it was a retrospective evaluation of a prospectively followed patient group. However, our cohort was a relatively large series of patients

12117 Ucpunar D.indd 318

Acta Orthopaedica 2018; 89 (3): 314–319

with a rare diagnosis. Furthermore, the main strength of the study was that we studied serum levels of many different parameters compared with a control group. In summary, a risk of type 2 DM with abnormal HbA1c levels was significantly higher in patients with SCFE when compared with obese adolescents without SCFE. Additionally, an abnormal serum lipid profile and vitamin D insufficiency was similar in patients with SCFE when compared with obese/overweight adolescents.

HU, MD was involved in study design, manuscript preparation, manuscript writing, and is the corresponding author. YC, MD and EU, MSc were involved in data collection, statistical analysis, and manuscript editing. SD, MD carried out data collection and analysis of literature. HS, MD was involved in language editing and performed measurements. AIB, MD was the senior author, and involved in last revision of the manuscript.

Acta thanks anonymous reviewers for help with peer review of this study.

Arkader A, Woon R P, Gilsanz V. Can subclinical rickets cause SCFE? A prospective, pilot study. J Pediatr Orthop 2015; 35(7): e72-5. Aversano M W, Moazzaz P, Scaduto A A, Otsuka N Y. Association between body mass index-for-age and slipped capital femoral epiphysis: the longterm risk for subsequent slip in patients followed until physeal closure. J Child Orthop 2016; 10(3): 209-13. Berenson G S, Srinivasan S R, Xu J H, Chen W. Adiposity and cardiovascular risk factor variables in childhood are associated with premature death from coronary heart disease in adults: The Bogalusa Heart Study. Am J Med Sci 2016; 352(5): 448-54. Bowen J R, Assis M, Sinha K, Hassink S, Littleton A. Associations among slipped capital femoral epiphysis, tibia vara, and type 2 juvenile diabetes. J Pediatr Orthop 2009; 29(4): 341-4. Buyukinan M, Ozen S. The relation of vitamin D deficiency with puberty and insulin resistance in obese children and adolescents. J Pediatr Endocrinol Metab 2012; 25(1-2): 83-7. Davies J H, Reed J M, Blake E, Priesemann M, Jackson A A, Clarke N M. Epidemiology of vitamin D deficiency in children presenting to a pediatric orthopaedic service in the UK. J Pediatr Orthop 2011; 31(7): 798-802. Cizmecioğlu F M, Hatun S, Kalaça S. Metabolic syndrome in obese Turkish children and adolescents: comparison of two diagnostic models. Turk J Pediatr 2008; 50(4): 359-65. Elmaogullari S, Tepe D, Ucakturk S A, Kara F K, Demirel F. Prevalence of dyslipidemia and associated factors in obese children and adolescents. J Clin Res Pediatr Endocrinol 2015; 7(3): 228-34. Escott B G, De La Rocha A, Jo C H, Sucato D J, Karol L A. Patient-reported health outcomes after in situ percutaneous fixation for slipped capital femoral epiphysis. J Bone Joint Surg Am 2015; 97(23): 1929-34. Holick M F. Vitamin D status: measurement, interpretation, and clinical application. Ann Epidemiol 2009; 19(2): 73-8. Judd J, Welch R, Clarke A, Reading I C, Clarke N M. Vitamin D deficiency in slipped upper femoral epiphysis: time to physeal fusion. J Pediatr Orthop 2015; 36(3): 247-52. Korsten-Reck U, Kromeyer-Hauschild K, Korsten K, Baumstark M W, Dickhuth H H, Berg A. Frequency of secondary dyslipidemia in obese children. Vasc Health Risk Manag 2008; 4(5): 1089-94. Loder R T, Aronsson D D, Weinstein S L, Breur G J, Ganz R, Leunig M. Slipped capital femoral epiphysis. Inst Course Lect 2008; 57: 473-98. Madhuri V, Arora S K, Dutt V. Slipped capital femoral epiphysis associated with vitamin D deficiency: a series of 15 cases. Bone Joint J 2013; 95-B(6): 851-4.

5/3/2018 11:43:39 AM

Acta Orthopaedica 2018; 89 (3): 314–319

Mann D C, Weddington J, Richton S. Hormonal studies in patients with slipped capital femoral epiphysis without evidence of endocrinopathy. J Pediatr Orthop 1988; 8(5): 543-5. Manoff E M, Banffy M B, Winell J J. Relationship between body mass index and slipped capital femoral epiphysis. J Pediatr Orthop 2005; 25(6): 744-6. Murray A W, Wilson N I. Changing incidence of slipped capital femoral epiphysis: a relationship with obesity? J Bone Joint Surg Br 2008; 90(1): 92-4. Nasreddine A Y, Heyworth B E, Zurakowski D, Kocher M S. A reduction in body mass index lowers risk for bilateral slipped capital femoral epiphysis. Clin Orthop Relat Res 2013; 471(7): 2137-44. Newmark C Y, Anhalt H. Type 2 diabetes in children and adolescents. Pediatr Ann 2007; 36(2): 109-13. NCEP Expert Panel on Blood Cholesterol Levels in Children and Adolescents. National Cholesterol Education Program: highlights of the Report of the Expert Panel on Blood Cholesterol Levels in Children and Adolescents. Pediatrics. 1992; 89(3): 495-501. Norman A C, Drinkard B, McDuffie J R, Ghorbani S, Yanoff L B, Yanovski J A. Influence of excess adiposity on exercise fitness and performance in overweight children and adolescents. Pediatrics 2005; 115(6): e690-6.25. Novais E N, Millis M B. Slipped capital femoral epiphysis: prevalence, pathogenesis and natural history. Clin Orthop Relat Res 2012; 470(12): 3432-8. Nowicka P, Santoro N, Liu H, Lartaud D, Shaw M M, Goldberg R, Guandalini C, Savoye M, Rose P, Caprio S. Utility of hemoglobin A1c for diagnosing prediabetes and diabetes in obese children and adolescents. Diabetes Care 2011; 34(6): 1306-11.

12117 Ucpunar D.indd 319

319

Ogden J A, Southwick W. Endocrine dysfunction and slipped capital femoral epiphysis. Yale J Biol Med 1977; 50(1): 1-16. Ogden C L, Flegal K M, Carroll M D, Johnson C L. Prevalence and trends in overweight among US children and adolescents, 1999–2000. JAMA 2002; 288(14): 1728-32. Pinhas-Hamiel O, Dolan L M, Daniels S R, Standiford D, Khoury P R, Zeitler P. Increased incidence of non-insulin-dependent diabetes mellitus among adolescents. J Pediatr 1996; 128(5 Pt 1): 608-15. Taylor E D, Theim K R, Mirch M C, Ghorbani S, Tanofsky-Kraff M, AdlerWailes D C, Brady S, Reynolds J C, Calis K A, Yanovski J A. Orthopedic complications of overweight in children and adolescents. Pediatrics 2006; 117(6): 2167-174. Torun E, Gönüllü E, Ozgen I T, Cindemir E, Oktem F. Vitamin D deficiency and insufficiency in obese children and adolescents and its relationship with insulin resistance. Int J Endocrinol 2013; 2013: 631845. Weiner D. Pathogenesis of slipped capital femoral epiphysis: current concepts. J Pediat Orthop Part B. 1996; 5: 67-73. Witbreuk M, van Kemenade F J, van der Sluijs J A, Jansma E P, Rotteveel J, van Royen B J. Slipped capital femoral epiphysis and its association with endocrine, metabolic and chronic diseases: a systematic review of the literature. J Child Orthop 2013; 7(3): 213-23.

5/3/2018 11:43:39 AM

320

Acta Orthopaedica 2018; 89 (3): 320–328

RSA migration of total knee replacements A systematic review and meta-analysis Bart G PIJLS 1, José W M PLEVIER 2, and Rob G H H NELISSEN 1

1 Department of Orthopaedics, Leiden University Medical Center, Leiden, 2 Walaeus Library, Leiden University Medical Center, Leiden, The Netherlands Correspondence: b.g.c.w.pijls@lumc.nl Submitted 2017-03-20. Accepted 2018-01-16.

Purpose — We performed a systematic review and meta-analyses to evaluate the early and long-term migration patterns of tibial components of TKR of all known RSA studies. Methods — Migration pattern was defined as at least 2 postoperative RSA follow-up moments. Maximal total point motion (MTPM) at 6 weeks, 3 months, 6 months, 1 year, 2 years, 5 years, and 10 years were considered. Results — The literature search yielded 1,167 hits of which 53 studies were included, comprising 111 study groups and 2,470 knees. The majority of the early migration occurred in the first 6 months postoperatively followed by a period of stability, i.e., no or very little migration. Cemented and uncemented tibial components had different migration patterns. For cemented tibial components there was no difference in migration between allpoly and metal-backed components, between mobile bearing and fixed bearing, between cruciate retaining and posterior stabilized. Furthermore, no difference existed between TKR measured with model-based RSA or marker-based RSA methods. For uncemented TKR there was some variation in migration with the highest migration for uncoated TKR. Interpretation — The results from this meta-analysis on RSA migration of TKR are in line with both the survival analyses results from joint registries of these TKRs as well as revision rates results from meta-analyses, thus providing further proof for the association between early migration and late revision for loosening. The pooled migration patterns can be used both as benchmarks and for defining migration thresholds for future evaluation of new TKR. ■

Worldwide, some several hundred thousand total knee replacements (TKR) are implanted every year. Aseptic loos-

ening is the major reasonfor revision in the long term, with rates between 5% and 10% in large national registry databases (AOANJRR 2016, SKAR 2016). Early migration of tibial components of TKR, measured with radiostereographic analysis (RSA), has been associated with long-term risk of revision for aseptic loosening (Pijls et al. 2012a). Furthermore, continuous migration during the second postoperative year has been associated with early onset of loosening (Ryd et al. 1995). One may consider early migration of an orthopedic implant (maximal total point motion (MTPM) 1 year) to be associated with achieving fixation and therefore to provide information on the bone–prosthesis or bone–cement–prosthesis interface. The better the fixation, the lower the risk of aseptic loosening in the future. The worse the initial fixation, the higher the risk of aseptic loosening in the future. One may consider continuous migration (MTPM over 1–2 years) to be associated with early manifestation of pathology, i.e., early loosening (Ryd et al. 1995). If the continuous migration progresses it will eventually be radiographically visible (i.e., radiolucencies around the implant) and give rise to symptoms in the patient. Although early migration and continuous migration are 2 completely different entities, they can be used in conjunction in a phased clinical introduction of new TKR designs and fixation techniques (e.g., new bone cements). The combination of this early and continuous migration defines a specific migration pattern for a specific implant (Pijls and Nelissen 2016). With an increasing number of studies evaluating early migration as well as long-term migration it has now become possible to evaluate the migration pattern between different TKR designs and modes of fixation. This is important since a particular migration pattern may be normal for one TKR design or fixation, but pathological for another TKR design or fixation. It also allows for comparison of the migration measured with

11469 Pijls D.indd 320

5/3/2018 11:44:49 AM

Acta Orthopaedica 2018; 89 (3): 320â&#x20AC;&#x201C;328

different RSA techniques: model-based RSA versus different types of marker-based RSA. The purpose of this systematic review and meta-analysis is therefore to evaluate the early and long-term migration patterns of tibial components of TKR of all known RSA studies.

Material and methods This systematic review is reported in accordance with the PRISMA statement (Liberati et al. 2009). Literature search We performed a thorough literature search together with a medical librarian (JP) to reduce bias by increasing the likelihood of retrieving all relevant studies (Vochteloo et al. 2010). The following bibliographies were searched up to July 2016: Pubmed, Embase, Web-of-Science, and the Cochrane Library. Articles in English, French, Italian, Spanish, Dutch, and German were considered. The search strategy consisted of the following components, each defined by a combination of controlled vocabulary and free text terms: (1) RSA, and (2) total knee joint replacement. This search strategy has been used in previously published meta-analyses (Pijls et al. 2012a, 2012b, van der Voort et al. 2015). Inclusion and exclusion analysis Initial screening on the basis of title and abstract of RSA studies was performed by BP to identify studies on patients treated with primary TKRs. When the information in the abstract did not suffice or where there was any doubt, the studies remained eligible. The full text of eligible studies was evaluated by BP with RN as consultant when needed. The inclusion criteria were: (1) primary TKR, and (2) MTPM. MTPM is the unit of measurement for the largest 3D migration of any point on the prosthesis surface (Ryd et al. 1995). Migration pattern was defined as at least 2 postoperative follow-up moments within the first 2 years of follow-up. Non-clinical studies (animal, phantom) were excluded. Data extraction BP extracted data from the included studies. Mean MTPM with corresponding standard deviation (SD) was extracted or calculated from reported median, inter-quartile range (IQR), or range using internationally accepted methodology (Hozo et al. 2005). Sometimes these values had to be estimated from graphs. In the very rare case that only the mean MTPM was given without range, IQR, SE, or SD, the SD was calculated as the average SD from similar studies. The rationale for this is that it is more important to include data that can later be subjected to sensitivity analyses, rather than to exclude data, which could lead to bias. MTPM at 6 weeks, 3 months, 6 months, 1 year, 2 years, 5 years, and 10 years were considered. Data concerning patient demograph-

11469 Pijls D.indd 321

321

ics, RSA technique, and prosthesis characteristics (i.e., type of prosthesis, fixation, and insert) were extracted to allow for sub-group analyses. Data synthesis and analysis Prostheses were classified according to prosthesis, fixation and insert (PFI) methodology, as previously used (Pijls et al. 2012a). A study group was defined as a group of patients in a study with the same PFI. Typically these represent the treatment groups of a trial: e.g., a trial comparing fixed versus mobile bearing has 2 study groups. Up to 2-year follow-up we determined and plotted the 10th, 25th, 50th, 75th, and 90th percentile of the means (there were not enough study groups with follow-up beyond 2 years to reliably determine the percentiles). A random effects model was employed to pool the MTPM of individual study groups in order to estimate an overall MTPM for each follow-up and its associated 95% confidence interval (CI). Random effects meta-regression on study level covariates such as component fixation or mobile versus fixed bearing was employed. MTPM 1 year was used for this metaregression since it was the most reported value. All analyses were performed using Metafor Package R statistics (Viechtbauer 2010). We assessed the potential effect of publication bias by comparing the results from the meta-analysis on migration with the results from national joint registries and meta-analyses on revision rates in the discussion.

Results The literature search yielded 1,167 hits of which 214 studies remained eligible (953 studies did not study primary TKR). After more detailed evaluation of these eligible studies 10 were excluded because they did not comprise primary TKR, 117 were excluded because they did not mention migration patterns, and 34 were excluded because they were doubles. This left 53 studies to be included, comprising 111 study groups and 2,470 knees (Adalberth et al. 1999, 2000, 2001, 2002, Albrektsson et al. 1990, 1992, Carlsson et al. 2005, Catani et al. 2004, Dalen and Nilsson 2005, Ejaz et al. 2015, Fukuoka et al. 2000, Hansson et al. 2005, 2008, 2009, Henricson et al. 2006, Henricson and Nilsson 2016, Hilding and Aspenberg 2006, 2007, Hilding et al. 1995, Hyldahl et al. 2005, Kienapfel et al. 2004, Ledin et al. 2012, Meunier et al. 2009, Molt et al. 2014, 2016, Molt and Toksvig-Larsen 2014a, 2014b, 2015, Nielsen et al. 1995, Nieuwenhuijse et al. 2013, Nilsson and Dalen 1998, Nilsson and Karrholm 1993, Nilsson et al. 1991, 1999; Petersen et al. 1999, Pijls et al. 2012c, 2012d, RegnĂŠr et al. 1998, 2000, Ryd et al. 1986, 1987, 1988, 1993, 1999, Saari et al. 2006, Schotanus et al. 2016, Teeter et al. 2016, Toksvig-Larsen et al. 1994, 1998, 2000, Uvehammer et al. 2007, von Schewelov et al. 2009, Wilson et al. 2012). TKR MTPM

5/3/2018 11:44:49 AM

322

Acta Orthopaedica 2018; 89 (3): 320–328

Table 1. Number of study groups and knees for each follow-up moment Follow-up: Study groups Knees

Baseline 6 weeks 3 months 6 months 1 year 111 2,470

53 932

80 1,587

70 1,318

2 years

111 2,210

105 2,070

Table 2. Precision as determined by double examinations from the included studies of the meta-analyses Factor X translation (mm) Y translation (mm) Z translation (mm) X rotation (°) Y rotation (°) Z rotation (°)

Studies, n

Mean

Low

11 19 11 18 18 18

0.14 0.13 0.20 0.24 0.34 0.19

0.06 0.03 0.10 0.06 0.11 0.06

5 years 10 years

High 0.3 0.22 0.4 0.6 0.8 0.6

MTPM was reported in only 4 studies with the following values: 0.1 mm, 0.2 mm, 0.2 mm, and 0.45 mm.

at 1 year was the most frequently reported follow-up time, which was reported by all studies. See Table 1 for a breakdown of study groups and knees for each follow-up moment. The median percentage of female patients in each study group was 65% (36–100%). The median percentage of patients with primary osteoarthritis (OA) was 100% (0–100%). 90 study groups were restricted to primary OA. The mean age of study groups varied between 54 years and 76 years with a median of 70 years. Double examinations Table 2 shows the precision as determined by double exami-

27 520

Migration results Early migration in percentiles of the 111 study groups is depicted in Figure 1. The pooled increase in migration between 6 months and 1 year in MTPM is 0.04 mm (CI 0.02–0.07) based on 70 study groups. The pooled increase in MTPM migration between 1 year and 2 years is 0.04 mm (CI 0.02–0.06) based on 105 study groups (6 studies did not report 2-year data). To validate our assumption, the MTPM migration pattern throughout follow-up of 2 known orthopedic implant disasters is also plotted in Figure 1: Boneloc cement (MG II TKR prosthesis) and Freeman–Samuelson uncoated and uncemented TKR. 8 study groups reported MTPM migration results up to 10 years’ follow-up (Figure 2). In these longterm studies, the majority of TKR stabilized during follow-up, although 2 uncemented types of TKR continued to migrate. The pooled MTPM 1 year of cemented TKR was less than that of uncemented TKR: 0.44 mm (CI 0.38–0.50) compared with 1.09 mm (CI 0.91–1.28) (Figure 3). The migration patterns of different fixation types of uncemented TKR are shown in Figure 3. There was some variation for MTPM 1 year with the most migration for uncoated TKR, but the confidence intervals overlapped: hydroxyapatite coated 0.87 mm (CI 0.54–1.20), trabecular metal 0.84 mm (CI 0–1.92), porous coated 1.13 mm (CI 0.87–1.38), and uncoated 1.38 mm (CI 0.95–1.82). Com-

Migration (mm)

MTPM (mm)

3.0

2.5

2.0

1.5

1.0

0.5

0.0 0.0

0.0 0.5

1.0

1.5

2.0

Years after index operation

Figure 1. Early migration in percentiles of 111 study groups, 2,470 knees. The migration of two known disasters is also plotted: Boneloc cement (MG II prosthesis) and Freeman–Samuelson all-poly uncoated and uncemented.

11469 Pijls D.indd 322

8 86

nations from the included studies. 19 of 53 included studies reported original double examinations. The studies that did not report original double examinations often referred to previous studies that reported double examinations or they used phantom experiments.

Years after index operation

Figure 2. Long-term migration of 8 study groups. Black for cemented TKR. Blue for cementless TKR. CR = cruciate retaining. PS = posterior stabilized.

5/3/2018 11:44:49 AM

Acta Orthopaedica 2018; 89 (3): 320–328

323

MTPM (mm)

3.0

2.5

2.0

1.5

1.0

0.5

0.0

Years after index operation

MTPM (mm)

3.0

2.5

2.0

1.5

1.0

0.5

0.0

Years after index operation

Figure 3. Migration patterns for cemented and cementless TKR, types of cementless TKR and migration according to decade in which the inclusion of the study started for cemented and cementless TKR separately. The number of RSA examinations is given for each follow-up in color and order corresponding to the legend.

paring TKR migration data throughout the last 35 years, for cemented TKR migration decreased during the last decades, with MTPM at 1 year in the 1980s being 0.70 mm (CI 0.44– 0.97), in the 1990s 0.48 mm (CI 0.37–0.59), and in the 2000s 0.41 mm (CI 0.33–0.49) (Figure 3). For the uncemented TKR the migration increased during these 35 years, MTPM at 1 year in the 1980s being 1.04 mm (CI 0.72–1.35), in the 1990s 1.06 mm (CI 0.79–1.33), and in the 2000s 1.33 mm (CI 0.81– 1.86) (Figure 3). There seemed to be no difference in mean migration between cemented all-poly and cemented metalbacked TKR with MTPM 1 year 0.36 mm (CI 0.22–0.49) and 0.46 mm (CI 0.39–0.53) (Figure 4). There seemed to be no difference in migration between cemented mobile bearing and cemented fixed bearing TKR, MTPM 1 year being 0.53 mm (CI 0.17–0.89) and 0.44 mm (CI 0.37–0.50) (Figure 4). There seemed to be no difference in migration between cemented cruciate retaining and cemented posterior stabilized TKR with MTPM 1 year 0.52 mm (CI 0.42–0.62) and 0.45 mm (CI 0.29–0.60) (Figure 4). For cemented TKR there seemed to be no difference in migration between RSA techniques used. Pooled MTPM 1 year was 0.56 mm (CI 0.29–0.83) for model-

11469 Pijls D.indd 323

based RSA (MBRSA), 0.43 mm (CI 0.34–0.51) for markerbased RSA with markers in the modular polyethylene (PE) insert, 0.43 mm (CI 0.27–0.60) for marker-based RSA with markers attached to the tibial baseplate or tibial stem and 0.45 mm (CI 0.31–0.60) for model-based RSA with markers in the PE of all-poly TKR or markers in the PE of non-modular TKR (Figure 4).

Discussion The results from this systematic review and meta-analysis of 2,470 TKRs show that the majority of early migration occurs in the first 6 postoperative months followed by a period of no or very little migration within the bone (i.e., plateau phase). From 6 months to 1 year there was on average 0.04 mm migration, which was similar to the migration between 1 year and 2 years of 0.04 mm. Given this similarity between MTPM 6 months and MTPM 1 year, we propose using the MTPM 6 months values (instead of MTPM 1 year values) for RSA threshold testing (Pijls et

5/3/2018 11:44:52 AM

324

Acta Orthopaedica 2018; 89 (3): 320–328

MTPM (mm)

3.0

2.5

2.0

1.5

1.0

0.5

0.0

Years after index operation

MTPM (mm)

3.0

2.5

2.0

1.5

1.0

0.5

0.0

Years after index operation

Figure 4. Migration patterns for cemented TKR. Comparisons of all poly versus metal backed, mobile bearing versus fixed bearing, cruciate retaining versus posterior stabilized and RSA techniques is made. The number of RSA examinations is given for each follow-up in color and order corresponding to the legend. MBRSA = model-based RSA. For a detailed description on RSA techniques see the text.

Migration Early migration

Stabilization 1

Stabilization 2

MTPM at 6 months a MTPM 6–12 months

< 0.2 mm stable > 0.2 mm unstable

< 0.5 mm acceptable 0.5–1.6 mm at risk > 0.5 mm unacceptable

MTPM at 12–24 months b

< 0.2 mm stable > 0.2 mm unstable

Months after index operation Figure 5. Proposed RSA migration evaluation for 2 years’ follow-up based on the results of the meta-analyses. Early migration is evaluated by MTPM at 6 months, stabilization 1 is evaluated by MTPM between 6 and 12 months, and stabilization 2 is evaluated by MTPM between 12 and 24 months. a modified according to Pijls et al. 2012a. b according to Ryd et al. 1995.

al. 2012a) (Figure 5). Since the TKR migrated only 0.04 mm between 6 months and 1 year, the MTPM 6 months is almost

11469 Pijls D.indd 324

identical to the MTPM 1 year (MTPM 6 months = MTPM 1 year—0.04 mm), which is currently used as a threshold for RSA studies (Pijls et al. 2012a). Therefore, the MTPM 6 months values can be subjected to the MTPM 1 year thresholds of less than 0.5 mm for acceptable migration and more than 1.6 mm for unacceptable migration. The migration between 6 months and 12 months can then be used to evaluate the presence of migration stabilization for the “at risk” and “unacceptable” groups, using the 0.2 mm values of Ryd et al. (1995). Nevertheless, the migration between 1 and 2 years can be used as an extra safeguard during the phased introduction of TKR implants. It is also important to have migration results before 6 months (e.g., 6 weeks and 3 months) to evaluate the migration pattern. Cemented TKR have very little early migration (MTPM 1 year 0.44 mm) most likely because they rely on primary bone fixation through cement interdigitation. The migration of Boneloc cement, a historical disaster, is very high with MTPM 1 year 1.25 mm compared with the mean 0.44 mm for all cemented TKR. Furthermore, TKR with Boneloc did not stabilize between 1 and 2 years. In retrospect, with the data

5/3/2018 11:44:54 AM

Acta Orthopaedica 2018; 89 (3): 320–328

from this meta-analysis it would already have been possible to identify Boneloc bone cement as a potential disaster at 6 months’ RSA follow-up and using sequential RSA migration follow-up data to evaluate whether a migration stabilization phase occurs. For cemented TKR there seemed to be no difference in migration between all-poly and metal-backed TKR for the first 6 postoperative months or for the period beyond this time point up to 2 years postoperatively. These data on the low migration of the all-poly tibial component are in line with Nouta et al. (2012) who found no difference in revision for aseptic loosening between all-poly and metal-backed TKR at long-term follow-up. For cemented TKR there seemed to be no difference in migration between mobile-bearing TKR and fixed-bearing TKR up to 5 years’ follow-up. These results are in line with a recent Cochrane review showing no differences in revision rates between fixed- and mobile-bearing TKR (Hofstede et al. 2015). Apparently, the theoretical advantage of mobile-bearing TKR in reducing the stress on the bone– cement–prosthesis interface does not translate into less migration of the tibial component or fewer revisions for aseptic loosening of cemented TKR. For cemented TKR there seemed to be no difference in migration between cruciate retaining and posterior stabilized TKR up to 5 years’ follow-up. These results are in line with a recent Cochrane review showing no differences in revision rates between cruciate retaining and posterior stabilized TKR (Verra et al. 2013). As for migration measurements of orthopedic implants within bone, 2 RSA measurement techniques exist, a manual, marker-based RSA method and a semi-automatic CAD model-based RSA (MBRSA) technique (Valstar et al. 2000, Hurschler et al. 2009). There have been concerns that modelbased RSA (MBRSA) might overestimate migration results compared with marker-based methods for the same analysis because MBRSA uses CAD models with thousands of points (triangles) compared with 3–5 tantalum markers for marker-based methods (Tjornild et al. 2015). The marker or point within a segment that has moved the most provides the MTPM. Our results show that, for the included TKR, there seemed to be no difference in migration measured with MBRSA or 3 types of marker-based methods (e.g., markers in tibial insert, attached to the tibial base plate). Between the 3 marker-based RSA methods there was no difference in migration for cemented TKR. We evaluated marker-based RSA that uses markers in the PE insert of modular TKR, marker-based RSA that uses markers attached to the tibial baseplate, and marker-based RSA that uses markers in the PE insert of nonmodular TKR or all-poly TKR. This is particularly important for marker-based RSA using markers in modular tibial inserts, since movement between the insert and tibial baseplate could appear as migration of the TKR relative to the bone when this is actually migration of the insert within the tibial base plate and thus backside wear and creep of the poly insert. This could theoretically give higher migration values for TKR with mark-

11469 Pijls D.indd 325

325

ers in modular inserts, whilst it is actually backside wear. Our results show that up to 5 years’ migration results, this effect is negligible and thus migration results are actually the migration of the tibial component within the bone. Uncemented TKR have high early migration (MTPM 1 year 1.09 mm) followed by stabilization of this migration. The latter is most likely due to the secondary fixation by bone in/ongrowth of these uncemented components. The migration of the uncemented uncoated FS AP, a historical disaster with early loosening, is very high and does not stabilize at all during follow-up, either at the 6 months’ mark or at the 2-year mark. It had one of the worst ever revision rates recorded in national registries (Knutson et al. 1986). In retrospect with the data from the current and previous meta-analysis it would already have been possible to identify the uncemented FS as a potential disaster at 6 months’ RSA follow-up using sequential RSA follow-up to evaluate whether stabilization of this progressive migration occurs. For uncemented TKR there was some variation in migration with the highest migration for the uncoated TKR, although the 95% confidence intervals overlapped. During the last 35 years the migration of uncemented TKR seemed to increase. A possible explanation for this phenomenon is the fact that 5 out of 8 PFIs from the 2000s are porous coated. Regarding precision of the RSA measurements, only 19 of 53 included studies reported precision as determined by original double examinations. It is important to know the precision of RSA measurements. Therefore future RSA studies should perform double examinations as is also required by the RSA guidelines (Valstar et al. 2005). A limitation of this meta-analysis may be that for the direct comparisons of different types of TKR (e.g., mobile versus fixed bearing) pooled results of different trials might have been more appropriate. However, continuous migration data or absence of a migration stabilization phase as shown in this meta-analysis are in line with the revision rates for these TKR types from meta-analyses and the survival analysis results of national joint registries. We focused on MTPM of the TKR, although translations and rotations migration measurements could have been of interest as shown by Gudnason et al. (2017). However, since translations and rotations were not reported in a uniform manner, a meaningful analysis is not possible. Future studies could benefit from further standardization regarding reporting of the results. The disadvantage of MTPM is that it gives information only on the magnitude of the migration, not on the direction of the migration. The latter could also be of interest and may even be better than MTPM (Gudnason et al. 2017). Nevertheless, MTPM does have advantages. It evaluates the fundamental assumption of loosening: is the prosthesis loose or not? Also it is the most reported parameter, going back to the beginning of RSA, and the present meta-analysis includes 2,470 knees with MTPM. In summary the results from this meta-analysis on RSA migration are in line with the results of national implant

5/3/2018 11:44:59 AM

326

registries as well as the results of meta-analyses on revision rates, providing further proof for the association between early implant migration and late revision for aseptic loosening of TKR. The pooled migration patterns can be used both as benchmarks as well as for defining migration thresholds for future evaluation of new TKR and fixations. Thus RSA has a place for a safe premarket approval evaluation and as such should be part of a phased introduction of new implants (Nelissen et al. 2011). With the data from this meta-analysis it appears possible (Figure 5) to have a first evaluation of the safety (i.e. implant–bone fixation) of the implant at 6 months.

BGP and RGN conceived the study. JWP designed the search strategy for the literature search. BGP performed the study selection, data extraction, and analyses. RGN ensure accuracy of data extraction, was the referee and helped with interpretation of the results. BGP, JWP, and RGN wrote the manuscript.

Data extraction of the RSA studies and results from the meta-analyses are available by contacting the corresponding author.

No conflicts of interest declared. No external funding.

Acta thanks Mogens Berg Laursen and Stephan Röhrl for help with peer review of this study.

Adalberth G, Nilsson K G, Bystrom S, Kolstad K, Mallmin H, Milbrink J. Stability assessment of a moderately conforming all-polyethylene tibial component in total knee arthroplasty: a prospective RSA study with. 2 years of follow-up of the Kinemax Plus design. Am J Knee Surg 1999; 12(4): 233-40. Adalberth G, Nilsson K G, Bystrom S, Kolstad K, Milbrink J. Low-conforming all-polyethylene tibial component not inferior to metal-backed component in cemented total knee arthroplasty: prospective randomized radiostereometric analysis study of the AGC total knee prosthesis. J Arthroplasty 2000; 15(6): 783-92. Adalberth G, Nilsson K G, Bystrom S, Kolstad K, Milbrink J. All-polyethylene versus metal-backed and stemmed tibial components in cemented total knee arthroplasty: a prospective randomised RSA study. J Bone Joint Surg Br 2001; 83(6): 825-31. Adalberth G, Nilsson K G, Karrholm J, Hassander H. Fixation of the tibial component using CMW-1 or Palacos bone cement with gentamicin: similar outcome in a randomized radiostereometric study of 51 total knee arthroplasties Acta Orthop Scand 2002; 73(5): 531-8. Albrektsson B E, Ryd L, Carlsson L V, Freeman M A, Herberts P, Regner L, Selvik G. The effect of a stem on the tibial component of knee arthroplasty: a roentgen stereophotogrammetric study of uncemented tibial components in the Freeman-Samuelson knee arthroplasty. J Bone Joint Surg Br 1990; 72(2): 252-8. Albrektsson B E, Carlsson L V, Freeman M A, Herberts P, Ryd L. Proximally cemented versus uncemented Freeman-Samuelson knee arthroplasty: a prospective randomised study. J Bone Joint Surg Br 1992; 74(2): 233-8. AOANJRR. Australian Orthopaedic Association National Joint Replacement Registry: Annual Report 2016. Carlsson A, Bjorkman A, Besjakov J, Onsten I. Cemented tibial component fixation performs better than cementless fixation: a randomized radiostereometric study comparing porous-coated hydroxyapatite-coated and cemented tibial components over 5 years Acta Orthop 2005; 76(3): 362-9 .

11469 Pijls D.indd 326

Acta Orthopaedica 2018; 89 (3): 320–328

Catani F, Leardini A, Ensini A, Cucca G, Bragonzoni L, Toksvig-Larsen S, Giannini S. The stability of the cemented tibial component of total knee arthroplasty: posterior cruciate-retaining versus posterior-stabilized design. J Arthroplasty 2004; 19(6): 775-82. Dalen T, Nilsson K G. VersaBond bone cement prospective randomized study of the clinical properties of a new bone cement in total knee replacement Knee 2005; 12(4): 311-17. Ejaz A, Laursen A C, Jakobsen T, Rasmussen S, Nielsen P T, Laursen M B. Absence of a tourniquet does not affect fixation of cemented TKA: A randomized RSA study of 70 patients. J Arthroplasty 2015; 30(12): 2128-32. Fukuoka S, Yoshida K, Yamano Y. Estimation of the migration of tibial components in total knee arthroplasty. J Bone Joint Surg Br 2000; 82(2): 222-7. Gudnason A, Adalberth G, Nilsson K G, Hailer NP. Tibial component rotation around the transverse axis measured by radiostereometry predicts aseptic loosening better than maximal total point motion. Acta Orthop 2017; 88(3): 282-7. Hansson U, Toksvig-Larsen S, Jorn L P, Ryd L. Mobile vs fixed meniscal bearing in total knee replacement: a randomised radiostereometric study. Knee 2005; 12(6): 414-18. Hansson U, Ryd L, Toksvig-Larsen S A. randomised RSA study of peri-apatite HA coating of a total knee prosthesis. Knee 2008; 15(3): 211-16. Hansson U, Toksvig-Larsen S, Ryd L, Aspenberg P. Once-weekly oral medication with alendronate does not prevent migration of knee prostheses: a double-blind randomized RSA study. Acta Orthop 2009; 80(1): 41-5. Henricson A, Nilsson K G. Trabecular metal tibial knee component still stable at 10 years. Acta Orthop 2016; 87(5): 504-10. Henricson A, Dalen T, Nilsson K G. Mobile bearings do not improve fixation in cemented total knee arthroplasty. Clin Orthop Relat Res 2006; 448: 114-21. Hilding M, Aspenberg P. Postoperative clodronate decreases prosthetic migration: 4-year follow-up of a randomized radiostereometric study of 50 total knee patients. Acta Orthop 2006; 77(6): 912-16. Hilding M, Aspenberg P. Local peroperative treatment with a bisphosphonate improves the fixation of total knee prostheses: a randomized double-blind radiostereometric study of 50 patients. Acta Orthop 2007; 78(6): 795-9. Hilding M B, Yuan X, Ryd L. The stability of three different cementless tibial components: a randomized radiostereometric study in 45 knee arthroplasty patients. Acta Orthop Scand 1995; 66(1): 21-7. Hofstede S N, Nouta K A, Jacobs W, van Hooff M L, Wymenga A B, Pijls B G, Nelissen R G, Marang-van de Mheen P J. Mobile bearing vs fixed bearing prostheses for posterior cruciate retaining total knee arthroplasty for postoperative functional status in patients with osteoarthritis and rheumatoid arthritis Cochrane Database Syst Rev 2015; 4(2): CD003130. Hozo S P, Djulbegovic B, Hozo I. Estimating the mean and variance from the median range and the size of a sample. BMC Med Res Methodol 2005; 5(13): 13. Hurschler C, Seehaus F, Emmerich J, Kaptein B L, Windhagen H. Comparison of the model-based and marker-based roentgen stereophotogrammetry methods in a typical clinical setting. J Arthroplasty 2009; 24(4): 594-606. Hyldahl H, Regner L, Carlsson L, Karrholm J, Weidenhielm L. All-polyethylene vs metal-backed tibial component in total knee arthroplasty: a randomized RSA study comparing early fixation of horizontally and completely cemented tibial components, Part 1: Horizontally cemented components: AP better fixated than MB. Acta Orthop 2005; 76(6): 769-77. Kienapfel H, Hildebrand R, Neumann T, Specht R, Koller M, Celik I, Mueller H H, Griss P, Klose K J, Georg C. The effect of Palamed G bone cement on early migration of tibial components in total knee arthroplasty. Inflamm Res 2004; 53(Suppl. 2): S159-S163. Knutson K, Lindstrand A, Lidgren L. Survival of knee arthroplasties: a nation-wide multicentre investigation of 8000 cases. J Bone Joint Surg Br 1986; 68(5): 795-803. Ledin H, Aspenberg P, Good L. Tourniquet use in total knee replacement does not improve fixation but appears to reduce final range of motion Acta Orthop 2012; 83(5): 499-503.

5/3/2018 11:44:59 AM

Acta Orthopaedica 2018; 89 (3): 320–328

Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med 2009; 6(7): e1000100. Meunier A P, Aspenberg P, Good L. Celecoxib does not appear to affect prosthesis fixation in total knee replacement: a randomized study using radiostereometry in 50 patients. Acta Orthop 2009; 80(1): 46-50 . Molt M, Toksvig-Larsen S. Similar early migration when comparing CR and PS in Triathlon TKA: A prospective randomised RSA trial. Knee 2014a; 21(5): 949-54. Molt M, Toksvig-Larsen S. Peri-apatite enhances prosthetic fixation in TKA: a prospective randomised RSA study. J Arthritis 2014b; 3(3): 1-6. Molt M, Toksvig-Larsen S. 2-year follow-up report on micromotion of a short tibia stem: a prospective randomized RSA study of 59 patients. Acta Orthop 2015; 86(5): 594-8. Molt M, Harsten A, Toksvig-Larsen S. The effect of tourniquet use on fixation quality in cemented total knee arthroplasty: a prospective randomized clinical controlled RSA trial. Knee 2014; 21(2): 396-401. Molt M, Ryd L, Toksvig-Larsen S. A randomized RSA study concentrating especially on continuous migration. Acta Orthop 2016; 87(3): 262-7. Nelissen R G, Pijls B G, Karrholm J, Malchau H, Nieuwenhuijse M J, Valstar E R. RSA and registries: the quest for phased introduction of new implants. J Bone Joint Surg Am 2011; 93(Suppl. 3): 62-5. Nielsen P T, Berg-Hansen E, Toksvig L S, Ryd L, Rechnagel K, Schroder H, Selvik G. AGC 2000 tibial components with hydroxyapatite coating: a randomized series followed 5 years with RSA. Acta Orthop 1995; (Suppl. 265): 77. Nieuwenhuijse M J, van der Voort P, Kaptein B L, van der Linden-van der Zwaag H M, Valstar E R, Nelissen R G. Fixation of high-flexion total knee prostheses: five-year follow-up results of a four-arm randomized controlled clinical and roentgen stereophotogrammetric analysis study. J Bone Joint Surg Am 2013; 95(19): e1411-11. Nilsson K G, Dalen T. Inferior performance of Boneloc bone cement in total knee arthroplasty: a prospective randomized study comparing Boneloc with Palacos using radiostereometry (RSA) in 19 patients. Acta Orthop Scand 1998; 69(5): 479-83. Nilsson K G, Karrholm J. Increased varus-valgus tilting of screw-fixated knee prostheses: stereoradiographic study of uncemented versus cemented tibial components. J Arthroplasty 1993; 8(5): 529-40. Nilsson K G, Karrholm J, Ekelund L, Magnusson P. Evaluation of micromotion in cemented vs uncemented knee arthroplasty in osteoarthrosis and rheumatoid arthritis: randomized study using roentgen stereophotogrammetric analysis. J Arthroplasty 1991; 6(3): 265-78. Nilsson K G, Karrholm J, Carlsson L, Dalen T. Hydroxyapatite coating versus cemented fixation of the tibial component in total knee arthroplasty: prospective randomized comparison of hydroxyapatite-coated and cemented tibial components with 5-year follow-up using radiostereometry. J Arthroplasty 1999; 14(1): 9-20. Nouta K A, Verra W C, Pijls B G, Schoones J W, Nelissen R G. All-polyethylene tibial components are equal to metal-backed components: systematic review and meta-regression. Clin Orthop Relat Res 2012; 470(12): 3549-59. Petersen M M, Nielsen P, Lebech A, Toksvig-Larsen S, Lund B. Preoperative bone mineral density of the proximal tibia and migration of the tibial component after uncemented total knee arthroplasty. J Arthroplasty 1999; 14(1): 77-81. Pijls B G, Nelissen R. Strategy for RSA migration thresholds. Acta Orthop 2016; 87(4): 432-3. Pijls B G, Valstar E R, Nouta K A, Plevier J W, Fiocco M, Middeldorp S, Nelissen R G. Early migration of tibial components is associated with late revision: a systematic review and meta-analysis of 21000 knee arthroplasties. Acta Orthop 2012a; 83(6): 614-24. Pijls B G, Nieuwenhuijse M J, Fiocco M, Plevier J W, Middeldorp S, Nelissen R G, Valstar E R. Early proximal migration of cups is associated with late revision in THA: a systematic review and meta-analysis of. 26 RSA studies and 49 survival studies. Acta Orthop 2012b; 83(6): 583-91.

11469 Pijls D.indd 327

327

Pijls B G, Valstar E R, Kaptein B L, Fiocco M, Nelissen R G. The beneficial effect of hydroxyapatite lasts: a randomized radiostereometric trial comparing hydroxyapatite-coated uncoated and cemented tibial components for up to 16 years. Acta Orthop 2012c; 83(2): 135-41. Pijls B G, Valstar E R, Kaptein B L, Nelissen R G. Differences in long-term fixation between mobile-bearing and fixed-bearing knee prostheses at ten to 12 years’ follow-up: a single-blinded randomised controlled radiostereometric trial. J Bone Joint Surg Br 2012d; 94(10): 1366-71. Regnér L, Carlsson L, Kärrholm J, Herberts P. Ceramic coating improves tibial component fixation in total knee arthroplasty. J Arthroplasty 1998; 13(8): 882-9. Regnér L, Carlsson L, Kärrholm J, Herberts P. Tibial component fixation in porous- and hydroxyapatite-coated total knee arthroplasty: a radiostereometric evaluation of migration and inducible displacement after 5 years. J Arthroplasty 2000; 15(6): 681-9. Ryd L, Lindstrand A, Rosenquist R, Selvik G. Tibial component fixation in knee arthroplasty. Clin Orthop Relat Res 1986; (213): 141-9. Ryd L, Lindstrand A, Rosenquist R, Selvik G. Micromotion of conventionally cemented all-polyethylene tibial components in total knee replacements: a roentgen stereophotogrammetric analysis of migration and inducible displacement. Arch Orthop Trauma Surg 1987; 106(2): 82-8. Ryd L, Albrektsson B E, Herberts P, Lindstrand A, Selvik G. Micromotion of noncemented Freeman-Samuelson knee prostheses in gonarthrosis: a roentgen-stereophotogrammetric analysis of eight successful cases. Clin Orthop Relat Res 1988; (229): 205-12. Ryd L, Carlsson L, Herberts P. Micromotion of a noncemented tibial component with screw fixation: an in vivo roentgen stereophotogrammetric study of the Miller-Galante prosthesis. Clin Orthop Relat Res 1993; (295): 21825. Ryd L, Albrektsson B E, Carlsson L, Dansgard F, Herberts P, Lindstrand A, Regner L, Toksvig-Larsen S. Roentgen stereophotogrammetric analysis as a predictor of mechanical loosening of knee prostheses. J Bone Joint Surg Br 1995; 77(3): 377-83. Ryd L, Hansson U, Blunn G, Lindstrand A, Toksvig-Larsen S. Failure of partial cementation to achieve implant stability and bone ingrowth: a longterm roentgen stereophotogrammetric study of tibial components. J Orthop Res 1999; 17(3): 311-20. Saari T, Uvehammer J, Carlsson L V, Regner L, Karrholm J. Influence of polyethylene constraint on tibial component fixation in total knee arthroplasty: follow-up report after 5 years. J Arthroplasty 2006; 21(7): 1032-7. Schotanus M G, Pilot P, Kaptein B L, Draijer W F, Tilman P B, Vos R, Kort N P. No difference in terms of radiostereometric analysis between fixedand mobile-bearing total knee arthroplasty: a randomized single-blind controlled trial. Knee Surg Sports Traumatol Arthrosc 2016; 25(9): 2978-85. SKAR. Swedish Knee Arthroplasty Register: Annual Report 2016. Teeter M G, Thoren J, Yuan X, McCalden R W, MacDonald S J, Lanting B A, Naudie D D. Migration of a cemented fixed-bearing polished titanium tibial baseplate (Genesis II) at ten years: a radiostereometric analysis. Bone Joint J 2016; 98-b(5): 616-21. Tjornild M, Soballe K, Hansen PM, Holm C, Stilling M. Mobile- vs fixedbearing total knee replacement. Acta Orthop 2015; 86(2): 208-14. Toksvig-Larsen S, Ryd L, Lindstrand A. Effect of a cooled saw blade on prosthesis fixation: randomized radiostereometry of 33 knee cases. Acta Orthop Scand 1994; 65(5): 533-7. Toksvig-Larsen S, Ryd L, Lindstrand A. Early inducible displacement of tibial components in total knee prostheses inserted with and without cement: a randomized study with roentgen stereophotogrammetric analysis. J Bone Joint Surg Am 1998; 80(1): 83-9. Toksvig-Larsen S, Jorn L P, Ryd L, Lindstrand A. Hydroxyapatite-enhanced tibial prosthetic fixation. Clin Orthop Relat Res 2000; (370): 192-200. Uvehammer J, Karrholm J, Carlsson L. Influence of joint area design on tibial component migration: comparison among a fixed symmetrical asymmetrical and moveable bearing. J Knee Surg 2007; 20(1): 20-6. Valstar E R, Vrooman HA, Toksvig-Larsen S, Ryd L, Nelissen R G. Digital automated RSA compared to manually operated RSA. J Biomech 2000; 33(12): 1593-9.

5/3/2018 11:44:59 AM

328

Valstar E R, Gill R, Ryd L, Flivik G, Borlin N, Karrholm J. Guidelines for standardization of radiostereometry (RSA) of implants. Acta Orthop 2005; 76(4): 563-72. van der Voort P, Pijls B G, Nieuwenhuijse M J, Jasper J, Fiocco M, Plevier J W, Middeldorp S, Valstar E R, Nelissen R G. Early subsidence of shapeclosed hip arthroplasty stems is associated with late revision: a systematic review and meta-analysis of 24 RSA studies and 56 survival studies. Acta Orthop 2015; 86(5): 575-85. Verra W C, van den Boom L G, Jacobs W, Clement D J, Wymenga A A, Nelissen R G. Retention versus sacrifice of the posterior cruciate ligament in total knee arthroplasty for treating osteoarthritis. Cochrane Database Syst Rev 2013; 11(10): CD004803.

11469 Pijls D.indd 328

Acta Orthopaedica 2018; 89 (3): 320â&#x20AC;&#x201C;328

Viechtbauer W. Conducting meta-analyses in R with the metafor package. J Statistical Software 2010; 36(3): 1-48. Vochteloo A J, Pijls B G, van der Heide H J. Sutures v staples: letâ&#x20AC;&#x2122;s add three other studies. BMJ 2010; 340(340): c2627. von Schewelov T, Besjakov J, Sanzen L, Carlsson A. A clinical and radiostereometric study of the cemented PFC-sigma prosthesis: a 5-year study of 29 cases with a fixed bearing. J Knee Surg 2009; 22(3): 231-6. Wilson D A, Richardson G, Hennigar A W, Dunbar M J. Continued stabilization of trabecular metal tibial monoblock total knee arthroplasty components at 5 years measured with radiostereometric analysis. Acta Orthop 2012; 83(1): 36-40.

5/3/2018 11:44:59 AM

Acta Orthopaedica 2018; 89 (3): 329–335

329

The effect of surgeon’s preference for hybrid or cemented fixation on the long-term survivorship of total knee replacement An analysis of 39,623 prostheses from the Australian Orthopaedic Association National Joint Replacement Registry Christopher J VERTULLO 1,2, Stephen E GRAVES 3, Yi PENG 3, and Peter L LEWIS 3

1 Knee

Research Australia, Gold Coast, Australia; 2 Gold Coast Orthopaedic Research and Educational Alliance, Menzies Health Institute, Griffith University, Gold Coast; Australia; 3 Australian Orthopaedic Association National Joint Replacement Registry, SAHMRI, Adelaide, Australia Correspondence: chris.vertullo@icloud.com Submitted 2017-09-24. Accepted 2017-11-27.

Background and purpose — Recent direct comparative reports suggest that hybrid fixation may have a similar or superior outcome to cemented fixation in total knee replacement (TKR); however, a paucity of long-term data exists. To minimize the confounders of a direct comparison, we performed an instrumental variable analysis examining the revision rate of 2 cohorts of patients based on their surgeon’s preference for cemented or hybrid fixation. Methods — Registry data were obtained from 1999 until 2015 for 2 cohorts of patients who received minimally stabilized TKR, defined as those treated by high-volume hybrid fixation preferring surgeons, designated routinely hybrid (RH), and those treated by high-volume cemented fixation preferring surgeons, designated routinely cemented (RC). Results — At 13 years, the cumulative percentage revision of the RC cohort was 4.8% (CI 4.1–5.7) compared with 5.5% (CI 3.5–8.7) for the RH cohort. The revision risk for each cohort was the same for all causes (HR = 1.0 (CI (0.84–1.20)), non-infective causes, and for infection. This finding was irrespective of patient age or sex, patella resurfacing, and with non-cross-linked polyethylene (NXLPE). The RH cohort who received cross-linked polyethylene (XLPE) had a lower revision risk than the RC cohort with XLPE (HR = 0.57 (0.37–0.88), p = 0.01). Interpretation — The risk of revision for the patients of surgeons who prefer cemented fixation in minimally stabilized TKR is the same as for the patients of surgeons who prefer hybrid fixation, except when used with XLPE, where hybrid fixation has a lower revision risk. ■

The optimum fixation in total knee replacement (TKR) is controversial, with cemented fixation remaining the most common method internationally (National Joint Registry, AOA National Joint Registry 2016), compared with hybrid fixation (cemented tibia and cementless femur) or cementless fixation of both components. Hybrid fixation was introduced to overcome the perceived concerns over cementless fixation of the tibia while attempting to minimize femoral bone loss, decrease operative time, and reduce the polymethylmethacrylate burden of the joint (Wright et al. 1990, Kraay et al. 1991, Faris et al. 2008). While recent reports suggest that hybrid fixation may have a similar or superior outcome to cemented fixation (Petursson et al. 2015), a paucity of long-term data exists concerning this method of fixation in TKR (Nakama et al. 2012). While cemented fixation of both components has excellent longterm survivorship in national registries (National Joint Registry, AOA National Joint Registry 2016), in the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR), hybrid fixation has the lowest revision risk overall when compared with cemented and cementless TKR. However, this revision risk is altered when prosthesis stability is considered. In posterior stabilized (PS) TKR, cemented fixation has the lowest revision risk. Conversely, in minimally stabilized (MS) TKR there is no difference between hybrid fixation and cemented fixation, and both have a lower revision risk compared with cementless fixation. The reasons why hybrid fixation has a lower risk for all prosthesis types, but not when PS or MS TKR are considered individually, are uncertain. Hybrid fixation may not be appropriate for all patients, particularly in patients with osteoporosis, osteonecrosis,

12130 Vertullo D.indd 329

5/3/2018 11:46:15 AM

330

complex deformity, rheumatoid arthritis, or inaccurate bone resection (Scott 2012). These factors may bias registry data against cemented fixation when directly compared. Conversely, hybrid fixation may be used more commonly in younger active patients, which may bias registry data against hybrid fixation. Previous registry studies have performed direct comparison of hybrid versus cemented TKR survivorship rates. In contradistinction, we performed an instrumental variable analysis based on surgeon preference for different prosthesis fixation options rather than the actual prosthesis received. This technique compares the revision rate of all primary minimally stabilized TKR undertaken by high-volume surgeons who preferred hybrid fixation TKR to those undertaken by highvolume surgeons who preferred cemented TKR. The rationale for this instrumental variable approach is that it has the capacity to remove the confounding by indication or disease severity between hybrid and cemented fixation that is not possible by directly comparing hybrid and cemented TKR implant registry revision rates (Vertullo et al. 2017). Our primary hypothesis was that there would be no difference in the revision rate when the 2 patient cohorts were compared. Our secondary hypothesis was that there would be no difference in the revision rate with sub-analysis based on age, sex, type of polyethylene, and patella resurfacing.

Methods Study design 2 groups of surgeons, who performed more than 50 TKR per year, differing by their femoral fixation preference were selected to perform an instrumental variable survivorship analysis (Newhouse and McClellan 1998; Stukel et al. 2007 references missing) with surgeon preference serving as the instrument, using data from the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR). A revealed preference for a femoral fixation option was defined as choosing to utilize it greater than 90% of the time, based on prior studies relating to knee implant choice by surgeons (Vertullo et al. 2017). Hence, a hybrid preferring (HP) surgeon used hybrid fixation at least 90% of the time, and the patient cohort treated by those surgeons has been termed routinely hybrid (RH). A cemented preferring (CP) surgeon used cemented fixation at least 90% of the time and the patient cohort treated by those surgeons has been termed routinely cemented (RC). This study included all MS primary TKR undertaken for osteoarthritis (OA) with fixed cemented tibial components and cemented or cementless femoral components, undertaken by the 2 groups of surgeons and reported to the registry, irrespective of patella resurfacing. PS TKR, mobile bearing TKR, cementless (cementless femur and tibia), reverse hybrid (cemented femur and cementless tibia) TKR, and TKR

12130 Vertullo D.indd 330

Acta Orthopaedica 2018; 89 (3): 329–335

with a higher than anticipated risk of revision in the AOANJRR were excluded, as were non-osteoarthritic patients such as those with rheumatoid arthritis, or osteonecrosis (AOA National Joint Registry 2016). Data for the 2 patient cohorts and their treating surgeons were obtained from the AOANJRR from September 1, 1999, until December 31, 2015. The AOANJRR commenced data collection in 1999 and includes data on more than 98% of arthroplasty procedures performed nationally since 2002 (AOA National Joint Registry 2016). The AOANJRR collects information on prosthesis type by catalogue and lot number, as well as cement used for each component by catalogue and lot number. Intended component fixation method is confirmed by linking component data to an internally developed comprehensive international prostheses library, validated with both manufacturers and other registries. If the actual component fixation method is not recorded at time of surgery (approximately less than 1% of TKR), the absent information is then obtained from the hospital. This linking of actual and intended fixation ensures almost complete accuracy in determining and verifying the fixation used in every procedure. The AOANJRR defines MS prostheses as those that have a flat or dished tibial articulation regardless of congruency, hence this group includes cruciate retaining and ultracongruent polyethylene options. PS prostheses provide additional posterior stability, most commonly using a peg and box design. Cross-linked-polyethylene (XLPE) was defined as ultra-highmolecular-weight polyethylene that has been irradiated with high-dose (≥ 50 kGy) radiation, regardless of re-melting or annealing (de Steiger et al. 2015). Time to first revision was the principal outcome measure, with revision being defined as any procedure that involves the insertion, removal, and/or replacement of a prosthesis. Reasons for revision and the type of revision were reported for procedures undertaken by both groups of surgeons. Further analyses based on patient’s age, sex, patella resurfacing, and the type of polyethylene were also undertaken. Analysis of surgeon practice public/private mix, years of contribution to registry, number of TKR in registry, and hospital arthroplasty volume was also undertaken. Statistics Kaplan–Meier estimates of survivorship were used to estimate the time to the first revision, with right censoring for death or closure of the database at the time of analysis. The unadjusted cumulative percentage revision (CPR) of the primary arthroplasty, along with 95% confidence intervals (CI), was calculated using unadjusted point-wise Greenwood estimates. Hazard ratios (HR), calculated using Cox proportional hazard models and adjusted for age and sex, were used to make statistical comparisons of the rate of revision between the 2 cohorts. All tests were 2-tailed at the 5% level of significance. The analysis was performed using SAS version 9.3 (SAS Institute Inc., Cary, NC, USA).

5/3/2018 11:46:15 AM

Acta Orthopaedica 2018; 89 (3): 329–335

331

Table 3. Demographic characteristics of the study cohorts

Total knee replacements in AOANJRR 1999–2015 n = 494,571 Excluded procedures (n = 454,948): – without diagnosis OA, 12,198 – not minimally stabilised, 134,357 – not with fixed bearing surface, 83,405 – with cementless tibial component, 58,941 – with unknown surgeon, 65,460 – by surgeons with < 50 procedures per year, 51,018 – with prosthesis with higher than anticipated revision risk, 4,622 – by surgeons who performed < 90% hybrid or cemented fixation, 44,947 Total knee replacements for OA, minimally stabilised, with fixed bearing surface and cemented tibial component n = 39,623

Female (%) Age, mean (SD) Male (%) Age, mean (SD) Type of hospital (%) Public Private

Routinely hybrid n = 9,079

57 68.8 (9.0) 43 68.4 (8.8)

58 69.0 (9.4) 42 68.7 (9.0)

33 67

77 23

Results

Figure 1. Flow diagram of total knee replacement exclusions.

Table 1. Total knee replacements (TKRs) included in the analysis by surgeon fixation preference Surgeon preference Cemented Hybrid fixation fixation No. of surgeons 94 TKR/surgeon, mean (SD) 325 (359) interquartile range [25–513] Type of TKR, n (%) Cemented 30,318 (99.3) Hybrid 226 (0.7) Total 30,544 (100)

Routinely cemented n = 30,544

14 649 (374) [311–903] 256 (2.8) 8,823 (97.2) 9,079 (100)

Total 108

30,574 9,049

Ethics, funding, and potential conflicts of interest The AOANJRR is approved by the Australian Federal Government as a Declaration of Quality Assurance Activity under section 124X of the Australian Federal Health Insurance Act, 1973. All investigations were conducted in accordance with ethical principles of research (the Helsinki Declaration II). No funding was received specific to this study and there are no competing interests to declare.

There were 39,623 primary TKR that met the inclusion criteria, undertaken by 108 surgeons, with 30,544 cemented TKRs and 9,079 hybrid TKRs (Figure 1). Most surgeons were cemented preferring (87%) and they undertook 77% of the included procedures. The hybrid preferring surgeons each performed on average more of the included TKR compared with the CP surgeons, with a mean of 649 TKR/surgeon compared with a mean of 325 TKR/surgeon, respectively (Table 1). The CP surgeons undertook hybrid fixation in 0.7% of their TKR and the HP surgeons undertook cemented fixation in 2.8% of their TKR. The demographics of each surgeon group had some differences, with the proportion of surgeons who worked in both private and public settings being higher in the HP surgeons (93%) compared with the CP surgeons (70%). The CP surgeons had contributed to the registry for more years (mean 9.6 years) compared with the HP surgeons (mean 6.2 years). Otherwise, the mean number of TKR in the registry, volume of all TKR/year and respective hospital arthroplasty volume was comparable for each group (Table 2). Both patient cohorts had similar mean age and gender demographics (Table 3), except that a greater proportion of the routinely hybrid cohort were undertaken at public hospitals (77%) compared with the routinely cemented cohort (33%). At 13 years, the CPR of the RC cohort was 4.8% (CI 4.1–5.7) compared with 5.5% (CI 3.5–8.7) for the RH cohort. The revision risk for each cohort was the same for all causes (HR = 1.0

Table 2. Surgeon practice data Surgeon preference Cemented fixation Hybrid fixation Type of practice: Public and private / private only / public only (%) Surgeons’ time in the registry a (year) No. of total knee replacements in registry a Included and excluded total knee replacements per year a Hospital volume per year of all lower-limb arthroplasties a a The

12130 Vertullo D.indd 331

70 / 25 / 5 6.2 (4.9) [2.1–7.8] 805 (435) [503–1040] 96 (41) [68–109] 501 (373) [235–618]

93 / 7 / 0 9.6 (4.3 ) [6.7–12.7] 781 (337) [581–987]) 85 (22) [66–106]) 616 (435) [276–840]

values are given as the mean (SD) [interquartile range].

5/3/2018 11:46:16 AM

332

Acta Orthopaedica 2018; 89 (3): 329–335

Cumulative revision (%)

Cumulative incidence (%) – routinely cemented

Cumulative incidence (%) – routinely hybrid

Routinely cemented

Infection

Routinely hybrid

Loosening/lysis

Patellofemoral pain

Pain

Instability

8 3

0 0

9 10 11 12 13 14 15

Years since primary procedure

Year 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

RC 30,544 25,107 20,336 16,260 12,755 9,541 6,901 4,594 3,021 2,421 1,919 1,433 1,009 618 335 127

9 10 11 12 13 14 15

Years since primary procedure

Figure 3. Cumulative incidence revision diagnosis of primary total knee replacement by surgeon fixation preference in patients with osteoarthritis.

Cumulative revision (%)

Cumulative revision for infection (%) 6

NXLPE – routinely cemented

Male – routinely cemented

NXLPE – routinely hybrid

Male – routinely hybrid

XLPE – routinely cemented

Female – routinely cemented

XLPE – routinely hybrid

Female – routinely hybrid

0 0

9 10 11 12 13 14 15

Years since primary procedure

Figure 5. Cumulative percentage revision of primary total knee replacement by polyethylene type and surgeon fixation preference in patients with osteoarthritis. HR adjusted for age and sex, entire period: NXLPE RC vs NXLPE RH: HR = 0.89 (0.73–1.08), p = 0.2 XLPE RH vs NXLPE RH: HR = 0.48 (0.31–0.75, p = 0.001 NXLPE RC vs XLPE RC: HR = 0.94 (0.79–1.12), p = 0.5 XLPE RH vs XLPE RC: HR = 0.57 (0.37–0.88), p = 0.01

(CI (0.84–1.20)) (Figure 2), non-infective causes (HR = 1.0 (0.81–1.24)), and for infection (HR = 1.0 (0.73–1.38)). The 5 most common diagnoses at revision were similar in each cohort (Figure 3) (Table 4, see Supplementary data). The types of revision were similar between the cohorts. When the effects of age were examined, the revision risk for the RC cohort who were less than 65 years was similar to the RH cohort who were less than 65 years (HR = 1.1 (1 = 0.83–1.42)). Similarly, the revision risk for patients older than

12130 Vertullo D.indd 332

9 10 11 12 13 14 15

RH 9,079 7,422 5,948 4,596 3,378 2,461 1,759 1,055 577 383 254 165 100 59 31 9

Years since primary procedure

Figure 2. Cumulative percentage revision of primary total knee replacement by surgeon fixation preference in patients with osteoarthritis. Gray zones represent 95% confidence intervals. HR adjusted for age and sex: Routinely cemented (RC) versus routinely hybrid (RH), entire period: HR = 1.00 (0.84–1.20). Number at risk (Figure 2)

9 10 11 12 13 14 15

Years since primary procedure

Figure 7. Cumulative percentage revision for infection of primary total knee replacement by patient gender and surgeon fixation preference in patients with osteoarthritis. HR adjusted for age, entire period: Male RC vs male RH: HR = 1.07 (0.71–1.63), p = 0.7 Male RH vs female RH: HR = 1.90 (1.07–3.37, p = 0.03 Male RC vs female RC: HR = 2.26 (1.65–3.09), p < 0.001 Female RC vs female RH: HR = 0.90 (0.54–1.49), p = 0.7

65 years (HR = 0.96 (0.76–1.21)) was similar between the 2 cohorts. Stratification of patients into males and females aged less than 65 years (Figure 4, see Supplementary data) and greater than 65 years revealed the same revision risk both for males in each cohort, and for females in each cohort. XLPE usage was more common in the RC cohort (47%) than in the RH cohort (29%). When the effects of XLPE were examined, the RH cohort with XLPE had a lower revision

5/3/2018 11:46:19 AM

Acta Orthopaedica 2018; 89 (3): 329–335

risk than the RC cohort with XLPE (HR = 0.57 (0.37–0.88)) (Figure 5). Revision risk with non-cross-linked-polyethylene (NXLPE) was the same between the cohorts. Patella resurfacing was more common in the RC cohort (61%) compared with the RH cohort (51%). When the effects of patella resurfacing were examined, both cohorts had the same revision risk with and without patella resurfacing (Figure 6, see Supplementary data). When the revisions for infection in each cohort were stratified by sex there was no difference between males and females in each cohort (Figure 7).

Discussion The advantages and disadvantages of cement fixation have been debated for decades; however, the choice of TKR fixation typically remains the preference of the surgeon, characteristically founded in efforts to maximize the long-term outcomes of their patients (Kobs and Lachiewicz 1993), with loosening and lysis remaining the dominant reasons for revision. In this analysis, patients of surgeons who preferred hybrid fixation had the same long-term risk of revision compared with the patients of those surgeons who preferred cemented fixation. This finding was irrespective of patient’s age or sex, and whether the patella was resurfaced or not. When reasons for revision were stratified into non-infective and infective, there was no difference between the 2 cohorts overall. There was also no difference between the 2 cohorts for those who received NXLPE, but the patients of surgeons who preferred hybrid fixation and received XLPE had a lower revision risk than the patients of surgeons who preferred cemented fixation and received XLPE. Cement disease (Jones and Hungerford 1987) was described in 1987, suggesting particles of polymethylmethacrylate were the primary cause of osteoclast-induced failure at the prosthesis–bone interface. More recently, the theory of cement disease has been discarded in favor of wear-particle induced lysis and loosening, which primarily focuses on bearing surface generated particles rather than those from the fixation interface (Harris 1994). Furthermore, cement has been suggested as a possible protective barrier, or seal, to wear particle-laden synovial fluid ingress into the prosthesis¬–bone interface (Harris et al. 1996). Our results are not in keeping with either the cement disease theory or cement as a seal theory. Despite a paucity of supporting clinical data, cementless fixation of the femur and tibia has been recommended as the optimum biologic TKR fixation solution for younger at-risk patients when compared with cemented fixation (Dorr 2002). However, in registry studies, the long-term revision risk of cementless fixation is higher than cemented and hybrid fixation and in smaller clinical series outcomes of cementless fixation of the femur and tibia remain similar, or inferior to, cemented fixation of the femur and tibia (Pulido et al. 2015,

12130 Vertullo D.indd 333

333

Dalury 2016). It is for this reason we did not examine cementless fixation of both components in this analysis. Cementless fixation of the tibia does not reduce the revision risk or migration when compared with cemented fixation in radiostereometric trials (Carlsson et al. 2005), in registry studies (Graves et al. 2016), or clinical series (Behery et al. 2016). Given that the tibia remains the component most at risk for failure (Voigt and Mosier 2011), hybrid fixation was introduced as a pragmatic alternative to employ the advantages of 2 differing fixation philosophies (Kraay et al. 1991, Petursson et al. 2015). Hybrid fixation has similar outcomes to cemented fixation in direct comparisons (Pelt et al. 2013), clinical series (Choi et al. 2012, McLaughlin and Lee 2014), and in some registry reports, a lower revision risk (Petursson et al. 2015, AOA National Joint Registry 2016). Petursson et al. (2015) performed a registry review of 3 different fixed and mobile bearing TKR designs, reporting that 1 of 3 three designs examined had a lower revision risk in the hybrid version. When this prosthesis, mainly performed at one high-volume hospital, was excluded, there was no difference between hybrid and cemented fixation in their direct comparative analysis, in keeping with our results. We specifically examined the effects of age and sex on the revision risk in each cohort. Cemented fixation for older females may have had some advantage due to lower femoral bone density; however, we found no difference with hybrid fixation in females over 65 years, consistent with other authors (Nakama et al. 2012, Dalury 2016). Similarly, hybrid fixation may be of advantage to younger active males, but there was no difference between fixation types for males under 65 years. We also assessed whether hybrid fixation lowered the infection risk given recent registry data suggesting a lower rate of revision for infection with certain TKR designs and bearing materials (Vertullo et al. 2017). Theoretically, a reduced burden of cement, cement particulate, and third-body wear could favorably alter the local immunomodulation of the joint environment (Spaan et al. 2013); however, our results suggest no advantage exists when only 1 major component is cemented. In registry studies, the use of cross-linked polyethylene in TKR lowers the risk of loosening and lysis when compared with non-cross-linked polyethylene (de Steiger et al. 2015), presumably via a reduction in particle-related osteolysis. In our series, when the effect of XLPE was examined, it resulted in a 43% lower revision risk in the HP cohort compared with the CP cohort. It remains uncertain if this is due to an additive effect of XLPE when used with a lesser volume of cement, or some other unrecognized confounders such as patient selection or femoral component design. When possible confounders were reduced using instrumented variable methodology, our analysis did not demonstrate superior survivorship with hybrid fixation overall. Consequently, it remains uncertain whether the extra cost of cementless femoral components is justified by the reduced

5/3/2018 11:46:32 AM

334

operative time (Petursson et al. 2015) and possibly decreased bone loss given the excellent long-term results of cemented fixation of the femur. By comparing revision risk based on surgeon fixation preference, we believe we have addressed concerns related to the potential for selection bias that may arise in a direct comparison of all cemented and hybrid TKR. To our knowledge this is the first time registry data have been used to investigate the outcome of surgeon preference in TKR fixation rather than directly comparing the long-term revision risk of hybrid and cemented TKR. This study was specifically designed to address major confounders that may introduce bias between the hybrid and cemented fixation. The impact of potential differences due to age, gender, and primary diagnosis have been considered. In addition, prosthesis-specific factors such as the use of posterior cruciate stabilization, mobile bearing, patellar resurfacing, and cross-linked polyethylene were also considered. Only patients with osteoarthritis were included as other diagnoses such as rheumatoid arthritis are more likely to have a higher incidence of osteoporosis and consequently a higher use of cemented TKR. Mobile bearing TKR were excluded as they have a known higher revision risk compared with fixed bearing, and potentially could have a detrimental interaction with cementless femoral fixation (AOA National Joint Registry 2016). PS TKR were excluded as they have a higher revision risk than MS TKR and have a higher revision risk with hybrid fixation than cemented fixation (Vertullo et al. 2017). As this is a registry analysis, some specific clarifications are important. First, a registry analysis differs from a clinical trial, in that while it can identify and monitor comparative national outcomes it cannot assign causality. Nonetheless, to optimize TKR survivorship, it is not vital to know why there is a difference between options, just that one exists, allowing all stakeholders to make shared informed decisions (Graves 2010). Second, another issue is unrecognized confounders. Unrecognized selection bias or confounding may have occurred, but, by focusing on the surgeon’s stability preference rather than the actual prosthesis used, this risk is minimized. Randomized controlled trials can reduce this selection bias; however, the current RCT literature showing no survivorship difference between hybrid and cemented fixation is underpowered to show a difference and has inadequate follow-up (Nakama et al. 2012). Surgeons with less surgical experience may prefer cemented fixation, but we restricted our analysis to surgeons who perform over 50 TKR per year to remove performance bias, as this has been previously cited as a large enough volume to exclude surgeon inexperience (Abdel et al. 2011). While the hybrid fixation preferring surgeons had on average more TKR per surgeon in the analysis than the cemented fixation preferring surgeons, the cemented fixation preferring surgeons performed more TKR per year on average and overall had performed a greater number of TKR. The AOANJRR collects level 1 and 2 data, hence comorbidities, patient-recorded

12130 Vertullo D.indd 334

Acta Orthopaedica 2018; 89 (3): 329–335

outcome measures, and prosthesis alignment data are not collected. The AOANJRR only recently commenced recording ASA and BMI, and hence these factors could not be included in this analysis. A possible limitation with any registry-based analysis is the data’s accuracy and validation. In the AOANJRR, after an initial capture rate of 96.8% (AOA National Joint Registry 2016), a sequential multi-level matching process against health department unit record data is undertaken, resulting in an almost complete dataset of primary and revision knee replacement in Australia. In summary, there was no overall difference in the revision risk for the patients of surgeons who prefer hybrid fixation in minimally stabilized TKR, compared with the patients of surgeons who prefer cemented fixation. Only when the effects of alternative bearing surfaces were examined had the patients of surgeons who preferred hybrid fixation and utilized XLPE a 43% reduction in revision risk. Supplementary data Table 4, Figure 4 and Figure 6 are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/17453674.2018.1449466

CJV designed the study, developed the methodology, performed the analysis, and wrote the manuscript. YP developed the methodology and performed the analysis. SEG and PLL collected the data, developed the methodology, performed the analysis, and wrote the manuscript.

The authors wish to thank Michelle Lorimer (BSc Hons) for technical assistance in the statistical methodology, the Australian Orthopaedic Association National Joint Replacement Registry, and the hospitals, orthopedic surgeons, and patients whose data made this work possible.

Acta thanks Gunnar Petursson and other anonymous reviewers for help with peer review of this study

Abdel M P, Morrey M E, Jensen M R, Morrey B F. Increased long-term survival of posterior cruciate-retaining versus posterior cruciate-stabilizing total knee replacements. J Bone Joint Surg Am 2011; 93(22): 2072-8. AOA National Joint Registry. 2016 Annual Report of the Australian Orthopaedic Association National Joint Replacement Registry. Adelaide: AOA; 2016. Behery O A, Kearns S M, Rabinowitz J M, Levine B R. Cementless vs cemented tibial fixation in primary total knee arthroplasty. J Arthroplasty [Internet]. 2016 Dec [cited 2017 Mar 12]. Carlsson A, Björkman A, Besjakov J, Onsten I. Cemented tibial component fixation performs better than cementless fixation: a randomized radiostereometric study comparing porous-coated, hydroxyapatite-coated and cemented tibial components over 5 years. Acta Orthop 2005; 76(3): 362-9. Choi Y J, Lee K W, Kim C H, Ahn H S, Hwang J K, Kang J H, Han H D, Cho W J, Park J S. Long-term results of hybrid total knee arthroplasty: minimum 10-years follow-up. Knee Surg Relat Res 2012; 24(2): 79-84. Dalury D F. Cementless total knee arthroplasty. Bone Joint J 2016; 98(7): 867-73.

5/3/2018 11:46:32 AM

Acta Orthopaedica 2018; 89 (3): 329–335

de Steiger R, Muratoglu O, Lorimer M, Cuthbert A R, Graves S E. Lower prosthesis-specific 10-year revision rate with crosslinked than with noncrosslinked polyethylene in primary total knee arthroplasty: 386,104 procedures from the Australian Orthopaedic Association National Joint Replacement Registry. Acta Orthop 2015; 86(6): 721-7. Dorr L D. Fixation of the millennium: the knee. J Arthroplasty 2002; 17(4 Suppl 1): 6-8. Faris P M, Keating E M, Farris A, Meding J B, Ritter M A. Hybrid total knee arthroplasty: 13-year survivorship of AGC total knee systems with average 7 years followup. Clin Orthop Rel Res 2008; 466(5): 1204-9. Graves S E. The value of arthroplasty registry data. Acta Orthop 2010; 81(1): 8-9. Graves S, Harris I, de Steiger R. Supplementary report: revision hip and knee arthroplasty from the Australian Orthopaedic Association National Joint Replacement Registry [Internet]. 2016. Harris W H. Osteolysis and particle disease in hip replacement: a review. Acta Orthop Scand 1994; 65(1): 113-23. Harris W H, Smith E J, Goetz D D. Bone cement as a seal protecting the femur from the ingress of particulate debris and femoral osteolysis. Instr Course Lect 1996; 45: 183-5. Jones L C, Hungerford D S. Cement disease. Clin Orthop Rel Res 1987; (225): 192-206. Kobs J K, Lachiewicz P F. Hybrid total knee arthroplasty: 2- to five-year results using the Miller-Galante prosthesis. Clin Orthop Rel Res 1993; (286): 78-87. Kraay M J, Meyers S A, Goldberg V M, Figgie H E, Conroy P A. “Hybrid” total knee arthroplasty with the Miller-Galante prosthesis: a prospective clinical and roentgenographic evaluation. Clin Orthop Rel Res 1991; (273): 32-41. McLaughlin J R, Lee K R. Hybrid total knee arthroplasty: 10- to 16-year follow-up. Orthopedics 2014 1; 37(11): e975-7. Nakama G Y, Peccin M S, Almeida G J, Lira Neto Ode A, Queiroz A A, Navarro R D. Cemented, cementless or hybrid fixation options in total knee arthroplasty for osteoarthritis and other non-traumatic diseases. Cochrane Database Syst Rev 2012; 10: CD006193.

12130 Vertullo D.indd 335

335

National Joint Registry. 13th Annual Report 2016 of the National Joint Registry for England, Wales, Northern Island and the Isle of Man [Internet]. Newhouse J P, McClellan M. Econometrics in outcomes research: the use of instrumental variables. Annu Rev Public Health.1998; 19(1): 17-34. Pelt C E, Gililland J M, Doble J, Stronach B M, Peters C L. Hybrid total knee arthroplasty revisited: midterm followup of hybrid versus cemented fixation in total knee arthroplasty. BioMed Res Int 2013; 2013: 854871. Petursson G, Fenstad A M, Havelin L I, Gøthesen Ø, Lygre S H, Röhrl S M, Furnes O. Better survival of hybrid total knee arthroplasty compared to cemented arthroplasty. Acta Orthop 2015; 86(6): 714-20. Pulido L, Abdel M P, Lewallen D G, Stuart M J, Sanchez-Sotelo J, Hanssen A D, Pagnano M W. The Mark Coventry Award: Trabecular metal tibial components were durable and reliable in primary total knee arthroplasty: a randomized clinical trial. Clin Orthop Rel Res 2015; 473(1): 34-42. Scott W N. Insall and Scott surgery of the knee. Philadelphia: Elsevier Churchill Livingstone; 2012. Spaan A N, Surewaard B G, Nijland R, van Strijp J A G. Neutrophils versus Staphylococcus aureus: a biological tug of war. Annu Rev Microbiol 2013; 67: 629-50. Stukel T A, Fisher E S, Wennberg D E, Alter D A, Gottlieb D J, Vermeulen M J. Analysis of observational studies in the presence of treatment selection bias: effects of invasive cardiac management on AMI survival using propensity score and instrumental variable methods. JAMA 2007; 297(3): 278-85. Vertullo C J, Lewis P L, Lorimer M, Graves S E. The effect on long-term survivorship of surgeon preference for posterior-stabilized or minimally stabilized total knee replacement: an analysis of 63,416 prostheses from the Australian Orthopaedic Association National Joint Replacement Registry. J Bone Joint Surg 2017; 99(13): 1129-39. Voigt J D, Mosier M. Hydroxyapatite (HA) coating appears to be of benefit for implant durability of tibial components in primary total knee arthroplasty: a systematic review of the literature and meta-analysis of 14 trials and 926 evaluable total knee arthroplasties. Acta Orthop 2011; 82(4): 448-59. Wright R J, Lima J, Scott R D, Thornhill T S. Two- to four-year results of posterior cruciate-sparing condylar total knee arthroplasty with an uncemented femoral component. Clin Orthop Rel Res 1990; (260): 80-6.

5/3/2018 11:46:32 AM

336

Acta Orthopaedica 2018; 89 (3): 336–344

Patient-reported symptoms and changes up to 1 year after meniscal surgery An observational cohort study of 641 adult patients with a meniscal tear Søren T SKOU 1,2, Kenneth PIHL 1, Nis NISSEN 3, Uffe JØRGENSEN 4, and Jonas Bloch THORLUND 1

1 Research Unit for Musculoskeletal Function and Physiotherapy, Department of Sports Science and Clinical Biomechanics, University of Southern Denmark, Odense; 2 Department of Physiotherapy and Occupational Therapy, Næstved-Slagelse-Ringsted Hospitals, Denmark, Region Zealand, Slagelse; 3 Department of Orthopaedic Surgery, Lillebælt Hospital in Kolding, Kolding; 4 Department of Orthopaedics and Traumatology, Odense University Hospital, Denmark Correspondence: stskou@health.sdu.dk Submitted 2017-08-24. Accepted 2018-01-22.

Background and purpose — Detailed information on the symptoms and limitations that patients with meniscal tears experience is lacking. This study was undertaken to map the most prevalent self-reported symptoms and functional limitations among patients undergoing arthroscopic meniscal surgery and investigate which symptoms and limitations had improved most at 1 year after surgery. Patients and methods — Patients aged 18–76 years from the Knee Arthroscopy Cohort Southern Denmark (KACS) undergoing arthroscopic meniscal surgery were included in this analysis of individual subscale items from the Knee Injury and Osteoarthritis Outcome Score and 1 question on knee stability. Severity of each item was scored as none, mild, moderate, severe, or extreme. Improvements were evaluated using Wilcoxon’s signed-rank test and effect size (ES). Results — The most common symptoms were knee grinding and clicking, knee pain in general, pain when twisting and bending the knee and climbing stairs (88–98%), while the most common functional limitations were difficulty bending to the floor, squatting, twisting, kneeling, and knee awareness (97–99%). Knee pain in general and knee awareness improved most 1 year after meniscal surgery (ES –0.47 and –0.45; p < 0.001), while knee instability and general knee difficulties improved least (ES 0.10 and –0.08; p < 0.006). Interpretation — Adults undergoing surgery for a meniscal tear commonly report clinical symptoms and functional limitations related to their daily activities. Moderate improvements were observed in some symptoms and functional limitations and small to no improvement in others at 1 year after surgery. These findings can assist the clinical discussion of symptoms, treatments, and patients’ expectations.

Meniscal tears are a common knee injury in adults, with an annual incidence of up to 172 injuries per 100,000 persons (Peat et al. 2014). Symptoms such as knee pain and mechanical symptoms (i.e., clicking, locking, or catching) are often considered to be related to meniscal tears (Niu et al. 2011, Yan et al. 2011). However, more detailed information on the clinical symptoms and functional limitations that these patients experience is lacking. Although meniscal surgery is not recommended in most patients with a degenerative meniscal tear (Siemieniuk et al. 2017), no consensus exists on when meniscal surgery is in fact indicated (Lyman et al. 2012). As such, an overview of the specific clinical symptoms and functional limitations that improve most following meniscal surgery would be helpful for clinicians and patients in a shared decision-making process discussing benefits, harms, and patients’ expectations of meniscal surgery. Importantly, meniscal tears often differ with regard to tear type and symptom onset (i.e., traumatic vs. slowly evolving) between younger and middle-aged to older patients (Poehling et al. 1990, Englund et al. 2008, Bergkvist et al. 2016). Hence, symptom patterns and improvements following meniscal surgery may also differ. The aim of this exploratory study was to determine the most common clinical symptoms and functional limitations and their severity as reported by patients with a meniscal tear undergoing meniscal surgery and in 2 subgroups based on age (40 years or younger and older than 40 years). Furthermore, we investigated which symptoms and functional limitations had improved most at 1 year after meniscal surgery.

■

12024 Skou D.indd 336

5/3/2018 11:47:59 AM

Acta Orthopaedica 2018; 89 (3): 336–344

Patients and methods We followed the STROBE guideline to report this observational cohort study. Patients Participants from the Knee Arthroscopy Cohort Southern Denmark (KACS) cohort were included in this study (Thorlund et al. 2013). KACS is a prospective cohort following adults undergoing arthroscopy for meniscal tears. Participants were consecutively recruited from 4 public hospitals in Denmark between February 1, 2013 and January 31, 2014, and at 1 of the original four hospitals from February 1, 2014 to January 31, 2015. Eligibility criteria In KACS, patients of at least 18 years of age referred for knee arthroscopy by an orthopedic surgeon on suspicion of a meniscal tear (based on clinical examination, injury history, and MRI if considered necessary) were included if they were able to read and understand Danish, had an email address and did not fulfill any of the following exclusion criteria: no meniscal tear at the later surgery; previous or planned reconstruction surgery of the anterior or posterior cruciate ligament in either knee; fractures to the lower extremities within the last 6 months; or inability to reply to questionnaires because of mental impairment (Thorlund et al. 2013). For the present study, all patients with baseline assessment were included. For the analysis of change from baseline to follow-up, only patients with both baseline assessment and 12-month followup data were included. Outcomes and other variables Patient characteristics and outcomes were collected using online questionnaires before surgery (median 7 days, interquartile range 3–10 days) and 3 and 12 months after arthroscopic meniscal surgery. Patient characteristics Patient characteristics included age, sex, height, weight, symptom onset, and duration of symptoms (Thorlund et al. 2013). Symptom onset was assessed by the question: “How did the knee pain/problems for which you are now having surgery develop?” with 3 response options: “The pain/problems have slowly developed over time,” “As a result of a less severe incident (i.e., kneeling, sliding, and/or twisting of the knee or the like),” and “As a result of a severe incident (i.e., during sports, a crash, or a collision or the like).” Duration of symptoms was assessed by the questions: “How long have you had your knee pain/knee problems for which you are now having surgery?” with response options ranging from “0–3 months” to “more than 24 months”.

12024 Skou D.indd 337

337

Outcomes KOOS. The Knee Injury and Osteoarthritis Outcome Score (KOOS) is a validated (Roos et al. 1998, Collins et al. 2016) and often used patient-reported outcome in studies concerning patients with meniscal tears (Herrlin et al. 2007, Katz et al. 2013). It consists of 5 subscales (i.e., Pain, Symptoms, Function in daily living (ADL), Function in sport and recreation (Sport/Rec), and knee related Quality of life (QoL)) constituted of several single-item questions with 5 response categories (0–4; typically ranging from “None” to “Extreme” symptoms) concerning specific perceived clinical symptoms, functional limitations, and quality of life (Roos et al. 1998). Hence, it is a possible source of detailed information on specific knee symptoms and functional limitations that patients with meniscal tears experience. Knee stability. As KOOS does not include specific items on knee instability and giving way, the following question was included: “In the last month, have you felt that your knee was unstable or about to buckle?” The patients responded on a 6-point Likert-like scale (0–5) ranging from “Never” to “All the time.” The question was adapted from the Oxford Knee Score (Dawson et al. 1998). Statistics The single-item outcomes were divided into clinical symptoms, defined as Pain and Symptoms subscales items from the KOOS and the knee instability question, and functional limitations and quality of life, defined as the individual items from the other three subscales of KOOS. Prevalence of clinical symptoms and functional limitations and quality of life were reported as the actual numbers (proportions) of the full study sample. Presence of symptoms was defined as response options 1–4 (leaving out the response option 0, corresponding to no symptoms) on KOOS and response options 1–5 (leaving out the response option 0, corresponding to no symptoms) on the stability question. Severity of symptoms is presented as the actual number of patients with that symptom severity (proportion). Only patients reporting having symptoms (responded 1–4 or 1–5 on the items described above) were included in this analysis. All outcome items were included in the analysis; however, to increase the readability of the manuscript only the 5 clinical symptoms and the 5 functional limitations and quality-of-life items with the highest prevalence were included in the results section, whereas the rest are presented in the Supplementary material. Subgroup analyses based on age (40 years or younger and older than 40 years) were conducted. We assessed the differences in single-item scores from baseline to 12 months’ follow-up by presenting the distribution of answers at baseline and 12 months for patients with complete data. Only the 5 clinical symptoms and functional limitations and quality-of-life items with the highest effect size are presented in the results, whereas the rest are presented in the Supplementary material. We made subgroup analyses of patients

5/3/2018 11:47:59 AM

338

Acta Orthopaedica 2018; 89 (3): 336–344

Table 1. Patient characteristics at baseline

Variables

All (n = 641)

Age, years (SD) 49 (13.0) Female, n (%) 280 (44) BMI (SD) 27 (4.4) Symptom onset, n (%) Developed slowly over time 208 (32) Developed as a result of less severe incident 260 (41) severe incident 173 (27) Duration of symptoms, n (%) 0–3 months 129 (20) 4–6 months 181 (28) 7–12 months 135 (21) 13–24 months 94 (15) Longer than 24 months 102 (16) KOOS subscale scores, mean (SD) Pain 54.9 (18) Symptoms 60.0 (19) ADL 63.7 (19) Sport/Rec 26.3 (22) QOL 41.6 (15) Type of surgery, no. (%) Resection 600 (94) Repair 33 (5.1) Both 8 (1.2)

≤ 40 years (n = 150)

> 40 years (n = 491)

31 (7.2) 50 (33) 26 (4.2)

55 (8.6) 230 (47) 28 (4.5)

29 (19)

179 (36)

51 (34) 70 (47)

209 (43) 103 (21)

41 (27) 24 (16) 31 (21) 20 (13) 34 (23)

88 (18) 157 (32) 104 (21) 74 (15) 68 (14)

58.9 (20) 60.6 (19) 69.8 (20) 31.1 (23) 40.2 (16)

53.6 (18) 59.8 (18) 61.8 (19) 24.9 (21) 42.0 (15)

118 (79) 24 (16) 8 (5.3)

482 (98.) 9 (1.8) 0 (0.0)

KOOS: The Knee injury and Osteoarthritis Outcome Score with scores ranging from 0 to 100 (worst to best scale); ADL: function in daily living; Sport/Rec: function in sport and recreation; QOL: quality of life.

40 years or younger and patients older than 40 years of age. The comparison was done using Wilcoxon’s signed-rank test. Effect sizes were calculated by the formula r = Z/√N, with Z being the z statistic output of the Wilcoxon signed-rank test and N being the number of observations. The distribution of answers at baseline and 12 months are presented alongside the effect sizes to allow for a comparison of how each item changed from baseline to 12 months follow-up. Due to the exploratory nature of this study, the significance level was set at p < 0.05, and all analyses were performed in IBM SPSS Statistics (Version 24, IBM Corporation, Armonk, NY, USA). Ethics, registration, funding, and potential conflicts of interest Written informed consent was obtained from all participants. The regional scientific ethics committee of Southern Denmark waived the need for ethical approval (Thorlund et al. 2013). The observational KACS cohort was pre-registered at ClinicalTrials.gov (NCT01871272). This study was funded by an individual postdoctoral grant (JBT) from the Danish Council for Independent Research and funds from the Region of Southern Denmark. The funder had no role in any part of the study other than to provide funding. The authors report no conflict of interest related to this study.

12024 Skou D.indd 338

Patients assigned for knee arthroscopy on clinical suspicion of a meniscal tear and assessed for eligibility for KACS n = 1,259 Excluded, did not ﬁt inclusion criteria (n = 138): – previous ACL/PCL surgery, 112 – fracture of lower extremities < 6 months before surgery, 5 – no email address, 18 – did not understand Danish, 2 – not mentally able to reply, 1 Excluded, other reasons (n = 213): – consented, but did not reply prior to surgery, 155 – declined or no reason, 50 – no time to participate, 8 Replied to baseline questionnaire n = 908 Excluded (n = 70): – surgery cancelled, 51 – re-scheduled to surgery at other hospital, 19 Had surgery n = 838 Excluded (n = 197): – no meniscal tear at surgery, 182 – ACL/PCL reconstruction at surgery, 15 Replied to baseline questionnaire and had a meniscal tear at surgery n = 641 No reply to 12 months’ questionnaire (n = 76) 12 months’ assessment n = 565

Figure 1. Study flow.

Results Baseline characteristics for the full group (n = 641) and the subgroups of patients based on age are presented in Table 1, while study flow is presented in Figure 1. 76 (12%) patients failed to reply to the 12 months questionnaire. Of those with complete follow-up, 60 (11%) patients self-reported additional surgery of their knee during the follow-up period. All patients (n = 641; n = 565 at 12 months’ follow-up) Increased awareness of the knee problem, knee pain in general, and difficulty twisting/pivoting the knee were the most prevalent symptoms and limitations. Severities of symptoms were mostly moderate to severe while severities of limitations were mostly severe to extreme for the most common symptoms and limitations (Tables 2 and 3, Figure 2 and Supplementary Tables 4 and 5). 12 months after meniscal surgery, knee pain in general and awareness of the knee problem had improved most, while

5/3/2018 11:47:59 AM

Acta Orthopaedica 2018; 89 (3): 336–344

339

Table 2. Prevalence and severity of the 5 most common clinical symptoms in the full group and the subgroups at baseline a ≤ 40 years (n = 150)

All (n = 641) Prevalence, n (%) [95% CI] S2. Grinding, clicking or other S7. Stiffness later in the day P1. Knee pain in general P2. Pain twisting/pivoting knee P4. Pain bending knee fully P6. Pain going up or down stairs Severity c (1 / 2 / 3 / 4), % S2. Grinding, clicking or other S7. Stiffness later in the day P1. Knee pain in general P2. Pain twisting/pivoting knee P4. Pain bending knee fully P6. Pain going up or down stairs

> 40 years (n = 491)

562 (88) [85–90] 546 (85) [82–88] b 631 (98) [97–99] 616 (96) [94–97] 567 (89) [86–91] 590 (92) [90–94]

135 (90) [84–94] 114 (76) [69–82] b 147 (98) [95–99] 141 (94) [89–97] 131 (87) [81–92] 131 (87) [81–92]

427 (87) [84–90] b 432 (88) [85–91] 484 (99) [97–99] 475 (97) [95–98] 436 (89) [86–91] 459 (93) [91–95]

12 / 31 / 42 / 15 34 / 43 / 22 / 1 5 / 11 / 65 / 19 14 / 31 / 43 / 12 24 / 33 / 31 / 11 24 / 36 / 32 / 9

12 / 26 / 43 / 19 43 / 33 / 22 / 2 8 / 20 / 53 / 19 18 / 28 / 38 / 16 25 / 31 / 26 / 18 34 / 34 / 21 / 11

12 / 26 / 43 / 19 31 / 46 / 22 / 1 4 / 9 / 69 / 18 13 / 32 / 44 / 11 24 / 34 / 33 / 9 21 / 36 / 35 / 8

95% CI = 95% confidence intervals. a Letters and numbers in front of each variable refer to item identification from the Knee Injury and Osteoarthritis Outcome Score (KOOS). b Not in top 5 for the group, only included as comparator for the other groups. c Severity: ranging from 1 (best) to 4 (worst) is the KOOS response categories for each individual item.

Table 3. Prevalence and severity of the 5 most common functional limitations and quality-of-life items in the full group and the subgroups at baseline a ≤ 40 years (n = 150)

All (n = 641) Prevalence, n (%) [95% CI] A5. Difficulty bending to floor SP1. Difficulty squatting SP4. Difficulty twisting/pivoting knee SP5. Difficulty kneeling Q1. Often aware of knee problem Q3. Lack of knee confidence Severity c (1 / 2 / 3 / 4), % A5. Difficulty bending to floor SP1. Difficulty squatting SP2. Difficulty running SP4. Difficulty twisting/pivoting knee SP5. Difficulty kneeling Q1. Often aware of knee problem Q3. Lack of knee confidence

> 40 years (n = 491)

619 (97) [95–98] 626 (98) [96–99] 628 (98) [97–99] 626 (98) [96–99] 637 (99) [98–100] 611 (95) [93–97] b

144 (96) [92–98] d 148 (99) [96–100] 144 (96) [92–98] d 145 (97) [93–99] 149 (99) [97–100] 146 (97) [94–99]

475 (97) [95–98] 478 (97) [96–98] 484 (99) [97–99] 481 (98) [96–99] 488 (99) [98–100] 465 (94) [92–96] b

16 / 28 / 39 / 16 8 / 18 / 37 / 37 6 / 14 / 38 / 41 8 / 14 / 36 / 42 9 / 19 / 35 / 38 1 / 4 / 63 / 32 18 / 33 / 40 / 9

17 / 35 / 33 / 15 11 / 18 / 36 / 34 10 / 17 / 34 / 39 10 / 17 / 35 / 37 13 / 20 / 34 / 32 3 / 9 / 64 / 25 14 / 27 / 47 / 12

15 / 27 / 41 / 17 8 / 17 / 37 / 38 5 / 13 / 40 / 42 7 / 13 / 37 / 43 8 / 18 / 35 / 40 1 / 3 / 63 / 34 19 / 34 / 38 / 8

knee instability (actually worsened) and general knee difficulties had improved least (Table 6, Figure 3 and Supplementary Table 7). Patients aged 40 years or younger (n = 150; n = 121 at 12 months’ follow-up) Increased awareness of the knee problem, difficulties squat-

12024 Skou D.indd 339

ting, and knee pain in general were the most prevalent symptoms and limitations. Severities of symptoms were mostly moderate to severe while severities of limitations were mostly severe to extreme for the most common symptoms and limitations (Tables 2 and 3 and Supplementary Tables 4 and 5). 12 months after meniscal surgery, knee pain in general and difficulty running had improved most, knee instability (no dif-

5/3/2018 11:48:00 AM

340

Acta Orthopaedica 2018; 89 (3): 336–344

Prevalence (%)

None Mild Moderate Severe Extreme

100

Often aware of knee problem (Q1)

Knee pain in general (P1)

SP4 P1

Difficulty twisting/pivoting knee (SP4) Difficulty kneeling (SP5) Difficulty squatting (SP1)

Difficulty bending to ﬂoor (A5)

A12

Pain twisting/pivoting knee (P2)

A10

Pain going up or down stairs (P6) Pain bending knee fully (P4)

Grinding, clicking or other (S2) 76

100

Prevalence (%) of clinical symptoms, disability, and quality of life items

Months after surgery

Figure 2. Prevalence (95% CI) and severity of the 5 most common clinical symptoms and five most common limitations and quality-of-life items in patients with a meniscal tear considered eligible for meniscal surgery (n = 641). Severity (color of the data points) is the most prevalent of the 5 levels of severity on the Knee Injury and Osteoarthritis Outcome Score (KOOS).

Figure 3. Prevalence (95% CI) of the 5 self-reported clinical symptoms and limitations and quality-of-life items mostly improved at 12 months after arthroscopic meniscal surgery at baseline, 3, and 12 months in patients with a meniscal tear considered eligible for meniscal surgery (n = 557). Letters and numbers in front of each variable refer to item identification from KOOS. Q1: Often aware of knee problem; SP4: Difficulty twisting/pivoting; P1: Knee pain in general; A12: Difficulty lying in bed; and A10: Difficulty rising from bed.

ference between baseline and 12 months follow-up) and general knee difficulties had improved least (Table 6 and Supplementary Table 7).

88–99% of all patients in the cohort and mostly of moderate to extreme severity. Furthermore, we found that patient-reported knee pain in general and knee awareness had improved most at 1 year after meniscal surgery, while knee instability and general knee difficulties improved least. Some differences did exist in subgroups of patients older or younger than 40 years of age.

Patients aged more than 40 years (n = 491; n = 444 at 12 months’ follow-up) Increased awareness of the knee problem, knee pain in general, and difficulty twisting/pivoting the knee were the most prevalent symptoms and limitations. Severities of symptoms were mostly moderate to severe while severities of limitations were mostly severe to extreme for the most common symptoms and limitations (Tables 2 and 3 and Supplementary Tables 4 and 5). 12 months after meniscal surgery, knee pain in general and difficulty lying in bed (turning over and maintaining knee position) had improved most, while knee instability (actually worsened) and general knee difficulties had improved least (Table 6 and Supplementary Table 7).

Discussion In this exploratory study we found that patient-reported grinding and clicking with knee movement, knee pain in general, and pain when twisting and bending the knee and going up and down stairs were the most common clinical symptoms experienced by patients about to undergo surgery for a meniscal tear. Difficulty bending to the floor, squatting, twisting, and kneeling, and awareness of the knee problem were the most common functional limitations and quality-of-life problems. These symptoms and functional limitations were present in

12024 Skou D.indd 340

Prevalence and severity of symptoms Meniscal injury is a common problem in the general population (Peat et al. 2014). However, detailed information on the most common self-reported symptoms, functional limitations, and quality-of-life problems, their severity, and improvements after treatment in patients with a meniscal tear is sparse despite its importance to guide the clinical discussion of symptoms, treatments, and patient expectations. A recent cross-sectional study of baseline data from 2 randomized controlled trials (RCT) of 199 middle-aged patients with a meniscal tear eligible for meniscal surgery found that knee pain in general, pain when twisting/pivoting the knee, when bending the knee fully, and when going up and down stairs, and lack of knee confidence were the 5 most common self-reported symptoms, functional limitations, and quality-of-life problems, all with at least moderate severity (Hare et al. 2017). In general, our results confirm these findings in a larger, prospective cohort that typically are more generalizable to the population undergoing meniscal surgery than participants recruited for RCTs with strict eligibility criteria (Bellomo and Bagshaw 2006). Furthermore, the clinical symptoms found to be most prevalent in our study are consistent with the meniscal symptoms that clinicians typically query patients about when diagnosing

5/3/2018 11:48:00 AM

Acta Orthopaedica 2018; 89 (3): 336–344

341

Table 6. Outcome from baseline to 12 months for the 5 clinical symptoms and limitations and quality-of-life items mostly improved after arthroscopic surgery in the full group and for patients of 40 years of age or younger and patients older than 40 years of age a All (n = 565)

≤ 40 years (n = 121)

> 40 years (n = 444)

Base12 Base12 Base12 line months Diff. line months Diff. line months Diff. Severity b n n p-value Effect size n n p-value Effect size n n p-value Effect size P1. Knee pain in general 0 10 124 < 0.001 –0.47 3 20 < 0.001 1 29 132 11 33 2 58 113 20 28 3 371 161 68 31 4 97 35 19 9 P6. Pain going up and down stairs 0 46 191 < 0.001 –0.42 c 17 46 < 0.001 1 125 158 38 32 2 185 134 37 26 3 169 71 22 13 4 40 11 7 4 A10. Difficulty rising from bed 0 131 347 < 0.001 –0.44 51 89 < 0.001 1 204 138 46 25 2 162 59 19 7 3 58 16 4 0 4 10 5 1 0 A12. Difficulty lying in bed (turning over, maintaining knee position) 0 89 302 < 0.001 –0.44 36 74 < 0.001 1 152 137 36 27 2 183 79 32 11 3 118 35 14 8 4 23 12 3 1 SP2. Difficulty running 0 23 106 < 0.001 –0.41 c 8 31 < 0.001 1 34 117 12 28 2 79 113 21 25 3 208 133 37 21 4 221 96 43 16 SP4. Difficulty twisting/pivoting knee 0 10 101 < 0.001 –0.43 4 27 < 0.001 1 43 122 13 27 2 85 121 25 29 3 198 123 39 22 4 229 98 40 16 Q1. Often aware of knee problem 0 3 51 < 0.001 –0.45 1 9 < 0.001 1 7 109 4 30 2 21 105 8 24 3 358 228 82 42 4 176 72 26 16 Q3. Lack of knee confidence 3 19 < 0.001 0 25 108 < 0.001 –0.39 c 1 94 206 14 35 2 183 120 36 33 3 216 116 56 30 4 47 15 12 4

–0.44

7 18 38 303 78

104 99 85 130 26

< 0.001

–0.48

–0.30 c

29 87 148 147 33

145 126 108 58 7

< 0.001

–0.45

–0.39 c

80 158 143 54 9

258 113 52 16 5

< 0.001

–0.46

–0.34 c

53 116 151 104 20

228 110 68 27 11

< 0.001

–0.47

–0.43

15 22 58 171 178

75 89 88 112 80

< 0.001

–0.41 c

–0.41

6 30 60 159 189

74 95 92 101 82

< 0.001

–0.44 c

–0.40

2 3 13 276 150

42 79 81 186 56

< 0.001

–0.46

–0.39

22 80 147 160 35

89 171 87 86 11

< 0.001

–0.39 c

a Letters

and numbers in front of each variable refer to item identification from the Knee Injury and Osteoarthritis Outcome Score (KOOS). b Severity: ranging from 0 (best) to 5 (worst) or 0 (best) to 4 (worst) is the response categories for each individual item. The comparison of the paired data was done using Wilcoxon’s signed-rank test. Effect sizes were calculated by the formula r = Z/√Nobservations. All comparisons between baseline and 12 months’ follow-up were significant (p-value < 0.05). c Not in top 5 for the group, only included as comparator for the other groups.

12024 Skou D.indd 341

5/3/2018 11:48:03 AM

342

a meniscal tear (Niu et al. 2011), highlighting the relevance of asking these questions in clinical practice. However, many of these symptoms are also typical for patients with early knee osteoarthritis (Hare et al. 2017). Niu et al. (2011) found that a Meniscal Symptom Index consisting of the combination of 4 symptoms (localized pain, clicking, catching, and giving way) were able to identify 76% of patients with symptoms consistent with a meniscal tear. As our cohort did not include patients without a meniscal tear, we were not able to assess the diagnostic value of a combination of symptoms to identify patients with a meniscal tear. However, a systematic review has shown that the accuracy of most commonly applied clinical tests of a meniscal tear is poor (Smith et al. 2015). Our findings may be used to develop and test a combined approach of clinical tests, medical history and self-reported symptoms for clinical diagnosing of a meniscal tear. Interestingly, we found some subgroup differences in the most common clinical symptoms and functional limitations and quality-of-life problems. Knee stiffness after resting was only among the most prevalent symptoms in patients above 40 years of age (88%). Knee stiffness is a typical symptom of knee osteoarthritis (OA) (Zhang et al. 2010) and a meniscal tear in middle-aged and elderly people has been suggested to be a signifying feature of incipient OA (Englund et al. 2012). This corresponds well to the previous finding that a large proportion of patients in the present cohort have early or more established OA (Pihl et al. 2017). Lack of knee confidence was prevalent in the full group (95%), but only among the 5 most common functional limitations and quality-of-life problems in patients of 40 years or younger (97%). Worse knee confidence has previously been found to predict functional decline in people with or at increased risk of having knee OA (Colbert et al. 2012). Furthermore, it has been suggested to be an important part of a downward spiral with worse knee confidence decreasing the level of activity, potentially leading to worse pain again affecting knee confidence (Skou et al. 2014). This highlights the need to address knee confidence along with other symptoms in the treatment of patients with a meniscal tear. Changes in symptoms 1 year after meniscal surgery Systematic reviews and meta-analyses of randomized controlled trials have failed to show superior effect of arthroscopic surgery compared with placebo surgery or in addition to exercise for middle-aged and older patients with degenerative meniscal tears (Khan et al. 2014, Thorlund et al. 2015, Brignardello-Petersen et al. 2017). Self-reported mechanical symptoms (i.e., the sensation of catching and/or locking) are typically considered an important indication for surgery (Stuart and Lubowitz 2006, Jevsevar et al. 2014, Krych et al. 2014). However, a recent study in patients with degenerative meniscal tears found no added benefit of surgery over that of placebo surgery in patients with preoperative mechanical

12024 Skou D.indd 342

Acta Orthopaedica 2018; 89 (3): 336–344

symptoms (Sihvonen et al. 2018). Furthermore, the proportion of patients with preoperative mechanical symptoms satisfied with their knee status and reporting improvements at 1 year after arthroscopic surgery is lower compared with those without preoperative mechanical symptoms (Sihvonen et al. 2016b). In our study, mechanical symptoms (S2 and S3 on KOOS) only improved to a small extent 1 year following meniscal surgery. To some extent, this supports the findings from a secondary analysis of an RCT demonstrating that meniscal surgery has no added benefit over sham surgery in relieving knee catching or occasional locking (Sihvonen et al. 2016a). Interestingly, although awareness of knee problem, difficulty twisting/pivoting knee, and knee pain in general were among the self-reported clinical symptoms and limitations and quality-of-life items that had improved most at 1 year after meniscal surgery, 77–91% of the patients still had the symptoms, suggesting that although patients can expect substantial improvements at 1 year after surgery, most will not be symptom free. We found some differences in the subgroups of patients older or younger than 40 years, including smaller improvements in knee joint stiffness in the morning and in pain at night in the younger subgroup, potentially explained by differences in the severity of the symptoms at baseline in the different subgroups. Based on data from the same cohort as used in this study, Thorlund et al. (2017) found a statistically larger, but clinically irrelevant, improvement in patient-reported outcomes following meniscal surgery in patients with a degenerative meniscal tear compared with patients who had a traumatic tear. However, evidence from RCTs supporting these findings is lacking. 2 ongoing trials will help shed light on the effects from meniscal surgery in a younger population, typically with a non-degenerative meniscal tear: 1 Dutch trial comparing arthroscopic resection and rehabilitation for a traumatic tear in adults aged 18–45 years of age (identifier www.trialregister. nl no. 17454) and 1 Danish trial of meniscal surgery versus exercise and education in patients of 40 or younger with a meniscal tear (Skou et al. 2017). Our results provide guidance to clinicians and patients in terms of which symptoms patients can expect will improve most, and which symptoms will improve the least at 1 year after meniscal surgery and whether the patient can expect the individual symptom to disappear or be relieved. Limitations Previous studies have suggested that the majority of the treatment effect from various treatments of knee pain is attributable to placebo or contextual factors (Bannuru et al. 2015, Zou et al. 2016). As our study is limited by the lack of a control group, the specific effect sizes found following meniscal surgery can only be used to compare changes in the different selfreported symptoms, functional limitations, and quality-of-life

5/3/2018 11:48:03 AM

Acta Orthopaedica 2018; 89 (3): 336â&#x20AC;&#x201C;344

problems and not as a measure of the effects of surgery per se. Furthermore, the exploratory nature of this study precludes any firm conclusions. 155 patients did not reply to the baseline questionnaire before surgery. However, this was mainly because patients did not have sufficient time to respond, as waiting time for surgery was very short. The age of patients not responding was a bit lower (46 years of age vs. 49), while the gender distribution was similar (43% females vs. 44%). In general, demographics with regard to sex and age for patients included in this study are similar to what has previously been reported for patients undergoing meniscal surgery in Denmark (Thorlund et al. 2014), thus we consider the external validity of the results from the cohort to be high. In summary, in patients undergoing meniscal surgery, grinding and clicking with knee movement, knee pain in general, and pain when twisting and bending the knee and going up and down stairs of mostly moderate to severe severity are common clinical symptoms. Difficulties bending to the floor, squatting, twisting, and kneeling, and awareness of the knee problem of mostly severe to extreme severity are common functional limitations and quality-of-life problems. At 1 year after surgery, moderate improvements were observed in some symptoms and functional limitations (knee pain in general and knee awareness improved most) and small to no improvement in others (knee instability and general knee difficulties improved least). Some differences did exist in the age-based subgroups, including worsened knee instability at 1 year after meniscal surgery in patients older than 40 years of age. Although the observational nature of our study precludes conclusions on the effects of surgery, the findings from our study can be used in the clinical discussion of symptoms, treatments, and patient expectations for patients with a meniscal tear. Supplementary data Supplementary Tables 4, 5 and 7 are available as supplementary data in the online version of this article, http://dx.doi.org/ 10.1080/17453674.2018.1447281

JBT, KP, and STS conceived and designed the study. NN and UJ participated in patient recruitment and data collection and gave feedback on the study design. STS conducted the analysis. STS drafted the first version of the manuscript. All authors helped in revising the manuscript.

The authors would like to acknowledge the study funder, all participating patients, and orthopedic surgeons, nurses, and secretaries at the Department of Orthopedics and Traumatology at Odense University Hospital (Odense and Svendborg) and Department of Orthopedics at Lillebaelt Hospital (Kolding and Vejle) for their assistance with patient recruitment and data collection. STS is supported by the Danish Council for Independent Research (DFF 6110-00045) and the Lundbeck Foundation.

343

Bannuru R R, McAlindon T E , Sullivan M C, Wong J B, Kent D M, Schmid C H. Effectiveness and implications of alternative placebo treatments: a systematic review and network meta-analysis of osteoarthritis trials. Ann Intern Med 2015; 163(5): 365-72. Bellomo R, Bagshaw S M. Evidence-based medicine: classifying the evidence from clinical trialsâ&#x20AC;&#x201D;the need to consider other dimensions Crit Care 2006; 10(5): 232. Bergkvist D, Dahlberg L E, Neuman P, Englund M. Knee arthroscopies: who gets them, what does the radiologist report, what does the surgeon find? An evaluation from southern Sweden. Acta Orthop 2016; 87(1): 12-16. Brignardello-Petersen R, Guyatt G H, Buchbinder R, Poolman R W, Schandelmaier S, Chang Y, Sadeghirad B, Evaniew N, Vandvik P O. Knee arthroscopy versus conservative management in patients with degenerative knee disease: a systematic review. BMJ Open 2017; 7(5): e016114. Colbert C J, Song J, Dunlop D, Chmiel J S, Hayes K W, Cahue S, Moisio K C, Chang A H, Sharma L. Knee confidence as it relates to physical function outcome in persons with or at high risk of knee osteoarthritis in the osteoarthritis initiative. Arthritis Rheum 2012; 64(5): 1437-46. Collins N J, Prinsen C A, Christensen R, Bartels E M, Terwee C B, Roos E M. Knee injury and Osteoarthritis Outcome Score (KOOS): systematic review and meta-analysis of measurement properties. Osteoarthritis Cartilage 2016; 24(8): 1317-29. Dawson J, Fitzpatrick R, Murray D, Carr A. Questionnaire on the perceptions of patients about total knee replacement. J Bone Joint Surg Br 1998; 80(1): 63-9. Englund M, Guermazi A, Gale D, Hunter D J, Aliabadi P, Clancy M, Felson D T. Incidental meniscal findings on knee MRI in middle-aged and elderly persons. N Engl J Med 2008; 359(11): 1108-15. Englund M, Roemer F W, Hayashi D, Crema M D, Guermazi A. Meniscus pathology, osteoarthritis and the treatment controversy. Nature Rev Rheumatol 2012; 8(7): 412-19. Hare K B, Lohmander L S, Kise N J, Risberg M A, Roos E M. Middle-aged patients with an MRI-verified medial meniscal tear report symptoms commonly associated with knee osteoarthritis: a cross sectional study of 199 patients. Acta Orthop 2017; 88(6): 664-9. Herrlin S, Hallander M, Wange P, Weidenhielm L, Werner S. Arthroscopic or conservative treatment of degenerative medial meniscal tears: a prospective randomised trial. Knee Surg Sports Traumatol Arthrosc 2007; 15(4): 393-401. Jevsevar D S, Yates Jr A J, Sanders J O. Arthroscopic partial meniscectomy for degenerative meniscal tear. N Engl J Med 2014; 370(13): 1260. Katz J N, Brophy R H, Chaisson C E, de Chaves L, Cole B J, Dahm D L, Donnell-Fink L A, Guermazi A, Haas A K, Jones M H, Levy B A, Mandl L A, Martin S D, Marx R G, Miniaci A, Matava M J, Palmisano J, Reinke E K, Richardson B E, Rome B N, Safran-Norton C E, Skoniecki D J, Solomon D H, Smith M V, Spindler K P, Stuart M J, Wright J, Wright R W, Losina E. Surgery versus physical therapy for a meniscal tear and osteoarthritis. N Engl J Med 2013; 368(18): 1675-84. Khan M, Evaniew N, Bedi A, Ayeni O R, Bhandari M. Arthroscopic surgery for degenerative tears of the meniscus: a systematic review and meta-analysis. CMAJ 2014; 186(14): 1057-64. Krych A J, Carey J L, Marx R G, Dahm D L, Sennett B J, Stuart M J, Levy B A. Does arthroscopic knee surgery work? Arthroscopy 2014; 30(5): 544-5. Lyman S, Oh L S, Reinhardt K R, Mandl L A, Katz J N, Levy B A, Marx R G. Surgical decision making for arthroscopic partial meniscectomy in patients aged over 40 years. Arthroscopy 2012; 28(4): 492-501.e491. Niu N N, Losina E, Martin S D, Wright J, Solomon D H, Katz J N. Development and preliminary validation of a meniscal symptom index. Arthritis Care Res(Hoboken) 2011; 63(2): 208-15. Peat G, Bergknut C, Frobell R, Joud A, Englund M. Population-wide incidence estimates for soft tissue knee injuries presenting to healthcare in southern Sweden: data from the Skane Healthcare Register. Arthritis Res Ther 2014; 16(4): R162.

Acta thanks Martin Englund and Rudolf Poolman for help with peer review of this study.

12024 Skou D.indd 343

5/3/2018 11:48:03 AM

344

Pihl K, Englund M, Lohmander L S, Jorgensen U, Nissen N, Schjerning J, Thorlund J B. Signs of knee osteoarthritis common in 620 patients undergoing arthroscopic surgery for meniscal tear. Acta Orthop 2017; 88(1): 90-5. Poehling G G, Ruch D S, Chabon S J. The landscape of meniscal injuries. Clin Sports Med 1990; 9(3): 539-49. Roos E M, Roos H P, Lohmander L S, Ekdahl C, Beynnon B D. Knee Injury and Osteoarthritis Outcome Score (KOOS): development of a self-administered outcome measure. J Orthop Sports Phys Ther 1998; 28(2): 88-96. Siemieniuk R A C, Harris I A, Agoritsas T, Poolman R W, BrignardelloPetersen R, Van de Velde S, Buchbinder R, Englund M, Lytvyn L, Quinlan C, Helsingen L, Knutsen G, Olsen N R, Macdonald H, Hailey L, Wilson H M, Lydiatt A, Kristiansen A. Arthroscopic surgery for degenerative knee arthritis and meniscal tears: a clinical practice guideline. BMJ 2017; 357: j1982. Sihvonen R, Englund M, Turkiewicz A. Jarvinen T L. Mechanical symptoms and arthroscopic partial meniscectomy in patients with degenerative meniscus tear: a secondary analysis of a randomized trial. Ann Intern Med 2016a; 164(7): 449-55. Sihvonen R, Englund M, Turkiewicz A. Jarvinen TL. Mechanical symptoms as an indication for knee arthroscopy in patients with degenerative meniscus tear: a prospective cohort study Osteoarthritis Cartilage 2016b; 24(8): 1367-75. Sihvonen R, Paavola M, Malmivaara A, Itala A, Joukainen A, Nurmi H, Kalske J, Ikonen A, Jarvela T, Jarvinen T A, Kanto K, Karhunen J, Knifsund J, Kroger H, Kaariainen T, Lehtinen J, Nyrhinen J, Paloneva J, Paivaniemi O, Raivio M, Sahlman J, Sarvilinna R, Tukiainen S, Valimaki V V, Aarimaa V, Toivonen P, Jarvinen T L. Arthroscopic partial meniscectomy versus placebo surgery for a degenerative meniscus tear: a 2-year follow-up of the randomised controlled trial. Ann Rheum Dis 2018; 77(2): 188-95. Skou S T, Wrigley T V, Metcalf B R, Hinman R S, Bennell K L. Association of knee confidence with pain, knee instability, muscle strength, dynamic varus–valgus joint motion in knee osteoarthritis. Arthritis Care Res 2014; 66(5): 695-701. Skou S T, Lind M, Hölmich P, Jensen H P, Jensen C, Afzal M, Jørgensen U, Thorlund J B. Study protocol for a randomised controlled trial of meniscal surgery compared with exercise and patient education for treatment of meniscal tears in young adults. BMJ Open 2017; 7(8): e017436.

12024 Skou D.indd 344

Acta Orthopaedica 2018; 89 (3): 336–344

Smith B E, Thacker D, Crewesmith A, Hall M. Special tests for assessing meniscal tears within the knee: a systematic review and meta-analysis. Evidence Based Med 2015; 20(3): 88-97. Stuart M J, Lubowitz J H. What, if any, are the indications for arthroscopic debridement of the osteoarthritic knee? Arthroscopy 2006; 22(3): 238-9. Thorlund J B, Christensen R, Nissen N, Jorgensen U, Schjerning J, Porneki J C, Englund M, Lohmander L S. Knee Arthroscopy Cohort Southern Denmark (KACS): protocol for a prospective cohort study. BMJ Open 2013; 3(10): e003399. Thorlund J B, Hare K B, Lohmander L S. Large increase in arthroscopic meniscus surgery in the middle-aged and older population in Denmark from 2000 to 2011. Acta Orthop 2014; 85(3): 287-92. Thorlund J B, Juhl C B, Roos E M, Lohmander L S. Arthroscopic surgery for degenerative knee: systematic review and meta-analysis of benefits and harms BMJ (Clin Res ed.) 2015; 350: h2747. Thorlund J B, Englund M, Christensen R, Nissen N, Pihl K, Jorgensen U, Schjerning J, Lohmander L S. Patient reported outcomes in patients undergoing arthroscopic partial meniscectomy for traumatic or degenerative meniscal tears: comparative prospective cohort study. BMJ 2017; 356: j356. Yan R, Wang H, Yang Z, Ji Z H, Guo Y M. Predicted probability of meniscus tears: comparing history and physical examination with MRI. Swiss Med Wkly 2011; 141: w13314. Zhang W, Doherty M, Peat G, Bierma-Zeinstra M A, Arden N K, Bresnihan B, Herrero-Beaumont G, Kirschner S, Leeb B F, Lohmander L S, Mazieres B, Pavelka K, Punzi L, So A K, Tuncer T, Watt I, Bijlsma J W. EULAR evidence-based recommendations for the diagnosis of knee osteoarthritis. Ann Rheum Dis 2010; 69(3): 483-9. Zou K, Wong J, Abdullah N, Chen X, Smith T, Doherty M, Zhang W. Examination of overall treatment effect and the proportion attributable to contextual effect in osteoarthritis: meta-analysis of randomised controlled trials. Ann Rheum Dis 2016; 75(11): 7.

5/3/2018 11:48:03 AM

Acta Orthopaedica 2018; 89 (3): 345–350

345

High risk for revision after shoulder arthroplasty for failed osteosynthesis of proximal humeral fractures A matched pair analysis of 285 cases from the Danish Shoulder Arthroplasty Registry Marc Randall KRISTENSEN 1, Jeppe Vejlgaard RASMUSSEN 1, Brian ELMENGAARD 2, Steen Lund JENSEN 3, Bo Sanderhoff OLSEN 1, and Stig BRORSON 4

1 Department of Orthopaedic Surgery, Herlev Hospital, University of Copenhagen, Herlev, Denmark; 2 Department of Orthopaedic Surgery, Aarhus University Hospital, Aarhus, Denmark; 3 Department of Orthopaedic Surgery, Aalborg University Hospital, Aalborg, Denmark; 4 Department of Orthopaedic Surgery, Zealand University Hospital, University of Copenhagen, Køge, Denmark Correspondence: marckristensen@hotmail.com Submitted 2017-07-03. Accepted 2018-01-25.

Background and purpose — It is unclear whether previous osteosynthesis is a risk factor for inferior outcome following shoulder arthroplasty for a proximal humeral fracture. We used data from the Danish Shoulder Arthroplasty Registry (DSR) to examine this question. Patients and methods — All 285 patients treated with a shoulder arthroplasty after failed osteosynthesis of a proximal humeral fracture reported to DSR from 2006 to 2013 were included. Each case was matched with 2 controls (570) treated with a primary shoulder arthroplasty for an acute proximal humeral fracture. Patient reported outcome was assessed using the Western Ontario Osteoarthritis of the Shoulder index (WOOS) and the relative risk of revision was reported. Results — The mean WOOS was 46 (SD 25) for a shoulder arthroplasty after failed osteosynthesis and 52 (27) after a primary shoulder arthroplasty. The relative risk of revision for a shoulder arthroplasty after failed osteosynthesis was 2 with a primary arthroplasty for fracture as reference. In a separate analysis of patients treated by locking plate the mean WOOS was 46 (24), with a relative risk of revision at 1.5 with a primary arthroplasty as reference. Interpretation — Compared with primary arthroplasty for proximal humeral fracture, we found an inferior patient-reported outcome and a substantial risk of revision for patients treated with a shoulder arthroplasty after failed osteosynthesis for a proximal humeral fracture. The risk and burdens of additional surgery should be accounted for when deciding on the primary surgical procedure. ■

Treatment modalities for a proximal humeral fracture vary from nonsurgical treatment including immobilization and physiotherapy to surgical treatment including operative fixation with head-preserving techniques or a primary shoulder arthroplasty. Locking-plate osteosynthesis has been popular in recent years (Jost et al. 2013, Sun et al. 2013), but many complications and reoperations have been reported (Clavert et al. 2010, Sproul et al. 2011), especially in complex fracture patterns (Brorson et al. 2011, 2012). The most commonly reported complications are avascular necrosis of the humeral head, screw penetration, and glenoid destruction (Spross et al. 2012, Brorson 2013, Jost et al. 2013). The reported reoperation rate of proximal humeral fractures treated with locking plates varies in the literature, ranging from 3% to 44% (Bjorkenheim et al. 2004, Clavert et al. 2010, Brorson et al. 2011, 2012, Spross et al. 2012). Recently, a large multicenter study found no better outcome after surgery compared with nonsurgical treatment (Rangan et al. 2015). Shoulder arthroplasty is often offered when revision of a failed osteosynthesis is required. Shoulder arthroplasty entails a risk of complications too, including persistent pain, infection, instability, neurologic injury, tuberosity migration and vanishing, rotator cuff tear, heterotopic ossification, glenoid erosion, stiffness, and periprosthetic fractures, which may eventually lead to revision surgery (Plausinis et al. 2005, Kontakis et al. 2008). It has been reported that the patient-reported outcome measurements (PROM) after early shoulder arthroplasty are significantly better than after late shoulder arthroplasty (Bosch et al. 1998). However, it has also been reported that improved outcome is achieved following treatment of fracture sequelae

11872 Kristensen D.indd 345

5/3/2018 11:49:03 AM

346

Acta Orthopaedica 2018; 89 (3): 345–350

by shoulder arthroplasty (Boileau et al. 2001, Jost et al. 2013, Alentorn-Geli et al. 2014). To our knowledge, only studies with a small number of patients have reported whether previous osteosynthesis is a risk factor for an inferior outcome of shoulder arthroplasty for proximal humeral fractures (Hussey et al. 2015, Dezfuli et al. 2016, Grubhofer et al. 2017). We hypothesized that previous osteosynthesis is a risk factor for inferior outcome following shoulder arthroplasty for a proximal humeral fracture. Thus, we evaluated patientreported outcome and risk of revision after arthroplasty for proximal humeral fracture.

Patients and methods Data were obtained from the Danish Shoulder Arthroplasty Registry (DSR). The DSR was established in 2004 with the purpose of systematically collecting data on all primary and revision arthroplasties on a national level. The registry is independent of commercial interests and financed by the Danish counties. Since 2006, reporting has been mandatory for all Danish hospitals and private clinics performing shoulder arthroplasty surgery and the completeness of reporting has been above 92% (Jensen et al. 2016). Data are reported online by the surgeon at the time of operation. A revision is defined as removal or exchange of the humeral component or the addition of a glenoid component (Rasmussen et al. 2012, 2014). Patients are identified in the register by their unique civil registration number (CPR). Thus, a revision arthroplasty is linked to the primary arthroplasty using the CPR. Patient-reported data are collected 10–14 months postoperatively for both primary prostheses and revision prostheses by mail, using the Western Ontario Osteoarthritis of the Shoulder (WOOS) index. WOOS is a patient-administrated questionnaire with 19 questions regarding the shoulder-related quality of life, resulting in a percentage score with 100 being the best. To improve the response rate, we sent a single reminder to non-responders and to those returning an incomplete questionnaire. If a patient dies or the shoulder arthroplasty is revised within 1 year postoperatively WOOS cannot be obtained. All patients reported to DSR from 2006 to 2013 with a shoulder arthroplasty after failed osteosynthesis of a proximal humeral fracture were identified. For comparison, we chose a group of patients with an acute proximal humeral fracture treated with a primary shoulder arthroplasty and reported to the DSR in the same time period (2006–2013). We matched each patient with a shoulder arthroplasty after a failed osteosynthesis with 2 patients treated by a primary shoulder arthroplasty. They were selected to match in terms of sex, age at the time of the arthroplasty operation, and whether the WOOS questionnaire was completed or not, resulting in the same WOOS response rate in the 2 groups. If more than 2 controls matched all three parameters we used those with a date of birth closest to that of the case. For patients younger than 50 or older than

11872 Kristensen D.indd 346

Shoulder arthroplasties after failed osteosynthesis reported to the Danish Shoulder Arthroplasty Registry 2006–2013 n = 299 Excluded (n = 14): – duplicate reports, 4 – misreported, 8 – missing CPR number, 1 – no matching control, 1 Shoulder arthroplasties included in the study n = 285

Figure 1. Flowchart of included cases. 299 were reported to the registry, 285 were included in this study.

90, the number of eligible patients with a primary shoulder arthroplasty were low. For these patients, we chose to match in age groups (21–30, 31–40, 41–50 and 91–100). 1 patient was reported to DSR with bilateral surgery and included as two independent cases. Statistics Data were analyzed using a chi-square test for binary outcomes and Student’s t-test for continuous outcomes (Lumley et al. 2002). A Cox regression model was used to calculate the relative risk of revision. Sex and age, as a binary variable with 65 years as the threshold, were included in the model. The Kaplan–Meier method was used to illustrate the unadjusted survival rates. SPSS was used for the statistical analysis (IBM Corp, Armonk, NY, USA). The level of statistical significance was set at p < 0.05 and p-values were 2-tailed. Ethics, funding, and potential conflicts of interest The Danish Data Protection Agency (journal number: HEH2015-008, date of issue February 4, 2015) and the Danish Health Authority (journal number: 3-3013-1075/1/, date of issue August 10, 2015) approved the study. No funding was recieved for this study. No conflicts of interest were declared.

Results Demographics Between 2006 and 2013, 299 shoulders were reported to DSR as having a shoulder arthroplasty after failed osteosynthesis of a proximal humeral fracture. 14 cases were excluded, leaving 285 cases and 570 controls (Figure 1). The data were extracted from the registry in March 2015. The follow-up time of the patients was 36 months (SD 27) among the patients with a secondary arthroplasty and 39 months (SD 25) among patients with a primary arthroplasty (p = 0.07).

5/3/2018 11:49:03 AM

Acta Orthopaedica 2018; 89 (3): 345â&#x20AC;&#x201C;350

Table 1. Type of osteosynthesis in 285 cases n Locking plate 208 K-wires 17 Non-locking plate 15 Screws only 10 Intramedullary nail 9 Helix wire 4 Intramedullary nail (thin) 3 Missing 19

% 73.0 6.0 5.3 3.5 3.2 1.4 1.1 6.7

347

Table 2. Mode of osteosynthesis failure and outcome

Mode

Patient-reported outcome 21 patients died and 5 patients were revised within 1 year. Of the remaining 259 patients 77% responded to the WOOS questionnaire. The mean WOOS of patients with a shoulder arthroplasty after failed osteosynthesis was 46 (SD 25). The mean WOOS of patients with a primary shoulder arthroplasty was 52 (27) (p = 0.005). The mean WOOS of the 2 groups was similar, except for the patients revised because of pain or infection, who had a mean WOOS score of 32. However, the numbers were too small for meaningful statistical comparison (Table 2). A subanalysis of the results after treatment with the different kind of prostheses in our study gives an average WOOS of 46 for both stemmed hemiarthroplasty and reverse arthroplasty. Risk of revision 31 (11%) of the 285 patients with a secondary shoulder arthroplasty after a failed osteosynthesis had a revision of the arthroplasty. The most common reasons for revision arthroplasty were dislocation, rotator cuff tears, and infection (Table 3). In

11872 Kristensen D.indd 347

Mean Revisions n (%) WOOS n (%)

Avascular necrosis 107 (38) Fracture displacement 49 (17) Pseudarthrosis/non-union 36 (13) Secondary arthrosis 12 (4) Infection 9 (3) Screw penetration 9 (3) Pain 8 (3) Rotator cuff tear 6 (2) Implant failure 6 (2) New fracture 2 (1) Missing 41 (14)

Mean age of patients with a secondary arthroplasty was 66 years (SD 12) and 213 (75%) patients were women. A locking plate was used in 73% of the patients (Table 1). In 7% of the patients it was not possible to find information regarding the type of osteosynthesis. In 39% of the patients avascular necrosis was registered as the reason for failure leading to shoulder arthroplasty. Fracture displacement (17%) and pseudarthrosis (13%) were other common reasons for revision (Table 2). 68% of the patients were treated with a stemmed hemiarthroplasty and 23% with a reverse shoulder arthroplasty. The remaining 9% were reoperated by a resurfacing prosthesis, total anatomical arthroplasty, or a bipolar shoulder arthroplasty. In the group of patients with a primary shoulder arthroplasty 94% were treated with a stemmed hemiarthroplasty. The remaining 6% were treated with a reverse shoulder arthroplasty, resurfacing prosthesis, or total anatomical arthroplasty.

Table 3. Causes of revision in cases (secondary arthroplasties) and controls (primary arthroplasties)

46 45 49 46 32 51 32 48 50 41 47

9 (29) 6 (19) 7 (23) 2 (6) 3 (10) 0 (0) 1 (3) 0 (0) 0 (0) 0 (0) 3 (10)

Cause Dislocation Rotator cuff tears Infection Pain Other Glenoid erosion Technical failure Loosening Missing

Cases n = 31

Controls n = 34

8 5 5 5 2 1 2 1 2

8 10 4 1 5 3 1 1 1

Cumulative implant survival 1.00

0.95

0.90

0.85

0.80

0.75

0.70 0

100

Months from surgery

Figure 2. Implant survival functions of primary arthroplasty (green) and arthroplasty after failed osteosynthesis (blue).

the group of patients treated with a primary shoulder arthroplasty 34 (6%) were revised. Overall, the revision rate due to infection was 1.8% (5 revisions) among the patients with a secondary arthroplasty and 0.7% (4 revisions) among patients with a primary arthroplasty. The number of patients is too small for meaningful statistical comparison. The mean time to revision in the group of patients with a shoulder arthroplasty after a primary failed osteosynthesis was 59 (27) months and 53 (24) months in the group of patients with a primary shoulder arthroplasty (p = 0.4) (Figure 2). The relative risk of revision was 2.0 (95% CI 1.2â&#x20AC;&#x201C;3.2) for an arthroplasty after failed osteosynthesis with a primary shoulder arthroplasty as reference. Subgroup: locking plate A locking plate was used in 208 (73%) of the patients with a primary osteosynthesis. The mean age in this subgroup was 66 (12) years and 155 (75%) patients were women. Avas-

5/3/2018 11:49:04 AM

348

Acta Orthopaedica 2018; 89 (3): 345–350

Table 4. Mode of failure of locking plates and outcome

Cumulative implant survival 1.00

Mode

n (%)

Mean WOOS

Revisions n (%)

Avascular necrosis Fracture displacement Pseudarthrosis/non-union Secondary arthrosis Infection Screw penetration Pain Rotator cuff tear Implant failure New fracture Missing

82 (39) 41 (20) 27 (13) 4 (2) 5 (2) 8 (4) 8 (4) 4 (2) 6 (3) 0 (0) 23 (11)

43 42 53 22 30 45 30 52 46 – 47

6 (33) 6 (33) 4 (22) 0 (0) 1 (6) 0 (0) 1 (6) 0 (0) 0 (0) 0 (0) 0 (0)

0.95

0.90

0.85

0.80

0.75

0.70 0

100

Months from surgery Figure 3. Implant survival functions of primary arthroplasty (green) and arthroplasty after failed locking plates (blue).

cular necrosis (39%) and fracture displacement (20%) were the major reasons for failure leading to shoulder arthroplasty (Table 4). The mean WOOS of patients with a shoulder arthroplasty after failed locking plate was 46 (24). The relative risk of revision was 1.5 (95% CI 0.8–2.6) for an arthroplasty after a failed locking plate with a primary shoulder arthroplasty as reference (Figure 3).

Discussion We found a difference in WOOS of 6 points and a relative risk of revision of 2.0 for an arthroplasty after failed osteosynthesis with a primary shoulder arthroplasty as reference. No value for minimal important change (MIC) regarding WOOS score in patients treated with shoulder arthroplasty has been reported. A study determining the MIC of 4 shoulderspecific PROMs (Simple Shoulder Test, Disabilities of the Arm, Shoulder and Hand (DASH and Quick DASH), and the Oxford Shoulder Score) found MIC values of 2.2, 12, 13, and 6 respectively (van Kampen et al. 2013), equivalent to 10–18% of the total score. Therefore, it is questionable whether the difference of 6 points in WOOS should be regarded as clinically relevant. A locking plate was the most common type of osteosynthesis (73%). The subgroup of patients treated with this kind of osteosynthesis was comparable to patients treated by all types of osteosynthesis, regarding demographics and patientreported outcome. However, the risk of revision among the patients treated with shoulder arthroplasty after failed locking plate was 1.5, indicating a lower revision rate among patients treated with a locking plate, compared with osteosynthesis in general. The response rate of WOOS was 77% for those patients who were available for follow-up. A systematic review of postal self-administrated questionnaires used in health research

11872 Kristensen D.indd 348

reported an average response rate of 65% (Nakash et al. 2006). The Danish version of WOOS has been translated and cross-culturally adapted (Guillemin et al. 1993) and validated using classical test theory (Rasmussen et al. 2013), but only for patients with osteoarthritis treated with a shoulder arthroplasty. When considering a revision arthroplasty several factors may influence the decision-making process, including patient age, co-morbidity, functional level, experience of the surgeon, the resources available, and the ability of the implant to be revised with an adequate outcome (Rasmussen et al. 2012). Thus, the indication for revision arthroplasty is not fully known and revision rates do not necessarily reflect clinical outcome. The risk and burdens of additional surgery should be accounted for when deciding on the primary surgical procedure. Several studies have analyzed possible risk factors for failure of osteosynthesis of proximal humeral fractures. Petrigliano et al. (2014) reported an age between 50 and 64 as the only significant factor for revision arthroplasty after primary osteosynthesis, whereas Hardeman et al. (2012) reported AO type C fractures as the only significant risk factor. Spross et al. (2012) reported that heavy smokers and fracture-dislocations had a significant increased risk of complications. We used a matched group of patients with a primary shoulder arthroplasty to have a reasonable understanding of what surgeons can expect after a primary shoulder arthroplasty. However, the groups were not matched by co-morbidity, arthroplasty type, fracture complexity, or other factors with a possible influence on outcome. Furthermore, the group of patients with a secondary shoulder arthroplasty represents only patients with failure after primary surgery. This could further contribute to the risk for poor outcome if the 2 groups are not comparable biologically, or they may have a different injury pattern. However, we were unable to correct for this in our study design.

5/3/2018 11:49:04 AM

Acta Orthopaedica 2018; 89 (3): 345–350

We only report postoperative outcome. Thus, we are not able to report any change in outcome related to the arthroplasty procedure. Only a few small studies have reported the outcome after secondary treatment of primary failed osteosynthesis of a proximal humeral fracture with a shoulder arthroplasty. Hussey et al. (2015), Dezfuli et al. (2016), and Grubhofer et al. (2017) only reported reoperations after primary failed osteosyntheses with a reverse shoulder arthroplasty. We included reoperations with all kind of prostheses, but a subgroup analysis showed an average WOOS of 46 for both stemmed hemiarthroplasty and reverse shoulder arthroplasty. In the studies by Hussey et al. (2015), Dezfuli et al. (2016), and Grubhofer et al. (2017) the clinical outcome was reported to be acceptable, but it was not possible for them to report the risk of revision because of small sample size (11–44 patients) compared with the 285 patients included in our study. Our large number of patients enabled a separate analysis on reasons for osteosynthesis failure. It seems likely patients who had a revision because of an infection had a markedly worse mean WOOS at 32 points and among these patients no less than one-third had a new revision arthroplasty; however, the number of patients was too small for a meaningful statistical comparison. Finally, it is important to stress that this study does not report or discuss the results of either osteosynthesis or other treatment modalities of failed osteosynthesis in general. In summary, compared with primary arthroplasty for proximal humeral fracture, we found an inferior patient-reported outcome and a substantial risk of revision for patients treated with a shoulder arthroplasty after failed osteosynthesis of a proximal humeral fracture. The risk and burdens of additional surgery should be accounted for when deciding on the primary surgical procedure. More knowledge on patient-reported outcome after non-revised osteosynthesis and nonsurgical approaches is needed to inform future evidence-based decision-making.

MK: study design, data collection, data analysis, writing of the draft paper, and revision of the paper. JR, SB: study design, data analysis, and revision of the paper. BO: study design and revision of the paper. BE, SJ: data collection and revision of the paper. Acta thanks Carl Ekholm and Christian Gerber for help with peer review of this study. Alentorn-Geli E, Guirro P, Santana F, Torrens C. Treatment of fracture sequelae of the proximal humerus: comparison of hemiarthroplasty and reverse total shoulder arthroplasty. Arch Orthop Trauma Surg 2014; 134(11): 1545-50. Bjorkenheim J M, Pajarinen J, Savolainen V. Internal fixation of proximal humeral fractures with a locking compression plate: a retrospective evaluation of 72 patients followed for a minimum of 1 year. Acta Orthop Scand 2004; 75(6): 741-5. Boileau P, Trojani C, Walch G, Krishnan S G, Romeo A, Sinnerton R. Shoulder arthroplasty for the treatment of the sequelae of fractures of the proximal humerus. J Shoulder Elbow Surg 2001; 10(4): 299-308.

11872 Kristensen D.indd 349

349

Bosch U, Skutek M, Fremerey R W, Tscherne H. Outcome after primary and secondary hemiarthroplasty in elderly patients with fractures of the proximal humerus. J Shoulder Elbow Surg 1998; 7(5): 479-84. Brorson S. Fractures of the proximal humerus. Acta Orthop 2013; 84 (Suppl. 351): 1-32. Brorson S, Frich L H, Winther A, Hrobjartsson A. Locking plate osteosynthesis in displaced 4-part fractures of the proximal humerus. Acta Orthop 2011; 82(4): 475-81. Brorson S, Rasmussen J V, Frich L H, Olsen B S, Hrobjartsson A. Benefits and harms of locking plate osteosynthesis in intraarticular (OTA Type C) fractures of the proximal humerus: a systematic review. Injury 2012; 43(7): 999-1005. Clavert P, Adam P, Bevort A, Bonnomet F, Kempf JF. Pitfalls and complications with locking plate for proximal humerus fracture. J Shoulder Elbow Surg 2010; 19(4): 489-94. Dezfuli B, King J J, Farmer K W, Struk A M, Wright T W. Outcomes of reverse total shoulder arthroplasty as primary versus revision procedure for proximal humerus fractures. J Shoulder Elbow Surg 2016; 25(7): 1133-7. Grubhofer F, Wieser K, Meyer D C, Catanzaro S, Schurholz K, Gerber C. Reverse total shoulder arthroplasty for failed open reduction and internal fixation of fractures of the proximal humerus. J Shoulder Elbow Surg 2017; 26(1): 92-100. Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of healthrelated quality of life measures: literature review and proposed guidelines. J Clin Epidemiol 1993; 46(12): 1417-32. Hardeman F, Bollars P, Donnelly M, Bellemans J, Nijs S. Predictive factors for functional outcome and failure in angular stable osteosynthesis of the proximal humerus. Injury 2012; 43(2): 153-8. Hussey M M, Hussey S E, Mighell M A. Reverse shoulder arthroplasty as a salvage procedure after failed internal fixation of fractures of the proximal humerus: outcomes and complications. Bone Joint J 2015; 97-B(7): 96772. Jensen S L, Pedersen A B, Hjelm A H. Danish Shoulder Arthroplasty Registry—Annual Report. 2016. Available from: http://www.sunhed.dk (in Danish). Jost B, Spross C, Grehn H, Gerber C. Locking plate fixation of fractures of the proximal humerus: analysis of complications, revision strategies and outcome. J Shoulder Elbow Surg 2013; 22(4): 542-9. Kontakis G, Koutras C, Tosounidis T, Giannoudis P. Early management of proximal humeral fractures with hemiarthroplasty: a systematic review. J Bone Joint Surg Br 2008; 90(11): 1407-13. Lumley T, Diehr P, Emerson S, Chen L. The importance of the normality assumption in large public health data sets. Annu Rev Public Health 2002; 23: 151-69. Nakash R A, Hutton J L, Jorstad-Stein E C, Gates S, Lamb S E. Maximising response to postal questionnaires: a systematic review of randomised trials in health research. BMC Med Res Methodol 2006; 6: 5. Petrigliano F A, Bezrukov N, Gamradt S C, SooHoo N F. Factors predicting complication and reoperation rates following surgical fixation of proximal humeral fractures. J Bone Joint Surg Am 2014; 96(18): 1544-51. Plausinis D, Kwon Y W, Zuckerman J D. Complications of humeral head replacement for proximal humeral fractures. Instr Course Lect 2005; 54: 371-80. Rangan A, Handoll H, Brealey S, Jefferson L, Keding A, Martin B C, Goodchild L, Chuang L H, Hewitt C, Torgerson D. Surgical vs nonsurgical treatment of adults with displaced fractures of the proximal humerus: the PROFHER randomized clinical trial. JAMA 2015; 313(10): 1037-47. Rasmussen J V, Olsen B S, Fevang B T, Furnes O, Skytta E T, Rahme H, Salomonsson B, Mohammed K D, Page R S, Carr A J. A review of national shoulder and elbow joint replacement registries. J Shoulder Elbow Surg 2012; 21(10): 1328-35. Rasmussen J V, Jakobsen J, Olsen B S, Brorson S. Translation and validation of the Western Ontario Osteoarthritis of the Shoulder (WOOS) index—the Danish version. Patient Relat Outcome Meas 2013; 4: 49-54.

5/3/2018 11:49:08 AM

350

Rasmussen J V, Polk A, Sorensen A K, Olsen B S, Brorson S. Outcome, revision rate and indication for revision following resurfacing hemiarthroplasty for osteoarthritis of the shoulder: 837 operations reported to the Danish Shoulder Arthroplasty Registry. Bone Joint J 2014; 96-B(4): 519-25. Spross C, Platz A, Rufibach K, Lattmann T, Forberger J, Dietrich M. The PHILOS plate for proximal humeral fractures: risk factors for complications at one year. J Trauma Acute Care Surg 2012; 72(3): 783-92. Sproul R C, Iyengar J J, Devcic Z, Feeley B T. A systematic review of locking plate fixation of proximal humerus fractures. Injury 2011; 42(4): 408-13.

11872 Kristensen D.indd 350

Acta Orthopaedica 2018; 89 (3): 345â&#x20AC;&#x201C;350

Sun J C, Li Y L, Ning G Z, Wu Q, Feng S Q. Treatment of three- and four-part proximal humeral fractures with locking proximal humerus plate. Eur J Orthop Surg Traumatol 2013; 23(6): 699-704. van Kampen D A, Willems W J, van Beers L W, Castelein R M, Scholtes V A, Terwee C B. Determination and comparison of the smallest detectable change (SDC) and the minimal important change (MIC) of four-shoulder patient-reported outcome measures (PROMs). J Orthop Surg Res 2013; 8: 40.

5/3/2018 11:49:08 AM

Acta Orthopaedica 2018; 89 (3): 351–356

351

9 years’ follow-up of 168 pin-fixed supracondylar humerus fractures in children Noora TUOMILEHTO 1, Antti SOMMARHEM 2, and Aarno Y NIETOSVAARA 2

1 Department of Orthopedics and Traumatology, Helsinki University; 2 Children’s Hospital, Helsinki University Central Hospital, Helsinki, Finland Correspondence: noora.tuomilehto@hus.fi Submitted 2017-11-03. Accepted 2018-01-17.

Background and purpose — The long-term outcome of pin-fixed supracondylar humerus fractures (SCHF) in children is not well known. We assessed the 7- to 12-year outcome in 168 children. Patients and methods — During 2002–2006, 210 domestic children (age 7 (1–14) years) with SCHF (Gartland III 79%, Gartland II 19%, and flexion type 2%) were pin fixed in Helsinki. 36 (17%) patients had a nerve palsy. Radiographic alignment was regarded as satisfactory in 81% of patients (Baumann angle (BA) within ±10˚ of normal range and whose anterior humeral line (AHL) crossed the capitulum). After a mean follow-up of 9 (7–12) years, 168 (80%) patients answered a questionnaire regarding elbow appearance (scale 0–10), function (scale 0–10), and pain (scale 0–10), and symmetry of range of motion (ROM) and carrying angle (CA). 65 (31%) patients also attended a clinical followup examination. Results — Mean subjective score for appearance was 8.7 (2–10) and for function 9.0 (2–10) (n = 168). Elbow ROM asymmetry was experienced by 28% and elbow CA asymmetry by 17% of the patients. Elbow pain was reported by 14%, and was more common in children with nerve injuries. Long-term outcome was good or excellent in 60/65 and CA in 56/65 of the follow-up visit patients using Flynn’s criteria. BA exceeding normal values by 10˚ was associated with lower subjective outcome; AHL crossing point with the capitulum was not associated with outcome. Interpretation — Long-term subjective outcome is satisfactory with few exceptions if elbow ROM and CA are restored within 10° of the uninjured elbow. Radiographs at fracture union have little prognostic value. Nerve injuries can cause long-term pain. ■

Displaced supracondylar humerus fractures in children (SCHF) have a high risk of fracture or treatment-related complications. Displaced SCHF is best treated with internal fixation and is thus the most common operatively treated fracture type in children (Cheng et al. 1999, Omid et al. 2008). Quality of SCHF treatment correlates with the experience of the treating institution: unsatisfactory standard of reduction and pin fixation are the most common reasons for treatment-related nerve injuries and malunion in displaced SCHF (Vallila et al. 2015). The risk of Volkmann’s ischemic contracture in displaced SCHF has fallen to nearly zero with internal fixation and adequate assessment of peripheral circulation and postoperative pain (Bashyal et al. 2009, Scannell et al. 2013). Displaced SCHFs carry the highest risk of nerve injury (up to 11%) among all pediatric fractures (Omid et al. 2008, Bashyal et al. 2009, Babal et al. 2010). Most nerve injuries are caused by the fracture itself, but up to a 4% incidence of iatrogenic injuries related to pin fixation have been reported (Gosens and Bongers 2003, Bashyal et al. 2009, Babal et al. 2010). The majority of nerve injuries associated with SCHF appear to resolve spontaneously within a few months (Dormans et al. 1995, Gosens and Bongers 2003, Ramachandran et al. 2006, Guner et al. 2013). Quality of reduction has traditionally been assessed by radiographs. Frontal alignment can be evaluated by measuring the Baumann angle (BA), which is the angle between the long axis of the humeral shaft and the growth plate of the lateral humeral condyle with reported normal values between 64° and 81° (Williamson et al. 1992, Dai 1999, Shank et al. 2011, Flynn et al. 2015). The most common way to register sagittal alignment is to record whether the anterior humeral line

12269 Tuomilehto D.indd 351

5/3/2018 11:50:12 AM

352

Acta Orthopaedica 2018; 89 (3): 351–356

14.1) years. Trauma mechanisms were falling from height in 53%, sporting activities in 30% and falling on the level in 15%. 10 Excluded: (5 %) of the fractures were open fractures. Living abroad The main surgeon was a consultant in 60 n=3 (29%) cases (8 consultants, mean number of operations per surgeon 8 (2–19)) and a regisGartland II (n = 40): Gartland III (n = 166): Flexion type (n = 4): trar in 150 (71%) cases (31 registrars, mean – questionnaire (Q) only, 20 – questionnaire (Q) only, 82 – questionnaire (Q) only, 1 – control visit and Q, 10 – control visit and Q, 53 – control visit and Q, 2 number of operations per surgeon 5 (1–20)). – no follow-up, 10 – no follow-up, 31 – no follow-up, 1 Registrars performed 59 (28%) operations alone without consultant supervision. Figure 1. Operatively treated supracondylar humerus fractures 2002–2006. Patients’ par14 (7%) patients had open reduction (anteticipation in the study and fracture classification by Gartland. rior 11, lateral 2, medial and lateral approach (AHL) passes through the anterior or middle third of the ossi- 1). Osteosynthesis was performed through a small skin incification center of the capitulum (Herman et al. 2009, Flynn et sion percutaneously with crossed pins in 200 patients and with al. 2015). Reliability of these radiographic indexes has been laterals pins in 10 patients. The brachial artery of 1 patient questioned (Silva et al. 2010), and it has been suggested that was repaired. Volkmann’s contracture did not develop in any it is better to record the quality of reduction peroperatively by of our patients. 2 patients were primary operated by a registrar clinical comparison with the healthy side (Simanovsky et al. alone and reoperated within a week because of unsatisfactory 2007, Tuomilehto et al. 2016). primary reduction. 3 patients had their primary operation elseThe remodeling capacity of the distal humerus is limited where and were reoperated in our institution within a week (Otsuka and Kasser 1997, Omid et al. 2008, Flynn et al. 2015). because of unsatisfactory primary reduction. 1 patient had a Malunion in the frontal plane has been considered predomi- deep infection and an additional 7 patients (together 4%) had nantly as a cosmetic disability, although elbow pain and dys- a superficial pin-track infection. Corrective osteotomy was function as well as an increased risk of lateral humeral con- performed in 2 patients (primarily operated by a registrar) due dyle fractures have been reported (Guven et al. 2015). Sagit- to gun-stock deformity after 3 and 8 years from the fracture. tal plane malunion can lead to permanent changes in elbow 36 (22%) of the 166 patients with Gartland III fractures range of motion (ROM) (Sinikumpu et al. 2016). Long-term had clinical findings of either median (19), ulnar (7), radial outcome of SCHF is usually assessed clinically by Flynn’s cri- (7), or median and ulnar (3) nerve injury at discharge (even teria (Flynn et al. 2015), which define unsatisfactory results minor findings were recorded). Normal motor and sensory as more than 15˚ asymmetry in elbow ROM or carrying angle findings were recorded in 13 of these 36 patients preopera(CA). The subjective outcome has been evaluated with vali- tively (1 median and ulnar, 3 ulnar, 3 radial, and 6 median dated scoring systems such as the Pediatric Outcome Data nerve palsies). Electromyography (EMG) was undertaken on Collection Instrument (PODCI), the Mayo Elbow Perfor- 18 of these 36 patients with no (11) or partial recovery within mance score, and QuickDASH. The limitation of these vali- 3–6 months. 1 patient’s (primary treated by a consultant) dated outcome measures is the absence of questions concern- median nerve was found partially entrapped in the fracture ing cosmetic outcome, which in our opinion is a weakness, gap, released and repaired with subtotal recovery 1 year after since the most common complication of SCHF, cubitus varus, fracture. All other patients’ sensorimotor functions recovered. is mainly considered a cosmetic problem. Permanent nerve injury rate as a complication of treatment Long-term outcome in displaced SCHF has not been well was thus 0.4%. documented, however. To our knowledge, this is the largPostoperative radiographs of 197 (94%) patients were anaest follow-up study of operatively treated SCHF in children. lyzed by a pediatric radiologist and an orthopedic registrar Second, our goal was to study whether primary treatment out- (NT). At fracture union alignment was regarded as satisfactory come parameters have a prognostic value of long-term out- (BA within ±10˚ of reported normal range and AHL crossed come of SCHF. capitulum) in 150/184 (82%) of the patients’ radiographs (13 patients did have either failed sagittal or coronal view radiographs). Grade II fractures healed in better alignment than grade III fractures (satisfactory result in 27/28 vs. 121/152, p Patients and methods = 0.03). The intra- (r = 0.91 and 0.91) and inter-observer (r = During the 5-year study period in 2002–2006, 210 domes- 0.92) reliability of Baumann angle measurement was considtic SCHF (Gartland III 79%, Gartland II 19%, and flexion ered excellent (Pearson’s correlation coefficient). type 2%) were pin fixed in the Children’s Hospital, Helsinki After a mean follow-up time of 9 (7–12) years, 168 (80%) (Figure 1). 115 (55%) patients were boys and 133 (63%) frac- patients answered a questionnaire regarding appearance, functures left-sided. Mean age at the time of fracture was 7.2 (1.8– tion, and pain (scale 0–10) of the fractured elbow and symPin fixation n = 213

12269 Tuomilehto D.indd 352

5/3/2018 11:50:12 AM

Acta Orthopaedica 2018; 89 (3): 351–356

353

Table 1. Questionnaire and answers (patients n = 168)

Questions

Answer options

1. How satisfied are you with the appearance of the fractured elbow? 2. How satisfied are you with the functional outcome of the fractured elbow? 3. Do you have any pain at rest in the fractured elbow? If you answered yes, how intense has the pain been in the past week? 4. Do you have any pain in motion in the fractured elbow? If you answered yes, how intense has the pain been in the past week? 5. Are the carrying angles of your elbows symmetrical? 6. Do your elbows flex symmetrically? 7. Do your elbows extend symmetrically?

0–10 0–10 no/yes 0–10 no/yes 0–10 no/yes no/yes no/yes

Answers “yes” mean (range)

Very unsatisfied—very satisfied Very unsatisfied—very satisfied

8.7 (2–10) 9.0 (2–10) 7

No pain—worst pain you can imagine

3.0 (1–8) 21

No pain—worst pain you can imagine

3.9 (1–8) 140 149 132

up to 5 years because of valgus deformity. The remaining 9 patients who did not answer the questionnaire were seen regularly until satisfactory recovery of their nerve palsy (mean follow-up 6 months (1 month–2 years)). There was no difference in radiological outcome at union between patients who answered only the questionnaire and those that also attended the clinical follow-up examination or between patients who participated in the study and those who did not. Statistics analysis was performed using Fisher’s exact and Wilcoxon’s test. P-values ≤ 0.05 are considered statistically significant. Ethics, funding, and potential conflicts of interest This study was approved by the local ethics committee (number 346/13/03/03/2012). Päivikki and Sakari Sohlberg Foundation, Finska Läkaresällskapet and Vappu Uuspää Foundation supported this study. No competing interests declared.

Results Figure 2. Carrying angle and elbow range of motion in 65 patients at mean 9 years after pin fixation of supracondylar humerus fracture.

metry of ROM and CA of their elbows (Table 1). 29 of the 36 nerve injury patents answered the questionnaire. 65 (31%) patients also attended for clinical examination performed by NT (Figure 1). ROM and CA of both elbows were measured with a goniometer (Figure 2). Radiographs of both elbows were taken if asymmetry in ROM or CA exceeded 10˚. Lengths of both arms were measured from the tip of acromion to the distal end of the lateral humeral condyle, and forearms from the tip of the olecranon to the distal end of the ulna, respectively. Of the 42 patients who did not participate in this study, 12 had postoperative follow-up elsewhere, 16 had no further controls after pin removal, 4 were followed up to 6–12 weeks to ensure fracture union and recovery of elbow motion and 1

12269 Tuomilehto D.indd 353

Mean subjective score for appearance was 8.7 (2–10) and for function 9.0 (2–10) according to the 168 answered questionnaires (Table 1). Fracture type, patient’s sex, and AHL in relation to capitulum did not affect the results. Mean functional scores were lower either in the 31 patients who were older than 10 years at the time of fracture (8.4, p = 0.01) and/or in the 29/36 patients who had nerve injuries (8.6, p = 0.05). Mean subjective scores for appearance were lower in the 14 patients who had had open reduction (7.8, p = 0.03) and in 9 patients who had BA values exceeding normal values by 10˚ (7.1, p = 0.02) (Figure 3, see Supplementary data). Open reduction patients’ AHL crossed the capitulum in 11/12 and BA was within 10˚ of the normal range in 11/12 cases. One or both subjective scores of 13 (8%) patients were below 6. 8 of these patients attended a control visit, and 5 of 8 had asymmetry of elbow ROM or CA or both of more than 10˚ (p = 0.002). 11 of these 13 patients were operated by a registrar (alone 5, under supervision 6) (p = 0.4).

5/3/2018 11:50:14 AM

354

Acta Orthopaedica 2018; 89 (3): 351–356

CA asymmetry to varus exceeding 10˚ was measured in 9/65 patients attending for clinical examination. None of these 9 Unrecognized Recognized Asymmetry asymmetry > 5° asymmetry > 5° patients had BA over 10˚ outside the normal ≤ 5˚ 6–10° 11–15° > 15° 6–10° 11–15° > 15° range or radiological evidence of malrotation at fracture union. CA varus asymmetry Flexion deficit 4 1 2 0 1 1 1 Extension deficit 0 1 0 0 1 1 0 exceeding 10˚ correlated with decreased hyper 9 11 0 0 4 1 0 subjective cosmetic (p = 0.005) and funcCarrying angle varus 2 8 2 1 4 4 2 tional (p = 0.004) outcomes (Figures 5 and valgus 0 0 0 0 2 0 0 6, see Supplementary data). All of these 9 patients were operated by registrars (alone 6, under supervision 3, p = 0.06). No asymmetry (> 5˚) in supination or pronation was found. Table 3. Long-term outcome by Flynn’s criteria at follow-up visit (n = 65) Lengths of arms and forearms were equal. Radiographs of both elbows were taken in 12/65 patients (10 due to varus Fracture type Loss of motion Loss of carrying angle deformity and 2 due to flexion restriction). Signs of avascular (Gartland) II III Flexion II III Flexion n = 10 n = 53 n = 2 n = 10 n = 53 n = 2 necrosis or degenerative changes were not found. Complication rate (iatrogenic nerve injury, re-reduction, Satisfactory deformity corrected by osteotomy, CA or ROM asymmetry > Excellent (0–5˚) 8 46 2 6 38 1 Good (6–10˚) 1 3 0 0 10 1 10˚) was greater in patients operated by a registrar (17/150 vs. Fair (11–15˚) 0 4 0 3 3 0 2/60, p = 0.1). Unsatisfactory Table 2. Patients’ subjective estimation compared with the control visit findings (n = 65)

Poor (> 15˚)

Elbow pain either at rest (PIR) or in motion (PIM) or both was reported by 14% of the 168 patients who answered the questionnaire (PIM by 13% (median score 4 (1–8)) and PIR by 4% (median score 2.5 (1–8)) (Table 1). Pain was not related to fracture type or sex, but was more common in patients with nerve injuries (8/29, p = 0.04), in patients older than 10 years of age at the time of fracture (7/31, p = 0.2), in patients with elbow flexion deficit (> 10˚; 2/4, p = 0.2), and in patients with elbow CA asymmetry exceeding 10˚ into varus (3/9, p = 0.4). Asymmetric elbow ROM was reported by 28% (extension 21%, flexion 11%) and CA by 17% of the 168 patients who answered the questionnaire (Table 1). Fracture type or radiographic findings at the time of fracture union did not affect the results, but asymmetric elbow ROM was more frequently experienced by girls (girls 29/75 vs. boys 18/93, p = 0.01). Subjective assessment of elbow ROM and CA of the 65 patients who attended the clinical examination proved to be unreliable in minor (< 11˚), but half as often recognized in more severe (> 10˚) asymmetry (Table 2). Long-term outcome of elbow ROM was good or excellent in 60/65 and CA in 56/65 of the follow-up patients using Flynn`s criteria (Table 3). ROM asymmetry exceeding 10˚ was registered in 5/65 of patients (4 elbow flexion deficit, 1 hyperextension) attending for clinical examination. AHL crossed capitulum in 4 of these 5 patients’ radiographs at fracture union (1 patient’s lateral view radiograph was unusable). Elbow flexion deficit exceeding 10˚ correlated with lower subjective functional outcome (mean score 5.5 (3–10) p = 0.03; Figure 4, see Supplementary data). 4 of these 5 patients were operated by a registrar (alone 1, under supervision 3, p = 1.0).

12269 Tuomilehto D.indd 354

Discussion We have assessed long-term outcome of pin-fixed SCHF in Helsinki University Children’s Hospital, which is the largest institution treating pediatric fractures in Finland and the only hospital in Helsinki treating displaced SCHF in children. The Children’s Hospital is also the largest teaching hospital of pediatric orthopedics in Finland and therefore the majority (71%) of the patients in these series were operated by registrars, which should be taken into account since the quality of fracture reduction and osteosynthesis obviously affects the outcome in treatment of SCHF. In fact, the risk of complication was higher and the long-term outcome was poorer in patients operated by registrars compared with consultants in our study, although the differences did not reach statistical significance. Volkmann’s contracture and permanent iatrogenic nerve injuries can be regarded as complications of treatment and their prevalence should be zero (Vallila et al. 2015). We consider a deep infection rate of less than 1% acceptable. Flynn’s criteria are generally used in order to assess the quality of treatment regarding elbow ROM and CA (Flynn et al. 2015). During the study period in 2002–6, the rate of Volkmann’s contracture was zero, and the risk of deep infection and permanent treatment-related nerve complications was low (< 1%) at our institution. Sinikumpu et al. (2016) have reported 12 years’ long-term subjective outcome of 81 patients assessed by Mayo elbow performance score, but most of the patients had non-displaced SCHFs that were treated non-operatively. The Mayo score in the 25 patients with grade III fractures was excellent (mean

5/3/2018 11:50:15 AM

Acta Orthopaedica 2018; 89 (3): 351–356

93), but range was not reported. In their minimum of 18 months’ follow-up study of a total of 154 Garland grade III SCHF, Wang et al. (2017) evaluated the long-term outcome of 33 neurovascular injury patients compared with patients with intact neurovascular findings. They used the Pediatric Outcomes Data Collection Instrument (PODCI) and Quick Disabilities of the Arm, Shoulder, and Hand (QuickDASH) outcome measures, both indicating excellent function. They found no statistically significant differences in outcome measures between the neurovascular injury patients and those with intact findings. No differences in outcomes were identified based on age, fracture site, sex, weight, direction of displacement, or operative technique in neurovascular injury patients. The authors speculated that a difference in the PODCI and DASH scores among groups could have been missed. In our study, considering subjective scores 0–5 unsatisfactory and 6–10 satisfactory, 92% of our patients who answered the questionnaire had satisfactory results. Subjective cosmetic outcome correlated negatively with open reduction, which is most likely due to the scar. The number of patients with unsatisfactory subjective results could have probably been smaller with better quality of fracture reduction and pin fixation (satisfactory alignment at fracture union in 82% of patients in our study). Radiographic measurements had little value predicting subjective or objective outcome. This is most likely due to relatively large normal variation (almost 20°) and intra- and inter-rater errors (up to 7°) of measuring BA (Silva et al. 2010), although good inter- and intra-rated reliability in our study, as well as difficulties in obtaining true lateral radiographs of SCHF (Simanovsky et al. 2007, Guven et al. 2015) rendered radiographic assessment of fracture alignment unreliable. Previous follow-up studies have not analyzed their quality of reduction at fracture union by comparing their radiological measurements with reported normal values (Guven et al. 2015, Sinikumpu et al. 2016). Sinikumpu et al. (2016) reported, also from Finland, that 9/25 of patients experienced pain at least 10 years after grade III fractures, whereas after grade I or II fractures the pain prevalence did not differ from normal age-matched controls. Fewer patients (14%) reported pain in our study without correlation to fracture grade but in association with nerve injury. Guven et al. (2015) reported good or excellent functional and cosmetic long-term outcome (mean 22 years of followup) using Flynn’s criteria in 33 and 43 of their 49 SCHF patients with open reduction and cross-pin fixation (low clinical follow-up rate of 12%). In the study by Sinikumpu et al. (2016) (follow-up rate 76%), good or excellent results were achieved in 22 of 33 patients with grade II and III SCHF treated by either percutaneous or open pin fixation after minimum follow-up of 10 years. Our results (clinical follow-up rate 31%) were clearly better with good or excellent cosmetic and functional outcome in 56 and 60 of the 65 patients who were clinically examined at mean follow-up of 9 years. It has been recently reported by Vallila et al. (2015) that quality of

12269 Tuomilehto D.indd 355

355

treatment in distal humerus fractures correlates positively with the experience of the treating institution. Our results support this finding, since the annual number of pin-fixed SCHFs in our institution is 3-fold compared with the number reported in the study of Guven et al. (2015) from Istanbul and 10-fold compared with the earlier report from the joint study of two other institutions in Finland (Sinikumpu et al. 2016). In Simanovsky’s study (2007) of 17 patients with malunited SCHF, 3 of 10 patients with elbow flexion restriction exceeding 5˚ had not recognized the ROM asymmetry and in 7 patients mild (up to 7˚) hyperextension of the elbow was registered. Only restricted elbow flexion was considered as a functional problem. In our study, only one-third of patients with minor elbow ROM deficiencies (less than 10˚ asymmetry) had noticed the asymmetry, whereas two-thirds of patients with more pronounced asymmetry were aware of CA or ROM discrepancy between their elbows (see Table 2). BA of more than 10˚ outside the normal reported range measured at fracture union was predictive of gun-stock deformity. Our results are in accordance with Simanovsky’s findings that decreased elbow flexion is a functional problem. We suggest that in the future only good and excellent results according to Flynn’s criteria should be classified as satisfactory, since asymmetry exceeding 10˚ in elbow ROM and CA was clearly associated with poorer subjective results in our study. Limitations of our study are its retrospective nature and that only 31% of the patients attended for clinical follow-up examination. On the other hand, 80% of the patients answered the questionnaire and there was no difference between fracture types or quality of primary treatment between patients who participated in either part of the study with those who did not. In summary, long-term cosmetic and functional outcome in SCHF are satisfactory with few exceptions if elbow ROM and CA can be restored within 10˚ of the uninjured elbow. Patients’ subjective estimation of minor (< 10˚) asymmetry in elbow ROM and CA is unreliable. Nerve injuries can cause long-term pain. Supplementary data Figures 3–6 are available in the online version of this article, http://dx.doi.org/10.1080/17453674.2018.1438765.

A pediatric radiologist, Reetta Kivisaari, analyzed postoperative and controlvisit radiographs.

NT participated in planning the study, collected patient data, examined patients at the control visit and analyzed all radiographs and stored fluoroscopic images with the pediatric radiologist. AS participated in planning the study and editing the manuscript. AY participated in planning the study and is the main writer of the manuscript together with NT.

Acta thanks Vera Halvorsen and Klaus Parsch for help with peer review of this study.

5/3/2018 11:50:15 AM

356

Babal J, Helan C, Klein G. Nerve injuries associated with pediatric supracondylar humeral fractures: a meta-analysis. J Pediatr Orthop 2010; 30(3): 253-63. Bashyal R, Chu J, Schoenecker P, Dobbs M, Luhmann S, Gordon E. Complications after pinning of supracondylar distal humerus fractures. J Pediatr Orthop 2009; 29(7): 704-8. Cheng J, Ng B, Ying S, Lam P. A 10-year study of the changes in the pattern and treatment of 6,493 fractures. J Pediatr Orthop 1999; 19(3): 344-50. Dai L. Radiographic evaluation of Baumann angle in Chinese children and its clinical relevance. J Pediatr Orthop B 1999; 8(3): 197-9. Dormans J, Squillante R, Sharf H, Wayne F. Acute neurovascular complications with supracondylar humerus fractures in children. J Hand Surg Am 1995; 20(1)1-4. Flynn J, Skaggs D, Waters P. Rockwood and Wilkins’ Fractures in Children. 8th ed. Wolters Kluwer Health, Philadelphia 2015. Gosens T, Bongers K. Neurovascular complications and functional outcome in displaced supracondylar fractures of the humerus in children. Injury 2003; 34(4): 267-73. Guner S, Guven N, Karadas S, Ceylan M, Turktas U, Gokalp M, Gozen A. Iatrogenic fracture-related nerve injuries in supracondylar humerus fracture: is treatment necessary for nerve injury? Eur Rew Med Pharmacol Sci 2013; 17(6): 815-19. Guven M F, Kaynak G, Inan M, Caliskan G, Unlu H B, Kesmezacar H. Results of displaced supracondylar humerus fractures treated with open reduction and internal fixation after mean 22.4 years of follow-up. J Shoulder Elbow Surg 2015; 24: 640-6. Herman M, Boardman M, Hoover J, Chafetz R. Relationship of the anterior humeral line to the capitellar ossific nucleus: variability with age. J Bone Joint Surg Am 2009; 91: 2188-93. Omid R, Choi P, Skaggs D. Current concepts review: supracondylar humeral fractures in children. J Bone Joint Surg Am 2008; 90: 1121-32. Otsuka N, Kasser J. Supracondylar fractures of the humerus in Children. J Am Acad Orthop Surg 1997; 5: 19-26.

12269 Tuomilehto D.indd 356

Acta Orthopaedica 2018; 89 (3): 351–356

Ramachandran M, Birch R, Eastwood D. Clinical outcome of nerve injuries associated with supracondylar fractures of the humerus in children. J Bone Joint Surg (Br) 2006; 88-B: 90-4. Scannell P, Jackson B, Bray C, Roush T, Brighton B, Frick S. The perfused, pulseless supracondylar humeral fracture: intermediate-term follow-up of vascular status and function. J Bone Joint Surg Am 2013; 95: 1913-19. Shank C, Wiater B, Pace J, Jinguji T, Schmale G, Bittner R, Bompadre V, Stulus J, Krengel W. The lateral capitellohumeral angle in normal children: mean, variation, and reliability in comparison to Baumann’s angle. J Pediatr Orthop 2011; 31: 266-71. Silva M, Pandarinath R, Farng E, Park S, Caneda C, Fong Y-J, Penman A. Inter- and intra-observer reliability of the Baumann angle of the humerus in children with supracondylar humeral fractures. Int Orthop 2010; 34: 553-7. Sinikumpu J-J, Victorzon S, Pokka T, Lindholm E-L, Peljo T, Serlo W. The long-term outcome of childhood supracondylar humeral fractures: a population-based follow up study with a minimum follow up of ten years and normal matched comparisons. Bone Joint J 2016; 98-B: 1410-17. Simanovsky N, Lamdan R, Mosheiff R, Simanovsky N. Underreduced supracondylar fractures of the humerus in children: clinical significance at skeletal maturity. J Pediatr Orthop 2007; 27: 733-8. Tuomilehto N, Kivisaari R, Sommarhem A, Nietosvaara A Y. Outcome after pin fixation of supracondylar humerus fractures in children: postoperative radiographic examinations are unnecessary. Acta Orthop 2016; 24: 1-7. Vallila N, Sommarhem A, Paavola M, Nietosvaara Y. Pediatric distal humeral fractures and complications of treatment in Finland: a review of compensation claims from 1990 through 2010. J Bone Joint Surg Am 2015; 97: 494-9. Wang S, Kwon T, Hwang H, Kim J, Jung R. Functional outcomes of Gartland III supracondylar humerus fractures with early neurovasvular complications in children: a retrospective observational study. Medicine 2017; 96: 25(e7148). Williamson D, Coates C, Miller R, Cole W. Normal characteristics of the Baumann (humerocapitellar) angle: an aid in assessment of supracondylar fractures. J Pediatr Orthop 1992; 12(5): 636-9.

5/3/2018 11:50:15 AM

Acta Orthopaedica 2018; 89 (3): 357–359

357

Low sensitivity of a-defensin (Synovasure) test for intraoperative exclusion of prosthetic joint infection Ruben SCHOLTEN, Jetze VISSER, Job L C VAN SUSANTE, and Corné J M VAN LOON

Department of Orthopedics, Rijnstate Hospital, Arnhem, The Netherlands Correspondence: cvanloon@rijnstate.nl Submitted 2017-12-19. Accepted 2018-01-16.

Background and purpose — The Synovasure lateral flow test was developed as a rapid test for the detection or exclusion of periprosthetic joint infection (PJI). 3 studies have reported promising results on its diagnostic value in total joint revision surgery. We aimed to assess the sensitivity and specificity of the Synovasure test to exclude infection in patients undergoing revision surgery for suspected early aseptic loosening of a total hip or knee arthroplasty. Patients and methods — In a prospective study design, 37 patients who underwent revision surgery for suspected early aseptic loosening (< 3 years after primary arthroplasty) were included. The Synovasure test was used intraoperatively to confirm the aseptic nature of the loosening and 6 tissue cultures were obtained in all cases. Exclusion criteria were patients with a preoperatively confirmed PJI, acute revisions (< 90 days after primary arthroplasty) and cases with malpositioning, wear, or instability of the prosthesis. Results — 5 of the 37 patients were diagnosed with a PJI based on the intraoperative tissue cultures. In only 1 out of these 5 cases this was confirmed by the intraoperative Synovasure test. No tests were falsely positive. Interpretation — In this case series the Synovasure lateral flow test had a low sensitivity to exclude PJI in patients with suspected aseptic loosening. The role of the Synovasure lateral flow test in the intraoperative exclusion of PJI during revision surgery for suspected early aseptic loosening appears to be more limited than previously indicated. ■

Periprosthetic joint infection (PJI) accounts for up to 25% of failed total knee arthroplasties (TKA) and up to 15% of failed total hip arthroplasties (THA) (Bozic et al. 2009, 2010). Distinguishing the septic from the aseptic failures in total joint arthroplasty (TJA) is critical in the successful treatment

of painful prosthetic joints, as they require different surgical strategies. To this end, the Musculoskeletal Infection Society has formulated criteria for decision-making in the work-up of suspected of PJI (Parvizi et al. 2011). However, this work-up remains far from straightforward, mainly since there is no uniform test for diagnosing PJI (Parvizi et al. 2011). Therefore, a simple diagnostic tool for PJI would be important. Recently, the presence of the α-defensin biomarker in synovial fluid was suggested as a possible marker of periprosthetic joint infection, since it is naturally released by neutrophils in the presence of synovial fluid pathogens (Ganz et al. 1985, Deirmengian et al. 2015b). 3 studies using quantitative measurements of α-defensin have reported a sensitivity and specificity above 96% for PJI (Deirmengian et al. 2014a, 2014b, 2015a). From these reported high sensitivity and specificity of α-defensin levels in detecting PJI, the Synovasure test (Zimmer Inc., Warsaw, IN, USA), a lateral flow test for the detection of α-defensin levels, has been made commercially available. The Synovasure lateral flow test is a rapid measure of the α-defensin levels in synovial fluid, and provides dichotomic results after 10 minutes. 2 studies reported a sensitivity (67–69%) and specificity (93–94%) lower than in the previously mentioned quantitative measurements (Kasparek et al. 2016, Sigmund et al. 2017). However, a recent study by Berger et al. (2017) reported higher sensitivity (97%) and specificity (97%). The previously mentioned studies were performed in nonspecific patient populations containing varying indications for revision surgery ranging from evident acute PJI to reimplantation in 2-stage PJI revisions. However, the true clinical value of this promising test may lie in its ability to intraoperatively distinguish the early (< 3 years following implantation) aseptic failure, where a 1-stage revision is indicated, from the early septic failure with less virulent micro-organisms where a staged revision would be more appropriate.

12284 Scholten D.indd 357

5/3/2018 11:52:36 AM

358

Acta Orthopaedica 2018; 89 (3): 357–359

In this study we evaluated the additional value of the intraoperative Synovasure lateral flow test in confirming the absence of PJI in a group of patients undergoing prosthetic joint revision surgery for suspected early aseptic loosening. The results of the Synovasure test were compared with intraoperative tissue cultures. Furthermore, we assessed the possible correlation of false-negative test results with the presence of metallosis, since previous studies have suggested that metallosis may predispose to false-positive results (Bonanzinga et al. 2017).

Table. Specification of false-negative cases

Patients and methods

Results

Since August 2015, the Synovasure test has been used in our clinic as an adjunct tool to exclude PJI intraoperatively in revision patients with suspected early aseptic loosening of an implanted THA or TKA. Cases were prospectively included in the presence of a chronically painful (> 90 days) prosthetic joint in those who underwent revision surgery due to suspected early aseptic loosening (< 3 years after primary arthroplasty) of the implant after TKA or THA between August 2015 and October 2017. During revision surgery on these patients in this period, the Synovasure test was used to aid in the exclusion of PJI. Excluded from this study were patients already diagnosed with PJI according to the MSIS criteria, acute revisions (< 90 days), revisions due to dislocations, revisions due to malpositioning, or cases where an insufficient amount of synovial fluid could be aspirated to perform the Synovasure lateral flow test. In all cases, synovial fluid was aspirated under aseptic conditions (after surgical dissection up to the joint capsule) from the affected joint whilst avoiding any contamination with blood. The Synovasure test was carried out according to the manufacturer’s instructions. In the case of a positive Synovasure test, we considered the joint to be infected and proceeded with the removal of the prosthesis and implantation of a spacer containing antibiotics. If the test was negative, we proceeded with a one-stage revision of the affected joint. In all cases, a total of 6 microbiologic cultures of synovial tissue and the interface membrane were collected. The tissue samples were cultured for 14 days in the microbiology laboratory. Tissue cultures were considered positive for PJI when at least 2 out of 6 cultures grew identical pathogens. During surgery, the presence of metallosis or macroscopic signs of infection (presence of pus) were noted.

37 patients (22 men), mean age 66 (51–81) years, planned for revision surgery met the inclusion criteria. Revision surgery was performed on 8 hips and 29 knees. 5 patients were diagnosed with PJI due to positive tissue cultures. Of these, 1 patient had a positive Synovasure test and 4 patients tested negative. Thus, there were 1 true-positive Synovasure test, and 4 false-negative Synovasure tests (Table). No false-positive Synovasure tests were observed, even in the presence of metallosis. 4 false-negative cases were observed; all these patients were treated as 1-stage revisions for PJI with adequate antibiotics over the course of 12 weeks. The only true positive test occurred in a patient with intraoperative macroscopic signs of infection due to the presence of pus. In this patient, previous tests did not indicate PJI according to the MSIS criteria; however, 6 intraoperative tissue cultures grew Staphylococcus epidermidis. 29 true negative tests were observed; in 1 of these cases metallosis was present during surgery.

Ethics, funding, and potential conflicts of interest The procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation and with the Helsinki Declaration of 1975, as revised in 2000. This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. No competing interests were declared.

12284 Scholten D.indd 358

False-negative case no. 1 2 3 4

Identified pathogens Staphylococcus epidermidis Staphylococcus epidermidis Staphylococcus epidermidis and Staphylococcus saccharolyticus Propionibacterium acnes

Positive cultures 2/6 3/6 2/6 2/6

Discussion The identified sensitivity (1/5) is clearly lower than reported in previous studies (Kasparek et al. 2016, Berger et al. 2017, Sigmund et al. 2017). These 3 earlier studies reported a sensitivity of 67–97% and specificity of 93–96% of the Synovasure test for the diagnosis of PJI In contrast to our study these studies included various kinds of procedures, namely: patients fulfilling the MSIS criteria for PJI preoperatively, 1-stage revisions, reimplantations at second-stage revision, explantations and spacer implantations, spacer exchanges, debridements with exchange of mobile parts and retention of the prosthesis and excision of a hip prosthesis (Sigmund et al. 2017). Kasparek et al. (2016) also included cases with varying indications: aseptic loosening, polyethylene wear with osteolysis, suspected chronic PJI, patients fulfilling the MSIS criteria for PJI preoperatively, instability and stiffness. Berger et al. (2017) included patients fulfilling a modified version of the MSIS criteria for PJI preoperatively. This modification, and the non-specific patient

5/3/2018 11:52:36 AM

Acta Orthopaedica 2018; 89 (3): 357–359

populations including varying indications for revision surgery, makes the reported values difficult to interpret and compare. Even more so in establishing the tests’ ability to distinguish early aseptic failure from septic failure in unclear cases that are not evidently infected (not fulfilling the modified MSIS criteria for PJI). In contrast to these earlier studies our study focused on the ability of the Synovasure test to exclude PJI in a uniform subgroup of patients undergoing revision surgery for suspected early aseptic loosening. To our knowledge, this is the first study to assess the Synovasure lateral flow test in this specific homogeneous subgroup of patients. From clinical practice this is an important strength of our study and as such is the finding of a rather low sensitivity in this particular subgroup. This strength has to be balanced against the limitation of a rather small number of patients included, which warrants caution in drawing firm conclusions. Another strength of our study, and that of Sigmund et al. (2017), is that there is no conflict of interest in relation to the manufacturer of the Synovasure test. Previous studies have suggested that the presence of metallosis may predispose to false-positive results of the Synovasure test. In our study, there was 1 case of metallosis, which yielded a negative Synovasure test. Our findings indicate that the Synovasure lateral flow test has limited additional value for the intraoperative exclusion of PJI from low virulent micro-organisms (i.e. Staphylococcus epidermidis and Proprionibacterium acnes) in a homogeneous subgroup of patients with suspected early aseptic failures of THA and TKA. This is an important limitation in the clinical use of the test, which initially promised to be ideal in simply confirming or excluding any PJI intraoperatively. The sensitivity may be improved by aiming future research at fine-tuning the thresholds of α-defensin in the presence of low-virulent micro-organisms. This may, however, be difficult to achieve since recent reviews failed to establish a more accurate cut-off value. The latter was due to the usage of different techniques among laboratories and a shortage of well-designed studies (Li et al. 2017, Yuan et al. 2017). It should also be noted that the dichotomic nature of the Synovasure lateral flow test, where the presence or absence of a PJI is claimed, is another limitation. Irrespective of the fact that a solution may be found to decrease the relatively high chance of a false-negative test outcome in case of low-virulent agents, the microbial agents and their resistance patterns would still have to be obtained from prolonged tissue cultures. For that reason future research should also continue to focus on advances in molecular microbiology and techniques for detecting microbial infections (e.g., susceptibility testing, DNA amplification assays) (Maurer et al. 2017). These techniques may offer increased diagnostic resolution and are not dichotomic by also providing information on the causative pathogen’s identity and resistance pattern. Further improve-

12284 Scholten D.indd 359

359

ment on these earlier mentioned microbial detection techniques may eventually bypass the dependency on tissue cultures for adequate antibiotic treatment.

RS: data collection and writing of the manuscript. JV: study setup and writing of the manuscript. JS: study design, writing of the manuscript. CL: study setup and design, manuscript review. Acta thanks Inge Skråmm and other anonymous reviewers for help with peer review of this study. Berger P, Van Cauter M, Driesen R, Neyt J, Cornu O, Bellemans J. Diagnosis of prosthetic joint infection with alpha-defensin using a lateral flow device: a multicentre study. Bone Joint J 2017; 99-b(9): 1176-82. Bonanzinga T, Zahar A, Dutsch M, Lausmann C, Kendoff D, Gehrke T. How reliable is the alpha-defensin immunoassay test for diagnosing periprosthetic joint infection? A prospective study. Clin Orthop Relat Res 2017; 475(2): 408-15. Bozic K J, Kurtz S M, Lau E, Ong K, Vail T P, Berry D J. The epidemiology of revision total hip arthroplasty in the United States. J Bone Joint Surg Am 2009; 91(1): 128-33. Bozic K J, Kurtz S M, Lau E, Ong K, Chiu V, Vail T P, Rubash H E, Berry D J. The epidemiology of revision total knee arthroplasty in the United States. Clin Orthop Relat Res 2010; 468(1): 45-51. Deirmengian C, Kardos K, Kilmartin P, Cameron A, Schiller K, Parvizi J. Combined measurement of synovial fluid alpha-defensin and C-reactive protein levels: highly accurate for diagnosing periprosthetic joint infection. J Bone Joint Surg Am 2014a; 96(17): 1439-45. Deirmengian C, Kardos K, Kilmartin P, Cameron A, Schiller K, Parvizi J. Diagnosing periprosthetic joint infection: has the era of the biomarker arrived? Clin Orthop Relat Res 2014b; 472(11): 3254-62. Deirmengian C, Kardos K, Kilmartin P, Cameron A, Schiller K, Booth R E, Jr, Parvizi J. The alpha-defensin test for periprosthetic joint infection outperforms the leukocyte esterase test strip. Clin Orthop Relat Res 2015a; 473(1): 198-203. Deirmengian C, Kardos K, Kilmartin P, Gulati S, Citrano P, Booth R E, Jr. The alpha-defensin test for periprosthetic joint infection responds to a wide spectrum of organisms. Clin Orthop Relat Res 2015b; 473(7): 2229-35. Ganz T, Selsted M E, Szklarek D, Harwig S S, Daher K, Bainton D F, Lehrer R I. Defensins: natural peptide antibiotics of human neutrophils. J Clin Invest 1985; 76(4): 1427-35. Kasparek M F, Kasparek M, Boettner F, Faschingbauer M, Hahne J, Dominkus M. Intraoperative diagnosis of periprosthetic joint infection using a novel alpha-defensin lateral flow assay. J Arthroplasty 2016; 31(12): 2871-4. Li B, Chen F, Liu Y, Xu G. Synovial fluid alpha-defensin as a biomarker for peri-prosthetic joint infection: a systematic review and meta-analysis. Surg Infect (Larchmt) 2017; 18(6): 702-10. Maurer F P, Christner M, Hentschke M, Rohde H. Advances in rapid identification and susceptibility testing of bacteria in the clinical microbiology laboratory: implications for patient care and antimicrobial stewardship programs. Curr Infect Dis Rep 2017; 9(1): 6839. Parvizi J, Jacovides C, Zmistowski B, Jung K A. Definition of periprosthetic joint infection: is there a consensus? Clin Orthop Relat Res 2011; 469(11): 3022-30. Sigmund I K, Holinka J, Gamper J, Staats K, Bohler C, Kubista B, Windhager R. Qualitative alpha-defensin test (Synovasure) for the diagnosis of periprosthetic infection in revision total joint arthroplasty. Bone Joint J 2017; 99-B(1): 66-72. Yuan J, Yan Y, Zhang J, Wang B, Feng J. Diagnostic accuracy of alpha-defensin in periprosthetic joint infection: a systematic review and meta-analysis. Int Orthop 2017; 41(12): 2447-55.

5/3/2018 11:52:36 AM

360

Acta Orthopaedica 2018; 89 (3): 360–366

Negative effect of zoledronic acid on tendon-to-bone healing In vivo study of biomechanics and bone remodeling in a rat model Geir Aasmund HJORTHAUG 1–4, Endre SØREIDE 1–3, Lars NORDSLETTEN 1–3, Jan Erik MADSEN 1–3, Finn P REINHOLT 5, Sanyalak NIRATISAIRAK 2,6, and Sigbjørn DIMMEN 2,3,7

1 Division of Orthopedic Surgery, Oslo University Hospital (OUS), Norway; 2 Institute of Clinical Medicine, Faculty of Medicine, University of Oslo (UIO); 3 Experimental Orthopedic Research, Institute for Surgical Research, OUS; 4 Department of Orthopedic Surgery, Martina Hansen’s Hospital; 5 Department of Pathology, OUS; 6 Biomechanics Laboratory, Division of Orthopedic Surgery, OUS; 7 Department of Orthopedic Surgery, Lovisenberg

Diaconal Hospital, Norway Correspondence: geir.hjorthaug@me.com Submitted 2017-09-08. Accepted 2018-01-18.

Background and purpose — Outcome after ligament reconstruction or tendon repair depends on secure tendon-to-bone healing. Increased osteoclastic activity resulting in local bone loss may contribute to delayed healing of the tendon–bone interface. The objective of this study was to evaluate the effect of the bisphosphonate zoledronic acid (ZA) on tendon-to-bone healing. Methods — Wistar rats (n = 92) had their right Achilles tendon cut proximally, pulled through a bone tunnel in the distal tibia and sutured anteriorly. After 1 week animals were randomized to receive a single dose of ZA (0.1 mg/kg IV) or control. Healing was evaluated at 3 and 6 weeks by mechanical testing, dual-energy X-ray absorptiometry and histology including immunohistochemical staining of osteoclasts. Results — ZA treatment resulted in 19% (95% CI 5–33%) lower pullout strength and 43% (95% CI 14–72%) lower stiffness of the tendon–bone interface, compared with control (2-way ANOVA; p = 0.009, p = 0.007). Administration of ZA did not affect bone mineral density (BMD) or bone mineral content (BMC). Histological analyses did not reveal differences in callus formation or osteoclasts between the study groups. Interpretation — ZA reduced pullout strength and stiffness of the tendon–bone interface. The study does not provide support for ZA as adjuvant treatment in tendon-to-bone healing. ■

Tendons and ligaments attach to bone through a transitional fibrocartilage tissue, the enthesis. This transitional tissue is complex in composition and organization and is not regenerated during healing of injuries or surgical repair. Tendon and ligament injuries often require surgical repair or reconstruc-

tion to regain function, and a favorable outcome depends on solid tendon-to-bone healing (Harryman et al. 1991, Gulotta and Rodeo 2007, Ekdahl et al. 2008). The osteointegration of the graft is the weak link in early tendon-to-bone tunnel healing. In a randomized controlled trial, transient decrease in local bone mineral density (BMD) was observed in the knee region of patients undergoing anterior cruciate ligament (ACL) reconstruction (Lui et al. 2012). Bone loss has also been observed in experimental tendon–bone repair studies and is probably due to increased ostoclastic activity (Galatz et al. 2005, Rodeo et al. 2007). Furthermore, a negative correlation between local bone loss and strength of the tendon–bone interface has been reported (Silva et al. 2002). Bone surfaces with high osteoclastic activity and peri-tunnel bone loss may therefore become a less suitable scaffold for healing of the tendon graft. Thus, prevention of early bone loss could enhance tendon-to-bone healing and improve clinical outcomes. Bisphosphonates (BPs) are rapidly incorporated into the crystal structure of apatite bone matrix (Cole et al. 2016) and bind to the bone present at the time of administration (Amanat et al. 2007). The main effect of BPs is inhibition of osteoclasts, resulting in increased BMD (Russell et al. 2008). Several experimental studies have reported that BP treatment may enhance fracture healing (Li et al. 2001, Amanat et al. 2005) and implant fixation (Andersson et al. 2010). Only a limited number of studies have assessed the effect of BPs using tendon-to-bone models, but improved healing has been reported for alendronate (Thomopoulos et al. 2007, Lui et al. 2013). Zoledronic acid (ZA) is the most potent BP available (Russell et al. 2008) and, to our knowledge, the effect of ZA on tendonto-bone tunnel healing has not been tested.

12072 Hjorthaug D.indd 360

5/3/2018 11:53:32 AM

Acta Orthopaedica 2018; 89 (3): 360–366

We hypothesized that ZA treatment would improve the mechanical properties at 3 and 6 weeks of tendon-to-bone tunnel healing in rats by reducing the local bone loss.

Materials and methods Animals Female Wistar rats (n = 92) (Taconic Europe, Lille Skensved, Denmark), skeletally mature, mean weight of 238 g (SD 9), were included in the study. The animals were acclimatized for 2 weeks before the surgical procedure, and were kept 2 per wire-topped plastic cage in an accredited animal facility with controlled temperature (21°C ± 1), humidity (55% ± 10), ventilation and 12 hours light/dark cycles. They were allowed free access to water and standard laboratory rodent nutrition. Analgesics were given according to protocol (Hjorthaug et al. 2015). Anesthesia was induced and maintained by intraperitoneal (IP) injection of ZRF cocktail. The induction dose was 1.6 mL/kg; a volume of 0.4 mL. Antibiotics were not used. Surgical procedure The skin on the medial side of the right Achilles tendon was incised with a 10 mm longitudinal incision. The Achilles tendon was released proximally from the calf muscle while the calcaneal insertion distally was left intact. A drill hole was made in the distal tibia in a dorsoventral direction, 3 mm proximal to the ankle joint. Predrilling was done with a 1.0 mm drill bit followed by drilling with a 1.5 mm drill bit at 3,200 rpm. A holding suture was added to the released tendon in a Kessler suture fashion using a 2-string monofilament nonabsorbable suture. The sutures were guided through the tunnel from posterior to anterior and the tendon gently passed through the tunnel and sutured to the anterior soft tissue with the ankle joint in a 90° flexed position. The skin incision was closed using resorbable sutures and sealed with spray dressing. No restriction on weight bearing was applied postoperatively. Groups 7 days postoperatively, single animals were allocated to either treatment (ZA group) or control group by computerized randomization. Under general anesthesia, induced by a low dose of ZRF cocktail (0.1 mL) IP and maintained by isoflurane (1.5–2.5%) inhalation anesthesia, the animals in the treatment group received a single ZA dose (0.1 mg/kg, total volume 0.5 mL) by IV infusion (Aclasta, Novartis, Frimley, UK) in the tail vein (Amanat et al. 2007). Control animals received an injection of saline in the same volume and route. The personnel giving the infusions were blinded to groups, as were observers before all further analyses. The animals were killed after 3 or 6 weeks. Mechanical assessment Animals allocated to mechanical assessments were killed by

12072 Hjorthaug D.indd 361

361

pentobarbital overdose. The right hind leg was disarticulated through the knee, and the fibula, talus, foot, and soft tissue including the long tendons were removed. The anterior suture was removed. The tibia, calcaneus, and tendon graft were carefully preserved and the specimens were tested and analyzed as described (Hjorthaug et al. 2015) in a material testing machine (Model 858 Mini Bionix with MTS FlexTest digital controller, MTS Systems Corporation, Eden Prairie, Minnesota, USA) in a custom-made jig, designed to firmly hold the tibia and calcaneus without applying pressure to the tendon insertion site. No pre-tensioning was done. A 250N load cell was pulled upwards at a constant speed of 0.1 mm/s. The sampling frequency was 20 Hz over 120 s. All specimens were pulled to complete failure. Data were analyzed for pullout strength, stiffness, energy, and elongation. Histological evaluation Tissues of 5 animals from each study group at each time point were fixed in vivo by vascular perfusion of 0.1 M phosphatebuffered 2% paraformaldehyde during deep anesthesia. The right leg containing the tendon-bone-interface was stripped of the skin only. The 3rd tail vertebra was debrided of all soft tissues. Specimens were immersed in the same fixative as above for 2 days, and decalcified in 7% EDTA in a 0.1 M phosphate buffer containing 0.5% paraformaldehyde. The specimens containing the tendon-bone-interface were embedded in paraffin and serial sectioned with 5 µm section thickness from anterior to posterior. Light microscopy For semiquantitative bone morphometric analyses of the bone–tendon interface, 1 representative section centrally in the coronal plane was analyzed. Digital images were analyzed using AnalySIS V (Olympus Soft Imaging Solutions GmbH, Münster, Germany). A circle of 1.6 mm in diameter was superimposed to the tunnel profile, the chosen diameter being 0.1 mm larger than the used drill bit to compensate for any elliptical shaped appearance of the bone tunnel due to manual drilling and the manual sectioning along the axis of the tunnel. Mineralized tissue formation inside the bone tunnel was measured. The bone–tendon ratio inside the tunnel was calculated by point counting using a square test grid (Figure 1). Histochemistry To identify and estimate the number of osteoclast profiles in the paraffin sections, a standard protocol for the osteoclast marker tartrate-resistant acid phosphatase (TRAP) was used as suggested by the provider of a commercially available kit (Acid Phosphatase Leukocyte, TRAP, 387A-1KT lot SLBB7136, Sigma Aldrich). The regions of interest (ROI) were the bone tunnel, the ipsilateral calcaneus and the 3rd tail vertebra. The analyses were performed both on unstained and on sections stained with hematoxylin as nuclear stain. The TRAP staining was graded as strong, weak, or none, and osteoclast profiles

5/3/2018 11:53:32 AM

362

Acta Orthopaedica 2018; 89 (3): 360–366

were counted. Osteoclasts were defined as TRAPpositive cells containing more than 1 nucleus (Figure 2). Tissue from a previous study was used as positive control (Solberg et al. 2015).

Figure 1. Medium power magnification (x10) light microscopy of a H&E stained histological section in a specimen from the 3 weeks ZA group from the middle of the bone tunnel showing (a) the superimposed circle of 1.6 mm and (b) the grid used to calculate the ratio of bone and tendon inside the tunnel. Scale bars = 2 mm.

Figure 2. (a) High-power magnification (x40) light microscopy of section of the calcaneus from a specimen in the ZA group after 3 weeks subjected to TRAP enzyme histochemistry hematoxylin nuclear staining. Arrows indicate counted osteoclasts. (b) High-power magnification (x40) light microscopy of H&E stained section of the tendon–bone interface from a specimen in the control group after 6 weeks with bone resorption lacunae (arrows) in the tunnel wall containing osteoclasts. There is adjacent new woven bone (Wb) formation. Scale bars = 500 µm.

Figure 3. Upper panel: A specimen at baseline, showing the region of interest used for BMD and BMC measurements over the bone tunnel (a) and 3rd tail vertebra (b). Lower panel: X-rays of two 6-weeks specimens demonstrating bone formation around the bone tunnel in a specimen from the control group (c) and the ZA group (d). Radiographs were obtained from Lunar PIXImus.

12072 Hjorthaug D.indd 362

Bone mineral measurement and radiographic evaluation Under anesthesia at each time point in all animals, BMD and bone mineral content (BMC) were measured over the bone tunnel using DEXA (Lunar PIXImus, software v. 2.10, Lunar, Madison, WI, USA). The ROI consisted of 19 x 19 pixels (≈ 4 x 4 mm) that was aligned along the axis of the bone tunnel and placed flush with the anterior bone cortex. To assess systemic bone mineral effects, we used a secondary ROI, 19 x 33 pixels (≈ 4 x 7 mm), that was aligned to the 3rd tail vertebra (Figure 3). Radiographs were reviewed to reveal any complications. Statistics The number of animals was based on previous studies, and estimated based on our main outcome (Dimmen et al. 2009, Hjorthaug et al. 2015). An a priori power calculation for the sample size was not performed. The main outcome was pullout strength. Other outcomes were stiffness, energy absorption, elongation, and differences in BMD and BMC from baseline. 2 independent observers analyzed these 6 variables, and the mean values were calculated and used for the statistical analyses. Normality of the distribution was evaluated using histograms with normality curves. Variances were tested by Levenes’ test, and in case of detected heterogeneity, comparisons for group mean differences with 95% confidence intervals (CI) were calculated on log-transformed data of all above outcomes except elongation. Differences between the groups were tested by 2-way Analysis of Variance (ANOVA), with time and treatment as fixed factors. If no interaction effects were found, analysis of the main effects for time and treatment was performed. Pairwise comparisons were run where effect estimates were expressed as a percentage of the mean of the corresponding group and the 95% confidence intervals (CI) and p-values were Bonferroni-adjusted to correct for multiple comparisons. The alpha level was set to 0.05. Results for semiquantitative data from histology are given as medians and ranges, and the groups were compared using the nonparametric Mann–Whitney test for independent samples, without correction of the p-values. Statistical analyses were performed using IBM SPSS® Statistics for Macintosh v. 23.0 (IBM Inc., Chicago, Illinois, USA).

5/3/2018 11:53:32 AM

Acta Orthopaedica 2018; 89 (3): 360–366

363

Results Animal inclusion The body weight of all animals increased during the study period by mean 13.9 g (SD 6.4) and 25.3 g (SD 8.5), at 3 weeks and at 6 weeks respectively. 7 animals died perioperatively and 2 animals were killed due to wound complications during the first week. 83 animals were randomized to study groups. 2 animals died in the second anesthesia prior to injections 1 week postoperatively. Thus, 81 animals were available for final analysis including DEXA measurements. Of the 81 animals, 20 were selected for histology and 61 for mechanical tests. No fractures were observed clinically or on radiographs.

Figure 4. Clustered boxplots of data from mechanical testing of the tendon–bone interface. Except for elongation, all variables increased with time. There was a negative effect of ZA treatment in pullout strength and stiffness, but no significant differences in energy or elongation.

Figure 5. Clustered boxplots of data from DEXA measurements of the ROI over the bone tunnel: BMD and BMC difference from baseline.

Ethics, funding, and potential conflicts of interest The experimental protocol was reviewed and approved by the Norwegian Animal Research Authority (2014, ID 5839), and the study conducted according to guidelines provided by Norecopa. The study was funded by internal hospital research grants. One author declares the following potential conflict of interest or source of funding: GAH received a minor research grant from Smith & Nephew in 2012.

12072 Hjorthaug D.indd 363

Mechanical assessment The tendons of all specimens were completely pulled out, and they all ruptured at the tendon–bone tunnel interface. No fractures occurred during testing. Overall, mean pullout strength was 5.7 N (SD 2.7) at 3 weeks and 9.4 N (SD 5.2) at 6 weeks. Interaction effects of “time” or “treatment” were not found in any of the mechanical variables (2-way ANOVA), therefore main effects were calculated. Pullout strength increased from 3 to 6 weeks by 22% (CI 9–40%) (p = 0.002) in ZA-treated animals and controls. Stiffness and energy increased similarly by 40% (CI 12–70%) (p = 0.007) and 18% (CI 4–35%) (p = 0.02), respectively. Elongation did not increase with time (p = 0.7). The main effect of ZA treatment was a reduction of pullout strength compared with control by 19% (CI 5–33%) (p = 0.009) at 3 and 6 weeks. ZA treatment reduced stiffness by 43% (CI 14–72%) (p = 0.004). No statistically significant differences for treatment between the groups were observed in energy (p = 0.1) or elongation (p = 0.9) (Figure 4). Bone mineral measurements Baseline BMD over the bone tunnels of the 81 included animals was mean 179 mg/cm2 (SD 24). Interaction effects between time and treatment were not found in any of the bone mineral variables, therefore main effects were calculated. BMD difference from baseline increased from 3 to 6 weeks by 15 % (CI 3–27%) (p = 0.01). However, administration of ZA did not affect BMD (9%) (CI –4% to 21%) (p = 0.2) or BMC (p = 0.9) (Figure 5). No systemic differences in bone mineral, as measured over the 3rd tail vertebra for time or treatment, were detected. Histological findings 1 specimen (6 weeks control group) was excluded due to technical problems with decalcification, leaving 19 specimens for histological evaluation. All bone tunnels were located in the metaphysis, no tunnels extended into the ankle joint, no fractures were observed and heat necrosis was not pronounced. A moderate amount of periosteal callus was observed around the tibia at both 3 and 6 weeks. New woven bone formation appeared also in the diaphyseal bone marrow canal and inside

5/3/2018 11:53:38 AM

364

Acta Orthopaedica 2018; 89 (3): 360–366

Bone/tendon ratio (%)

Figure 6. Low-power magnification (x4) light microscopy of H&E stained sections of distal leg showing anatomy and changes over time. The location of the bone tunnel (Tu) containing tendon graft was aimed 3 mm above the ankle joint in the transition between the tibia diaphysis (D) and metaphysis (M). The ankle joint (A) was visible in most sections. (a) 3-weeks ZA group: sharp tunnel edges. Callus/Wb. (b) 6-weeks control-group: Callus/Wb mature. The arrows point at large capillaries inside the tendon graft. Scale bars = 2 mm.

Total number of TRAP-positive osteoclast counts (median, range) in distal tibia containing the bone tunnel and the calcaneus of the ipsilateral leg. Comparison of median values by independent samples Mann–Whitney test. 1 outlier (a) of the 3-week controls was excluded from the above analyses due to a very high number of osteoclasts (> 200) in the calcaneus

3 weeks Control a (n = 4) Zoledronic acid (n = 5) p-value 6 weeks Control (n = 4) Zoledronic acid (n = 5) p-value

Distal tibia

Calcaneus

Not measured Not measured

2 (1–4) 5 (4–15) 0.1

6 (5–23) 17 (0–45) 1.0

4 (1–4) 6 (0–23) 1.0

the tunnel, sometimes replacing the tendon graft. The 6 weeks specimens demonstrated increased vascularity of the tendon graft in addition to more mature new bone formation (Figure 6). The inner surfaces of the bone tunnels were more irregular in the 6 weeks specimens. No evident tunnel widening or peri-tunnel bone loss was seen. Bone resorption lacunae were seen in both groups (see Figure 2). However, the ZA specimens were not histologically distinguishable from the control group in the H&E sections. The bone morphometric evaluation of new bone formation in the bone tunnels demonstrated no statistically significant differences between the groups in bone–tendon ratio at 3 or 6 weeks (Figure 7). At 6 weeks, we noted both weak and strong TRAP-stained osteoclasts in the sections containing the bone tunnels. However, we could not detect differences between the ZA group and the controls, either regarding the number (Table) or the histologic distribution of osteoclasts. In the sections of 3rd tail vertebra, TRAP-positive osteoclasts were hardly present in any group.

12072 Hjorthaug D.indd 364

Figure 7. Clustered box-plots showing the bone– tendon ratio in the tunnel as measured by histomorphometry. No statistically significant effect of ZA treatment was observed at 3 weeks (p = 0.9, n = 8) or 6 weeks (p = 0.7, n = 9). Comparison of median values by independent samples Mann– Whitney test.

Discussion Our study demonstrated an unfavorable effect of a single dose of ZA on pullout strength and stiffness of the tendon–bone interface. We were not able to detect any differences in bone remodeling by DEXA, or histology with specific osteoclast staining and bone morphometric evaluation of mineralized tissue inside the bone tunnels, but the sample size of specimens evaluated by histology was small. The early phase of tendon to bone healing is associated with regional bone loss and decreased mechanical strength, especially if a bone-tunnel technique is used, as described in a canine model of toe flexor injury and repair (Silva et al. 2006). In a subsequent study, the same group was able to prevent the regional bone loss and improve tendon-to-bone repair strength at 3 weeks by administration of the BP alendronate (Thomopoulos et al. 2007). In a study in rats, alendronate was reported to reduce peri-tunnel bone loss as measured by CT (Lui et al. 2013). This prevention of local bone loss correlated with increased mechanical strength, but no differences in stiffness were observed. We used an unrestricted model and preserved the fibrous and callus tissue surrounding the tendon to increase external validity. The suture was removed prior to mechanical testing, to ensure that we assessed the healing and not the fixation and/or healing. We did not confirm the positive BP effects reported in these two studies (Thomopoulos et al. 2007, Lui et al. 2013). ZA treatment increased the native supraspinatus tendon failure stress and improved bone density at the rotator cuff footprint in ovariectomized rats after 12 weeks (Cadet et al. 2010), but may not be directly comparable to the tendon-tobone tunnel healing in our study. We chose the ZA dose, route, and delayed administration based on a study of fracture healing in rats (Amanat et al. 2007). They argued that delayed administration of the anticatabolic ZA would allow bone formation to establish, and

5/3/2018 11:53:40 AM

Acta Orthopaedica 2018; 89 (3): 360â&#x20AC;&#x201C;366

increase the net difference between anabolism and catabolism by an increased uptake of ZA in the forming callus. Delayed administration of ZA has been used also in clinical bone-healing studies for the same reasons (Harding et al. 2010, 2011). However, the optimal dose and timing of ZA treatment in the presence of a fracture is not firmly established, and for tendonâ&#x20AC;&#x201C;bone healing it has not been much investigated. There are some limitations to our model. The distal location for the bone tunnel on the tibia may provide limited healing potential with a reduced availability of marrow-derived mesenchymal stem cells due to the higher fat content of the marrow in the ankle region compared with the knee region. Nevertheless, the model may be relevant also for tendon transfers in foot/ankle surgery. No clinical studies of BPs on tendon-to-bone healing exist. BMD was found to be one of the independent factors predicting tendon-to-bone healing in rotator cuff surgery (Chung et al. 2011), and local bone loss in the humeral head probably leads to a further decrease in cuff tendon insertion strength (Chen et al. 2015). Thus, the rationale for using bone resorption inhibitors as adjuvant treatment in tendon-to-bone healing remains plausible. BPs in studies of fracture healing produce a larger callus and improved early mechanical fracture properties (Turker et al. 2016). In the early fracture-healing phase, a larger callus may explain an increased ability to withstand load. In the tendon-to-bone tunnel interface, the anatomical space for callus volume formation is limited. We were not able to observe delayed bone remodeling or prevention of local bone loss in the ZA-treated animals. If osteoclasts were inhibited by ZA a possible effect could be delayed transition from woven bone to lamellar bone, resulting in reduced tendon-to-bone healing. In summary, our study of early tendon-to-bone tunnel healing in rats demonstrated that a single dose of ZA reduced pullout strength and stiffness of the tendonâ&#x20AC;&#x201C;bone interface. Local bone loss was not evident, and no difference in BMD between ZA-treated animals and controls at 3 or 6 weeks was noted. Our study does not provide evidence supporting the use of ZA as adjuvant treatment in tendon-to-bone healing.

The authors would like to thank the Department of Comparative Medicine, Oslo University Hospital, Rikshospitalet for providing excellent animal facilities and enthusiastic personnel at our disposal. Senior Engineer Linda T. Dorg is acknowledged for excellent help with the histological work.

GAH, ES, LN, JEM, FPR, and SD were responsible for the design of the study. GAH performed surgery, DEXA measurements, mechanical testing, histological analyses, independent DEXA analyses, statistical calculations, and wrote the draft manuscript. ES performed surgery, DEXA measurements, and independent DEXA analyses. JEM and SD performed surgery and DEXA measurements. FPR supervised histological analyses. SN performed independent mechanical analyses. All authors approved the final version of the manuscript.

12072 Hjorthaug D.indd 365

365

Amanat N, Brown R, Bilston L E, Little D G. A single systemic dose of pamidronate improves bone mineral content and accelerates restoration of strength in a rat model of fracture repair. J Orthop Res 2005; 23(5): 102934. Amanat N, McDonald M, Godfrey C, Bilston L, Little D. Optimal timing of a single dose of zoledronic acid to increase strength in rat fracture repair. J Bone Miner Res 2007; 22(6): 867-76. Andersson T, Agholme F, Aspenberg P, Tengvall P. Surface immobilized zoledronate improves screw fixation in rat bone: a new method for the coating of metal implants. J Mater Sci Mater Med 2010; 21(11): 3029-37. Cadet E R, Vorys G C, Rahman R, Park S H, Gardner T R, Lee F Y, Levine W N, Bigliani L U, Ahmad C S. Improving bone density at the rotator cuff footprint increases supraspinatus tendon failure stress in a rat model. J Orthop Res 2010; 28(3): 308-14. Chen X, Giambini H, Ben-Abraham E, An K N, Nassr A, Zhao C. Effect of bone mineral density on rotator cuff tear: an osteoporotic rabbit model. PLoS One 2015; 10(10): e0139384. Chung S W, Oh J H, Gong H S, Kim J Y, Kim S H. Factors affecting rotator cuff healing after arthroscopic repair: osteoporosis as one of the independent risk factors. Am J Sports Med 2011; 39(10): 2099-107. Cole L E, Vargo-Gogola T, Roeder R K. Targeted delivery to bone and mineral deposits using bisphosphonate ligands. Adv Drug Deliv Rev 2016; 99(Pt A): 12-27. Dimmen S, Nordsletten L, Engebretsen L, Steen H, Madsen J E. The effect of parecoxib and indomethacin on tendon-to-bone healing in a bone tunnel: an experimental study in rats. J Bone Joint Surg Br 2009; 91(2): 259-63. Ekdahl M, Wang J H, Ronga M, Fu F H. Graft healing in anterior cruciate ligament reconstruction. Knee Surg Sports Traumatol Arthrosc 2008; 16(10): 935-47. Galatz L M, Rothermich S Y, Zaegel M, Silva M J, Havlioglu N, Thomopoulos S. Delayed repair of tendon to bone injuries leads to decreased biomechanical properties and bone loss. J Orthop Res 2005; 23(6): 1441-7. Gulotta L V, Rodeo S A. Biology of autograft and allograft healing in anterior cruciate ligament reconstruction. Clin Sports Med 2007; 26(4): 509-24. Harding A K, W-Dahl A, Geijer M, Toksvig-Larsen S, Tagil M. A single bisphosphonate infusion does not accelerate fracture healing in high tibial osteotomies. Acta Orthop 2011; 82(4): 465-70. Harding A K, Toksvig-Larsen S, Tagil M, W-Dahl A. A single dose zoledronic acid enhances pin fixation in high tibial osteotomy using the hemicallotasis technique: a double-blind placebo controlled randomized study in 46 patients. Bone 2010; 46(3): 649-54. Harryman D T, 2nd, Mack L A, Wang K Y, Jackins S E, Richardson M L, Matsen F A, 3rd. Repairs of the rotator cuff: correlation of functional results with integrity of the cuff. J Bone Joint Surg Am 1991; 73(7): 982-9. Hjorthaug G A, Madsen J E, Nordsletten L, Reinholt F P, Steen H, Dimmen S. Tendon to bone tunnel healing: a study on the time-dependent changes in biomechanics, bone remodeling, and histology in a rat model. J Orthop Res 2015; 33(2): 216-23. Li C, Mori S, Li J, Kaji Y, Akiyama T, Kawanishi J, Norimatsu H. Long-term effect of incadronate disodium (YM-175) on fracture healing of femoral shaft in growing rats. J Bone Miner Res 2001; 16(3): 429-36. Lui P P, Cheng Y Y, Yung S H, Hung A S, Chan K M. A randomized controlled trial comparing bone mineral density changes of three different ACL reconstruction techniques. The Knee 2012; 19(6): 779-85. Lui P P, Lee Y W, Mok T Y, Cheuk Y C, Chan K M. Alendronate reduced peri-tunnel bone loss and enhanced tendon graft to bone tunnel healing in anterior cruciate ligament reconstruction. Eur Cell Mater 2013; 25: 78-96. Rodeo S A, Kawamura S, Ma C B, Deng X H, Sussman P S, Hays P, Ying L. The effect of osteoclastic activity on tendon-to-bone healing: an experimental study in rabbits. J Bone Joint Surg Am 2007; 89(10): 2250-9. Russell R G, Watts N B, Ebetino F H, Rogers M J. Mechanisms of action of bisphosphonates: similarities and differences and their potential influence on clinical efficacy. Osteoporos Int 2008; 19(6): 733-59.

5/3/2018 11:53:43 AM

366

Silva M J, Boyer M I, Ditsios K, Burns M E, Harwood F L, Amiel D, Gelberman R H. The insertion site of the canine flexor digitorum profundus tendon heals slowly following injury and suture repair. J Orthop Res 2002; 20(3): 447-53. Silva MJ, Thomopoulos S, Kusano N, Zaegel MA, Harwood FL, Matsuzaki H, Havlioglu N, Dovan TT, Amiel D, Gelberman RH. Early healing of flexor tendon insertion site injuries: tunnel repair is mechanically and histologically inferior to surface repair in a canine model. J Orthop Res 2006; 24(5): 990-1000.

12072 Hjorthaug D.indd 366

Acta Orthopaedica 2018; 89 (3): 360â&#x20AC;&#x201C;366

Solberg L B, Stang E, Brorson SH, Andersson G, Reinholt F P. Tartrateresistant acid phosphatase (TRAP) co-localizes with receptor activator of NF-KB ligand (RANKL) and osteoprotegerin (OPG) in lysosomal-associated membrane protein 1 (LAMP1)-positive vesicles in rat osteoblasts and osteocytes. Histochem Cell Biol 2015; 143(2): 195-207. Thomopoulos S, Matsuzaki H, Zaegel M, Gelberman R H, Silva M J. Alendronate prevents bone loss and improves tendon-to-bone repair strength in a canine model. J Orthop Res 2007; 25(4): 473-9. Turker M, Aslan A, Cirpar M, Kochai A, Tulmac O B, Balci M. Histological and biomechanical effects of zoledronate on fracture healing in an osteoporotic rat tibia model. Eklem Hastalik Cerrahisi 2016; 27(1): 9-15.

5/3/2018 11:53:43 AM

Acta Orthopaedica 2018; 89

369 367

Information to authors Open Access/ Article Publication Charge Acta Orthopaedica, owned by the Nordic Orthopaedic Federation, is a non-profit, immediate Open Access journal, meaning your work is free for every-one to access online as soon as it is published (Gold OA). The standard research article publishing charge (APC) is EURO 1,750 (USD 2,000/GBP 1,520). For Case Reports the APC is EURO 875 (USD 1,000/GBP 760). Depending on your location these charges may be subject to local taxes. Note: There is no APC for members in the national orthopedic associations who are paying members of the Nordic Orthopaedic Federation (Denmark, Estonia, Finland, Iceland, Lithuania, Norway, and Sweden) Submission of manuscripts Manuscripts should be submitted online, www.actaorthop. org. Authors are encouraged to provide the names, addresses, and e-mail of potential reviewers. Note the following points: a. The manuscript should be prepared according to Uniform Requirements for Manuscripts submitted to Biomedical Journals (www.icmje.org/). American spelling is preferred. A manuscripts total word count should not exceed 3,300 (including Abstract (max. 250 words) but excluding title page, figure and table legends, and references). The word count should be added on the title page. b. Authors submitting a paper do so on the understanding that it has not been published and is not being considered for publication elsewhere. The authors should provide a statement about previous publications that are similar to the submitted study. Copies of such studies should be submitted with the paper. c. Authors should provide a description of their individual contributions to the study. d. Interventional studies (defined by human participants being prospectively assigned to one or more health-related interventions to evaluate the effects on health outcomes) should be registered in a public trials register before the enrolment of the first participant. The registration ID should be presented under ‘Ethics’. In addition, manuscripts presenting randomized trials should comply with the CONSORT statement (www.consort-statement.org). The study protocol, approved by the appropriate ethics committee, a completed CONSORT checklist and flowchart should be submitted together with the manuscript. Animal in vivo experiments should comply with ARRIVE and a completed checklist (www.nvcr3s.org.uk/ arrive-guidelines) and a copy of the study protocol should be submitted together with the manuscript. e. Systematic reviews and meta-analyses of interventional studies should comply with the PRISMA statement (www. prisma-statement.org), and the PRISMA checklist and flow-

chart should be appended to the manuscript. For systematic reviews and meta-analyses of observational studies the MOOSE checklist and flowchart (www.equator-network. org) should be used. f. Manuscripts presenting results on diagnostic accuracy should comply with the STARD-statement (www.stardstatement.org), and the STARD checklist and flowchart should be appended to the manuscript. g. Reports of observational studies should comply with the STROBE-statement (www.strobe-statement.org) and the STROBE checklist should be appended to the manuscript. h. Extensive tables, complementary figures, and other explanatory material should be given as Supplementary data which will be available in the online version of the article. i. All original data of relevance to the submitted manuscript should be stored and be retrievable on request for a minimum of 10 years. Authors are encouraged to make these data available via the internet. j. Acta does not publish studies that only assess completeness, coverage, accuracy, and validity of data in registers. Such studies are considered part of normal register administration, and the results should be presented in the register’s annual report. k. Acta does not publish studies reporting merely the existence or organization of a new register. Furthermore, to qualify for publication an epidemiologic/register study must include a question assessed by use of the register data, for example: outcome related to different treatments, treatment changes over time, or incidence changes over time. Manuscript layout Title. The title should include information on the results of the investigation, including the number of patients, average follow-up, and animal or cadaver experiments. The first name, middle initial(s), and last name of each author should be given with department affiliations as well as the name and e-mail of the corresponding author. Abstract. The abstract should not exceed 250 words. It should consist of 4 parts: Background and purpose, Patients/ material/animals and methods, Results, and Interpretation. Introduction. The introduction should focus on the state of knowledge at the beginning of the study. The aims and main hypothesis of the study should be stated clearly. Only in exceptional cases should it exceed 1 typewritten page. Case reports. These should be short, include the Case history and Discussion without an Introduction. Patients/material/animals and methods. The selection of subjects and the inclusion or exclusion criteria should be described. Subjects who declined to participate, withdrawals

IORT_GN_600744.indd 369

03-05-2018 17:31:41

370 368

and those with incomplete follow-up should be accounted for. Describe in detail how the measurements were made and the techniques used. Statistics. All statistical methods should be mentioned. Unusual methods should be referenced. The tests should be 2-sided, unless the use of 1-sided tests is justified. No data should be removed, imputed, weighted, adjusted or trimmed unless they are specifically described and justified and its consequences are given. Hypothesis tests (p-values). These tests should be used, in combination with a defined effect size and when statistical power has been considered. Give p-values with real numbers, if these are greater than 0.001 (1 digit except zeros), otherwise use ‘p < 0.001’. Do not use ‘ns’, ‘p > 0.05’ or asterisks. In most instances, the 95% confidence intervals should be given, especially in exploratory analyses and when estimating the effects or differences. Ethics, registration, funding, and potential conflicts of interest. Approval by the local ethics committee should be stated with date of issue and registration number when appropriate declare registration in a public trials register. Funding and potential conflicts of interest should be declared. Results. Do not give the same data in more than one way. When summarizing the data, always include measures of variability and the number of subjects. Give the median and range—e.g., 60 (35–70) years, the mean and standard deviation—59 (SD 15) years, and the frequencies for nominal data. The results of matched data should be given in relevant form (e.g., the distribution of pairwise differences). Percentages should not be used if the sample size is less than 100. Discussion. This section should contain a critical discussion of the results—e.g., the quality of the data (selection and information bias) and adequacy of the statistical analysis (confounding bias). It should also assess the relevant literature for or against the findings and if possible, the conclusions as regards clinical application or further research. Discuss, do not recapitulate, your results. Tables. Each table should be self-explanatory, with an adequate title and a logical presentation of the data. Abbreviate words in the columns and explain in footnotes. Each

IORT_GN_600744.indd 370

Acta Acta Orthopaedica Orthopaedica 2018; 2018; 89 89

column heading for numerical data should include the unit of measurement. Use SI units. Avoid unnecessary decimals! It is seldom advisable to use more than 2 digits for biologic measurements. Digital illustrations should be used. A scanned picture or digital photo should usually be about 85 mm (i.e. 1000 pixels) wide with a resolution of 300 DPI (dots per inch), which gives a square picture an uncompressed file size of 4.5 MiB in color and 1 MiB in gray-scale. A black and white drawing or graph may be scanned in 800 DPI bitmap, i.e. 1 bit TIFF. The preferred format is a TIFF-file. The resolution of WEB illustrations (gif) is usually too poor. Color photos should be in CMYK colors. The cost of color will be paid by Acta. Highlights, arrows and letters may be added to digital photos but keep a clean version of the photo for the layout process. Acta prefers to add these as a separate layer. For digital graphs, use a graphics program that can export vectorized graphs as EMF, EPS or PDF files. Avoid frames around diagrams, diagrams with perspective drawing, and bar graphs or histograms (use tables). Symbols should be consistent throughout a series of figures. Each axis should be horizontally labeled, with a description of the variable it represents. Use sans serif letters (Arial or Helvetica). Make diagrams in black-and-white with gray or color areas, but avoid complex patterns. Author contributions. Describe in short what each author did (initials). Acknowledgments. Technical help or other help may be acknowledged. References. We prefer the references to be cited by name and year (chronologically) of publication in the text instead of giving their numbers, i.e. 1 author, 2 authors or author et al. and year. Thus, the list of references should be given in alphabetical order. In the reference list all authors should be included. References with same author and year should be separated by adding a letter to the year (a, b, c...) in the list and the text. Submission Submit a double line spaced and line numbered Word file including tables and figures at the end of the manuscript.

03-05-2018 17:31:42