Issuu on Google+


haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation

Editor-in-Chief Jan Cools (Leuven)

Deputy Editor Luca Malcovati (Pavia)

Managing Director Antonio Majocchi (Pavia)

Associate Editors Hélène Cavé (Paris), Ross Levine (New York), Claire Harrison (London), Pavan Reddy (Ann Arbor), Andreas Rosenwald (Wuerzburg), Juerg Schwaller (Basel), Monika Engelhardt (Freiburg), Wyndham Wilson (Bethesda), Paul Kyrle (Vienna), Paolo Ghia (Milan), Swee Lay Thein (Bethesda), Pieter Sonneveld (Rotterdam)

Assistant Editors Anne Freckleton (English Editor), Cristiana Pascutto (Statistical Consultant), Rachel Stenner (English Editor), Kate O’Donohoe (English Editor), Ziggy Kennell (English Editor)

Editorial Board Omar I. Abdel-Wahab (New York); Jeremy Abramson (Boston); Paolo Arosio (Brescia); Raphael Bejar (San Diego); Erik Berntorp (Malmö); Dominique Bonnet (London); Jean-Pierre Bourquin (Zurich); Suzanne Cannegieter (Leiden); Francisco Cervantes (Barcelona); Nicholas Chiorazzi (Manhasset); Oliver Cornely (Köln); Michel Delforge (Leuven); Ruud Delwel (Rotterdam); Meletios A. Dimopoulos (Athens); Inderjeet Dokal (London); Hervé Dombret (Paris); Peter Dreger (Hamburg); Martin Dreyling (München); Kieron Dunleavy (Bethesda); Dimitar Efremov (Rome); Sabine Eichinger (Vienna); Jean Feuillard (Limoges); Carlo Gambacorti-Passerini (Monza); Guillermo Garcia Manero (Houston); Christian Geisler (Copenhagen); Piero Giordano (Leiden); Christian Gisselbrecht (Paris); Andreas Greinacher (Greifswals); Hildegard Greinix (Vienna); Paolo Gresele (Perugia); Thomas M. Habermann (Rochester); Claudia Haferlach (München); Oliver Hantschel (Lausanne); Christine Harrison (Southampton); Brian Huntly (Cambridge); Ulrich Jaeger (Vienna); Elaine Jaffe (Bethesda); Arnon Kater (Amsterdam); Gregory Kato (Pittsburg); Christoph Klein (Munich); Steven Knapper (Cardiff); Seiji Kojima (Nagoya); John Koreth (Boston); Robert Kralovics (Vienna); Ralf Küppers (Essen); Ola Landgren (New York); Peter Lenting (Le Kremlin-Bicetre); Per Ljungman (Stockholm); Francesco Lo Coco (Rome); Henk M. Lokhorst (Utrecht); John Mascarenhas (New York); Maria-Victoria Mateos (Salamanca); Simon Mendez-Ferrer (Madrid); Giampaolo Merlini (Pavia); Anna Rita Migliaccio (New York); Mohamad Mohty (Nantes); Martina Muckenthaler (Heidelberg); Ann Mullally (Boston); Stephen Mulligan (Sydney); German Ott (Stuttgart); Jakob Passweg (Basel); Melanie Percy (Ireland); Rob Pieters (Utrecht); Stefano Pileri (Milan); Miguel Piris (Madrid); Andreas Reiter (Mannheim); Jose-Maria Ribera (Barcelona); Stefano Rivella (New York); Francesco Rodeghiero (Vicenza); Richard Rosenquist (Uppsala); Simon Rule (Plymouth); Claudia Scholl (Heidelberg); Martin Schrappe (Kiel); Radek C. Skoda (Basel); Gérard Socié (Paris); Kostas Stamatopoulos (Thessaloniki); David P. Steensma (Rochester); Martin H. Steinberg (Boston); Ali Taher (Beirut); Evangelos Terpos (Athens); Takanori Teshima (Sapporo); Pieter Van Vlierberghe (Gent); Alessandro M. Vannucchi (Firenze); George Vassiliou (Cambridge); Edo Vellenga (Groningen); Umberto Vitolo (Torino); Guenter Weiss (Innsbruck).

Editorial Office Simona Giri (Production & Marketing Manager), Lorella Ripari (Peer Review Manager), Paola Cariati (Senior Graphic Designer), Igor Ebuli Poletti (Senior Graphic Designer), Marta Fossati (Peer Review), Diana Serena Ravera (Peer Review)

Affiliated Scientific Societies SIE (Italian Society of Hematology, www.siematologia.it) SIES (Italian Society of Experimental Hematology, www.siesonline.it)


haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation

Information for readers, authors and subscribers Haematologica (print edition, pISSN 0390-6078, eISSN 1592-8721) publishes peer-reviewed papers on all areas of experimental and clinical hematology. The journal is owned by a non-profit organization, the Ferrata Storti Foundation, and serves the scientific community following the recommendations of the World Association of Medical Editors (www.wame.org) and the International Committee of Medical Journal Editors (www.icmje.org). Haematologica publishes editorials, research articles, review articles, guideline articles and letters. Manuscripts should be prepared according to our guidelines (www.haematologica.org/information-for-authors), and the Uniform Requirements for Manuscripts Submitted to Biomedical Journals, prepared by the International Committee of Medical Journal Editors (www.icmje.org). Manuscripts should be submitted online at http://www.haematologica.org/. Conflict of interests. According to the International Committee of Medical Journal Editors (http://www.icmje.org/#conflicts), “Public trust in the peer review process and the credibility of published articles depend in part on how well conflict of interest is handled during writing, peer review, and editorial decision making”. The ad hoc journal’s policy is reported in detail online (www.haematologica.org/content/policies). Transfer of Copyright and Permission to Reproduce Parts of Published Papers. Authors will grant copyright of their articles to the Ferrata Storti Foundation. No formal permission will be required to reproduce parts (tables or illustrations) of published papers, provided the source is quoted appropriately and reproduction has no commercial intent. Reproductions with commercial intent will require written permission and payment of royalties. Detailed information about subscriptions is available online at www.haematologica.org. Haematologica is an open access journal. Access to the online journal is free. Use of the Haematologica App (available on the App Store and on Google Play) is free. For subscriptions to the printed issue of the journal, please contact: Haematologica Office, via Giuseppe Belli 4, 27100 Pavia, Italy (phone +39.0382.27129, fax +39.0382.394705, E-mail: info@haematologica.org). Rates of the International edition for the year 2017 are as following: Print edition

Institutional Euro 500

Personal Euro 150

Advertisements. Contact the Advertising Manager, Haematologica Office, via Giuseppe Belli 4, 27100 Pavia, Italy (phone +39.0382.27129, fax +39.0382.394705, e-mail: marketing@haematologica.org). Disclaimer. Whilst every effort is made by the publishers and the editorial board to see that no inaccurate or misleading data, opinion or statement appears in this journal, they wish to make it clear that the data and opinions appearing in the articles or advertisements herein are the responsibility of the contributor or advisor concerned. Accordingly, the publisher, the editorial board and their respective employees, officers and agents accept no liability whatsoever for the consequences of any inaccurate or misleading data, opinion or statement. Whilst all due care is taken to ensure that drug doses and other quantities are presented accurately, readers are advised that new methods and techniques involving drug usage, and described within this journal, should only be followed in conjunction with the drug manufacturer’s own published literature. Direttore responsabile: Prof. Edoardo Ascari; Autorizzazione del Tribunale di Pavia n. 63 del 5 marzo 1955. Printing: Tipografia PI-ME, via Vigentina 136, Pavia, Italy. Printed in January 2017.


haematologica calendar of events

Journal of the European Hematology Association Published by the Ferrata Storti Foundation

EHA Scientific Meeting on Anemias Diagnosis and Treatment in the Omics Era Chair: A Iolascon February 2-4, 2017 Barcelona, Spain

43rd Annual Meeting of the EBMT European Society for Blood and Marrow Transplantation (EBMT) March 26-29, 2017 Marseille, France

EuroClonality Workshop: “Clonality assessment in Pathology” European Scientific foundation for Laboratory Hemato Oncology (ESLHO) Chairs: P Groenen, J van Krieken, A Langerak February 13-15, 2016 Nijmegen, The Netherlands

22nd Congress of the European Hematology Association European Hematology Association June 22 - 25, 2017 Madrid, Spain

EHA Hematology Tutorial on Lymphoid malignancies, Multiple myeloma and Bone Marrow Failure Chairs: R Foà , K Wickramaratne February 23-24, 2017 Colombo, Sri Lanka

EHA Hematology Tutorial on Lymphoid Malignancies Chairs: R Foà, I Hus, T Robak March 17-18, 2017 Warsaw, Poland

EHA Scientific Meeting on Advances in Biology and Treatment of B Cell Malignancies with a Focus on Rare Lymphoma Subtypes Chairs: M Kersten and M Dreyling March 10-12, 2017 Barcelona, Spain

EHA Scientific Meeting on Challenges in the Diagnosis and Management of Myeloproliferative Neoplasms Chairs: J Kiladjian and C Harrison October 12-14, 2017 Location: TBC

EHA Scientific Meeting on Shaping the Future of Mesenchymal Stromal Cells Therapy Chair: W Fibbe November 23-25, 2017 Location: TBC

Calendar of Events updated on January 6, 2017


haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation

Table of Contents Volume 102, Issue 2: February 2017 Cover Figure Artistic view on neutrophil extracellular trap(NET)-mediated thrombosis. (Image created by www.somersault1824.com)

Editorials 203

Research at the heart of hematology: thrombocytopenias and platelet function disorders Carlo L. Balduini and Federica Melazzini

Review Articles 206

Propagation of thrombosis by neutrophils and extracellular nucleosome networks Susanne Pfeiler et al.

214

Cure for thalassemia major – from allogeneic hematopoietic stem cell transplantation to gene therapy Alok Srivastava and Ramachandran V. Shaji

224

Anti-thymocyte globulin as graft-versus-host disease prevention in the setting of allogeneic peripheral blood stem cell transplantation: a review from the Acute Leukemia Working Party of the European Society for Blood and Marrow Transplantation Frédéric Baron et al.

Guideline Article 235

Guidelines for diagnosis, prevention and management of central nervous system involvement in diffuse large B-cell lymphoma patients by the Spanish Lymphoma Group (GELTAMO) Francisco-Javier Peñalver et al.

Articles Red Cell Biology & Its Disorders

246

ARQ 092, an orally-available, selective AKT inhibitor, attenuates neutrophil-platelet interactions in sickle cell disease Kyungho Kim et al.

Iron Metabolism & Its Disorders

260

Hemolytic anemia repressed hepcidin level without hepatocyte iron overload: lesson from Günther disease model Sarah Millot et al.

Coagulation & Its Disorders

271

The C1 and C2 domains of blood coagulation factor VIII mediate its endocytosis by dendritic cells Bagirath Gangadharan et al.

Haematologica 2017; vol. 102 no. 2 - February 2017 http://www.haematologica.org/


haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation Platelet Biology & Its Disorders

282

Germline variants in ETV6 underlie reduced platelet formation, platelet dysfunction and increased levels of circulating CD34+ progenitors Marjorie Poggi et al.

Bone Marrow Failure

295

Necroptosis in spontaneously-mutated hematopoietic cells induces autoimmune bone marrow failure in mice Junping Xin et al.

Myelodysplastic Syndrome

308

Immunophenotypic analysis of erythroid dysplasia in myelodysplastic syndromes. A report from the IMDSFlow working group Theresia M. Westers, et al.

320

Implementation of erythroid lineage analysis by flow cytometry in diagnostic models for myelodysplastic syndromes Eline M.P. Cremers et al.

Myeloproliferative Disorders

327

Primary analysis of a phase II open-label trial of INCB039110, a selective JAK1 inhibitor, in patients with myelofibrosis John O. Mascarenhas et al.

Acute Myeloid Leukemia

336

Distinct global binding patterns of the Wilms tumor gene 1 (WT1) -KTS and +KTS isoforms in leukemic cells Tove Ullmark et al.

Acute Lymphoblastic Leukemia

346

The role of ZAP70 kinase in acute lymphoblastic leukemia infiltration into the central nervous system Ameera Alsadeq et al.

Non-Hodgkin Lymphoma

356

Early treatment intensification with R-ICE and 90Y-ibritumomab tiuxetan (Zevalin)-BEAM stem cell transplantation in patients with high-risk diffuse large B-cell lymphoma patients and positive interim PET after 4 cycles of R-CHOP-14 Mark Hertzberg et al.

364

Bone marrow findings in autoimmune lymphoproliferative syndrome with germline FAS mutation Yi Xie et al.

373

Inhibition of demethylase KDM6B sensitizes diffuse large B-cell lymphoma to chemotherapeutic drugs Rohit Mathur et al.

Plasma Cell Disorders

381

Due to interleukin-6 type cytokine redundancy only glycoprotein 130 receptor blockade efficiently inhibits myeloma growth Renate Burger et al.

Cell Therapy & Immunotherapy

391

Comparable composite endpoints after HLA-matched and HLA-haploidentical transplantation with post-transplantation cyclophosphamide Shannon R. McCurdy et al.

Cell therapy and Immunotherapy

401

Post-transplant cyclophosphamide versus anti-thymocyte globulin as graft-versus-host disease prophylaxis in haploidentical transplant Annalisa Ruggeri et al.

Haematologica 2017; vol. 102 no. 2 - February 2017 http://www.haematologica.org/


haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation

Letters to the Editor Letters are available online only at www.haematologica.org/content/102/2.toc

e33

Carboxy-terminal fragment of fibroblast growth factor 23 induces heart hypertrophy in sickle cell disease Marie Courbebaisse et al. http://www.haematologica.org/content/102/2/e33

e36

Decreased numbers of dense granules in fetal and neonatal platelets Denisa Urban et al. http://www.haematologica.org/content/102/2/e36

e39

Response dynamics of pediatric patients with chronic myeloid leukemia on imatinib therapy Rick Proschmann et al. http://www.haematologica.org/content/102/2/e39

e43

Association between the TP53 Arg72Pro polymorphism and clinical outcomes in acute myeloid leukemia Matheus F. Bezerra et al. http://www.haematologica.org/content/102/2/e43

e47

Low-dose clofarabine in combination with a standard remission induction in patients aged 18-60 years with previously untreated intermediate and bad-risk acute myeloid leukemia or high-risk myelodysplastic syndrome: combined phase I/II results of the EORTC/GIMEMA AML-14A trial Dominik Selleslag et al. http://www.haematologica.org/content/102/2/e47

e52

Evolution of disease activity and biomarkers on and off rapamycin in 28 patients with autoimmune lymphoproliferative syndrome Christian Klemann et al. http://www.haematologica.org/content/102/2/e52

e57

Comprehensive translocation and clonality detection in lymphoproliferative disorders by next-generation sequencing DĂśrte Wren et al. http://www.haematologica.org/content/102/2/e57

e61

Higher adherence to the Mediterranean diet is associated with lower levels of D-dimer: findings from the MOLI-SANI study Augusto Di Castelnuovo et al. http://www.haematologica.org/content/102/2/e61

e65

The chemokine CXCR2 antagonist (AZD5069) preserves neutrophil-mediated host immunity in non-human primates Mohib Uddin et al. http://www.haematologica.org/content/102/2/e65

Case Report This case report is available online only at www.haematologica.org/content/102/2.toc

e69

Human RAD52 – a novel player in DNA repair in cancer and immunodeficiency Sujal Ghosh et al. http://www.haematologica.org/content/102/2/e69

Haematologica 2017; vol. 102 no. 2 - February 2017 http://www.haematologica.org/


EDITORIALS Research at the heart of hematology: thrombocytopenias and platelet function disorders Carlo L. Balduini and Federica Melazzini Department of Medicine, Università of Pavia and Fondazione IRCCS Policlinico San Matteo, Pavia, Italy E-mail: c.balduini@smatteo.pv.it

doi:10.3324/haematol.2016.158055

N

atural selection drove the evolution of all living beings for millions of years until mankind acquired the ability to interfere with this mechanism by counteracting the natural hazards that tend to eliminate the weakest humans and select the strongest ones for the continuation of the species. Medicine, which has acquired the ability to cure many common diseases that were previously lethal, is an important result of the attempt to oppose natural selection. Medical progress was initially very slow, but with the advent of the scientific revolution in the 18th century, this has continued to get faster and has played a role in doubling the life expectancy in Western countries since the beginning of the 19th century.1 Strenuous research, brilliant insights and, much more rarely, serendipity, were at the basis of medical progress. Importantly, some major advances in medicine derived from a widening of knowledge in other disciplines. For example, the development of the microscope (the “occhiolino” or “little eye” of Galileo Galilei) in the early 17th century was the prerequisite for recognizing blood cells and starting to identify blood disorders. Thus, a major advance in optical science made the birth of hematology possible. Platelets, because of their small size and the limited resolution of early microscopes, escaped identification for a long time, and when, in 1735, the German physician and poet Paul Gottlieb Werlhof provided the first detailed description of ‘morbus maculosus haemorrhagicus’,2 now known as immune thrombocytopenia (ITP), these blood cells were unknown (Figure 1). The discovery of platelets had to wait until 1882, when the Italian pathologist Giulio Bizzozzero, also thanks to a major technical improvement in microscope technology (correction of chromatic aberration), described in detail these small elements and the relationship between platelet adhesion and aggregation, and the subsequent fibrin formation and deposition.3 A year after the brilliant insight of Bizzozero, the link between thrombocytopenia and ITP was identified by Brohm.4 So, 150 years after its description, we finally had an explanation for the 'morbus maculosus haemorrhagicus'. Also subsequent advances in knowledge of ITP were made thanks to ingenious intuition and laborious research. The intuition of Kaznelson, in 1916, that the spleen was responsible for platelet destruction led to the identification of splenectomy as an effective treatment for this disease.5 A few decades later, in 1951, Harrington observed a child with transient purpura born to a mother with ITP and suspected that the passage of a humoral factor from mother to child was responsible for platelet destruction.6 In the same year, Evans hypothesized that the ITP had an immune genesis,7 paving the way to immunosuppressants as an effective treatment for this disorders.8 Finally, the search for the humoral substance responsible for causing the platelet count to rise in response to thrombocytopenia, which had involved many groups of researchers for many decades, led in 1994 to haematologica | 2017; 102(2)

the purification and cloning of thrombopoietin (TPO).9 This achievement opened the way to the development of TPO mimetics, which have proven effective not only in ITP, but also in other forms of thrombocytopenia and in bone marrow aplasia. Can we assume that, after three centuries of investigation, the history of ITP is over and there is no longer a need to allocate resources to the study of this disease? We believe that the answer is “surely not”! Just think that the diagnosis of ITP is still a process of exclusion because we have no sensitive and specific laboratory tests for this condition. As a consequence, both patients with ITP and those with other forms of isolated thrombocytopenia are at risk of misdiagnosis and unnecessary treatments. For instance, patients with inherited

Figure 1. The long history of immune thrombocytopenia (ITP). After nearly 300 years of research, we now know the main pathogenetic mechanisms of ITP and have several effective treatments. However, further investigation is required to make the diagnosis of this disease easier and more reliable, to better understand the pathogenic mechanism in the individual patient and personalize accordingly the therapeutic approach.

203


Editorials

thrombocytopenias (ITs) are often misdiagnosed with ITP and many of them receive useless immunosuppressant drugs or even splenectomy.10 Another goal yet to be achieved is personalization of treatment. We currently have different treatment options, each of them effective in a variable proportion of patients, but we are still unable to predict to which one(s) each patient will respond. Moreover, we are unable to recognize subjects whose disease will go into remission spontaneously and therefore might benefit from treatment as limited as possible. Finally, we are also still unable to identify patients requiring immediate treatment because they are at risk of clinically relevant bleeding episodes, since the degree of thrombocytopenia is not always effective in this respect. These uncertainties are primarily derived from the fact that the pathogenic mechanisms of ITP, probably different in different patients, are not yet fully known. So, as recognized by the “EHA Roadmap for European Hematology Research” recently published in this Journal,11 both basic research and clinical studies are required to further improve care for ITP patients. While advances in the field of ITP occurred slowly over centuries, knowledge of ITs have been increasing exponentially in recent years due to the huge improvement in technologies for gene sequencing. Up to 20 years ago, only a few ITs were well defined from a clinical point of view and only for 4 of them was the genetic defect known. Now, we know more than 30 diseases caused by mutations in different genes, with half of these disorders identified in the last five years (Figure 2). However, nearly 50% of patients with ITs have forms that do not fit the criteria for any known disorder,12 and identification of these 'new' disorders is required for several reasons. First of all, we recently realized that nearly 50% of patients with known ITs are at risk of because of acquiring additional illnesses during life, such as bone marrow aplasia, kidney failure or hematologic malignancies, which endanger patients' lives much more than thrombocytopenia itself. Of note, the 2016 revision of the WHO classification of myeloid neoplasms and acute leukemias introduced a new category of diseases defined as 'Myeloid neoplasms with germ line predisposition and pre-existing platelet disorders', which includes neoplasms developing in patients with ITs caused by mutations in the genes RUNX1, ANKRD26 and ETV6. We do not know whether some of the yet unknown forms of IT expose patients to risks other than those deriving from thrombocytopenia, and only their identification and characterization will allow us to answer this important question. Another reason to keep searching for new diseases is that affected subjects could benefit from treatments that are effective in some known ITs. For instance, it is possible that, as already shown for Wiskott-Aldrich syndrome,13 ANKRD26-related thrombocytopenia and MYH9-related disease,14 TPO mimetics increase the number of platelets even in some yet unknown forms of IT and can therefore be used instead of platelet transfusions to prepare patients to surgery. Huge improvements have been obtained in recent years also for other forms of thrombocytopenia, but, as for ITP and ITs, further research is required to respond to the 204

Figure 2. Identification of the etiology of inherited thrombocytopenias. Until the 1980s, only a very few forms of inherited thrombocytopenia had been identified on the basis of their peculiar clinical pictures. Subsequently, the availability of more and more effective gene sequencing techniques resulted in the unveiling of the etiology of the already known forms and the identification of an increasing number of new diseases. Of note, 14 new causative genes have been identified since 2010. We presently know more than 30 forms of inherited thrombocytopenias deriving from mutations in different genes, but despite this explosion of knowledge, almost 50% of patients still remain without a diagnosis because their illnesses have not yet been identified. AML: acute myeloid leukemia; related disorder (RD) and related thrombocytopenias (RT) suffixes refer to syndromic and non-syndromic inherited thrombocytopenias (respectively) deriving from mutations in the genes indicated in italics.

many unanswered questions and improve patient care. We refer interested readers to the “EHA Roadmap for European Hematology Research”, where this matter is discussed in detail by leading experts in different thrombocytopenic disorders.11 At variance with thrombocytopenias, past and recent achievements in the field of defects of platelet function have been limited and knowledge of these conditions is still very unsatisfactory. This is surprising, since both acquired and inherited platelet dysfunctions are more prevalent than the corresponding forms of thrombocytopenia. For instance, it has been shown that many common liver and kidney disorders affect platelet function and may result in bleeding tendency,15 but we have little information on this matter, and both diagnosis and treatment of these conditions are poorly defined. Concerning inherited forms, most patients still receive diagnoses, as 'primary secretion defect' or 'granule disorder', that are merely haematologica | 2017; 102(2)


Editorials

descriptive, because the underlying genetic defects, as well as the pathogenic mechanisms, are unknown.16 A major problem that hindered progress in the field is that the definition itself of platelet dysfunction has remained vague. This has been because, despite recent and important efforts to reach a consensus among experts,16,17 we do not have validated criteria for recognizing and classifying these conditions. Moreover, our ability to predict the risk of bleeding in the individual patient remains limited and, therefore, we are often in doubt as to whether or not to administer prophylactic treatments in situations of hemostatic challenges, such as giving birth or surgery. Finally, also the therapeutic armamentarium for these conditions is still very limited and, in many cases, there is no clear evidence of efficacy. Knowledge of functional platelet defects, therefore, requires a big leap forward to be made. A major advance would be the identification of a comprehensive laboratory approach for diagnosing these disorders. Acquiring the ability to identify the bleeding risk of each patient by in vitro techniques would be even better. These are very ambitious goals, the achievement of which requires the joint evaluation of the clinical and laboratory features of large series of patients. Organizing these kind of studies is challenging, especially for the inherited forms that need international collaboration for achieving a sufficient sample size. However, reaching these objectives would represent the starting point for further studies designed to identify both the pathogenic mechanisms of these conditions and new effective treatments. Moreover, the application of next generation sequencing techniques to large series of well characterized patients is expected to identify a large number of inherited platelet dysfunctions that are still waiting to be recognized. Initial evidence of the effectiveness of this approach has been recently published.18,19 In conclusion, as in many other fields of medicine, also in the area of thrombocytopenias and platelet function disorders we are observing the apparent paradox that the advance of knowledge increases the number of questions to be answered instead of reducing it. So, theoretically, it is to be expected that medical research and the consequent improvement of human health is a never-ending story.

haematologica | 2017; 102(2)

References 1. Oeppen J, Vaupel JW. Demography. Broken limits to life expectancy. Science. 2002;296(5570):1029-1031. 2. Werlhof PG. Disquisitio medica et philologica de variolis et anthracibus, signis differentiis, medelis disserit etc. Nicolai Foersteri. Hannover, 1735. 3. Bizzozero G. [Ueber einen neuen Forrnbestandteil des Blutes und dessen Rolle bei der Thrombose und Blutgerinnung. Virchows Archiv fur Pathologische Anatomie und Physiologie und fur Klinische Medizin]. 1882;90:261-332. 4. Brohm F. Quoted by Krauss, E. Über purpura. Inaugural dissertation. Heidelberg, 1883. 5. Kaznelson P. [Verschwinden der ha¨morrhagischen Diathese bei einem Falle von ‘‘essentieller Thrombopenie’’ (Frank) nach Milzexstirpation: splenogene thrombolytische Purpura]. Wiener Klinische Wochenschrift. 1916;29:1451-1454. 6. Harrington WJ, Minnich V, Hollingsworth JW, Moore CV. Demonstration of a thrombocytopenic factor in the blood of patients with thrombocytopenic purpura. J Lab Clin Med. 1951;38:1-10. 7. Evans RS, Takahashi K, Duane RT, Payne R, Liu C. Primary thrombocytopenic purpura and acquired hemolytic anemia; evidence for a common etiology. AMA Arch Intern Med. 1951;87(1):48-65. 8. Wintrobe MM, Cartwright GE, Palmer JG, Kuhns WJ, Samuels LT. Effect of corticotrophin and cortisone on the blood in various disorders in man. AMA Arch Intern Med. 1951;88(3):310-336. 9. Schick BP. Hope for treatment of thrombocytopenia. N Engl J Med. 1994;331(13):875-876. 10. Balduini CL, Savoia A, Seri M. Inherited thrombocytopenias frequently diagnosed in adults. J Thromb Haemost. 2013;11(6):1006-1019. 11. Engert A, Balduini C, Brand A et al. The European Hematology Association Roadmap for European Hematology Research: a consensus document. Haematologica. 2016;101(2):115-208. 12. Balduini CL, Noris P. Innovation in the field of thrombocytopenias: achievements since the beginning of the century and promises for the future. Haematologica. 2016;101(1):2-4. 13. Gerrits AJ, Leven EA, Frelinger AL 3rd, et al. Effects of eltrombopag on platelet count and platelet activation in Wiskott-Aldrich syndrome/Xlinked thrombocytopenia. Blood. 2015;126(11):1367-1378. 14. Pecci A, Gresele P, Klersy C, et al. Eltrombopag for the treatment of the inherited thrombocytopenia deriving from MYH9 mutations. Blood. 2010;116(26):5832-5837. 15. Jalal DI, Chonchol M, Targher G. Disorders of hemostasis associated with chronic kidney disease. Semin Thromb Hemost. 2010;36:34-40. 16. Gresele P; Subcommittee on Platelet Physiology of the International Society on Thrombosis and Hemostasis. Diagnosis of inherited platelet function disorders: guidance from the SSC of the ISTH. J Thromb Haemost. 2015;13(2):314-322. 17. Cattaneo M, Cerletti C, Harrison P, et al. Recommendations for the Standardization of Light Transmission Aggregometry: A Consensus of the Working Party from the Platelet Physiology Subcommittee of SSC/ISTH. J Thromb Haemost. 2013;11:1183-1189. 18. Simeoni I, Stephens JC, Hu F, et al. A high-throughput sequencing test for diagnosing inherited bleeding, thrombotic, and platelet disorders. Blood. 2016;127(23):2791-2803. 19. Johnson B, Lowe GC, Futterer J, et al; UK GAPP Study Group. Whole exome sequencing identifies genetic variants in inherited thrombocytopenia with secondary qualitative function defects. Haematologica. 2016;101(10):1170-1179.

205


REVIEW ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Propagation of thrombosis by neutrophils and extracellular nucleosome networks Susanne Pfeiler,1 Konstantin Stark,2 Steffen Massberg2 and Bernd Engelmann1

1 Institut fßr Laboratoriumsmedizin and 2Medizinische Klinik und Poliklinik I, LudwigMaximilians-Universität, Munich, Germany

Haematologica 2017 Volume 102(2):206-213

ABSTRACT

N

Correspondence: Susanne.Pfeiler@med.uni-muenchen.de or Konstantin.Stark@med.uni-muenchen.de

Received: April 29, 2016. Accepted: August 17, 2016. Pre-published: December 7, 2016.

eutrophils, early mediators of the innate immune defense, are recruited to developing thrombi in different types of thrombosis. They amplify intravascular coagulation by stimulating the tissue factor-dependent extrinsic pathway via inactivation of endogenous anticoagulants, enhancing factor XII activation or decreasing plasmin generation. Neutrophil-dependent prothrombotic mechanisms are supported by the externalization of decondensed nucleosomes and granule proteins that together form neutrophil extracellular traps. These traps, either in intact or fragmented form, are causally involved in various forms of experimental thrombosis as first indicated by their role in the enhancement of both microvascular thrombosis during bacterial infection and carotid artery thrombosis. Neutrophil extracellular traps can be induced by interactions of neutrophils with activated platelets; vice versa, these traps enhance adhesion of platelets via von Willebrand factor. Neutrophil-induced microvascular thrombus formation can restrict the dissemination and survival of blood-borne bacteria and thereby sustain intravascular immunity. Dysregulation of this innate immune pathway may support sepsis-associated coagulopathies. Notably, neutrophils and extracellular nucleosomes, together with platelets, critically promote fibrin formation during flow restriction-induced deep vein thrombosis. Neutrophil extracellular traps/extracellular nucleosomes are increased in thrombi and in the blood of patients with different vaso-occlusive pathologies and could be therapeutically targeted for the prevention of thrombosis. Thus, during infections and in response to blood vessel damage, neutrophils and externalized nucleosomes are major promoters of intravascular blood coagulation and thrombosis.

doi:10.3324/haematol.2016.142471

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/206

Š2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

206

Introduction Blood neutrophils are among the first immune cells that are recruited to pathogen-infected tissues and sterile injuries.1 After extravasation, and with the assistance of complement, neutrophils ingest pathogens and kill them inside intracellular phagolysosomal granules or, as in the case of sterile injuries, engulf the cellular debris to degrade it. Besides their well-documented functions in combating tissue-based infections and injuries, neutrophils are also involved in restricting blood-based infections. Recent evidence suggests that neutrophils can be central components of intravascular immunity in response to circulating pathogens. Indeed, neutrophils help to prevent circulating pathogens from spreading and are also involved in intravascular microbicidal activities.2,3 To protect the host against pathogens, neutrophils expel neutrophil extracellular traps (NET).4 NET are mainly formed by decondensed nucleosomes and proteins derived from intracellular granules, such as neutrophil elastase (NE) and myeloperoxidase. Thus, apart from their known capacity to restrict infections by intracellular mechanisms, neutrophils also use extracellular tools to protect the host from infection-induced damage. haematologica | 2017; 102(2)


Thrombosis propagation by neutrophils and extracellular nucleosome

Activation of intravascular blood coagulation and the formation of thrombi in microvessels (microvascular thrombosis) under certain conditions support a distinct mechanism of intravascular immunity named immunothrombosis.2 This biological form of “protective� thrombosis immobilizes circulating bacteria, restricts tissue invasion, and limits the survival of circulating bacteria in organs such as the liver and spleen. NET/extracellular nucleosomes were identified as major effectors of intravascular immunity supported by microvascular thrombosis in mice in vivo.5 In parallel, it was shown that NET derived from neutrophils are present at sites of pathological thrombus formation in large arteries in experimental mouse models as well as in coronary thrombi of patients.6 Notably, a substantial fraction of neutrophilderived extracellular nucleosomes did not exhibit the typical morphology of NET and were present in fragmented forms. Apart from being causally involved in microvascular thrombosis as part of the physiological host response to bacterial infection, extracellular nucleosomes, the major constituents of NET, were shown to critically promote the development of arterial thrombosis in mouse models in vivo.5 Besides this, NET were also detected in animal models of deep vein thrombosis and were shown to enhance thrombus formation in vivo.7,8 Thus, neutrophils and NET/extracellular nucleosomes are crucial promoters of thrombosis under physiological and pathological conditions, in different vascular beds, and under diverse conditions of vessel injury and infection. This review summarizes the mechanisms supporting propagation of thrombosis by neutrophils. In particular, we discuss how NET/nucleosomes are generated/externalized at the cellular level, demonstrate the molecular events supporting their procoagulant functions, as well as emphasize the critical role of neutrophils in the activation of various types of thrombosis in vivo.

Mechanisms of chromatin release from activated/dead cells and host defense functions of neutrophil extracellular traps Various triggers such as cytokines, bacterial components, the experimental agonist phorbol-12-myristate-13acetate, or activated platelets can stimulate the activation of neutrophils (Table 1), which in turn leads to degranula-

Table 1. Triggers of NETosis.

Micro-organisms as trigger S. aureus67 S. pyogenes11 M. tuberculosis68 S. flexneri4 E. coli70 Apicomplexan species (e.g. T. gondii)71 Leishmania species73 C. albicans74 Asperigillus species76 Influenza77 Human immunodeficiency virus-149

Sterile trigger High mobility group box 114 Heme45 Tumor necrosis factor69 Interleukin-84 Interleukin-5 + C5a31 GM-CSF + C5a72 Interferon + C5a72 ANCAS75 PMA4 H2O278

C5a: complement component 5a; GM-CSF: granulocyte-macrophage colony-stimulating factor; ANCAS: anti-neutrophil cytoplasmic antibody; PMA: phorbol 12-myristate 13-acetate.

haematologica | 2017; 102(2)

tion and the concomitant release of decondensed chromatin fibers (designated as NET) into extracellular compartments.9 Such fibers, which can also be formed by eosinophils and mast cells and greatly exceed the size of the neutrophils themselves, enable trapping, and eventually, killing of micro-organisms. The process of NET formation, also called NETosis, is dependent on the enzyme peptidylarginine deiminase-4 (PAD4).10 PAD4 catalyzes the conversion of histone-associated arginine residues into the non-canonical amino acid citrulline. This is mediated by the conversion of the imino group of arginine into a keto-group. Replacement of arginine residues by citrullines in turn causes dissociation of histones from the tightly packed DNA backbone that is wrapped around unmodified histones, and thereby induces chromatin decondensation. In parallel, nuclear membranes begin to vesiculate and neutrophil granules disintegrate. Thereby, granule proteins, including myeloperoxidase, come into contact with nuclear components such as chromatin. Myeloperoxidase is a major trigger for NET generation in addition to PAD4.9 Further molecules involved in NET formation remain to be identified. Disintegration of the nuclei and granules allows the fusion of different intracellular membranes. The cells round up and in an abrupt event, the decondensed chromatin, together with various granule components, is expelled to form NET. NET were found to capture Grampositive and Gram-negative bacteria and to be responsible for microbicidal activities in vitro.4 In line with this, neutrophils isolated from PAD4-deficient mice (PAD4-/-) showed reduced NET formation and microbicidal activities. In addition, these mice were found to be more susceptible to infection with S. pyogenes in vivo.11 Formation of NET does not necessarily result in neutrophil death. Indeed, following NET formation the generated neutrophil fragments have been described to be present in abscesses in vivo.12 Apart from activated cells, also dying or dead cells, in particular apoptotic cells, can release nucleosomes.13 Extracellular nucleosomes, especially when complexed with high mobility group box protein 1, can activate antigen-presenting cells, including dendritic cells, and thereby disrupt tolerance against nucleosomes/double-stranded DNA which might favor autoimmune diseases such as systemic lupus erythematosus.14 Stimulation of NETosis in lupus, which might be supported by reactive oxygen species originating from mitochondria, could well contribute to the increased risk of both arterial and venous thrombosis in this disease.15 In the case of tumor cells, nucleosome release has been shown to occur from both apoptotic and necrotic cells.16 The molecular mechanisms supporting release of nucleosomes from apoptotic and late apoptotic cells are largely unknown. It has been suggested that the serine protease factor seven activating protease (FSAP or hyaluronic acid binding protein-2) plays a major role in the release reaction,17 whereby FSAP is found associated with the released nucleosomes. Similarly to RNA, the DNA components of nucleosomes could also promote auto-activation of FSAP, which in turn might contribute to some of the prothrombotic actions of extracellular nucleosomes, given that FSAP supports arterial thrombosis.18, 19

Extracellular nucleosomes enhance blood coagulation and platelet activation Platelets are increasingly recognized as critical players in 207


S. Pfeiler et al.

immune responses.20-22 Their immunoregulatory effects are in part related to platelet interactions with innate immune cells such as neutrophils and monocytes. In line with this, activated platelets are substantially involved in the formation of NET by neutrophils.5,23 This can be mediated by high mobility group box protein 1 exposed on the surface of activated platelets and by adhesive interactions of platelet P-selectin with neutrophil P-selectin glycoprotein ligand-1.24,25 The expelled NET in turn allow the adhesion of additional platelets and promote their activation, mediated in particular by histones.7,26 NET likely also bind von Willebrand factor as well as fibrinogen. Interactions between activated platelets and activated neutrophils can result in an activation of the extrinsic pathway of coagulation both in human and mouse systems.5,27 Fibrin formation via the extrinsic pathway is centrally initiated by the cell membrane protein tissue factor (TF). P2Y12-mediated platelet activation by the platelet agonist ADP plays a critical role in induction of TF activity by platelet-neutrophil interactions.28 Such procoagulant activity can originate in principle from neutrophil TF, platelet TF or both.27,29 Nonetheless, so far no consensus has been reached regarding the functional relevance of neutrophil and platelet TF. Neutrophils and platelets release microparticles that have been suggested to express TF and to be trapped by NET.30 In line with this, TF has been detected in association with NET inside venous thrombi in vivo.8 Eosinophils, which expel DNA-histone complexes upon activation as well, serve as a major intravascular pool of TF.31,32 Thus, under certain conditions, neutrophils and eosinophils might co-expose TF and extracellular DNA. To prevent pathological vessel occlusion due to intravascular blood coagulation, it is essential that the potent procoagulant role of TF can be silenced under physiological conditions. Suppression of TF activation is mediated by different mechanisms, including proteins that prevent the translocation of phosphatidylserine to the outer leaflet of the plasma membrane, given that phosphatidylserine exposure activates TF.33 Protein disulfide isomerase serves as another mechanism to promote the extrinsic pathway by oxidation (internal disulfide bond formation) of TF.34 Protein disulfide isomerase, which can be released from activated platelets and from endothelial cells at sites of vascular injury, has been shown to be associated with NET in deep vein thrombosis.8 Activation of the extrinsic pathway of blood coagulation by TF is controlled by the natural anticoagulant protein tissue factor pathway inhibitor (TFPI), which directly inhibits coagulation factors VIIa and Xa involving the Kunitz 1 and 2 domains of TFPI. To disinhibit TF-driven blood coagulation, neutrophil serine proteases such as NE and cathepsin G, locally degrade and thereby inactivate TFPI.5 Since these proteases and TFPI both bind to NET, they greatly enhance inactivation of TFPI.5 Isolated nucleosomes (derived from insects and mammalian cells) mimic this effect, and can thus also act as a stimulus of TF-dependent fibrin formation. While the role of NET in promoting intravascular fibrin formation and thrombosis in vivo has been relatively well documented (see below), the question as to whether nucleosomes released from normal apoptotic cells and/or tumor cells also act as prothrombotic triggers requires future investigation. Apart from activating the extrinsic pathway of blood coagulation, NET can also promote factor XII activation, 208

and thereby stimulate the contact pathway of blood coagulation.8 Moreover, NET might enhance fibrin formation by inhibiting tissue plasminogen activator-mediated fibrinolysis.35 The procoagulant mechanisms induced by NET, in particular their ability to increase blood coagulation via TFPI degradation, are likely mediated in part by nucleic acids, which provide a polyanionic surface that, like other polyanions (including, for example, RNA and polyphosphates), allows the proteolytic activation of coagulation factors such as factor XII.36-38 NE-dependent degradation of endogenous anticoagulants, activation of the contact pathway, and inhibition of fibrinolysis, support propagation of blood coagulation, rather than triggering its initiation. Overall, neutrophils and NET thus operate through multiple pathways that propagate fibrin formation and enhance recruitment and activation of platelets.

Immunothrombosis: intravascular fibrin formation as part of the innate immune defense Systemic bacterial infections can be a lethal threat to an organism. The mechanisms of host defense against infections by circulating bacteria are still not fully understood. Recently, it was shown that neutrophil serine proteases (NE, cathepsin G) and NET/extracellular nucleosomes trigger the formation of immunothrombosis in liver and spleen sinusoids during systemic bacterial infection in vivo.5 The ability of NET to promote fibrin formation enables microvascular thrombi to limit the dissemination, tissue invasion as well as the survival of circulating E. coli. The procoagulant role of neutrophils and their released NET during immunothrombosis critically depends on neutrophil serine proteases such as NE. NET can thus serve as a platform for NE-mediated activation of intravascular coagulation in vivo. Consistent with the role of NE-induced TFPI cleavage for the antimicrobial activity of intravascular blood coagulation, infusion of a TFPI mutant specifically resistant to cleavage by NE and cathepsin G (T87F/L89A), which almost completely suppressed microvascular fibrin formation, markedly increased tissue invasion of E.coli and enhanced bacterial survival.5 However, apart from its beneficial role in combating circulating pathogens, NET-induced microvascular thrombosis under certain conditions can become detrimental to the host.39,40 This is particularly true for disseminated intravascular coagulation, a serious complication of sepsis, which is likely a direct pathological consequence of immunothrombosis. In line with this, NET have been shown to foster the development of sepsis.41 In particular, NET have been detected in several organs during sepsis, including lungs, or even circulating in the systemic blood stream.42 Thus, in severe sepsis, the prothrombotic actions of NET may have deleterious side effects on the blood supply and functions of multiple organs. In line with a role of NET in pathological microvascular thrombosis in humans, patients with acute thrombotic microangiopathies show impaired DNase-mediated NET degradation.43 Moreover, NET, predominantly via their histone components, can directly induce endothelial (and epithelial) cell death.44 In addition to their role in microbial infections, NET are also main regulators of microvascular thrombosis in sterile inflammatory processes and tumor cell metastasis. Indeed, NET are involved in veno-occlusive crises of sickle cell disease and contribute significantly to the mortality associated with this disease.45 Interestingly, heme released from haematologica | 2017; 102(2)


Thrombosis propagation by neutrophils and extracellular nucleosome

lysed erythrocytes was identified as a new trigger for NETosis under these conditions. In sickle cell crises, NET do not only cause microvascular thrombosis, but also generate excessive damage to pulmonary tissue, the main cause of mortality in this setting, which could be reversed by DNase I treatment. Similarly, NET have been detected within the pulmonary microcirculation during transfusion-related acute lung injury, and contribute significantly to morbidity and mortality by increasing endothelial permeability.46 Another pathological side-effect of NET formation in the microcirculation may be promotion of tumor metastasis, whereby NET formed in the liver sinusoids in response to infection have been shown to support the adhesion and trapping of circulating tumor cells.47 NETosis can also be detected in viral infections and various viruses (such as influenza and human immunodeficiency virus-1) are able to induce the formation of NET, which may bind and thereby neutralize viruses.48,49 A new host-protective effect of NET has been described in fungal infections: neutrophils that are exposed to a micro-organism such as C. albicans, which cannot be phagocytosed because of its large size, initiate NETosis to capture the pathogen. Vice versa, if the microbe can be phagocytosed by the neutrophil, NET formation is inhibited.50 Interestingly, NET may not only stimulate, but may also restrict inflammatory reactions. Once they have formed densely packed aggregates, NET can degrade neutrophilderived inflammatory mediators by means of their own serine proteases, thereby limiting inflammatory reactions during gout.51 However, it is still not clear whether this mechanism is also relevant for resolution of microbial infections and microvascular thrombosis. Hence, neutrophils and the procoagulant mechanisms supported by them can be seen to be as efficient tools in fighting bacterial and viral infections, but can also cause substantial collateral damage to host tissues.

Detection and functional role of neutrophil extracellular traps in arterial thrombosis While the development of microvascular thrombosis in general is relatively slow in nature and triggered by diverse stimuli, arterial thrombosis is a fast process with a uni-

form trigger, especially disruption of the endothelial cell layer and exposure of the subendothelial matrix, following rupture at sites of atherosclerotic plaques. The sudden loss of blood supply induced by thrombosis in coronary arteries results in myocardial infarction and stroke. TF plays a central role in inducing arterial thrombosis. TF is highly concentrated in atherosclerotic plaques in both cellular and acellular regions. Plaque rupture leads to the exposure of TF to the blood and together with platelet adhesion, activation and aggregation at sites of turbulent flow these interactions lead to the rapid development of arterial occlusions.52 Notably, it could be demonstrated that neutrophils and NET/extracellular nucleosomes are also of major relevance for the development of arterial thrombosis. Using a model of ligation-induced thrombosis of the carotid artery, NET/extracellular nucleosomes were detected in association with neutrophils adhering to the damaged endothelium in vivo.5 The NET detected in arterial thrombi were intact in some cases; in other cases, they were fragmented. These fragmented NET were clearly derived from neutrophils, as they stained positive for myeloperoxidase. Blocking NET with anti-H2A/H2B-DNA antibody decreased fibrin formation at the site of injury. Moreover, this treatment delayed the time to vessel occlusion and strongly reduced the duration of occlusion without affecting firm adhesion of platelets.5 These findings established for the first time a causal role for NET in large vessel thrombosis in mouse models. The procoagulant mechanisms induced by neutrophils in arterial thrombosis partially overlapped with the mechanisms supporting the development of microvascular thrombosis. Accordingly, fibrin formation and arterial vessel occlusion were strongly reduced in mice deficient for neutrophil serine proteases.5 As mentioned, TFPI inactivation by NET-associated neutrophilic proteases participates in arterial thrombosis since TFPI was degraded at the site of vessel injury and thrombus formation in wild-type mice, but not in neutrophil serine protease-deficient animals. Moreover, the TFPI mutant T87F/L89A, which is resistant to cleavage by neutrophil serine proteases, decreased arterial thrombosis more efficiently than did native TFPI.

Figure 1: Neutrophil extracellular traps (blue; arrows) and neutrophil elastase (red) in thrombi recovered from human coronary arteries. Left and right images show thrombi from two different patients.

haematologica | 2017; 102(2)

209


S. Pfeiler et al.

The importance of these findings is underscored by the detection of NET in association with neutrophils in specimens of human coronary thrombi (Figure 1) and at lesion sites of patients with acute myocardial infarction and stent thrombosis.6,53,54 NET could only be detected in early stages of coronary thrombosis, but not in organized thrombi, which could be attributed to digestion of NET by plasma DNases or disruption of NET by heparin treatment applied during cardiac catheterization.53 Increased NET burden in coronary thrombi correlated with infarct size, and DNase treatment (in particular together with tissue plasminogen activator) promoted resolution of coronary thrombi ex vivo, suggesting that a combination of established antithrombotic and NET-disrupting therapy might be beneficial in the therapy of acute coronary syndromes.6 In addition, neutrophils are by far the most important innate leukocyte subtype in stent thrombosis.54 Apart from neutrophils, eosinophils were present in all types of stent thrombosis, indicating a highly sensitive determinant that could induce thrombosis via eosinophil-associated TF.32

Neutrophils propagate venous thrombosis Deep vein thrombosis and its complication pulmonary embolism are frequent disorders that contribute considerably to the mortality associated with cardiovascular diseases. One of the main triggers of venous thrombosis is decreased or stagnant blood flow as in the setting of immobilization. Using a mouse model of reduced blood flow in the inferior vena cava it was demonstrated that the processes leading to occlusion of venous vessels are reminiscent of microvascular thrombosis.2,8,55 The development of deep vein thrombosis was found to be driven by a tight co-operation between platelets, monocytes and neutrophils resulting in both the initiation and propagation of fibrin formation. In particular, neutrophil-derived extracellular nucleosomes were detected in venous thrombi in mice in vivo.7,8 Overall, NET profoundly enhanced deep vein thrombosis by a yet to be analyzed mechanism.8 Potentially, NET might play different roles in the development of venous thrombosis. For example, NET could damage the endothelium in large veins through the cytotoxic potential of histones. Platelet-neutrophil interactions at the site of deep vein thrombosis formation were found to induce NETosis and to be of substantial relevance for thrombogenesis in the context of deep vein thrombosis in general.8 NET in turn support propagation of blood coagulation as indicated by the inhibition of fibrin formation following infusion of anti-H2A/H2B-DNA antibody and DNase I. Notably, NEdependent TFPI degradation, which contributes to arterial thrombosis, is not an essential element for deep vein thrombosis, suggesting that alternate, yet unknown pathways support the NET-induced thrombotic process in veins. (K. Stark, B. Engelmann, S. Massberg, unpublished data, 2017;56) Fibrin formation during development of deep vein thrombosis was dependent both on the extrinsic and on the contact pathways of blood coagulation. Correspondingly, intravascular TF most likely expressed by monocytes, but to a lesser extent exposed on NET, as well as factor XIIa critically mediated venous thrombosis. This was suggested by experiments performed with mice lacking hematopoietic TF or myeloid TF and in FXII-deficient mice.8 FXII can bind to extracellular chromatin and thus be activated, which contributes to the propagation of intravascular clot formation. NET, apart from activating 210

the coagulation system during venous thrombosis also bind platelets, a process mediated by von Willebrand factor.7,8 In addition, NET can bind erythrocytes, and histones can induce a procoagulant phenotype in these anucleated cells by inducing the exposure of phosphatidylserine; nevertheless, the relevance of this observation for deep vein thrombosis in vivo is unclear.8,57 Recently, it has been described that NET and circulating nucleosomes are present in human thromboembolism, suggesting that extracellular nucleosomes may be of relevance to deep vein thrombosis in patients.58,59 Apart from immobilization, cancer is another important risk factor for venous thrombosis and is associated with hypercoagulability, which could in part be explained by an increased activation of neutrophils and their enhanced ability to form NET in tumor-bearing mice.60

Neutrophil extracellular traps/extracellular chromatin as a marker and therapeutic target of thrombosis In line with the central role of neutrophils and extracellular chromatin in different types of experimental thrombosis, extracellular nucleosomes and distinct components of them such as citrullinated histones have been shown to be increased in plasma of patients with sepsis, arterial thrombosis, atherosclerosis, and in deep vein thrombosis.58,41,53,58,59,61-63 Since nucleosomes are not only externalized by neutrophils but also by apoptotic and necrotic cells, and since the plasma levels of nucleosomes have been shown to be increased under various pathological conditions (e.g. ischemia/reperfusion, cancer), the diagnostic evaluation of nucleosome-driven thrombosis requires the use of additional markers.64 Additional markers might include, for example, plasma markers of neutrophil activation such as NE as well as Ddimer levels.6,58 Inhibition of the prothrombotic functions of NET/extracellular nucleosomes by specific antibodies, such as anti-H2A/H2B-DNA antibody, or their degradation by DNase I robustly inhibits thrombosis in different vascular beds in animal models.5,7,8 Inhibition of NET formation in PAD4-deficient mice does not change bleeding times.65 Future studies will need to address in more detail whether blocking/degrading NET is associated with bleeding complications. Overall, neutralizing antibodies targeting nucleosomes and their histone components, DNases as well as PAD4 inhibitors, are interesting candidate molecules for the prevention of various types of thrombosis in humans.66

Conclusions During infections and inflammatory responses, neutrophils promote intravascular blood coagulation and thrombosis. A major mechanism allowing neutrophils to shape platelet activation and fibrin formation is via extrusion of NET/extracellular nucleosomes (Figure 2). Platelets are critically involved in neutrophil-regulated thrombosis since they promote NET formation and are themselves activated by extracellular chromatin. NET-induced blood coagulation is probably a key mediator of intravascular immunity which, supported by microvascular thrombosis, restricts the dissemination and survival of circulating bacteria. However, the defense against circulating pathogens by neutrophil-induced prothrombotic mechanisms most likely comes at a high cost. Indeed, NET markedly promote vessel haematologica | 2017; 102(2)


Thrombosis propagation by neutrophils and extracellular nucleosome

Figure 2. Mechanisms of neutrophil extracellular trap(NET)-mediated thrombosis (model). PSGL-1: P-selectin glycoprote in ligand-1; HMGB1: high mobility group box 1; RAGE: receptor for advanced glycation end-products; PAD4: peptidylarginine deiminases 4; NE: neutrophil; MPO: myeloperoxidase; TF: tissue factor; MP: microparticle; vWF: von Willebrand factor.

occlusion in experimental models of arterial and deep vein thrombosis. Moreover, high numbers of neutrophils and extracellular nucleosomes have been detected in thrombi and blood of patients with arterial and deep vein thrombosis. Overall, this suggests that neutrophils and their NET may contribute to cardiovascular diseases induced by thrombosis, such as myocardial infarction, stroke, and venous thromboembolism. As illustrated in Figure 2, NETosis can be induced by interactions of activated platelets (red) with neutrophils (blue), which result in the formation of intact and fragmented NET in different vascular beds in vivo. The externalized nucleosomes promote the propagation of intravas-

References 1. Mantovani A, Cassatella MA, Costantini C, Jaillon S. Neutrophils in the activation and regulation of innate and adaptive immunity. Nat Rev Immunol. 2011;11(8):519-531. 2. Engelmann B, Massberg S. Thrombosis as an intravascular effector of innate immunity. Nat Rev Immunol. 2013;13(1):34-45. 3. Kolaczkowska E, Kubes P. Neutrophil

haematologica | 2017; 102(2)

cular blood coagulation, von Willebrand factor binding, and platelet adhesion/activation, which fosters thrombosis both in the microcirculation and in large vessels. During infection and inflammation, TF may be exposed on NET-embedded microparticles, whereby platelet-neutrophil conjugates could also participate in the initiation of intravascular fibrin formation. Acknowledgment Work performed in the laboratories of Bernd Engelmann and Steffen Massberg mentioned in this review has been supported by grants from the DFG, including SFB 1123, and from Deutsche Krebshilfe.

recruitment and function in health and inflammation. Nat Rev Immunol. 2013;13 (3):159-175. 4. Brinkmann V, Reichard U, Goosmann C, et al. Neutrophil extracellular traps kill bacteria. Science. 2004;303(5663):1532-1535. 5. Massberg S, Grahl L, von Bruehl ML, et al. Reciprocal coupling of coagulation and innate immunity via neutrophil serine proteases. Nat Med. 2010;16(8):887-896.

6. Mangold A, Alias S, Scherz T, et al. Coronary neutrophil extracellular trap burden and deoxyribonuclease activity in STelevation acute coronary syndrome are predictors of ST-segment resolution and infarct size. Circ Res. 2015;116(7):1182-1192. 7. Fuchs TA, Brill A, Duerschmied D, et al. Extracellular DNA traps promote thrombosis. Proc Natl Acad Sci USA. 2010;107(36): 15880-15885.

211


S. Pfeiler et al. 8. von Bruhl ML, Stark K, Steinhart A, et al. Monocytes, neutrophils, and platelets cooperate to initiate and propagate venous thrombosis in mice in vivo. J Exp Med. 2012;209(4):819-835. 9. Brinkmann V, Zychlinsky A. Neutrophil extracellular traps: is immunity the second function of chromatin? J Cell Biol. 2012; 198(5):773-783. 10. Wang Y, Li M, Stadler S, et al. Histone hypercitrullination mediates chromatin decondensation and neutrophil extracellular trap formation. J Cell Biol. 2009;184(2):205-213. 11. Li P, Li M, Lindberg MR, Kennett MJ, Xiong N, Wang Y. PAD4 is essential for antibacterial innate immunity mediated by neutrophil extracellular traps. J Exp Med. 2010;207 (9):1853-1862. 12. Yipp BG, Petri B, Salina D, et al. Infectioninduced NETosis is a dynamic process involving neutrophil multitasking in vivo. Nat Med. 2012;18(9):1386-1393. 13. Pisetsky DS. The origin and properties of extracellular DNA: from PAMP to DAMP. Clin Immunol. 2012;144(1):32-40. 14. Urbonaviciute V, Furnrohr BG, Meister S, et al. Induction of inflammatory and immune responses by HMGB1-nucleosome complexes: implications for the pathogenesis of SLE. J Exp Med. 2008;205(13):3007-3018. 15. Lood C, Blanco LP, Purmalek MM, et al. Neutrophil extracellular traps enriched in oxidized mitochondrial DNA are interferogenic and contribute to lupus-like disease. Nat Med. 2016;22(2):146-153. 16. Jahr S, Hentze H, Englisch S, et al. DNA fragments in the blood plasma of cancer patients: quantitations and evidence for their origin from apoptotic and necrotic cells. Cancer Res. 2001;61(4):1659-1665. 17. Zeerleder S, Zwart B, te Velthuis H, et al. Nucleosome-releasing factor: a new role for factor VII-activating protease (FSAP). FASEB J. 2008;22(12):4077-4084. 18. Nakazawa F, Kannemeier C, Shibamiya A, et al. Extracellular RNA is a natural cofactor for the (auto-)activation of factor VII-activating protease (FSAP). Biochem J. 2005;385(Pt 3):831-838. 19. Subramaniam S, Thielmann I, Morowski M, et al. Defective thrombus formation in mice lacking endogenous factor VII activating protease (FSAP). Thromb Haemost. 2015;113(4):870-880. 20. Duerschmied D, Bode C, Ahrens I. Immune functions of platelets. Thromb Haemost. 2014;112(4):678-691. 21. Semple JW, Italiano JE Jr., Freedman J. Platelets and the immune continuum. Nat Rev Immunol. 2011;11(4):264-274. 22. Yeaman MR. Platelets: at the nexus of antimicrobial defence. Nat Rev Microbiol. 2014;12(6):426-437. 23. Clark SR, Ma AC, Tavener SA, et al. Platelet TLR4 activates neutrophil extracellular traps to ensnare bacteria in septic blood. Nat Med. 2007;13(4):463-469. 24. Maugeri N, Campana L, Gavina M, et al. Activated platelets present high mobility group box 1 to neutrophils, inducing autophagy and promoting the extrusion of neutrophil extracellular traps. J Thromb Haemost. 2014;12(12):2074-2088. 25. Sreeramkumar V, Adrover JM, Ballesteros I, et al. Neutrophils scan for activated platelets to initiate inflammation. Science. 2014;346(6214):1234-1238. 26. Semeraro F, Ammollo CT, Morrissey JH, et al. Extracellular histones promote thrombin generation through platelet-dependent mechanisms: involvement of platelet TLR2 and TLR4. Blood. 2011;118(7):1952-1961.

212

27. Muller I, Klocke A, Alex M, et al. Intravascular tissue factor initiates coagulation via circulating microvesicles and platelets. FASEB J. 2003;17(3):476-478. 28. Leon C, Alex M, Klocke A, et al. Platelet ADP receptors contribute to the initiation of intravascular coagulation. Blood. 2004;103 (2):594-600. 29. Maugeri N, Brambilla M, Camera M, et al. Human polymorphonuclear leukocytes produce and express functional tissue factor upon stimulation. J Thromb Haemost. 2006;4(6):1323-1330. 30. Kambas K, Mitroulis I, Apostolidou E, et al. Autophagy mediates the delivery of thrombogenic tissue factor to neutrophil extracellular traps in human sepsis. PLoS One. 2012;7(9):e45427. 31. Yousefi S, Gold JA, Andina N, et al. Catapult-like release of mitochondrial DNA by eosinophils contributes to antibacterial defense. Nat Med. 2008;14(9):949-953. 32. Moosbauer C, Morgenstern E, Cuvelier SL, et al. Eosinophils are a major intravascular location for tissue factor storage and exposure. Blood. 2007;109(3):995-1002. 33. Langer F, Ruf W. Synergies of phosphatidylserine and protein disulfide isomerase in tissue factor activation. Thromb Haemost. 2014;111(4):590-597. 34. Reinhardt C, von Bruhl ML, Manukyan D, et al. Protein disulfide isomerase acts as an injury response signal that enhances fibrin generation via tissue factor activation. J Clin Invest. 2008;118(3):1110-1122. 35. Varju I, Longstaff C, Szabo L, et al. DNA, histones and neutrophil extracellular traps exert anti-fibrinolytic effects in a plasma environment. Thromb Haemost. 2015; 113(6):1289-1298. 36. Kannemeier C, Shibamiya A, Nakazawa F, et al. Extracellular RNA constitutes a natural procoagulant cofactor in blood coagulation. Proc Natl Acad Sci USA. 2007;104(15):63886393. 37. Muller F, Mutch NJ, Schenk WA, et al. Platelet polyphosphates are proinflammatory and procoagulant mediators in vivo. Cell. 2009;139(6):1143-1156. 38. Smith SA, Mutch NJ, Baskar D, Rohloff P, Docampo R, Morrissey JH. Polyphosphate modulates blood coagulation and fibrinolysis. Proc Natl Acad Sci USA. 2006;103 (4):903-908. 39. Pfeiler S, Massberg S, Engelmann B. Biological basis and pathological relevance of microvascular thrombosis. Thromb Res. 2014;133 (Suppl 1)S35-37. 40. Gould TJ, Lysov Z, Liaw PC. Extracellular DNA and histones: double-edged swords in immunothrombosis. J Thromb Haemost. 2015;13 (Suppl 1)S82-91. 41. Camicia G, Pozner R, de Larranaga G. Neutrophil extracellular traps in sepsis. Shock. 2014;42(4):286-294. 42. Tanaka K, Koike Y, Shimura T, et al. In vivo characterization of neutrophil extracellular traps in various organs of a murine sepsis model. PLoS One. 2014;9(11):e111888. 43. Jimenez-Alcazar M, Napirei M, Panda R, et al. Impaired DNase1-mediated degradation of neutrophil extracellular traps is associated with acute thrombotic microangiopathies. J Thromb Haemost. 2015;13(5):732-742. 44. Saffarzadeh M, Juenemann C, Queisser MA, et al. Neutrophil extracellular traps directly induce epithelial and endothelial cell death: a predominant role of histones. PLoS One. 2012;7(2):e32366. 45. Chen G, Zhang D, Fuchs TA, Manwani D, Wagner DD, Frenette PS. Heme-induced neutrophil extracellular traps contribute to

46.

47.

48.

49.

50.

51.

52. 53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

the pathogenesis of sickle cell disease. Blood. 2014;123(24):3818-3827. Caudrillier A, Kessenbrock K, Gilliss BM, et al. Platelets induce neutrophil extracellular traps in transfusion-related acute lung injury. J Clin Invest. 2012;122(7):2661-2671. Cools-Lartigue J, Spicer J, McDonald B, et al. Neutrophil extracellular traps sequester circulating tumor cells and promote metastasis. J Clin Invest. 2013;123(8):3446-3458. Narasaraju T, Yang E, Samy RP, et al. Excessive neutrophils and neutrophil extracellular traps contribute to acute lung injury of influenza pneumonitis. Am J Pathol. 2011;179(1):199-210. Saitoh T, Komano J, Saitoh Y, et al. Neutrophil extracellular traps mediate a host defense response to human immunodeficiency virus-1. Cell Host Microbe. 2012;12(1):109-116. Branzk N, Lubojemska A, Hardison SE, et al. Neutrophils sense microbe size and selectively release neutrophil extracellular traps in response to large pathogens. Nat Immunol. 2014;15(11):1017-1025. Schauer C, Janko C, Munoz LE, et al. Aggregated neutrophil extracellular traps limit inflammation by degrading cytokines and chemokines. Nat Med. 2014;20(5):511517. Jackson SP. Arterial thrombosis--insidious, unpredictable and deadly. Nat Med. 2011;17(11):1423-1436. de Boer OJ, Li X, Teeling P, et al. Neutrophils, neutrophil extracellular traps and interleukin-17 associate with the organisation of thrombi in acute myocardial infarction. Thromb Haemost. 2013;109(2):290-297. Riegger J, Byrne RA, Joner M, et al. Histopathological evaluation of thrombus in patients presenting with stent thrombosis. A multicenter European study: a report of the prevention of late stent thrombosis by an interdisciplinary global European effort consortium. Eur Heart J. 2016;37(19):1538-1549. Schulz C, Engelmann B, Massberg S. Crossroads of coagulation and innate immunity: the case of deep vein thrombosis. J Thromb Haemost. 2013;11 (Suppl 1)233-241. Martinod K, Witsch T, Farley K, Gallant M, Remold-O'Donnell E, Wagner DD. Neutrophil elastase-deficient mice form neutrophil extracellular traps in an experimental model of deep vein thrombosis. J Thromb Haemost. 2016;14(3):551-558. Semeraro F, Ammollo CT, Esmon NL, Esmon CT. Histones induce phosphatidylserine exposure and a procoagulant phenotype in human red blood cells. J Thromb Haemost. 2014;12(10):1697-1702. van Montfoort ML, Stephan F, Lauw MN, et al. Circulating nucleosomes and neutrophil activation as risk factors for deep vein thrombosis. Arterioscler Thromb Vasc Biol. 2013;33(1):147-151. Savchenko AS, Martinod K, Seidman MA, et al. Neutrophil extracellular traps form predominantly during the organizing stage of human venous thromboembolism development. J Thromb Haemost. 2014;12(6):860870. Demers M, Krause DS, Schatzberg D, et al. Cancers predispose neutrophils to release extracellular DNA traps that contribute to cancer-associated thrombosis. Proc Natl Acad Sci USA. 2012;109(32):13076-13081. Zeerleder S, Zwart B, Wuillemin WA, et al. Elevated nucleosome levels in systemic inflammation and sepsis. Crit Care Med. 2003;31(7):1947-1951. Megens RT, Vijayan S, Lievens D, et al.

haematologica | 2017; 102(2)


Thrombosis propagation by neutrophils and extracellular nucleosome

63.

64. 65.

66.

67.

68.

Presence of luminal neutrophil extracellular traps in atherosclerosis. Thromb Haemost. 2012;107(3):597-598. Borissoff JI, Joosen IA, Versteylen MO, et al. Elevated levels of circulating DNA and chromatin are independently associated with severe coronary atherosclerosis and a prothrombotic state. Arterioscler Thromb Vasc Biol. 2013;33(8):2032-2040. Holdenrieder S, Stieber P. Clinical use of circulating nucleosomes. Crit Rev Clin Lab Sci. 2009;46(1):1-24. Martinod K, Demers M, Fuchs TA, et al. Neutrophil histone modification by peptidylarginine deiminase 4 is critical for deep vein thrombosis in mice. Proc Natl Acad Sci USA. 2013;110(21):8674-8679. Wang S, Wang Y. Peptidylarginine deiminases in citrullination, gene regulation, health and pathogenesis. Biochim Biophys Acta. 2013;1829(10):1126-1135. Pilsczek FH, Salina D, Poon KK, et al. A novel mechanism of rapid nuclear neutrophil extracellular trap formation in response to Staphylococcus aureus. J Immunol. 2010;185(12):7413-7425. Ramos-Kichik V, Mondragon-Flores R,

haematologica | 2017; 102(2)

69.

70.

71.

72.

73.

Mondragon-Castelan M, et al. Neutrophil extracellular traps are induced by Mycobacterium tuberculosis. Tuberculosis (Edinb). 2009;89(1):29-37. Gupta AK, Joshi MB, Philippova M, et al. Activated endothelial cells induce neutrophil extracellular traps and are susceptible to NETosis-mediated cell death. FEBS Lett. 2010;584(14):3193-3197. Grinberg N, Elazar S, Rosenshine I, Shpigel NY. Beta-hydroxybutyrate abrogates formation of bovine neutrophil extracellular traps and bactericidal activity against mammary pathogenic Escherichia coli. Infect Immun. 2008;76(6):2802-2807. Abi Abdallah DS, Lin C, Ball CJ, King MR, Duhamel GE, Denkers EY. Toxoplasma gondii triggers release of human and mouse neutrophil extracellular traps. Infect Immun. 2012;80(2):768-777. Martinelli S, Urosevic M, Daryadel A, et al. Induction of genes mediating interferondependent extracellular trap formation during neutrophil differentiation. J Biol Chem. 2004;279(42):44123-44132. Guimaraes-Costa AB, Nascimento MT, Froment GS, et al. Leishmania amazonensis

74.

75.

76.

77.

78.

promastigotes induce and are killed by neutrophil extracellular traps. Proc Natl Acad Sci USA. 2009;106(16):6748-6753. Urban CF, Reichard U, Brinkmann V, Zychlinsky A. Neutrophil extracellular traps capture and kill Candida albicans yeast and hyphal forms. Cell Microbiol. 2006;8(4):668676. Kessenbrock K, Krumbholz M, Schonermarck U, et al. Netting neutrophils in autoimmune small-vessel vasculitis. Nat Med. 2009;15(6):623-625. Bruns S, Kniemeyer O, Hasenberg M, et al. Production of extracellular traps against Aspergillus fumigatus in vitro and in infected lung tissue is dependent on invading neutrophils and influenced by hydrophobin RodA. PLoS Pathog. 2010;6(4):e1000873. Hemmers S, Teijaro JR, Arandjelovic S, Mowen KA. PAD4-mediated neutrophil extracellular trap formation is not required for immunity against influenza infection. PLoS One. 2011;6(7):e22043. Fuchs TA, Abed U, Goosmann C, et al. Novel cell death program leads to neutrophil extracellular traps. J Cell Biol. 2007;176 (2):231-241.

213


REVIEW ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Cure for thalassemia major – from allogeneic hematopoietic stem cell transplantation to gene therapy Alok Srivastava1*and Ramachandran V. Shaji1 Department of Haematology & Centre for Stem Cell Research (a unit of inStem, Bengaluru), Christian Medical College, Vellore- 632004, Tamil Nadu, India

1

Haematologica 2017 Volume 102(2):214-223

ABSTRACT

A

Correspondence: aloks@cmcvellore.ac.in

Received: March 20, 2016. Accepted: October 12, 2016. Pre-published: December 1, 2016. doi:10.3324/haematol.2015.141200

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/214

Š2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

214

llogeneic hematopoietic stem cell transplantation has been well established for several decades as gene replacement therapy for patients with thalassemia major, and now offers very high rates of cure for patients who have access to this therapy. Outcomes have improved tremendously over the last decade, even in high-risk patients. The limited data available suggests that the long-term outcome is also excellent, with a >90% survival rate, but for the best results, hematopoietic stem cell transplantation should be offered early, before any end organ damage occurs. However, access to this therapy is limited in more than half the patients by the lack of suitable donors. Inadequate hematopoietic stem cell transplantation services and the high cost of therapy are other reasons for this limited access, particularly in those parts of the world which have a high prevalence of this condition. As a result, fewer than 10% of eligible patients are actually able to avail of this therapy. Other options for curative therapies are therefore needed. Recently, gene correction of autologous hematopoietic stem cells has been successfully established using lentiviral vectors, and several clinical trials have been initiated. A gene editing approach to correct the b-globin mutation or disrupt the BCL11A gene to increase fetal hemoglobin production has also been reported, and is expected to be introduced in clinical trials soon. Curative possibilities for the major hemoglobin disorders are expanding. Providing access to these therapies around the world will remain a challenge. Introduction Thalassemias are the most common human monogenic disorders related to the deficiency of the production of either the Îą- or b-globin chains.1 b-thalassemia is a larger clinical problem because its homozygous form, thalassemia major, leads to severe morbidity and mortality due to very low endogenous hemoglobin levels which are incompatible with life.2 Even though educating the society at large, mass screening, cohort counselling and prenatal diagnosis can be applied to very effectively reduce the incidence of b-thalassemia major, this has only been achieved in some countries.3 More than 50,000 children with this disease are born worldwide each year, adding to the disease burden of this condition.4 Hypertransfusion and iron chelation have been the mainstay of therapy for thalassemia major for nearly 50 years.5 However, these therapies are often ineffective due to complications related to repeated transfusions and inadequate iron chelation. The main reasons for this are the lack of compliance with the planned therapy due to its logistic demands as well as the ongoing costs of chelation therapy. This approach, therefore, leads to significant morbidity and mortality with up to 50% of these patients developing significant organ dysfunction by the time they are adults, even in Western countries.6 This figure is even higher for patients in developing countries.7 The need for curative therapy for thalassemia major was addressed with the success of allogeneic hematopoietic stem cell transplantation (alloHSCT), which was initiated in the early 1980s as a way to replace the defective gene in such patients.8 AlloHSCT haematologica | 2017; 102(2)


Cure for thalassemia major

remains the only widely available curative therapy for this condition at present. The best results are seen when alloHSCT is offered early, before complications related to iron overload or transfusion-transmitted infections set in, with survival rates of over 90% being reported in these patients.9 The outcomes of higher risk patients have also continuously improved over the last two decades, so that 80-90% long-term survival rates have been obtained even in this group of patients.10,11 However, there are many challenges in offering alloHSCT as a therapy for these patients all over the world.12 More recently, gene replacement in autologous hematopoietic stem cells (autoHSCs) using viral vectors has become a reality, witGST M1 null status'h several clinical trials showing its success and potential for wider use.13 This article will address the status of alloHSCT for b-thalassemia major and also briefly review the newer options for gene correction in autoHSCs using different approaches.

Allogeneic Hematopoietic Stem Cell Transplantation While the classical principles of alloHSCT remain the same in treating patients with thalassemia major, early results showed that there were special challenges related to pre-existing organ dysfunction, particularly involving the liver as well as the hyperactive immune system, probably related to repeated transfusions.14 Given that most of these patients were children with a non-malignant disease, busulfan (14-16mg/kg total dose) and cyclophosphamide (160-200 mg/kg total dose) based conditioning was chosen, even though many of them had liver dysfunction related to iron overload or transfusion-related viral hepatitis.8 Significant regimen-related toxicities (RRT) were observed if the doses of the drugs were intensified, while there was a high incidence of graft rejections if the doses were reduced. In high-risk patients, this led to mortality rates of up to 35% and rejection in nearly 30% of patients, resulting in long-term event-free survival of only about 50%.15 High RRT was attributed to the sequential use of these two drugs, often with severe sinusoidal obstruction syndrome, particularly in those with significant liver dysfunction.16 RRTs were also shown to be associated with certain genetic polymorphisms in the glutathione S transferase M1 (GSTM1) and the cytochrome P450 (CYP450) genes. Detailed pharmacogenetic studies showed that the GST M1 null status and the CYP2C9 gene polymorphisms affected the pharmacokinetics of busulfan and cyclophosphamide, with a possible impact on RRTs.17,18 Therefore, modifications in the approach to alloHSCT for these patients were needed. Over the last decade, the results of alloHSCT have improved significantly. This has been possible due to better risk stratification, more effective targeted dose adjustment of intravenous busulfan during conditioning, a modified conditioning regimen and continually improving supportive care.19,20 Patients over 7 years of age with hepatomegaly of more than 5cm have been identified as an especially high-risk group who are candidates for/require novel approaches, particularly with regard to their conditioning regimen and preparation for transplant.21,22 Two approaches have been taken in this regard, one based on pre-transplant immunosuppression using fludarabine, and the other by introducing a longer gap haematologica | 2017; 102(2)

between the use of busulfan and cyclophosphamide during conditioning23,24 or using less toxic myeloablative agents such as treosulfan and avoiding cyclophosphamide completely.11 All these modifications have led to significantly improved survival rates of nearly 80-90% in these high-risk patients (Table 1).10,11,15,19,20,23-27 While we have to appreciate the impact of successful alloHSCT on the lives of these patients, we need to continue to recognize the many challenges that persist with respect to the still significant morbidity and mortality associated with this procedure. All over the world a major constraint is the lack of access to this therapy related to the lack of a suitable donor. Matched related donors are generally available only for a third of patients with thalassemia major,28 but in countries with larger families or communities with consanguineous marriages up to two-thirds of patients may have suitable related donors.29 However, only a very small minority of the global patient population falls into this category. The need for alternative donors has therefore been explored in several ways - partially mismatched related donors,30 related haploidentical transplants,31,32 and matched unrelated donors, the latter of which could be adult humans or cryopreserved cord blood units in blood banks.20,33 While good results have been achieved with alternative donors in some studies, experience is still limited and the outcomes variable. Therefore these approaches have not yet been translated into becoming the standard of care for thalassemia major.28 Persisting concerns regarding high rejection rates and graft versus host disease (GvHD) need to be addressed through the evaluation of novel protocols in future studies. If these challenges can be addressed, haploidentical alloHSCTs could become a major alternative to matched unrelated donors or cord blood transplants. This could be particularly advantageous in resource-constrained countries with large numbers of these patients, where access to unrelated donors is severely restricted due to lack of services and the associated costs.32-34 The aim of curative therapy for any disease is to eradicate it and also enable a normal life afterwards. With high survival rates in patients with thalassemia major undergoing alloHSCT, the next important issue relates to the management of pre-existing and late transplant-related complications and their impact on long-term survival.6,15 Over 90% of patients who survive the first two years after alloHSCT are generally expected to become long-term survivors.35 Late (more than 2 years) complications after HSCT for hematological malignancies and bone marrow failure syndromes are accounted for mostly by disease relapses, chronic GvHD, infections and second cancers.36 In addition, particularly in children, growth retardation, multiple endocrine and other organ dysfunction and cognitive changes have also been noted, depending on their pre-existing co-morbidities at the time of alloHSCT, or if they had developed major post-transplant complications such as significant chronic GvHD. Since most of the alloHSCTs for b-thalassemia major involve children and adolescents, data on long-term outcomes becomes particularly important as they would be expected to have several decades of life post-HSCT. This is even more relevant in patients receiving repeated transfusions because many of them already have systemic complications related to chronic anemia, iron overload and resultant endocrine and metabolic dysfunction, in addition to transfusion-transmitted infections.37 Organ dysfunction exists in almost 215


A. Srivastava et al.

75% of patients undergoing alloHSCT for thalassemia major with endocrine, hepatic and cardiovascular disease accounting for nearly 30-40% each, along with a host of other organ dysfunctions.34 Unfortunately, information on late complications in ex-thalassemic patients is very limited. The only very long-term (over 20 years) study of outcome after alloHSCT for thalassemia major has shown that patients who were well-managed before transplant can have a quality of life comparable to the normal population, and certainly better than those being managed conservatively with hypertransfusion and iron chelation alone.38 Older age at alloHSCT and the development of significant chronic GvHD were found to be factors leading to a worse long-term outcome. More data with a systematic evaluation of organ dysfunction as well as an analysis of possible pre-transplant contributing factors is therefore needed in ex-thalassemic patients. An important issue will be the extent of iron overload at HSCT and the effectiveness of iron chelation post-HSCT, as this is the major cause of organ dysfunction in these patients. This is another aspect of the long-term outcome in these patients which has not been adequately addressed.39,40 Varied approaches to iron chelation have been reported in small numbers of patients after alloHSCT, ranging from phlebotomies if serum ferritin was >2000 ng/ml,39 to using only iron chelators such as desferrioxamine at 40mg/kg or deferasirox at

20-30mg/kg.41,42 The time at which chelation was initiated has also varied from shortly after engraftment to up to two years after HSCT. A more considered and coordinated approach is needed to initiate intensive iron depletion therapy as soon as possible after alloHSCT in order to reduce or stop continuing organ damage. As the vast majority of these patients come from the Mediterranean, Middle-Eastern and Asia-Pacific regions, where there is poor access to optimal transfusion-chelation therapy pre-HSCT, particular attention should be paid to iron depletion therapy in the post-HSCT period,7,40 as many patients develop significant organ dysfunction at an early age, often within the first decade. A clear strategy needs to be implemented for improving pre-HSCT management with hypertransfusion and iron chelation, selecting patients early for HSCT before any major organ dysfunction sets in, and for the post-HSCT management of these residual complications. This is particularly important for all those patients who were poorly chelated before HSCT, and are closer to puberty. Without such an approach, while better survival may be achieved even in older higher risk patients, it is unlikely that these patients will have a normal quality of life because of the pre-existing irreversible morbidities. In fact, they will end up requiring multidisciplinary management of those dysfunctions post-HSCT. An additional feature after HSCT for major hemoglobin

Table 1. Major reported clinical studies that have attempted to improve the outcome of patients with class 3 thalassemia major$.

Year

N

Median Proportion Proportion Age (yrs) / in Class in Class (range) 3 (%) 3HR (%)

Lucarelli et al.15*

1996

115

11 (3-16)

100

NA

Sodani et al.25

2004

33

11 (5-16)

100

NA

Gaziev et al.19

2010

71

9 (1.6–27)

57.3

NA

Chiesa et al.26

2010

53

8 (1-17)

47

NA

Chiesa et al.26#

2010

25

NA

100

NA

Bernardo et al.10 Li et al.20

2012 2012

60 82

7 (1-37) 6 (0.5-15)

27^ NA

NA NA

Choudhary et al.27 Anurathapan et al.23

2013 2013

28 18

9.6 (2-18) 14 (10-18)

75 100

39 NA

Mathews et al.11 Mathews et al.11@ Gaziev et al.24

2013 2013@ 2016

50 24 37

11 (2-21) 12 (3-21) 10 (5-17)

100 100 100

48 100 NA

Major defining feature of change in protocol

Treatment Graft EFS related rejection mortality (%) (%)

Bu / Cy based regimen with reduction in Cy total dose from 200 mg/kg to 160 mg/kg Reduction in Cy dose to ≤160 mg/kg with addition of Flu. Additional therapy from day -45 with immunosuppression with Azt and suppression of erythropoiesis with HU Intravenous Bu, dose adjustments with therapeutic drug monitoring Intravenous Bu, dose adjustments with therapeutic drug monitoring Intravenous Bu, dose adjustments with therapeutic drug monitoring Treo based conditioning regimen Conditioning with age adjusted PK based IV Bu, Cy (110mg/kg), high-dose Flu (200mg/kg), Thio. Additional therapy from day -45 with immunosuppression with Azt and suppression of erythropoiesis with HU Treo based conditioning regimen Conditioning regimen of Flu &IV Bu Pre-conditioning immunosuppression therapy with Flu and Dexa for 1-2 months. Treo based conditioning regimen with PBSC graft in 74% Treo based conditioning regimen with PBSC graft in 74% As in Sodani et al.25 but with higher dose of Flu (150 mg/kg) and addition of Thio(10 mg/kg)

OS (%)

24

35

49

74

6

6

85

93

7

5

87

91

4

15

79

96

4

34

66

96

7 8.5^

9 4^

84 88^

93 91^

21 5

7 0

71 89

79 89

12 13 8

8 8 0

79 78 92

87 87 92

$ Adapted from Mathews et al.10; *Only patients <17 years included in this table; #Subset of high-risk cases from same paper; ^Includes all adult cases as well (assumed to be Class 3); ^Includes lowrisk patients also; @Subset of high-risk cases from same paper. Cy: Cyclophosphamide; Flu: Fludarabine; Dexa: Dexamethasone; Bu: Busulfan; Treo: Treosulfan; Azt: Azathioprine; HU: Hydroxyurea; Thio: Thiotepa. HR: high-risk; EFS: event-free survival; OS: overall survival; NA: not applicable; PK: pharmacokinetics; PBSC: peripheral blood stem cell.

216

haematologica | 2017; 102(2)


Cure for thalassemia major

disorders is the state of mixed chimerism. This refers to the persistence of significant levels of residual host cells (RHC) of the hematopoietic system post-HSCT classified as level 1 (<10%), level 2 (10-25%) and level 3 (>25%). Transient mixed chimerism (TMC) is not uncommon in the immediate post-HSCT period and can occur in 30-45% of patients in the first 100 days.43 Depending on the level of RHC, patients with TMC may later become complete chimeras in 60-75% of cases for those at level 1, but in only 5-30% of cases of those with level 3. (Figure 1) However, these predictions are based on limited data. Therefore careful monitoring of TMC and correlation with the hemoglobin levels and other blood counts is necessary post-HSCT in patients with major hemoglobin disorders. It should also be noted that some patients develop

persistent mixed chimerism (PMC), in which a significant component of RHC persist long-term without rejection of the graft.44 In some patients, the RHC component could be much higher than 25% with an acceptable hemoglobin level. If stable, then it is obviously a case of PMC of different hematopoietic elements and immune tolerance in such a way that it does not affect the erythroid lineage adversely. T regulatory cells could play a role, but the exact mechanism of this phenomenon is not fully understood.45 The assessment of subset chimerism, particularly of the erythroid and lymphoid lineage in more patients, as shown in Figure 2, could shed more light on the mechanisms involved.46 While RHC levels below 50% in T cells has been reported to be associated with a very low risk of rejection, high levels of RHC in both T and NK cells are

A

B

Figure 1. Evolution of chimerism after hematopoietic stem cell transplantation (HSCT). Early mixed chimerism is associated with higher risk of rejection while late chimerism often persists with a stable graft. RHCs: residual host cells. MC: mixed chimerism.

haematologica | 2017; 102(2)

217


A. Srivastava et al. associated with rejections in up to 90% of patients.47 These parameters also need further evaluation. More HSCTs are now being done in the Asia-Pacific region for hemoglobin disorders than in any other part of the world,48 given the high need and greater support from health care authorities as well as an increasing number of centers offering these services. It is therefore of the utmost importance and necessity to better assess the elements that will give the best long-term outcomes. This requires the complete evaluation and documentation of non-hematological complications pre-HSCT as well as the late posttransplant complications, in order that they can be appropriately managed to allow these ex-thalassemic patients to lead as normal a life as possible.38 To this end, attention should be especially directed towards educating these families and the health care providers early about the potential complications and their prevention, as well as the importance of safe and effective transfusion-chelation programs. This will help keep patients in the best condition possible for alloHSCT to be undertaken. Indeed, our goal should be to offer alloHSCT to all patients with major hemoglobin disorders well before they become high-risk due to end-organ damage. As we persist with our efforts to increase access to alloHSCT and its outcome, it is very important to note its limitations from a public health perspective. While more than 50,000 children with b-thalassemia major are added to the world population every year, less than 5,000 alloHSCTs have been reported so far for this condition worldwide over the last 30 years.28 Even accounting for under-reporting, the actual number is likely to be well below 10,000 during this period. With the majority of these patients being in the Mediterranean, Middle-Eastern

and Asia-Pacific regions of the world, access to this therapy remains highly restricted due to the lack of centers with trained personnel and facilities for HSCT services, and the inability of patients to access care due to the high costs, which very often have to be met by the patients or their families.7,49 Recent data from the Asia Pacific Blood and Marrow Transplantation Registry shows that only about 450 alloHSCTs are being reported from the region every year, and perhaps less than 1,000 worldwide (personal communication, lida M, APBMT Registry) The enormity of this mismatch means that a different strategy is needed for effectively managing the large numbers of such patients in the world. Two approaches are necessary to address this problem. First, a preventative strategy through effective population screening, genetic counselling and prenatal diagnosis should be initiated in all countries where the prevalence of these conditions are high.3 Second, the development of a more practical curative therapy, both in terms of technology, logistics and cost, if possible, such as gene correction in autoHSCs.50 This will perhaps be the best way by which large numbers of patients who cannot avail of alloHSCT for various reasons will be able to access a curative therapy without the inherent risks and potential complications of alloHSCT.

Gene therapy for b-thalassemia major Much effort has been put into developing gene therapy for several monogenic hematological disorders over the last several decades.51 For those disorders where the defect is due to mutations in single genes that affect hemoglobin synthesis, the approach has been to introduce a normal

Figure 2. Methods to detect red cell chimerism. In patients with late persistent mixed chimerism, erythroid precursors may be predominantly donor while the nonerythroid cells may be more recipient in origin. Red cell chimerism can be measured by analyzing donor specific RBC antigens, STRs in erythroid DNA or nucleotide variations in erythroid transcripts. BFU-E: burst forming unit â&#x20AC;&#x201C; erythroid; CFU-E: colony forming unit â&#x20AC;&#x201C; erythroid; STR: short tandem repeats; RBC: red blood cell.

218

haematologica | 2017; 102(2)


Cure for thalassemia major

gene into autoHSCs and, more recently, to actually repair such genes with various genome editing tools.52,53 The initial success achieved with the retroviral vectors used for efficient gene transfer into autoHSCs for different forms of severe combined immunodeficiency (SCID) disorders53 was tempered by the development of clonal diseases and overt leukemia.54 These vectors had intact long terminal repeats (LTRs) which included efficient and ubiquitous enhancers. While this aided the desirable high expression of the target transgene, it also led to undesirable genotoxicity.53,54 The activation of oncogenes such as LMO2 in the SCID trial led to the development of acute lymphoblastic leukemia in some of these patients. However, soon thereafter, the development of lentiviral vectors devoid of their pathogenic elements and with a self-inactivating (SIN) design provided a better alternative, not only in terms of its safety but also its ability to infect quiescent HSCs and carry larger transgene cassettes, which are needed for more effective expression in order to produce gram quantities of hemoglobin.55 These advances in HSC-based gene therapy laid the foundation for its application in the major hemoglobin disorders. The first successful gene therapy for a major b-globin disorder was in 2007 when a patient with bE/b0-thalassemia was treated.56 The process involved the harvesting of autologous HSCs from the recipient, followed by

ex vivo transduction of these cells with the lentiviral vector carrying the transgene. (Figure 3) If these transduced HSCs were found suitable following quality assessment to determine the number transduced as well as the vector copy numbers per HSC, they were used to perform an autologous HSCT after appropriate conditioning therapy to destroy existing HSCs. With a transduction efficiency of about 30% and only 10-20% of HSCs showing the transgene expression, and a busulfan-based myeloablative conditioning regimen in the initial patient, there was only limited expression of b-globin in the first year after gene therapy. This patient, therefore, required several transfusions during this time. In the second year, however, the transgene expression gradually improved and led to about 2-3 g/dL of transgene-associated HbA, resulting in the overall hemoglobin stabilizing at 8.5 to 9.0 g/dL and the patient becoming transfusion-independent.57 However, this was also associated with a clonal expansion of erythroid cells (10-12%), with the insertion site being the high mobility group AT-hook 2 (HMGA2) locus. This clone peaked at 4% of hematopoietic cells at about 4 years, but has since declined to about 1% at 5 years without a reduction in total hemoglobin.58 More recent data from this group has shown that with further improvements in the lentiviral vector leading to greater transduction efficiency and higher vector copy numbers, more than 15 patients have now

Figure 3. Overview of current approaches to gene therapy for the major hemoglobin disorders. Gene modifications may be through viral vectors or genome editing technologies to achieve the desired therapeutic effect. HSC: Hematopoietic stem cell; BM: Bone marrow; PB: Peripheral blood; ZFN: zinc finger nucleases; TALEN: transcription activator-like effectors with Fokl nuclease; CRISPR: clustered regularly interspaced short palindromic repeats.

haematologica | 2017; 102(2)

219


A. Srivastava et al. been treated.59,60 Most of these patients achieved much higher transgene expression resulting in the hemoglobin increasing by 4-6 g/dl within the 2-5 months following gene therapy.58,61 Significantly, no clonal events have been described thus far in these patients. The level of transgeneassociated hemoglobin expression is related not only to the quality of transgene construct, but also to the efficiency of transduction itself with any vector, which in turn could be dependent on certain intrinsic properties of the particular HSC.62 While much more needs to be learnt about effective gene therapy for major hemoglobin disorders, these results have certainly set the stage for further clinical trials to be undertaken. Several more studies are now open for patient recruitment.63-66 Initial results continue to be encouraging and there have not been any significant safety concerns so far. While this method of gene therapy for the major hemoglobin disorders relies on substituting a functional b-globin gene for a defective one, another approach could be to increase fetal hemoglobin (HbF) production by switching on the γ-globin gene expression which becomes gradually suppressed in the months following birth.67 A major regulator of this phenomenon has been shown to be the BCL11A transcription factor which could itself be disrupted to reinitiate γ-globin gene expression through cellintrinsic mechanisms.68 This could then combine with the normal α−globin to produce enough HbF in these patients to effectively correct their anemia.69 This approach has indeed been shown to be effective in animal models and cultured human HSC derived erythroid cells, where inactivation of the BCL11A gene increased HbF production to reverse sickle cell disease.70,71 This has also now found clinical applications as described below. There are many challenges ahead related to several aspects of gene therapy for the major hemoglobin disor-

A

B

ders. Not least of these is the vector design, which is of utmost importance in order to ensure efficient transduction in HSCs to produce clinically meaningful high expression of the required globin chain that can then result in near-normalization of the hemoglobin in these patients. The success of these challenges needs to be achieved without dangerous genetic perturbations from the unavoidable random integration of the lentiviral vectors, which can lead to clonal hematopoietic disturbances with their own serious implications. Another aspect that can restrict success is the survival advantage of the transduced HSCs, given the fact that even with myeloablative conditioning, untransduced HSCs will also find their way back into the hematopoietic compartment in the patient, given the variable ex vivo transduction efficiency of the lentiviral vectors. Unlike alloHSCT, there is no donor immune system in this form of autologous stem cell therapy to give a survival advantage to the transduced transplanted HSCs with the functional b-globin gene. The dynamics that will determine the sustained presence of the transduced HSCs in the long-term are still unknown.

Genome editing for gene correction Another approach to correct mutations in autoHSCs is through genome editing techniques using targeted nucleases, which can specifically target these sites and replace them with the normal sequence to bring back the wildtype functional configuration. These technologies include the zinc fingers nucleases (ZFN) and the transcription activator-like effector nucleases (TALENS) and more recently, the clustered regularly interspaced short palindromic repeats (CRISPR) with Cas9 nuclease system.52 Conceptually, these powerful technologies allow for a

C

D

Figure 4. Genome editing techniques for correction of molecular defects. The three different gene editing techniques allow fixing of single mutations in the hemoglobin gene in patients with b-thalassemia major. ZFNs: zinc finger nucleases; TALENs: transcription activator-like effectors with Fokl nuclease; CRISPR: clustered regularly interspaced short palindromic repeats.'

220

haematologica | 2017; 102(2)


Cure for thalassemia major

molecular â&#x20AC;&#x2DC;cut and pasteâ&#x20AC;&#x2122; approach to gene correction (Figure 4),72 which could transform the entire approach to gene therapy for many diseases. In fact, the speed with which these applications are entering the clinic is remarkable.73 The ability to efficiently and safely edit the genome to modify genes has indeed taken both therapeutics and diagnostics by storm. It has also given a new dimension to the potential of stem cell-based therapies to treat different diseases. A clinical trial has already been reported showing the safety and limited efficacy of the ZFN based system in altering the CCR5 receptor on lymphocytes from patients with HIV infection on highly active antiretroviral therapy (HAART) therapy.74 Thus far however, the major challenge in using gene editing strategies in the correction of bâ&#x2C6;&#x2019;thalassemia is the low efficiency of these methods in targeted gene correction in HSCs. This is necessary for obtaining a large number of gene corrected HSCs to produce therapeutically significant levels of hemoglobin. However, a recent study that used improved methods to deliver ZFNs, showed that integrase-deficient lentiviral vectors to deliver ZFN messenger ribonucleic acids (mRNAs) and an oligonucleotide as a gene correction template had a very high efficiency of gene correction in HSCs obtained from patients with sickle cell disease.75,76 A similar approach can be applied in b-thalassemia major. The correction of b-globin gene mutations in induced pluripotent stem cells (iPSCs) derived from patients with thalassemia major is also possible.77-81 One advantage of using iPSCs for gene correction is that it could then be possible to obtain a completely corrected clone of pluripotent stem cells, from which a large number of stem cells or other cells of interest could be derived for transplantation. However, so far it has not been possible to use iPSCderived HSCs to engraft and provide sustained hematopoiesis.82 On the other hand, an encouraging proof of concept comes from the report of using ZFN technology to correct the interleukin 2 receptor (IL2R) gene in diseased human HSCs and progenitors, and the demonstration in animal models of their ability to sustain adequate hematopoiesis after transplantation.83 In fact, Sangamo BioSciences, Richmond, CA, USA, has recently obtained approval from the U.S. Food and Drug Administration (FDA) for the Investigational New Drug11 application for SB-BCLmR-HSPC for the treatment of b-thalassemia,

References 1. Weatherall DJ. The inherited diseases of hemoglobin are an emerging global health burden. Blood. 2010;115(22):4331-4336. 2. Higgs DR, Engel JD, Stamatoyannopoulos G. Thalassaemia. Lancet. 2012;379(9813): 373-383. 3. Cao A, Kan YW. The prevention of thalassemia. Cold Spring Harb Perspect Med. 2013;3(2):a011775. 4. Modell B, Darlison M. Global epidemiology of haemoglobin disorders and derived service indicators. Bull World Health Organ. 2008;86(6):480-487. 5. Giardina PJ. THALASSEMIA SYNDROMES. In: Hoffman R, ed. Hematology: Basic Principles and Practice. 5th Edition ed. Philadelphia, PA: Elsevier, 2008:535-563. 6. Borgna-Pignatti C, Rugolotto S, De Stefano

haematologica | 2017; 102(2)

7. 8.

9.

10.

developed in collaboration with Biogen, Cambridge, MA, USA.84 Data presented more recently by this group claimed about 70% efficiency of gene knockout using this approach in HSCs manipulated ex vivo.85 Clearly, these are very significant developments that have created more options for gene replacement or correction therapies for b-thalassemia major and other serious hemoglobin disorders (Figure 3). It may then be possible to move from alloHSCs as vehicles for the transfer of the normal gene to viral vectors or genome editing techniques being used to correct the gene defect in autoHSCs. One cannot underestimate the enormity of the work ahead in establishing the safety and efficacy of the latter approaches, but given the nature of the technology, it is not difficult to imagine that they hold great promise for providing platforms that could be much more amenable to developing cost-effective therapies for applications across the globe. At least they will not be limited by the much more difficult task of finding suitable donors for alloHSCT and the significant treatment-related immediate and long-term complications associated with it. However, these are very early days with regards to gene therapy, and many more clinical trials and follow-up studies will be needed to establish the long-term safety of the different approaches to gene therapy for the major hemoglobin disorders. As with any technology-based options to therapeutics, intellectual property rights based restrictions on the use of effective technologies along with the cost and marketing policy related issues can lead to hugely restricted access to such therapies around the world.86 Solutions will need to be found for those challenges if such situations do arise. Even with all these unresolved issues, there is no doubt that these are very exciting times for patients, physicians and scientists working in this field because after years of great effort, not only have the results of alloHSCT improved significantly for these patients, but several innovative gene correction therapies are also on the horizon. Acknowledgments The authors would like to like to thank Mr. Tamil Vanan J. from the Centre for Stem Cell Research, Christian Medical College Campus, Bagayam, Vellore, India for his assistance with the preparation of this manuscript.

P, et al. Survival and complications in patients with thalassemia major treated with transfusion and deferoxamine. Haematologica. 2004;89(10):1187-1193. Weatherall DJ. The challenge of thalassemia for the developing countries. Ann N Y Acad Sci. 2005;1054:11-17. Lucarelli G, Polchi P, Galimberti M, et al. Marrow transplantation for thalassaemia following busulphan and cyclophosphamide. Lancet. 1985;1(8442):1355-1357. Lucarelli G, Galimberti M, Polchi P, et al. Marrow transplantation in patients with thalassemia responsive to iron chelation therapy. N Engl J Med. 1993;329(12):840844. Bernardo ME, Piras E, Vacca A, et al. Allogeneic hematopoietic stem cell transplantation in thalassemia major: results of a reduced-toxicity conditioning regimen based on the use of treosulfan. Blood.

2012;120(2):473-476. 11. Mathews V, George B, Viswabandya A, et al. Improved clinical outcomes of high risk beta thalassemia major patients undergoing a HLA matched related allogeneic stem cell transplant with a treosulfan based conditioning regimen and peripheral blood stem cell grafts. PLoS One. 2013;8(4):e61637. 12. Mathews V, Srivastava A, Chandy M. Allogeneic stem cell transplantation for thalassemia major. Hematol Oncol Clin North Am. 2014;28(6):1187-1200. 13. Dong A, Rivella S, Breda L. Gene therapy for hemoglobinopathies: progress and challenges. Transl Res. 2013;161(4):293-306. 14. Lucarelli G, Galimberti M, Polchi P, et al. Bone marrow transplantation in patients with thalassemia. N Engl J Med. 1990;322(7):417-421. 15. Lucarelli G, Clift RA, Galimberti M, et al. Marrow transplantation for patients with

221


A. Srivastava et al.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

222

thalassemia: results in class 3 patients. Blood. 1996;87(5):2082-2088. Hassan M, Ljungman P, Ringden O, et al. The effect of busulphan on the pharmacokinetics of cyclophosphamide and its 4hydroxy metabolite: time interval influence on therapeutic efficacy and therapy-related toxicity. Bone Marrow Transplant. 2000; 25(9):915-924. Srivastava A, Poonkuzhali B, Shaji RV, et al. Glutathione S-transferase M1 polymorphism: a risk factor for hepatic venoocclusive disease in bone marrow transplantation. Blood. 2004;104(5):1574-1577. Balasubramanian P, Desire S, Panetta JC, et al. Population pharmacokinetics of cyclophosphamide in patients with thalassemia major undergoing HSCT. Bone Marrow Transplant. 2012;47(9):1178-1185. Gaziev J, Nguyen L, Puozzo C, et al. Novel pharmacokinetic behavior of intravenous busulfan in children with thalassemia undergoing hematopoietic stem cell transplantation: a prospective evaluation of pharmacokinetic and pharmacodynamic profile with therapeutic drug monitoring. Blood. 2010; 115(22):4597-4604. Li C, Wu X, Feng X, et al. A novel conditioning regimen improves outcomes in beta-thalassemia major patients using unrelated donor peripheral blood stem cell transplantation. Blood. 2012;120(19):3875-3881. Mathews V, George B, Deotare U, et al. A new stratification strategy that identifies a subset of class III patients with an adverse prognosis among children with beta thalassemia major undergoing a matched related allogeneic stem cell transplantation. Biol Blood Marrow Transplant. 2007;13(8):889894. Sabloff M, Chandy M, Wang Z, et al. HLAmatched sibling bone marrow transplantation for beta-thalassemia major. Blood. 2011;117(5):1745-1750. Anurathapan U, Pakakasama S, Mekjaruskul P, et al. Outcomes of thalassemia patients undergoing hematopoietic stem cell transplantation by using a standard myeloablative versus a novel reduced-toxicity conditioning regimen according to a new risk stratification. Biol Blood Marrow Transplant. 2014;20(12):2066-2071. Gaziev J, Isgro A, Sodani P, et al. Optimal outcomes in young class 3 patients with thalassemia undergoing HLA-identical sibling bone marrow transplantation. Transplantation. 2016;100(4):925-932. Sodani P, Gaziev D, Polchi P, et al. New approach for bone marrow transplantation in patients with class 3 thalassemia aged younger than 17 years. Blood. 2004; 104(4):1201-1203. Chiesa R, Cappelli B, Crocchiolo R, et al. Unpredictability of intravenous busulfan pharmacokinetics in children undergoing hematopoietic stem cell transplantation for advanced beta thalassemia: limited toxicity with a dose-adjustment policy. Biol Blood Marrow Transplant. 2010;16(5):622-628. Choudhary D, Sharma SK, Gupta N, et al. Treosulfan-thiotepa-fludarabine-based conditioning regimen for allogeneic transplantation in patients with thalassemia major: a single-center experience from north India. Biol Blood Marrow Transplant. 2013; 19(3):492-495. Angelucci E, Matthes-Martin S, Baronciani D, et al. Hematopoietic stem cell transplantation in thalassemia major and sickle cell disease: indications and management recommendations from an international expert

panel. Haematologica. 2014;99(5):811-820. 29. Jawdat DM, Al Saleh S, Sutton P, et al. Chances of finding an HLA-matched sibling: The Saudi experience. Biol Blood Marrow Transplant. 2009;15(10):1342-1344. 30. Sodani P, Isgro A, Gaziev J, et al. Purified Tdepleted, CD34+ peripheral blood and bone marrow cell transplantation from haploidentical mother to child with thalassemia. Blood. 2010;115(6):1296-1302. 31. Gaziev J, Marziali M, Isgro A, et al. Bone marrow transplantation for thalassemia from alternative related donors: improved outcomes with a new approach. Blood. 2013;122(15):2751-2756. 32. Anurathapan U, Hongeng S, Pakakasama S, et al. Hematopoietic stem cell transplantation for homozygous beta-thalassemia and beta-thalassemia/hemoglobin E patients from haploidentical donors. Bone Marrow Transplant. 2016;51(6):813-818. 33. Ruggeri A, Eapen M, Scaravadou A, et al. Umbilical cord blood transplantation for children with thalassemia and sickle cell disease. Biol Blood Marrow Transplant. 2011; 17(9):1375-1382. 34. Sruamsiri R, Chaiyakunapruk N, Pakakasama S, et al. Cost utility analysis of reduced intensity hematopoietic stem cell transplantation in adolescence and young adult with severe thalassemia compared to hypertransfusion and iron chelation program. BMC health services research. 2013; 13:45. 35. Majhail NS, Rizzo JD, Lee SJ, et al. Recommended screening and preventive practices for long-term survivors after hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2012;18(3):348371. 36. Wingard JR, Majhail NS, Brazauskas R, et al. Long-term survival and late deaths after allogeneic hematopoietic cell transplantation. J Clin Oncol. 2011;29(16):2230-2239. 37. Khalil A, Zaidman I, Elhasid R, et al. Factors influencing outcome and incidence of late complications in children who underwent allogeneic hematopoietic stem cell transplantation for hemoglobinopathy. Pediatr Hematol Oncol. 2012;29(8):694-703. 38. La Nasa G, Caocci G, Efficace F, et al. Longterm health-related quality of life evaluated more than 20 years after hematopoietic stem cell transplantation for thalassemia. Blood. 2013;122(13):2262-2270. 39. Angelucci E, Muretto P, Lucarelli G, et al. Phlebotomy to reduce iron overload in patients cured of thalassemia by bone marrow transplantation. Italian Cooperative Group for Phlebotomy Treatment of Transplanted Thalassemia Patients. Blood. 1997;90(3):994-998. 40. Musallam K, Cappellini MD, Taher A. Challenges associated with prolonged survival of patients with thalassemia: transitioning from childhood to adulthood. Pediatrics. 2008;121(5):e1426-1429. 41. Giardini C, Galimberti M, Lucarelli G, et al. Desferrioxamine therapy accelerates clearance of iron deposits after bone marrow transplantation for thalassaemia. Br J Haematol. 1995;89(4):868-873. 42. Yesilipek MA, Karasu G, Kazik M, Uygun V, Ozturk Z. Posttransplant oral iron-chelating therapy in patients with beta-thalassemia major. Pediatr Hematol Oncol. 2010; 27(5):374-379. 43. Andreani M, Testi M, Lucarelli G. Mixed chimerism in haemoglobinopathies: from risk of graft rejection to immune tolerance. Tissue Antigens. 2014;83(3):137-146.

44. Andreani M, Testi M, Gaziev J, et al. Quantitatively different red cell/nucleated cell chimerism in patients with long-term, persistent hematopoietic mixed chimerism after bone marrow transplantation for thalassemia major or sickle cell disease. Haematologica. 2011;96(1):128-133. 45. Andreani M, Gianolini ME, Testi M, et al. Mixed chimerism evolution is associated with T regulatory type 1 (Tr1) cells in a betathalassemic patient after haploidentical haematopoietic stem cell transplantation. Chimerism. 2014;5(3-4):75-79. 46. Hsieh MM, Wu CJ, Tisdale JF. In mixed hematopoietic chimerism, the donor red cells win. Haematologica. 2011;96(1):13-15. 47. Breuer S, Preuner S, Fritsch G, et al. Early recipient chimerism testing in the T- and NK-cell lineages for risk assessment of graft rejection in pediatric patients undergoing allogeneic stem cell transplantation. Leukemia. 2012;26(3):509-519. 48. Angelucci E, Baronciani D. Allogeneic stem cell transplantation for thalassemia major. Haematologica. 2008;93(12):1780-1784. 49. Gratwohl A, Baldomero H, Gratwohl M, et al. Quantitative and qualitative differences in use and trends of hematopoietic stem cell transplantation: a Global Observational Study. Haematologica. 2013;98(8):12821290. 50. Drakopoulou E, Papanikolaou E, Georgomanoli M, Anagnou NP. Towards more successful gene therapy clinical trials for beta-thalassemia. Curr Mol Med. 2013; 13(8):1314-1330. 51. Nienhuis AW. Development of gene therapy for blood disorders: an update. Blood. 2013; 122(9):1556-1564. 52. Gupta RM, Musunuru K. Expanding the genetic editing tool kit: ZFNs, TALENs, and CRISPR-Cas9. J Clin Invest. 2014; 124(10):4154-4161. 53. Hacein-Bey-Abina S, Le Deist F, Carlier F, et al. Sustained correction of X-linked severe combined immunodeficiency by ex vivo gene therapy. N Engl J Med. 2002; 346(16):1185-1193. 54. Hacein-Bey-Abina S, Von Kalle C, Schmidt M, et al. LMO2-associated clonal T cell proliferation in two patients after gene therapy for SCID-X1. Science. 2003;302(5644):415419. 55. Naldini L, Blomer U, Gallay P, et al. In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science. 1996;272(5259):263-267. 56. Cavazzana-Calvo M, Payen E, Negre O, et al. Transfusion independence and HMGA2 activation after gene therapy of human betathalassaemia. Nature. 2010;467(7313):318322. 57. Payen E, Leboulch P. Advances in stem cell transplantation and gene therapy in the beta-hemoglobinopathies. Hematology Am Soc Hematol Educ Program. 2012;2012(276283. 58. Negre O, Eggimann AV, Beuzard Y, et al. Gene Therapy of the betaHemoglobinopathies by Lentiviral Transfer of the beta(A(T87Q))-Globin Gene. Hum Gene Ther. 2016;27(2):148-165. 59. Marina Cavazzana M e-AR, Emmanuel Payen, et al. Outcomes of gene therapy for severe sickle disease and beta-thalassemia major via transplantation of autologous hematopoietic stem cells transduced ex vivo with a lentiviral beta AT87Q-globin vector. Blood. 2015;126(23):202. 60. Mark C. Walters M JR, SuradejHongeng, et al. Update of results from the Northstar

haematologica | 2017; 102(2)


Cure for thalassemia major

61.

62.

63.

64.

65.

66.

Study (HGB-204): A Phase 1/2 study of gene therapy for beta-thalassemia major via transplantation of autologous hematopoietic stem cells transduced ex-vivo with a lentiviral beta AT87Q-globin vector (LentiGlobin BB305 Drug Product). Blood. 2015; 126(23):201. Soni S, Thompson AA, Walters M, Lebouche P, Calvo M. Early Results of a phase I/ II study of gene therapy for b-thalassemia major via transplantation of autologous hematopoietic stem cells transduced ex-vivo with a lentiviral bAT87Q-globin vector. Biology of Blood and Marrow Transplantation. 2015;21(2, Supplement): S28-S29. Boulad F, Wang X, Qu J, et al. Safe mobilization of CD34+ cells in adults with beta-thalassemia and validation of effective globin gene transfer for clinical investigation. Blood. 2014;123(10):1483-1486. Bluebird bio: A study evaluating the efficacy and safety of lentiglobin BB305 drug product in beta-thalassemia major and sickle cell disease. Clinicaltrials.gov Identifier:02151526 [updated: October 17 2016]. Available from: https://clinicaltrials.gov/ct2/show/NCT021 51526?term=NCT02151526&rank=1. Last accessed: 21st October 2016 Bluebird bio: A study evaluating the efficacy and safety of the lentiglobinÂŽ BB305 drug product in subjects with transfusion-dependent b-thalassemia. Clinicaltrials.gov Identifier:02906202 [updated: September 14 2016]. Available from: https://clinicaltrials.gov/ct2/show/NCT029 06202?term=Thalassemia+gene+therapy&r ank=7. Last accessed: 21st October 2016 Memorial Sloan Kettering Cancer Center: b-thalassemia major with autologous CD34+ hematopoietic progenitor cells transduced with TNS9.3.55 a lentiviral vector encoding the normal human b-globin gene. Clinicatrials.gov Identifier:01639690, [updated: August 2 2016]. Available from: https://clinicaltrials.gov/ct2/show/NCT016 39690?term=NCT01639690&rank=1. Last accessed: 21st October 2016 Alessandro Aiuti, IRCCS San Raffaele: Gene therapy for transfusion dependent beta-tha-

haematologica | 2017; 102(2)

67.

68.

69.

70.

71.

72. 73.

74.

75.

76.

lassemia (TIGET-BTHAL). Clinicaltrials.gov Identifier:02453477 [updated: June 21 2016]. Available from: https://clinicaltrials.gov/ct2/show/NCT024 53477?term=Thalassemia+gene+therapy&r ank=1. Last accessed: 21st October 2016 Stamatoyannopoulos G. Control of globin gene expression during development and erythroid differentiation. Exp Hematol. 2005;33(3):259-271. Sankaran VG, Menne TF, Xu J, et al. Human fetal hemoglobin expression is regulated by the developmental stage-specific repressor BCL11A. Science. 2008;322(5909):18391842. Sankaran VG. Targeted therapeutic strategies for fetal hemoglobin induction. Hematology Am Soc Hematol Educ Program. 2011;2011(459-465. Xu J, Peng C, Sankaran VG, et al. Correction of sickle cell disease in adult mice by interference with fetal hemoglobin silencing. Science. 2011;334(6058):993-996. Wilber A, Hargrove PW, Kim YS, et al. Therapeutic levels of fetal hemoglobin in erythroid progeny of beta-thalassemic CD34+ cells after lentiviral vector-mediated gene transfer. Blood. 2011;117(10):28172826. Gammon K. Gene therapy: editorial control. Nature. 2014;515(7526):S11-13. Lombardo A, Naldini L. Genome editing: a tool for research and therapy: targeted genome editing hits the clinic. Nat Med. 2014;20(10):1101-1103. Tebas P, Stein, D.,Tang, W. W. et al. . Gene editing of CCR5 in autologous CD4 T cells of persons infected with HIV. N Engl J Med. 2014;370(10):901-910. Hoban MD, Cost GJ, Mendel MC, et al. Correction of the sickle cell disease mutation in human hematopoietic stem/progenitor cells. Blood. 2015;125(17):2597-2604. Siyuan Tan P K-HC, Sarah Smith, et al. Genome editing of the Bcl11A erythroid specific enhancer in bone marrow derived hematopoietic stem and progenitor cells for the treatment of sickle cell disease. Blood Abstract 203: 57th Annual Meeting Abstracts. 3 Dec 2015;

77. Ma N, Liao B, Zhang H, et al. Transcription activator-like effector nuclease (TALEN)mediated gene correction in integration-free beta-thalassemia induced pluripotent stem cells. J Biol Chem. 2013;288(48):3467134679. 78. Xie F, Ye L, Chang JC, et al. Seamless gene correction of beta-thalassemia mutations in patient-specific iPSCs using CRISPR/Cas9 and piggyBac. Genome Res. 2014; 24(9):1526-1533. 79. Ma N, Shan Y, Liao B, et al. Factor-induced reprogramming and zinc finger nucleaseaided gene targeting cause different genome instability in beta-thalassemia induced pluripotent stem cells (iPSCs). J Biol Chem. 2015;290(19):12079-12089. 80. Song B, Fan Y, He W, et al. Improved hematopoietic differentiation efficiency of gene-corrected beta-thalassemia induced pluripotent stem cells by CRISPR/Cas9 system. Stem Cells Dev. 2015;24(9):1053-1065. 81. Xu P, Tong Y, Liu XZ, et al. Both TALENs and CRISPR/Cas9 directly target the HBB IVS2-654 (C > T) mutation in beta-thalassemia-derived iPSCs. Sci Rep. 2015; 5(12065. 82. Szabo E, Rampalli S, Risueno RM, et al. Direct conversion of human fibroblasts to multilineage blood progenitors. Nature. 2010;468(7323):521-526. 83. Genovese P, Schiroli G, Escobar G, et al. Targeted genome editing in human repopulating haematopoietic stem cells. Nature. 2014;510(7504):235-240. 84. AdisInsight: Study of SB-BCLmR-HSPC (BCL11A enhancer approach) for the treatment of transfusion-dependent beta-thalassemia major; 2015 [updated 13 May 2015; cited 20 October 2016]. Available from: http://adisinsight.springer.com/trials/70025 3680 . Last accessed: 21st October 2016. 85. Urnov F D RA, Vierstra J et al. Clinical-scale genome editing of the BCL11a human erythroid enhancer for treatment of the hemoglobinopathies. Blood. 2015;126(23):204. 86. Lopes Gde L, Jr., de Souza JA, Barrios C. Access to cancer medications in low- and middle-income countries. Nat Rev Clin Oncol. 2013;10(6):314-322.

223


REVIEW ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2017 Volume 102(2):224-234

Anti-thymocyte globulin as graft-versus-host disease prevention in the setting of allogeneic peripheral blood stem cell transplantation: a review from the Acute Leukemia Working Party of the European Society for Blood and Marrow Transplantation Frédéric Baron,1 Mohamad Mohty,2-4 Didier Blaise,5 Gérard Socié,6 Myriam Labopin,2,4 Jordi Esteve,7 Fabio Ciceri,8 Sebastian Giebel,9 Norbert Claude Gorin,2 Bipin N Savani,10 Christoph Schmid11 and Arnon Nagler12,13

Giga-Hematology University of Liège, Belgium; 2Hopital Saint-Antoine, AP-HP, Paris, France,3Université Pierre & Marie Curie, Paris, France; 4INSERM UMRs U938, Paris, France; 5Aix Marseille Univ, CNRS, INSERM, CRCM, Institut Paoli-Calmettes, Marseille, France; 6AP-HP, Hematology Transplantation, Hospital Saint-Louis, Paris, France; 7 Department of Hematology, Hospital Clinic, Barcelona, Spain; 8Department of Hematology, Ospedale San Raffaele, Università degli Studi, Milano, Italy; 9Maria Sklodowska-Curie Cancer Center and Institute of Oncology, Gliwice Branch, Gliwice, Poland; 10Long term Transplant Clinic, Vanderbilt University Medical Center, Nashville, TN, USA; 11Klinikum Augsburg, Department of Hematology and Oncology, University of Munich, Augsburg, Germany; 12Division of Hematology and Bone Marrow Transplantation, The Chaim Sheba Medical Center, Tel-Hashomer, Ramat-Gan, Israel and 13EBMT Paris Office, Hospital Saint Antoine, Paris, France 1

ABSTRACT

A

Correspondence: f.baron@ulg.ac.be

Received: April 29, 2016. Accepted: August 24, 2016. Pre-published: December 7, 2016. doi:10.3324/haematol.2016.148510

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/224

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

224

llogeneic hematopoietic stem cell transplantation is increasingly used as treatment for patients with life-threatening blood diseases. Its curative potential is largely based on immune-mediated graftversus-leukemia effects caused by donor T cells contained in the graft. Unfortunately, donor T cells are also the cause of graft-versus-host disease. The vast majority of human leukocyte antigen-matched allogeneic hematopoietic stem cell transplants are nowadays carried out with peripheral blood stem cells as the stem cell source. In comparison with bone marrows, peripheral blood stem cells contain more hematopoietic stem/progenitor cells but also one log more T cells. Consequently, the use of peripheral blood stem cells instead of bone marrow has been associated with faster hematologic recovery and a lower risk of relapse in patients with advanced disease, but also with a higher incidence of chronic graftversus-host disease. These observations have been the basis for several studies aimed at assessing the impact of immunoregulation with antithymocyte globulin on transplantation outcomes in patients given human leukocyte antigen-matched peripheral blood stem cells from related or unrelated donors. After a brief introduction on anti-thymocyte globulin, this article reviews recent studies assessing the impact of antithymocyte globulin on transplantation outcomes in patients given peripheral blood stem cells from human leukocyte antigen-matched related or unrelated donors as well as in recipients of grafts from human leukocyte antigen haploidentical donors.

Introduction Allogeneic hematopoietic stem cell transplantation (HCT) is a potentially curative treatment for a wide range of hematologic malignancies.1,2 Its anti-tumor activity relies in large part on immune-mediated graft-versus-leukemia (GvL) effects.3,4 However, donor immune cells contained in the graft can also target healthy host haematologica | 2017; 102(2)


In vivo T-cell depletion with ATG

tissues causing graft-versus-host disease (GvHD).5 GvHD can be divided into two syndromes, acute GvHD which is an inflammatory disease causing maculopapular erythematous rash, gastrointestinal symptoms, and/or cholestatic hepatitis, and chronic GvHD which generally occurs beyond day 100, is non-inflammatory, and shares many clinical features with autoimmune diseases.6 Acute GvHD can be further divided into classical acute GvHD occurring within the first 100 days after transplantation, and late acute GvHD occurring later.7,8 Chronic GvHD can also be subdivided, in this case between classical chronic GvHD in which only typical signs of chronic GvHD are present, and overlap syndrome in which typical manifestations of both acute and chronic GvHD coexist.7,8 While chronic GvHD has been associated with GvL effects,3,8-11 it is also the leading cause of mortality/morbidity in long-term transplant recipients12,13 and impairs quality of life. Although the pathogeneses of acute and chronic GvHD are distinct and involve several types of cells,5,14 donor T cells play a pivotal role in both syndromes as demonstrated by the very low incidence of GvHD observed when profound ex vivo T-cell depletion is performed,15 even in the HLA-mismatched setting.16 The vast majority of HLA-matched allogeneic HCT performed as treatment for acute leukemia in 2013 were carried out with peripheral blood stem cells (PBSC) as the stem cell source.17 The use of PBSC instead of bone marrow (BM) in patients receiving grafts from HLA-matched donors has been associated with faster hematologic recovery and a lower risk of relapse in patients with advanced disease (due to greater GvL effects), but also with higher incidences of each of severe acute and extensive chronic GvHD.18-20 These observations served as the basis for studies combining the use of PBSC with in vivo T-cell depletion of the graft with the aim of benefiting from the faster engraftment associated with the use of PBSC without exposing patients to high risks of severe GvHD. In this article, after briefly discussing mechanisms of action of the different brands of anti-thymocyte globulin (ATG), we review recent studies assessing the impact of immunoregulation with ATG on transplantation outcomes in patients given PBSC from HLA-matched donors as well as in those given grafts (PBSC plus granulocyte colony-stimulating factormobilized BM) from HLA-haploidentical donors, and propose indications for the use of ATG in those settings.

Anti-thymocyte globulin Three preparations of ATG are currently available (Table 1).21 ATGAM (ATG-h) consists of polyclonal IgG obtained

from hyperimmune sera of horses immunized with human thymic cells. The two other brands of ATG consist of polyclonal IgG obtained from hyperimmune sera of rabbits immunized either with human thymocytes recovered from patients undergoing cardiac surgery (Thymoglobuline, ATG-T) or with the human Jurkat leukemic T-cell line (which was derived from the peripheral blood of a 14-year old boy suffering from acute T-cell leukemia22) [ATG Fresenius/Neovii (ATG-F)]. Although ATG-h is still currently used for in vivo T-cell depletion in the USA, two prospective randomized studies (including one performed almost 4 decades ago) failed to demonstrate its efficacy at preventing acute or chronic GvHD after HLA-matched BM transplantation (BMT).23,24 Furthermore, a retrospective study from the Brazilian National Cancer Institute in a cohort of 40 patients with aplastic anemia receiving BMT from HLA-identical siblings observed higher incidences of grade II-IV acute GvHD (35% versus 0%, P=0.009) and moderate/severe chronic GvHD (3-year rate: 34% versus 0%, P=0.04) in the 20 patients conditioned with cyclophosphamide plus ATG-h (90 mg/kg total dose) than in those conditioned with cyclophosphamide plus ATG-T (8 mg/kg total dose).25 These findings might be due to the fact that, in comparison to rabbit ATG, ATG-h induces less profound and less durable lymphopenia, even if it is administered at higher doses.25,26 Antigens targeted by ATG-T have been well described and include antigens expressed on T cells (such as CD2, CD3, CD4, CD7, or CD8), B cells, natural killer cells, macrophages and dendritic cells, as well as HLA class 1 and HLA-DR.27 Recognition by ATG-T of B cells and dendritic-cell antigens can also be attributed to the presence of antigen-presenting cells, thymic stromal cells and B cells in thymus fractions, although they are composed mainly of T cells.27 Furthermore, ATG-T also contains antibodies targeting antigens involved in cell adhesion and cell trafficking, as well as antigens involved in inflammation, apoptosis and cell proliferation.27 Since ATG-F is produced by immunizing rabbits with a homogenous Jurkat cell line, and since during ATG-F (but not ATG-T) production rabbit IgG are adsorbed on human placental cells in addition to adsorption on human erythrocytes, the spectrum of antigens recognized by ATG-F is narrower than that recognized by ATG-T. For example, ATG-F does not contain or contains significantly fewer antibodies directed against CD3, CD4 or HLA-DR28,29 (Figure 1). However, in contrast, compared to ATG-T, ATG-F contains more antibodies directed against CD107 (an antigen expressed on T cells during degranulation following antigenic stimulation).29 Competitive binding experiments have demonstrated that

Table 1. Types of ATG.

Name

Type of antibodies

Antithymocyte globulin (ATG) ATGAM (ATG-h) ATG-Thymoglobuline (ATG-T) ATG-Fresenius / Neovii (ATG-F)

Lympho-depletion in vivo

GvHD prevention (total dose administered)

Polyclonal IgG from horses immunized with human thymocytes Polyclonal IgG from rabbits immunized with human thymocytes

+/+

Polyclonal IgG from rabbits immunized with human Jurkat T leukemia cell line

+

â&#x20AC;&#x201C;24* + (2.5-10 mg/kg) + (15-60 mg/kg)

*Champlin et al.24

haematologica | 2017; 102(2)

225


F. Baron et al.

ATG-T has stronger reactivity than ATG-F to activated peripheral blood mononuclear cells.28 Furthermore, complement-mediated cytotoxic effects of ATG-T on peripheral blood mononuclear cells were stronger than those mediated by ATG-F28 and ATG-T induced dendritic-cell apoptosis more effectively than similar doses of ATG-F.30 Consequently, doses of ATG given for GvHD prohylaxis have been typically higher for ATG-F (15-60 mg/kg) than for ATG-T (2.5-10 mg/kg) (Table 1). After in vivo infusion, all forms of ATG induce depletion of both T and antigen-presenting cells by complementdependent lysis or antibody-dependent cellular cytotoxicity, apoptosis of activated T cells, and maintenance of dendritic cells in a tolerogeneic state.27 Furthermore, rabbit (but not horse) ATG induces the generation of regulatory T cells (Treg), both in vitro and in vivo.31-33 This is clinically relevant given accumulating evidence showing an important role for Treg in GvHD prevention after allogeneic HCT.34-37 The next paragraphs review the impact of both forms of rabbit ATG on transplantation outcomes in various transplantation settings. However since there has been no controlled head-to-head comparison between ATG-T and ATG-F reported thus far in the allogeneic HCT setting, one should be cautious when extrapolating data from medical literature referring to one of these two drugs.

Role of rabbit anti-thymocyte globulin in patients given HLA-matched peripheral blood stem cells after myeloablative conditioning Pharmacokinetics of anti-thymocyte globulin and impact of post-transplant anti-thymocyte globulin serum levels on transplantation outcomes Several studies have assessed ATG pharmacokinetics after PBSC transplantation. These studies demonstrated that residual total rabbit IgG levels or residual free T-cellspecific rabbit IgG levels were each influenced by the dose of ATG administered, although there was a considerable inter-patient variability. In detail, Forcina et al. assessed specific ATG-F pharmacokinetics in 22 patients who underwent allogeneic HCT after a myeloablative conditioning combining fludarabine and treosulfan.38 ATG-F was administered at a dose of either 10 mg/kg/day (n=17) or 20 mg/kg/day (n=5) on days -4, -3 and -2 before transplantation. T-cell-specific rabbit IgG levels peaked at the end of the last dose of ATG-F administration and were four times higher in patients given the 20 mg/kg dose. These differences persisted on day 0. Furthermore, while patients given the 10 mg/kg dose reached sub-therapeutic specific rabbit IgG levels on day +10 after transplantation, those given the 20 mg/kg dose kept supra-therapeutic specific rabbit IgG levels beyond day +21 after the allogeneic transplant. Waller et al. assessed ATG-T pharmacokinetics in 19 patients with high-risk hematologic malignancies who received CD34+-selected, lymphocyte-depleted PBSC from partially HLA-matched related donors.39 ATG-T was administered at a dose of 2.5 mg/kg/day (n=2, 10 mg/kg total dose) or 1.5 mg/kg/day (n=17, 6 mg/kg total dose) for 4 consecutive days (the last 4 days of the conditioning regimen). In comparison to patients given ATG-T at the 6 mg/kg total dose, those receiving a total dose of 10 mg of ATG-T had comparable total rabbit IgG levels (77±14 ver226

Figure 1. Quantification of ATG antibodies targeting specific human antigens. (adapted from Table 1 from Popow et al.29) Black bars represent ATG-F data and white bars ATG-T data.

sus 62±22 μg/mL, P=0.4) but higher specific rabbit IgG levels on day 0 (25±8 versus 9±6 μg/mL, P=0.002). Furthermore, the number of days required to reach infratherapeutic levels of specific rabbit IgG levels was 35±4 versus 17±9 days in patients given 10 or 6 mg/kg ATG-T total dose, respectively (P=0.01). Remberger et al. prospectively assessed total ATG-T levels in 76 patients given PBSC (n=60) or BM (n=16) after myeloablative (n=37) or reduced-intensity (n=39) conditioning.40 All patients received ATG-T at the dose of 2 mg/kg/day for 2-4 days (total dose 4 mg/kg to 8 mg/kg) with the last dose given on day -1. Day 0 and 7 total rabbit IgG levels were 49 and 26 μg/mL, respectively in patients receiving 6 mg/kg ATG-T total dose (n=46), versus 63 and 42 μg/mL respectively in patients receiving 8 mg/kg ATGT total dose (n=26). The estimated half-life of total rabbit IgG was 9 days. Analyses of the impact of ATG serum levels on transplantation outcomes were pioneered by the group of Jan Storek at the University of Calgary. This group correlated ATG serum levels on days 0 (immediately before graft infusion), 7 and 28 after allogeneic HCT with transplantation outcomes in a large cohort of patients conditioned with a myeloablative regimen combining fludarabine (250 mg/m2 total dose), busulfan [12.8 mg/kg IV – pharmacokinetics adjusted - total dose], and ATG-T [given at the doses of 0.5 mg/kg on day −2, 2.0 mg/kg on day −1, and 2.0 mg/kg on day 0 (total, 4.5 mg/kg)]. Patients with leukemia also received 4 Gy total body irradiation. The authors demonstrated that high levels of serum ATG-T on days 7 and 28 each correlated with a low incidence of acute and chronic GvHD, but also with a high incidence of post-transplant lymphoproliferative disorder.41,42 In contrast, high ATG levels on day 0 were associated only with a low incidence of chronic GvHD.42 Importantly, no associations were observed between ATG levels and relapse or non-relapse mortality. Subsequently, the same group demonstrated that high levels of ATG-T specificities capable of binding to Treg and invariant natural killer T cells on day 7 were associated with a low incidence of relapse.43 haematologica | 2017; 102(2)


In vivo T-cell depletion with ATG

Impact of anti-thymocyte globulin on immune recovery The first large study assessing the impact of ATG-T on immune recovery was reported by Bosch et al.44 The authors compared immune recovery the first 2 years after transplantation in patients given PBSC either in Calgary after the fludarabine/busulfan/ATG-T regimen described above (n=125), or in Seattle after conditioning with cyclophosphamide 120 mg/kg and 12 Gy total body irradiation (and no ATG) (n=46). Post-grafting immunosuppression and flow-cytometry analyses were similar in the two cohorts of patients. Main observations were that, 1 month after allogeneic HCT, ATG-T patients had lower counts of B cells, as well as CD4+ and CD8+ T cells, while, for at least 1 year after transplantation, counts of naïve and memory CD4+ T cells as well as of naïve CD8+ T cells remained lower in patients given ATG-T than in those not given ATG. Comparable qualitative observations have been recently made in a cohort of 65 patients given PBSC after various myeloablative conditioning regimens at the

University of Liège and who were given ATG-F (45 mg/kg total dose, n=37) or not (n=28).45

Non-randomized studies comparing anti-thymocyte globulin versus no anti-thymocyte globulin A number of phase II studies have assessed the impact of rabbit ATG in patients given unmodified grafts after myeloablative conditioning46-52 (Table 2). Collectively, these studies suggested that ATG decreased the incidence of grade III-IV acute and chronic GVHD without increasing non-relapse mortality (some studies even found lower non-relapse mortality with ATG), increased the incidence of post-transplantation lymphoproliferative disorders, and improved quality of life. A recent study from La Societe Francophone de Greffe de Moelle et de Thérapie Cellulaire (SFGM-TC) compared the impact of ATG according to stem cell source (PBSC versus BM) in patients given grafts from 10/10 HLA-matched unrelated donors following myeloablative conditioning regimens. Data from 356

Table 2. Selected non-randomized study comparing myeloablative allogeneic stem cell transplantation outcomes in patients given ATG or not.

Study

N. of ATG patients/ N. of control patients

Unrelated donors Zander et al.46

145 / 188

Study design

Main observations. In comparison to control patients, ATG patients had:

Retrospective analysis of CML patients given ATG-F (average dose 90 mg/kg) or not before unrelated BMT or PBSCT.

Faster leukocyte enraftment (P<0.001) Lower incidence of grade III-IV acute GvHD (P=0.03) Lower NRM (P=0.03) A trend for higher relapse (P=0.09) Better OS (P=0.03) Lower NRM (P=0.004) Better DFS (P=0.005) Better OS (P=0.002) Lower incidence of grade III-IV acute GvHD (P=0.03) Lower incidence of chronic GvHD (P=0.002) Higher incidence of infections (P=0.02) Better global quality of live (P<0.01) Less fatigue (P<0.01) Lower incidence of grade III-IV acute GvHD (P=0.02) Lower incidence of chronic GvHD (P<0.001) Lower NRM (P=0.01) Better OS (P=0.01) Slower neutrophil engraftment (P<0.001) Higher EBV infection Higher viral infection Lower incidence of chronic GvHD (P=0.03) Lower NRM and better OS with lower ATG doses Lower incidence of grade III-IV acute GvHD (P=0.04) Lower incidence of chronic GvHD (P=0.03) Higher GvHD-free/relapse-free survival (P<0.01) No statistically significant impact of ATG on outcomes

Schattenberg et al.47

34 / 22

Yu et al.48

54 / 42

Ratanatharathorn et al.50*

76 / 121

Retrospective study in patients given ATG-T (4.5 mg/kg) or not before unrelated HCT (mainly PBSCT)

Binkert et al.51

120 / 145

Retrospective study in patients given ATG-F (35 or 60 mg/kg) or not before related (n=117) or unrelated (n=148) PBSCT

Ravinet et al.53

47 / 92

Retrospective study in patients given ATG-T or not before unrelated PBSCT

Ravinet et al.53

64 / 153

Retrospective study in patients given ATG-T or not before unrelated BMT

Related donors Bonifazi et al.52

47 / 146

Wolschke et al.49*

79 / 159

Retrospective study of patients given ATG-F (15 or 30 mg/ kg) before PBSCT from HLA-identical siblings Retrospective study in patients receiving PBSCT from sibling donors and given ATG-F (median dose of 30 mg/kg) or not.

Retrospective study in patients given ATG-T (8 mg or 16 mg/kg) or not before partially TCD unrelated BMT. Retrospective study in patients given ATG-F (16 mg/kg) or not before related or unrelated HCT.

Lower incidence of extensive chronic GvHD (P=0.03)

Slower leukocyte enraftment (P=0.001) Higher incidence of PTLD (P=0.05) Lower incidence of grade II-IV acute GvHD (P=0.04) Lower incidence of chronic GvHD (P=0.002)

iTCD: in vivo T-cell depletion; ATG-T: ATG-Thymoglobuline; ATG-F: ATG-Fresenius/Neovii; Alem: alemtuzumab; BMT: bone marrow transplantation; HCT: hematopoietic cell transplantation; PBSCT: peripheral blood stem cell transplantation; CML: chronic myeloid leukemia; GvHD: graft-versus-host disease; OS: overall survival; DFS: disease-free survival; NRM: non-relapse mortality; *also includes patients given grafts after reduced intensity conditioning.

haematologica | 2017; 102(2)

227


F. Baron et al.

patients with acute myeloid leukemia or myelodysplastic syndrome were included in the analyses. Among patients given PBSC (n=139), those given ATG (n=47) had lower cumulative incidences of grade III-IV acute GvHD (HR 0.17, P=0.04) and chronic GvHD (HR 0.31, P=0.03) in comparison to those not given ATG.53 Interestingly, the patients who received ATG had a significantly better GvHD-free/relapse-free survival (HR 0.48, P<0.01) than patients who did not. Notably, these correlations were not statistically significant in the group of patients who received BM as their source of stem cells (n=217).

Anti-thymocyte globulin dose-finding study The best dose of ATG to use in patients who have been given PBSC after myeloablative conditioning has remained undefined. In an effort to address this issue, Deeg et al. performed a dose-finding study of ATG-T in patients with myeloid malignancies given PBSC after targeted busulfan plus cyclophosphamide conditioning.54 The starting dose of ATG-T was 4.5 mg/kg total dose (0.5 mg/kg on day -3, and 2 mg/kg on each of days -2 and -1) and escalation was dependent on the occurrence of acute GvHD on the one hand, and Epsteinâ&#x20AC;&#x201C;Barr virus (EBV) reactivation on the other hand. The authors identified a total dose of 6 mg/kg as the highest tolerable dose (in term of EBV reactivation). The incidences of grade II-IV acute and extensive chronic GvHD were comparable in patients given 4.5 or 6.0 mg/kg ATG-T, but were both lower in ATG-T patients (n=56) than in concurrent control patients (n=27) (50% versus 82% for grade II-IV acute GvHD, and 34% versus 82% for extensive chronic GvHD, respectively). This study suggests that administration of ATG-T at a total dose of 4.5 mg/kg to 6 mg/kg is safe and might successfully prevent GvHD in patients receiving PBSC from HLA-matched donors after myeloablative conditioning.

Randomized studies comparing anti-thymocyte globulin versus no anti-thymocyte globulin Four randomized studies have assessed the impact of rabbit ATG on allogeneic HCT outcomes.55-58 None of these studies was double-blinded and thus the possibility of a certain bias in the grading of GvHD cannot be ruled out. - Unrelated donors The first randomized study assessing the use of rabbit ATG was carried out by the Gruppo Italiano Trapianti Midollo Osseo (GITMO) in patients who underwent BMT from unrelated donors after myeloablative conditioning.55

It is important to note that this study was performed before HLA high-resolution typing was available. In a first part of the study, 54 patients were randomized between no ATG or ATG-T 7.5 mg/kg total dose, while in the second part, 55 patients were randomized between no ATG or ATG-T 15 mg/kg total dose (Table 3). Patients not given ATG and those given 7.5 mg/kg ATG-T had similar incidences of grade III-IV acute GvHD, while patients given ATG-T 15 mg/kg had a lower incidence of grade III-IV acute GvHD but also a higher incidence of infections. Importantly, the incidence of extensive chronic GvHD was significantly lower among the patients given ATG-T 7.5 mg (38%) and ATG-T 15 mg (41%) than among the control patients (62%; P=0.04). However, ATG-T failed to improve overall survival, even with a long follow-up.59 Finke et al. and Socie et al. conducted a phase III randomized study comparing standard GvHD prophylaxis with cyclosporine and short-course methotrexate with or without added ATG-F.56,60 The study included 201 patients who underwent unrelated BMT (n=37) or PBSC transplantation (n=164) after myeloablative conditioning. Patients were randomized between ATG-F (20 mg/kg on days -3, -2 and -1) or no ATG. The primary endpoint was grade III-IV acute GvHD or death within the first 100 days after the allogeneic HCT. Although the primary endpoint was not statistically different between the two groups (P=0.13), the study demonstrated that the patients given ATG had lower incidences of grade II-IV acute GvHD (33% versus 51%, P=0.01),56 chronic GvHD (30% versus 60%, P<0.001), and extensive chronic GvHD (12% versus 45%, P<0.001).60 However, ATG significantly delayed both neutrophil engraftment (26 versus 19 days to achieve 1x109 neutrophils/L, P<0.001) and platelet engraftment (20 versus 37 days to achieve 50x109 platelets/L, P<0.001), while five patients in the ATG arm but none in the control arm developed a post-transplant lymphoproliferative disorder.56 Importantly, although there was only a statistically non-significant survival advantage for ATG patients (at 3 years 55% versus 43%, P=0.39), the 3-year probability of being alive and free of immunosuppressive drugs was three times higher among ATG patients than among control patients (53% versus 17%, P<0.001).60 Another randomized study in the unrelated allogeneic HCT setting was recently reported by Walker et al.57 The main inclusion criteria were HLA-matched or 1/8 single HLA-mismatched unrelated donor, hematologic malignancy, myeloablative conditioning or reduced intensity conditioning (RIC), and PBSC or BM as the stem cell

Table 3. Randomized studies of rabbit ATG as GvHD prevention in patients given allogeneic hematopoietic cell transplantation.

N. of patients

Bacigalupo et al.55 Bacigalupo et al.55 Finke & Socie et al.56 60 Kroger et al.58 Walker et al.57

54 55 201 155 196

ATG brand / total dose (mg/kg)

T / 7.5 T / 15 F / 60 F / 30 T / 4.5

Acute GvHD II-IV

Chronic GvHD

Non-relapse mortality

Relapse

Overall survival

% ATG/ % no ATG (P)

% ATG/ % no ATG (P)

% ATG/ % no ATG (P)

% ATG/ % no ATG (P)

% ATG/ % no ATG (P)

69 / 72 (0.6) 37 / 79 (0.001) 33 / 51 (0.01) 11 / 18 (0.13) 50 / 65 (0.01)d

38 / 65 (0.08) 41 / 59 (0.3) 30 / 60 (<0.001)a 32 / 69 (<0.001)c 22 / 33 (0.06)b

43 / 39 (0.7)a 47 / 49 (0.9)b 19 / 34 0.18)a 14 / 12 (0.6)c 23 / 24 (NS)b

10 / 12 (0.6)a 36 / 18 (0.8)b 33 / 28 (0.5)a 32 / 26 (0.17)c 11 / 16 (NS)b

56 / 55 (0.8)a 43 / 43 (0.8)b 55 / 43 (0.39)a 74 / 78 (0.5)c 75 / 65 (0.24)b

at 3 years; bat 1 year; cat 2 years; dgrade I-IV at day 100. F: ATG-Fresenius; T: ATG-Thymoglobuline.

a

228

haematologica | 2017; 102(2)


In vivo T-cell depletion with ATG

source. The primary endpoint was freedom from GvHD and systemic immunosuppressive treatment without resumption up to 12 months after transplantation. Of 203 randomized patients, 196 were eligible. These included 99 patients randomized to ATG-T (0.5 mg/kg on day -2 and 2 mg/kg on days -1 and +1), and 97 randomized to no ATG. Sixty-seven percent of patients received myeloablative conditioning (and 33% RIC), and in 88% of cases peripheral blood was the stem cell source. The main observations were that the primary endpoint was met in 37% of patients given ATG versus 16% of the patients not given ATG (P<0.001). Furthermore, ATG-treated patients had a significantly lower incidence of grade I-IV acute (50% versus 65% at 100 days, P=0.01) and chronic (22 versus 33, P=0.06) GVHD, without an increased risk of relapse (11% versus 16% at 1 year). However, there was a higher incidence of EBV reactivation among the patients given ATG (33% versus 3%). Finally, there was a non-statistically significant survival advantage for the ATG-treated patients (at 1 year: 75% versus 65%, P=0.2). Longer follow-up is needed to determine the impact of ATG on late relapses and on late non-relapse deaths. The study by Walker et al. also prospectively assessed the impact of ATG-T on quality of life.57 Patients assigned to the ATG-T arm had better Atkinson Life Happiness Scale Scores at 12 months and lower chronic GvHD symptom burden (assessed by the Lee scale) at 6 and 12 months after transplantation. These data suggest a beneficial impact of ATG on quality of life. - HLA-identical sibling donors The impact of ATG on transplantation outcomes has recently been tested in the setting of PBSC transplantation from HLA-identical sibling donors in patients suffering from acute leukemia in complete remission.58 One hundred fifty-five eligible patients were randomized between ATG-F (n=83, administered at 10 mg/kg/day on days -3, 2 and -1) or no ATG (n=72). The 2-year cumulative incidence of chronic GvHD (primary endpoint of the study) was 32% in ATG patients versus 69% in controls (P<0.001). Other observations included slower leukocyte and platelet engraftment in ATG patients, while other transplantation outcomes (infections, acute GvHD, relapse incidence, non-relapse mortality, overall survival and progression-free survival) were not statistically significantly different between the two groups of patients (Table 3). Since this study only included patients in complete remission at the time of transplantation, further studies are needed to assess the impact of ATG in patients with advanced leukemia.

Role of rabbit anti-thymocyte globulin in patients given HLA-matched peripheral blood stem cells after reduced-intensity conditioning RIC allogeneic HCT relies mainly on GvL effects for tumor eradication.61-63 The biology of GvL effects remains poorly defined but has been thought to involve reactions to polymorphic minor histocompatibility antigens expressed either specifically on hematopoietic cells or more widely on a number of tissue cells.64 As mentioned above, several studies have demonstrated a close relationships between GvHD and GvL responses after RIC or low-intensity immunosuppressive conditioning,8-11,65 suggesting that the use of in vivo T-cell depletion might be haematologica | 2017; 102(2)

harmful in that setting. However achievement/maintenance of complete remission has been observed in many patients given grafts after RIC/low-intensity conditioning who did not develop GvHD,8-11,66 suggesting that clinical manifestations of GvHD are not universally required in order to achieve/maintain remissions. Unfortunately, the randomized studies on the use or not of ATG were conducted mostly in patients undergoing myeloablative transplants. However, 33% of the patients included in the study by Walker et al. described above were given RIC or low intensity conditioning and, interestingly, the relapse incidence was similar between patients who were given ATG (n=33) and those who were not (n=31) also in this subgroup of patients.57 Since it is hazardous to draw definitive conclusions based on data from a subgroup analysis of 64 patients, we also have to rely on phase II prospective and registry studies to estimate the impact of ATG specifically in the RIC allogeneic HCT setting.

Data from single center studies The Marseille group conducted a number of phase II studies aimed at optimizing ATG-T dosing in patients given grafts after RIC conditioning.67-70 They used the RIC regimen combining fludarabine and busulfan.61 Post-grafting immunosuppression consisted of cyclosporine alone in the case of HLA-identical sibling donors, or cyclosporine plus mycophenolate mofetil in the case of unrelated donors. In a first study, they compared transplantation outcomes in patients given high (7.5-10 mg/kg total dose, n=46) or low (2.5 mg/kg total dose, n=55) dose ATG-T before transplantation from HLA-identical siblings. Incidences of grade II-IV acute GvHD (P=0.001) and chronic GvHD (P=0.02) were significantly lower in patients given the higher dose of ATG. However, this benefit was offset by a higher incidence of relapse in the group of patients given high-dose ATG. In a second study, the authors retrospectively compared ATG-T 2.5 mg/kg total dose (n=124) to ATG-T 5 mg/kg total dose (n=105).68 All patients received PBSC from either HLA-identical siblings (n=187) or unrelated donors (n=42). The main observations were that patients given 5 mg of ATG had lower incidences of grade II-IV acute GvHD (23% versus 42%, P=0.002) and chronic GvHD (35% versus 69%, P<0.001) than those given 2.5 mg/kg ATG. However, importantly, other transplantation outcomes (relapse, non-relapse mortality, overall survival and leukemia-free survival) were comparable between the two groups of patients. Similar observations were made when the analyses were restricted to patients transplanted as treatment for myeloid malignancies.69 Investigators from Karolinska University performed a retrospective study including data from 110 patients given unrelated PBSC (n=95) or BM (n=15) following chemotherapy-based RIC.71 Transplantation outcomes were compared between patients given a total ATG-T total dose of 6 mg/kg (n=66) or 8 mg/kg (n=44). The authors observed a higher incidence of relapse (P=0.04) and lower leukemia-free survival rate (P=0.04) in patients given the higher dose of ATG. More recently, Langston et al. reported the results of a retrospective study of 85 patients given PBSC (n=74) or BM (n=11) from unrelated donors at Emory University after fludarabine plus melphalan conditioning.72 Patients with 10/10 HLA-matched unrelated donors were not 229


F. Baron et al.

given ATG (n=54), while patients given grafts from HLAmismatched donors were given ATG-T 6 mg/kg total dose. Remarkably, the authors observed comparable outcomes in the two cohorts of patients, suggesting that the addition of ATG was able to offset the negative impact of HLA-mismatch without increasing the incidence of relapse or infection.

Data from registry studies A first registry study assessing the impact of in vivo T-cell depletion on outcomes in the RIC setting was reported by Soiffer et al. on behalf of the Center for International Blood and Marrow Transplant Research (CIBMTR).73 The study included data from 1676 adult patients given BM (n=203) or PBSC (n=1473) after various RIC regimens as treatment for a heterogeneous group of hematologic malignancies. Patients were divided into three groups: one group was not given either ATG or alemtuzumab (controls, n=879), one group was given ATG (n=584; including 160 patients who received horse ATG and 405 patients who received ATG-T at a median total dose of 7 mg/kg), and one group was given alemtuzumab (n=213). In multivariate analyses, in comparison to control patients, those given ATG had a similar incidence of grade III-IV acute GvHD (HR 0.9,

A

P=0.2), a lower incidence of chronic GvHD (HR 0.7, P<0.001), a higher incidence of non-relapse mortality (HR 1.3, P=0.01), and a higher incidence of relapse (HR 1.5, P<0.001). This translated into significantly worse overall survival (HR 1.3, P=0.002) and disease-free survival (HR 1.5, P<0.001) in patients given ATG. The Acute Leukemia Working Party (ALWP) of the European Society for Blood and Marrow Transplantation (EBMT) revisited this issue in a more homogeneous cohort of 1,250 patients with acute myeloid leukemia in first complete remission given PBSC from HLA-identical siblings.74 A total of 554 patients did not receive any form of in vivo T-cell depletion (control group), whereas ATG and alemtuzumab were given in 444 and 252 patients, respectively. In multivariate analyses, the use of ATG was associated with a lower risk of chronic GvHD (HR=0.6, P<0.001) and a lower risk of extensive chronic GvHD (HR=0.5, P<0.001). Furthermore, in contrast to what was observed in the CIBMTR study, ATG patients had a similar risks of relapse (HR=1.1, P=0.40) and non-relapse mortality (HR=0.9, P=0.6), and similar overall survival (HR=0.9, P=0.6) and leukemia-free survival (HR=1.0, P=0.8) in comparison to control patients (Figure 2). The impact of ATG on the risk of relapse and on leukemia-free

B

C

Figure 2. Impact of ATG on transplantation outcomes in patients given grafts after reduced-intensity conditioning. (A) Forest plots showing the results of multivariate analyses from two large registry studies assessing the impact of ATG on transplantation outcomes either in patients with various hematologic malignancies [study from the Center for International Blood and Marrow Transplant Research (CIBMTR)73] or in patients with acute myeloid leukemia in first complete remission [study from the Acute Leukemia Working Party of the European Society for Blood and Marrow Transplantation (EBMT)74]. (B) Relapse incidence and (C) progressionfree survival in the subroup of patients from the EBMT study given peripheral blood stem cells (PBSC) after busulfan-based RIC (n=674, including 389 patients given ATG) 74.

230

haematologica | 2017; 102(2)


In vivo T-cell depletion with ATG

survival in the subgroup of patients transplanted following a busulfan-based RIC is shown in Figure 2B,C. Within this subgroup of patients, the relapse incidence was similar in those given <6 mg/kg ATG and in those not given ATG (HR=1.1, P=0.9), while there was a suggestion of a higher incidence of relapses in patients given ATG at a dose ≥6 mg/kg (HR 1.4, P=0.08). Taken together, these phase II and registry study data suggest that a total dose of ATG-T around 5 mg/kg might be optimal in the RIC setting. This assumption should be tested in an appropriate phase III trial with a uniform conditioning regimen and post-grafting immunosuppression.75

Role of rabbit anti-thymocyte globulin in patients given HLA-matched peripheral blood stem cells after low-intensity immunosuppressive conditioning The impact of ATG has also been investigated in the setting of PBSC transplantation following low-intensity immunosuppressive conditioning. The Stanford group developed a low-intensity conditioning regimen combining ATG-T (7.5 mg total dose given at the dose of 1.5 mg/kg from day -11 to day -7) and 8 Gy total lymphoid irradiation in ten fractions followed by PBSC transplantation (TLI-ATG regimen).76,77 This conditioning allowed sustained engraftment and GvL effects (mainly in patients who achieved full donor T-cell chimerism) with a very low incidence of GvHD.76,77 In murine models of transplantation, this was achieved through Th2 polarization of donor T cells by recipient invariant natural killer/T cells (still present at transplantation given their relative resistance to ionizing radiation),78 and through expansion of donor Treg by recipient invariant natural killer/T cells.78 The Belgian Society of Hematology conducted a phase II randomized study comparing low-intensity transplantation following a fludarabine plus 2 Gy total body irradiation regimen (n=49) developed by the Seattle group or following TLI-ATG (n=45).79 The main observations were that overall survival rates were comparable in the two study arms (53% in patients treated with fludarabine and total body irradiation versus 54% in TLI-ATG patients) although the TLI-ATG patients had a significantly lower incidence of moderate/severe chronic GvHD (18% versus 41%, P=0.02) but a higher risk of relapse/progression at 4 years (50% versus 22%, P=0.02). The low incidence of chronic GvHD in TLI-ATG recipients could be attributed in part to higher Treg /naïve CD4 T-cell ratios early after transplantation.80

Role of rabbit anti-thymocyte globulin in patients given HLA-haploidentical peripheral blood stem cells Although only 20-30% of patients have an HLA-identical sibling donor, virtually all patients have a potential HLA-haploidentical donor. Historically, HLA-haploidentical transplants were performed by transplanting a “megadose” of in vitro-selected CD34+ cells.81-83 This technique resulted in extensive in vitro T-cell depletion that prevented severe GvHD but also affected immune reconstitution leading to high incidences of both infections and disease relapse/progression.84 Fortunately, several novel approaches have been developed in the last decades and have resulted in more favorhaematologica | 2017; 102(2)

able outcomes.85 These approaches include co-infusion of megadoses of CD34+ cells with donor Treg and conventional T cells at a 2:1 ratio,36 and T-cell-repleted HLA-haploidentical transplantation with either high-dose posttransplantation cyclophosphamide,86 or intensive pharmacological immunosuppression including ATG. Using the latter strategy, the Bejing group developed a protocol combining granulocyte-colony-stimulating factor-mobilized BM and PBSC, as well as administration of ATG-T to prevent both graft rejection and GvHD.87 ATG-T was administered from days -5 to -2 at the dose of either 2.5 mg/kg/day or 1.5 mg/kg/day. The conditioning regimen consisted of cytarabine (4 g/m2/day on days -10 to -9), busulfan (4 mg/kg/day orally or 3.2 mg/kg i.v. on days –8 to –6), cyclophosphamide (1.8 g/ m2/day on days -5 to -4), or semustine (250 mg/m2 on day -3), while post-grafting immunosuppression was obtained with cyclosporine, mycophenolate mofetil, and short-course methotrexate. The outcomes of 1,210 consecutive patients offered HLAhaploidentical transplantation following this strategy were reported recently. The incidence of grade III-IV acute GvHD was 12%, while the 3-year cumulative incidence of chronic GvHD was 50%. At 3 years, progression-free and overall survival rates were 67% and 70%, respectively. The same group of authors recently reported the results of a phase III non-inferiority trial investigating two doses of ATG-T (6 mg/kg total dose versus 10 mg/kg total dose) before infusion of granulocyte-colony-stimulating factormobilized BM and PBSC from HLA-haploidentical donors.88 The primary endpoint was the incidence of grade III-IV acute GvHD. In comparison to patients given ATG-T 10 mg/kg total dose, those given 6 mg/kg total dose had a higher incidence of grade III-IV acute GvHD (16% versus 4%, P=0.005). Furthermore, the 1.5-year cumulative incidence of chronic GvHD was 65% versus 45% in the ATG-T 6 mg/kg and ATG-T 10 mg/kg groups, respectively (P=0.01). However, the incidences of septicemia (5% versus 12%, P=NS), EBV reactivation (10% versus 25% at 1 year, P<0.001) and post-transplant lymphoproliferative disorder (2% versus 8%, P=0.03) were each lower in the ATG-T 6 mg/kg group than in the ATGT 10 mg/kg group. This could be attributed to slower immune recovery in patients given ATG-T 10 mg/kg total dose. Relapse, progression-free and overall survival rates were similar in the two arms. The Milan group developed another protocol of HLAhaploidentical transplantation. The study included 121 patients, most with advanced disease. The conditioning regimen combined treosulfan (14 g/m2/day on days -6 to -4) and fludarabine (30 mg/m2/day on days -6 to -2), and post-grafting immunosuppression combining ATG-F 10 mg/kg on days -4 to -2, sirolimus and mycophenolate mofetil.89 The incidence of grade III-IV acute GvHD was 22%, while the 2-year cumulative incidence of chronic GvHD was 47%. At 3 years, the cumulative incidences of non-relapse mortality, progression-free survival and overall survival were 31%, 20% and 25%, respectively.

Summary: possible indications for anti-thymocyte globulin in patients transplanted with peripheral blood stem cells In summary, three prospective randomized studies have demonstrated that ATG decreases the incidence of 231


F. Baron et al. Table 4. Proposed indications for immunoregulation with ATG in patients given PBSC from allogeneic donors.

Recommendation for ATG 58

Myeloablative PBSCT from matched sibling donors Myeloablative PBSCT from HLA-matchedunrelated donors56,60,57

standard of care standard of care

RIC-PBSCT fludarabine-busulfan68 Non-myeloablative PBSCT HLA-haplo-identical stem cell transplantation (Bejing approach)88

recommended developmental standard of care

Dose and timing of ATG ATG-F 10 mg/kg/day on days -3, -2 and -1. ATG-F 20 mg/kg/day on days -3, -2 and -1*. ATG-T 0.5 mg/kg on day -2 and 2 mg/kg on days -1 and +1. ATG-T 2.5 mg/kg/day on days -2 and -1. / ATG-T 2.5 mg/kg/day from days -5 to -2.

* some centers use smaller doses such as 15 mg/kg total dose.

chronic GvHD without increasing the risk of relapse or non-relapse mortality in patients given HLA-matched related90 or unrelated PBSC after myeloablative conditioning.60 This suggests that ATG might become a standard of care in that setting (Table 4) although ATG has also been associated with infusion reactions, delayed hematopoietic and immune recovery, and increased risks of cytomegalovirus and EBV infections. Potential limitations of these studies are that they allowed different conditioning regimens (based on total body irradiation or chemotherapy), that they included patients with various risks of disease relapse, and that they lacked statistical power to assess the impact of ATG on disease relapse in high-risk patients. Future phase III studies should deal with these limitations and also, ideally, compare ATG administration with new methods of GvHD prophylaxis such as post-transplant cyclophosphamide.91 Another important comparison that should be addressed in a phase III study is that between a combination of PBSC with ATG versus BM without ATG. In the RIC setting, a retrospective study from the CIBMTR observed that in vivo T-cell depletion with ATG increased the risk of relapse and decreased disease-free survival in a cohort of patients transplanted for various hematologic malignancies.73 However, several phase II studies68,69 as well as a retrospective study by the ALWP of the EBMT74 suggested that low doses of ATG efficiently prevented chronic GvHD without impairing leukemiafree or overall survival, while the use higher doses of ATG

References 1. Baron F, Storb R. Allogeneic hematopoietic cell transplantation as treatment for hematological malignancies: a review. Springer Semin Immunopathol. 2004;26(1-2):71-94. 2. Sureda A, Bader P, Cesaro S, et al. Indications for allo- and auto-SCT for haematological diseases, solid tumours and immune disorders: current practice in Europe, 2015. Bone Marrow Transplant. 2015;50(8):1037-1056. 3. Weiden PL, Sullivan KM, Flournoy N, Storb R, Thomas ED. Antileukemic effect of chronic graft-versus-host disease: contribution to improved survival after allogeneic marrow transplantation. N Engl J Med. 1981;304(25):1529-1533. 4. Kolb HJ, Schmidt C, Barrett AJ, Schendel DJ. Graft-versus-leukemia reactions in allogeneic chimeras. Blood. 2004;103(3):767-776. 5. Blazar BR, Murphy WJ, Abedi M. Advances

232

(>6 mg/kg ATG-T) was associated with a higher risk of relapse. Prospective randomized studies are thus needed to confirm the impact of in vivo T-cell depletion on outcomes specifically in the RIC setting.

Conclusions Three brands of ATG are currently commercialized. ATG-h (ATGAM) is currently available almost only in the USA. ATG-h induces less lymphopenia than the two brands of rabbit ATG.26 However, its impact on GvHD prevention has remained uncertain and is not supported by data from phase III studies. Although the two brands of rabbit ATG (ATG-T and ATG-F) share some similarities (such as inducing a profound and more durable lymphopenia than ATGh), they diverge by the nature and intensity of antigens recognized.29 Large studies comparing the impact of ATG-T versus ATG-F on immune recovery and post-transplant lymphoproliferative disorder are lacking. Consequently, data observed with one rabbit ATG product cannot be automatically extended to the other rabbit ATG formulation. Proposed indications/doses for immunoregulation with ATG in patients given PBSC from allogeneic donors are summarized in Table 4. Acknowledgments FB is senior research associate of the national fund for scientific research (F.R.S., FNRS), Belgium.

in graft-versus-host disease biology and therapy. Nat Rev Immunol. 2012;12(6):443458. 6. Duarte RF, Greinix H, Rabin B, et al. Uptake and use of recommendations for the diagnosis, severity scoring and management of chronic GVHD: an international survey of the EBMT-NCI Chronic GVHD Task Force. Bone Marrow Transplant. 2014;49(1):49-54. 7. Filipovich AH, Weisdorf D, Pavletic S, et al. National Institutes of Health consensus development project on criteria for clinical trials in chronic graft-versus-host disease: I. Diagnosis and Staging Working Group report. Biol Blood Marrow Transplant. 2005;11(12):945-956. 8. Thepot S, Zhou J, Perrot A, et al. The graftversus-leukemia effect is mainly restricted to NIH-defined chronic graft-versus-host disease after reduced intensity conditioning before allogeneic stem cell transplantation. Leukemia. 2010;24(11):1852-1858.

9. Baron F, Maris MB, Sandmaier BM, et al. Graft-versus-tumor effects after allogeneic hematopoietic cell transplantation with nonmyeloablative conditioning. J Clin Oncol. 2005;23(9):1993-2003. 10. Baron F, Labopin M, Niederwieser D, et al. Impact of graft-versus-host disease after reduced-intensity conditioning allogeneic stem cell transplantation for acute myeloid leukemia: a report from the Acute Leukemia Working Party of the European Group for Blood and Marrow Transplantation. Leukemia. 2012;26(12):2462-2468. 11. Storb R, Gyurkocza B, Storer BE, et al. Graftversus-host disease and graft-versus-tumor effects after allogeneic hematopoietic cell transplantation. J Clin Oncol. 2013;31(12): 1530-1538. 12. Pidala J, Kurland B, Chai X, et al. Patientreported quality of life is associated with severity of chronic graft-versus-host disease as measured by NIH criteria: report on base-

haematologica | 2017; 102(2)


In vivo T-cell depletion with ATG

13.

14. 15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27. 28.

line data from the Chronic GVHD Consortium. Blood. 2011;117(17):46514657. Wingard JR, Majhail NS, Brazauskas R, et al. Long-term survival and late deaths after allogeneic hematopoietic cell transplantation. J Clin Oncol. 2011;29(16):2230-2239. Socie G, Ritz J. Current issues in chronic graft-versus-host disease. Blood. 2014;124 (3):374-384. Ho VT, Soiffer RJ. The history and future of T-cell depletion as graft-versus-host disease prophylaxis for allogeneic hematopoietic stem cell transplantation. Blood. 2001;98 (12):3192-3204. Bertaina A, Merli P, Rutella S, et al. HLAhaploidentical stem cell transplantation after removal of alphabeta+ T and B cells in children with nonmalignant disorders. Blood. 2014;124(5):822-826. Passweg JR, Baldomero H, Bader P, et al. Hematopoietic SCT in Europe 2013: recent trends in the use of alternative donors showing more haploidentical donors but fewer cord blood transplants. Bone Marrow Transplant. 2015;50(4):476-482. Stem Cell Trialists' Collaborative Group. Allogeneic peripheral blood stem-cell compared with bone marrow transplantation in the management of hematologic malignancies: an individual patient data meta-analysis of nine randomized trials. J Clin Oncol. 2005;23(22):5074-5087. Anasetti C, Logan BR, Lee SJ, et al. Peripheral-blood stem cells versus bone marrow from unrelated donors. N Engl J Med. 2012;367(16):1487-1496. Savani BN, Labopin M, Blaise D, et al. Peripheral blood stem cell graft compared to bone marrow after reduced intensity conditioning regimens for acute leukemia: a report from the ALWP of the EBMT. Haematologica. 2016;101(2):256-262. Storek J, Mohty M, Boelens JJ. Rabbit anti-T cell globulin in allogeneic hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2015;21(6):959-970. Schneider U, Schwenk HU, Bornkamm G. Characterization of EBV-genome negative "null" and "T" cell lines derived from children with acute lymphoblastic leukemia and leukemic transformed non-Hodgkin lymphoma. Int J Cancer. 1977;19(5):621-626. Weiden PL, Doney K, Storb R, Thomas ED. Antihuman thymocyte globulin for prophylaxis of graft-versus-host disease. A randomized trial in patients with leukemia treated with HLA-identical sibling marrow grafts. Transplantation. 1979;27(4):227-230. Champlin RE, Perez WS, Passweg JR, et al. Bone marrow transplantation for severe aplastic anemia: a randomized controlled study of conditioning regimens. Blood. 2007;109(10):4582-4585. Atta EH, de Sousa AM, Schirmer MR, Bouzas LF, Nucci M, Abdelhay E. Different outcomes between cyclophosphamide plus horse or rabbit antithymocyte globulin for HLA-identical sibling bone marrow transplant in severe aplastic anemia. Biol Blood Marrow Transplant. 2012;18(12):1876-1882. Scheinberg P, Nunez O, Weinstein B, et al. Horse versus rabbit antithymocyte globulin in acquired aplastic anemia. N Engl Med. 2011;365 (5):430-438. Mohty M. Mechanisms of action of antithymocyte globulin: T-cell depletion and beyond. Leukemia. 2007;21(7):1387-1394. Popow I, Leitner J, Majdic O, et al. Assessment of batch to batch variation in polyclonal antithymocyte globulin prepara-

haematologica | 2017; 102(2)

tions. Transplantation. 2012;93(1):32-40. 29. Popow I, Leitner J, GrabmeierPfistershammer K, et al. A comprehensive and quantitative analysis of the major specificities in rabbit antithymocyte globulin preparations. Am J Transplant. 2013;13(12): 3103-3113. 30. Naujokat C, Berges C, Fuchs D, Sadeghi M, Opelz G, Daniel V. Antithymocyte globulins suppress dendritic cell function by multiple mechanisms. Transplantation. 2007;83(4): 485-497. 31. Feng X, Kajigaya S, Solomou EE, et al. Rabbit ATG but not horse ATG promotes expansion of functional CD4+CD25highFOXP3+ regulatory T cells in vitro. Blood. 2008;111(7):3675-3683. 32. Shimony O, Nagler A, Gellman YN, et al. Anti-T lymphocyte globulin (ATG) induces generation of regulatory T cells, at least part of them express activated CD44. J Clin Immunol. 2012;32(1):173-188. 33. Ehx G, Hannon M, Beguin Y, HumbletBaron S, Baron F. Validation of a multicolor staining to monitor phosphoSTAT5 levels in regulatory T-cell subsets. Oncotarget. 2015;6(41):43255-43266. 34. Cohen JL, Trenado A, Vasey D, Klatzmann D, Salomon BL. CD4(+)CD25(+) immunoregulatory T cells: new therapeutics for graft-versus-host disease. J Exp Med. 2002;196(3):401-406. 35. Hannon M, Lechanteur C, Lucas S, et al. Infusion of clinical-grade enriched regulatory T cells delays experimental xenogeneic graft-versus-host disease. Transfusion. 2014;54(2):353-363. 36. Martelli MF, Di Ianni M, Ruggeri L, et al. HLA-haploidentical transplantation with regulatory and conventional T cell adoptive immunotherapy prevents acute leukemia relapse. Blood. 2014; 124(4):638-644. 37. Edinger M, Hoffmann P, Ermann J, et al. CD4+CD25+ regulatory T cells preserve graft-versus-tumor activity while inhibiting graft-versus-host disease after bone marrow transplantation. Nat Med. 2003;9(9):11441150. 38. Forcina A, Seitz I, Martinius H, et al. Pharmacokinetics (PK) study of antithymocyte globulins fresenius (ATG-F) prior to allogeneic stem cell transplantation: implications for timing of graft and early adoptive immunotherapy infusions. Blood. 2013;122 (21):4536. 39. Waller EK, Langston AA, Lonial S, et al. Pharmacokinetics and pharmacodynamics of anti-thymocyte globulin in recipients of partially HLA-matched blood hematopoietic progenitor cell transplantation. Biol Blood Marrow Transplant. 2003;9(7):460-471. 40. Remberger M, Sundberg B. Low serum levels of total rabbit-IgG is associated with acute graft-versus-host disease after unrelated donor hematopoietic stem cell transplantation: results from a prospective study. Biol Blood Marrow Transplant. 2009;15(8):996999. 41. Podgorny PJ, Ugarte-Torres A, Liu Y, Williamson TS, Russell JA, Storek J. High rabbit-antihuman thymocyte globulin levels are associated with low likelihood of graftvs-host disease and high likelihood of posttransplant lymphoproliferative disorder. Biol Blood Marrow Transplant. 2010;16(7):915926. 42. Chawla S, Dharmani-Khan P, Liu Y, et al. High serum level of antithymocyte globulin immediately before graft infusion is associated with a low likelihood of chronic, but not acute, graft-versus-host disease. Biol

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

Blood Marrow Transplant. 2014;20(8):11561162. Hoegh-Petersen M, Amin MA, Liu Y, et al. Anti-thymocyte globulins capable of binding to T and B cells reduce graft-vs-host disease without increasing relapse. Bone Marrow Transplant. 2013;48(1):105-114. Bosch M, Dhadda M, Hoegh-Petersen M, et al. Immune reconstitution after anti-thymocyte globulin-conditioned hematopoietic cell transplantation. Cytotherapy. 2012;14 (10):1258-1275. Servais S, Menten-Dedoyart C, Beguin Y, et al. Impact of pre-transplant anti-T cell globulin (ATG) on immune recovery after myeloablative allogeneic peripheral blood stem cell transplantation. PloS One. 2015;10(6): e0130026. Zander AR, Kroger N, Schleuning M, et al. ATG as part of the conditioning regimen reduces transplant-related mortality (TRM) and improves overall survival after unrelated stem cell transplantation in patients with chronic myelogenous leukemia (CML). Bone Marrow Transplant. 2003;32(4):355361. Schattenberg A, van der Meer A, Preijers F, et al. Addition of ATG to the conditioning regimen is a major determinant for outcome after transplantation with partially lymphocyte-depleted grafts from voluntary unrelated donors. Bone Marrow Transplant. 2004;33(11):1115-1121. Yu ZP, Ding JH, Wu F, et al. Quality of life of patients after allogeneic hematopoietic stem cell transplantation with antihuman thymocyte globulin. Biol Blood Marrow Transplant. 2012;18(4):593-599. Wolschke C, Zabelina T, Ayuk F, et al. Effective prevention of GVHD using in vivo T-cell depletion with anti-lymphocyte globulin in HLA-identical or -mismatched sibling peripheral blood stem cell transplantation. Bone Marrow Transplant. 2014;49(1):126130. Ratanatharathorn V, Deol A, Ayash L, et al. Low-dose antithymocyte globulin enhanced the efficacy of tacrolimus and mycophenolate for GVHD prophylaxis in recipients of unrelated SCT. Bone Marrow Transplant. 2015;50(1):106-112. Binkert L, Medinger M, Halter JP, et al. Lower dose anti-thymocyte globulin for GvHD prophylaxis results in improved survival after allogeneic stem cell transplantation. Bone Marrow Transplant. 2015;50 (10):1331-1336. Bonifazi F, Bandini G, Arpinati M, et al. Intensification of GVHD prophylaxis with low-dose ATG-F before allogeneic PBSC transplantation from HLA-identical siblings in adult patients with hematological malignancies: results from a retrospective analysis. Bone Marrow Transplant. 2012;47(8): 1105-1111. Ravinet A, Cabrespine A, Socie G, et al. Impact of thymoglobulin by stem cell source (peripheral blood stem cell or bone marrow) after myeloablative stem cell transplantation from HLA 10/10-matched unrelated donors. A report from the Societe Francaise de Greffe de Moelle et de Therapie Cellulaire. Transplantation. 2015; 100(8):1732-1739. Deeg HJ, Storer BE, Boeckh M, et al. Reduced incidence of acute and chronic graft-versushost disease with the addition of thymoglobulin to a targeted busulfan/cyclophosphamide regimen. Biol Blood Marrow Transplant. 2006;12(5):573-584. Bacigalupo A, Lamparelli T, Bruzzi P, et al. Antithymocyte globulin for graft-versus-host

233


F. Baron et al.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

234

disease prophylaxis in transplants from unrelated donors: 2 randomized studies from Gruppo Italiano Trapianti Midollo Osseo (GITMO). Blood. 2001;98(10):2942-2947. Finke J, Bethge WA, Schmoor C, et al. Standard graft-versus-host disease prophylaxis with or without anti-T-cell globulin in haematopoietic cell transplantation from matched unrelated donors: a randomised, open-label, multicentre phase 3 trial. Lancet Oncol. 2009;10(9):855-864. Walker I, Panzarella T, Couban S, et al. Pretreatment with anti-thymocyte globulin versus no anti-thymocyte globulin in patients with haematological malignancies undergoing haemopoietic cell transplantation from unrelated donors: a randomised, controlled, open-label, phase 3, multicentre trial. Lancet Oncol. 2016; 17(2):164-173. Kroger N, Solano C, Wolschke C, et al. Antilymphocyte globulin for prevention of chronic graft-versus-host disease. New Engl J Med. 2016;374(1):43-53. Bacigalupo A, Lamparelli T, Barisione G, et al. Thymoglobulin prevents chronic graftversus-host disease, chronic lung dysfunction, and late transplant-related mortality: long-term follow-up of a randomized trial in patients undergoing unrelated donor transplantation. Biol Blood Marrow Transplant. 2006;12(5):560-565. Socie G, Schmoor C, Bethge WA, et al. Chronic graft-versus-host disease: long-term results from a randomized trial on graft-versus-host disease prophylaxis with or without anti-T-cell globulin ATG-Fresenius. Blood. 2011;117(23):6375-6382. Slavin S, Nagler A, Naparstek E, et al. Nonmyeloablative stem cell transplantation and cell therapy as an alternative to conventional bone marrow transplantation with lethal cytoreduction for the treatment of malignant and nonmalignant hematologic diseases. Blood. 1998;91(3):756-763. Giralt S, Estey E, Albitar M, et al. Engraftment of allogeneic hematopoietic progenitor cells with purine analog-containing chemotherapy: harnessing graft-versusleukemia without myeloablative therapy. Blood. 1997;89(12):4531-4536. Gill S, Porter DL. Reduced-intensity hematopoietic stem cell transplants for malignancies: harnessing the graft-versus-tumor effect. Annu Rev Med. 2013;64(101-117. Warren EH, Deeg HJ. Dissecting graft-versus-leukemia from graft-versus-host-disease using novel strategies. Tissue Antigens. 2013;81(4):183-193. Valcarcel D, Martino R, Caballero D, et al. Sustained remissions of high-risk acute myeloid leukemia and myelodysplastic syndrome after reduced-intensity conditioning allogeneic hematopoietic transplantation: chronic graft-versus-host disease is the strongest factor improving survival. J Clin Oncol. 2008;26(4):577-584. McSweeney PA, Niederwieser D, Shizuru JA, et al. Hematopoietic cell transplantation in older patients with hematologic malignancies: replacing high-dose cytotoxic therapy with graft-versus-tumor effects. Blood. 2001;97(11):3390-3400. Mohty M, Faucher C, Blaise D. Graft-versus-host-disease and granulocyte colonystimulating factor administration after allo-

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

geneic stem cell transplantation. Leukemia. 2005;19(4):500-503. Crocchiolo R, Esterni B, Castagna L, et al. Two days of antithymocyte globulin are associated with a reduced incidence of acute and chronic graft-versus-host disease in reduced-intensity conditioning transplantation for hematologic diseases. Cancer. 2013;119(5):986-992. Devillier R, Crocchiolo R, Castagna L, et al. The increase from 2.5 to 5 mg/kg of rabbit anti-thymocyte-globulin dose in reduced intensity conditioning reduces acute and chronic GVHD for patients with myeloid malignancies undergoing allo-SCT. Bone Marrow Transplant. 2012;47(5):639-645. Crocchiolo R, Esterni B, Castagna L, et al. Two days of antithymocyte globulin are associated with a reduced incidence of acute and chronic graft-versus-host disease in reduced-intensity conditioning transplantation for hematologic diseases. Cancer. 2013;119(5):986-992. Remberger M, Ringden O, Hagglund H, et al. A high antithymocyte globulin dose increases the risk of relapse after reduced intensity conditioning HSCT with unrelated donors. Clin Transplant. 2013;27(4):E368-374. Langston AA, Prichard JM, Muppidi S, et al. Favorable impact of pre-transplant ATG on outcomes of reduced-intensity hematopoietic cell transplants from partially mismatched unrelated donors. Bone Marrow Transplant. 2014;49(2):185-189. Soiffer RJ, Lerademacher J, Ho V, et al. Impact of immune modulation with anti-Tcell antibodies on the outcome of reducedintensity allogeneic hematopoietic stem cell transplantation for hematologic malignancies. Blood. 2011;117(25):6963-6970. Baron F, Labopin M, Blaise D, et al. Impact of in vivo T-cell depletion on outcome of AML patients in first CR given peripheral blood stem cells and reduced-intensity conditioning allo-SCT from a HLA-identical sibling donor: a report from the Acute Leukemia Working Party of the European Group for Blood and Marrow Transplantation. Bone Marrow Transplant. 2014;49(3):389-396. Rubio MT, Labopin M, Blaise D, et al. The impact of graft-versus-host disease prophylaxis in reduced-intensity conditioning allogeneic stem cell transplant in acute myeloid leukemia: a study from the Acute Leukemia Working Party of the European Group for Blood and Marrow Transplantation. Haematologica. 2015;100(5):683-689. Lowsky R, Takahashi T, Liu YP, et al. Protective conditioning for acute graft-versus-host disease. N Engl J Med. 2005;353(13):1321-1331. Kohrt HE, Turnbull BB, Heydari K, et al. TLI and ATG conditioning with low risk of graft-versus-host disease retains antitumor reactions after allogeneic hematopoietic cell transplantation from related and unrelated donors. Blood. 2009;114(5):1099-1109. Pillai AB, George TI, Dutt S, Strober S. Host natural killer T cells induce an interleukin-4dependent expansion of donor CD4+CD25+Foxp3+ T regulatory cells that protects against graft-versus-host disease. Blood. 2009;113(18):4458-4467. Baron F, Zachee P, Maertens J, et al. Nonmyeloablative allogeneic hematopoietic cell

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

transplantation following fludarabine plus 2 Gy TBI or ATG plus 8 Gy TLI: a phase II randomized study from the Belgian Hematological Society. J Hematol Oncol. 2015;8(1):4. Hannon M, Beguin Y, Ehx G, et al. Immune recovery after allogeneic hematopoietic stem cell transplantation following Flu-TBI versus TLI-ATG conditioning. Clin Cancer Res. 2015;21(14):3131-3139. Aversa F, Tabilio A, Velardi A, et al. Treatment of high-risk acute leukemia with T-cell-depleted stem cells from related donors with one fully mismatched HLA haplotype. N Engl J Med. 1998;339(17): 1186-1193. Kanakry CG, Fuchs EJ, Luznik L. Modern approaches to HLA-haploidentical blood or marrow transplantation. Nat Rev Clin Oncol. 2016;13(1):10-24. Apperley J, Niederwieser D, Huang XJ, et al. Haploidentical hematopoietic stem cell transplantation: a global overview comparing Asia, the European Union, and the United States. Biol Blood Marrow Transplant. 2016;22(1):23-26. Ciceri F, Labopin M, Aversa F, et al. A survey of fully haploidentical hematopoietic stem cell transplantation in adults with high-risk acute leukemia: a risk factor analysis of outcomes for patients in remission at transplantation. Blood. 2008;112(9):3574-3581. Rubio MT, Savani BN, Labopin M, et al. Impact of conditioning intensity in T-replete haplo-identical stem cell transplantation for acute leukemia: a report from the acute leukemia working party of the EBMT. J Hematol Oncol. 2016;9(1):25. O'Donnell PV, Luznik L, Jones RJ, et al. Nonmyeloablative bone marrow transplantation from partially HLA-mismatched related donors using posttransplantation cyclophosphamide. Biol Blood Marrow Transplant. 2002;8(7):377-386. Wang Y, Chang YJ, Xu LP, et al. Who is the best donor for a related HLA haplotype-mismatched transplant? Blood. 2014;124(6): 843-850. Wang Y, Fu HX, Liu DH, et al. Influence of two different doses of antithymocyte globulin in patients with standard-risk disease following haploidentical transplantation: a randomized trial. Bone Marrow Transplant. 2014;49(3):426-433. Peccatori J, Forcina A, Clerici D, et al. Sirolimus-based graft-versus-host disease prophylaxis promotes the in vivo expansion of regulatory T cells and permits peripheral blood stem cell transplantation from haploidentical donors. Leukemia. 2015;29(2): 396-405. Bonifazi F, Solano C, Wolschke C, et al. Prevention of chronic GvHD after HLAidentical sibling peripheral hematopietic stem cell transplantation with or without anti-lymphocyte globulin (ATG). Results from a prospective, multicenter randomized phase III trial (ATGfamilystudy). Blood. 2014;124(21):37. Kanakry CG, Tsai HL, Bolanos-Meade J, et al. Single-agent GVHD prophylaxis with posttransplantation cyclophosphamide after myeloablative, HLA-matched BMT for AML, ALL, and MDS. Blood. 2014;124(25): 3817-3827.

haematologica | 2017; 102(2)


GUIDELINE ARTICLE

Guidelines for diagnosis, prevention and management of central nervous system involvement in diffuse large B-cell lymphoma patients by the Spanish Lymphoma Group (GELTAMO)

Francisco-Javier Peñalver,1 Juan-Manuel Sancho,2 Adolfo de la Fuente,3 María-Teresa Olave,4 Alejandro Martín,5 Carlos Panizo,6 Elena Pérez,7 Antonio Salar8 and Alberto Orfao5 on behalf of the Spanish Lymphoma Group (GELTAMO)

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2017 Volume 102(2):235-245

1 Department of Hematology, Hospital Universitario Fundación Alcorcón, Madrid; 2Clinical Hematology Department, ICO-IJC Hospital Germans Trias i Pujol, Badalona, Barcelona; 3 Department of Hematology, MD Anderson Cancer Center, Madrid; 4Department of Hematology, Hospital Clínico Universitario Lozano Blesa, Zaragoza; 5Department of Hematology, Hospital Universitario de Salamanca, Department of Medicine, Cytometry Service (NUCLEUS) and Cancer Research Center (IBMCC-USAL-CSIC) and IBSAL, University of Salamanca; 6Department of Hematology, Clínica Universidad de Navarra, Pamplona; 7Department of Hematology, Hospital General Universitario Morales Meseguer, Murcia and 8Department of Hematology, Hospital del Mar, Barcelona, Spain

ABSTRACT

D

iffuse large B-cell lymphoma patients have a 5% overall risk of central nervous system events (relapse or progression), which account for high morbidity and frequently fatal outcomes,1 and shortened overall survival of <6 months.2 Early diagnosis of central nervous system events is critical for successful treatment and improved prognosis. Identification of patients at risk of central nervous system disease is critical to accurately identify candidates for central nervous system prophylaxis vs. therapy.3–5 This report by the Spanish Lymphoma Group (GELTAMO) aims to provide useful guidelines and recommendations for the prevention, diagnosis, and treatment of central nervous system diffuse large B-cell lymphoma patients with, or at risk of, leptomeningeal and/or brain parenchyma lymphoma relapse. A panel of lymphoma experts working on behalf of GELTAMO reviewed all data published on these topics available in PubMed up to May 2016. Recommendations were classified according to the Grading of Recommendations Assessment Development and Evaluation (GRADE) approach.6 A practical algorithm based on the proposed recommendations was then developed (Figure 1). Initial discussions among experts were held in May 2014, and final consensus was reached in June 2016. The final manuscript was reviewed by all authors and the Scientific Committee of GELTAMO.

Correspondence: fjpenalver@fhalcorcon.es

Received: May 11, 2016. Accepted: October 7, 2016. Pre-published: October 20, 2016. doi:10.3324/haematol.2016.149120

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/235

Risk factors for central nervous system involvement in diffuse large B-cell lymphoma Several factors hinder the identification of risk factors for central nervous system (CNS) involvement in diffuse large B-cell lymphoma (DLBCL), including the retrospective nature of most studies, the relatively low frequency of CNS relapse in DLBCL, and the heterogeneity of CNS prophylaxis methods used in these studies. Moreover, the impact of newly developed diagnostic tools (such as multiparameter flow cytometry [FCM]) and new treatments introduced in the last decade, in particular rituximab, has still not been fully clarified. Several studies4,5,7–10 and a recent meta-analysis1 have described a decrease in rates haematologica | 2017; 102(2)

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

235


F.J. Peñalver et al.

of CNS relapse in the post-rituximab era (probably due to improved control of systemic lymphoma), in addition to a change in the pattern of CNS relapse, with predominance of parenchymal over leptomeningeal relapse, isolated over combined (systemic plus CNS) relapses, and delayed CNS relapses. Similarly, recently published British guidelines11 have concluded that the incidence of CNS relapse decreased after the introduction of rituximab (Table 1). The identification of risk factors has been the major goal of many studies of CNS involvement. Several large retrospective studies conducted in the pre-rituximab era12–15 reported higher rates of CNS relapse in patients with increased serum lactate dehydrogenase (LDH) levels and/or involvement of >1 extranodal site, although these factors failed to predict CNS relapse in more than half of all cases.12 In addition to the involvement of >1 extranodal site and increased LDH, International Prognostic Index (IPI) score was also identified as an independent predictor for CNS relapse in other studies.13,16 A post-rituximab era study of 399 DLBCL patients, randomized to R-CHOP or CHOP chemotherapy,3 identified an age-adjusted IPI (aaIPI) >1 as the only risk factor for CNS involvement, although a high aaIPI score was recorded for more than 60% of the patients. When aaIPI was excluded from the

analysis, elevated LDH and a poor performance status (PS >1) were identified as independent predictive factors for CNS relapse. Similarly, in the randomized RICOVER-60 trial,4 the combination of increased LDH levels, the involvement of >1 extranodal site, and PS >1 (recorded for 4.8% of patients) was associated with a probability of CNS relapse of 33.5% as compared with 2.8% in the remaining patients. Elevated LDH levels, the involvement of >1 extranodal site, and an intermediate-high or high IPI score have also been cited as risk factors in other retrospective studies, reviews, and meta-analyses of the postrituximab era (Table 1).1,2,7,9,17,18 Accumulated evidence from studies of extranodal involvement have shown that testicular or breast involvement (particularly as primary lymphoma, but also as secondary involvement) is clearly associated with a higher rate of CNS relapse.5,19–21 A growing body of evidence indicates a higher CNS relapse rate among patients with renal involvement by lymphoma. Villa and colleagues22 reported CNS involvement in 36% of patients with DLBCL with renal involvement. Similarly, Tai and colleagues23 found that renal involvement was the primary risk factor for CNS relapse, ahead of even breast or testis involvement. The association of other extranodal sites with CNS

Table 1. Influence of systemic rituximab treatment on the incidence of CNS relapse in DLBCL and risk factors for CNS disease.

Study (year)

N

Incidence of CNS relapse

Risk factors for CNS relapse

Use and type of CNS prophylaxis

Feugier et al. (2004)3

399 DLBCL 1217 B-cell lymphomas (944 DLBCL)

Shimazu et al. (2009)7

403 DLBCL

13.3% CHOP 8.4% R-CHOP

Villa et al. (2010)5

435 DLBCL

9.7% CHOP 6.4% R-CHOP (P=0.085)

aaIPI>1 LDH and PS after exclusion aaIPI Overall series: >1 extranodal site B symptoms LDH (not significant) Patients treated with R-CHOP: >1 extranodal site LDH ECOG PS >1 >60 years LDH >1 extranodal site BM No rituximab Testis Kidney Stage IV No rituximab

NA

Boehme et al. (2009)4

4.6% CHOP 5.4% R-CHOP 6.9% CHOP 4.1% R-CHOP

Yamamoto et al. (2010)18 375 DLBCL

2.9% CHOP 3.9% R-CHOP (P=0.71)

Chihara et al. (2011)64

386 DLBCL

Tai et al. (2011)23

499 DLBCL

7.3% CHOP 5.3% R-CHOP (P=0.42) 5.1% CHOP 6% R-CHOP

IT MTX (days 1, 5) in first 2 cycles

Criteria for CNS prophylaxis -

BM Testis Upper neck or head

IT MTX (18 patients)

IT MTX or cytarabine × 6 doses (alternating)

Nasal sinuses Testis Vertebra

Before 2002: BM, peripheral blood, epidural disease, testicular or nasal sinus After 2002: nasal sinus -

Multivariate analysis: NA no risk factors Univariate analysis: LDH, high IPI, BM, systemic relapse Bulky disease IT MTX or cytarabine Testis (after 1999) Lymphocyte count <1000/mm3 × 4 doses Extranodal involvement ECOG PS >1 IT prophylaxis (82 patients): >1 extranodal No CR physician discretion Orbital sinus, Testicular and patient preference posterior nasal space Kidney Breast Breast Testicular, BM

continued in the next page

236

haematologica | 2017; 102(2)


Guidelines for management of CNS involvement in DLBCL

relapse is less clear. Epidural space involvement has been proposed as a risk factor in very old studies,24 but CNS prophylaxis is recommended for these patients in recently published British guidelines,11 potentially because of the anatomical proximity. Regarding extranodal craniofacial involvement, a recent review of 4,155 patients from 11 consecutive trials by the German High-Grade NonHodgkin Lymphoma Study Group25 reported no differences in the 2-year cumulative rate of CNS disease between rituximab-treated patients with and without craniofacial involvement (1.6% vs. 3.4%, P=0.682), in line with the findings of another more recent study.26 Based on all the above evidence, a new prognostic model to assess the risk of CNS disease in DLBCL (CNSIPI) has been proposed.27 This model has been validated in other series from the British Columbia Cancer Agency,28 and includes the 5 IPI factors in addition to kidney/adrenal gland involvement, and it stratifies patients into 3 risk groups for CNS relapse: low risk (0-1 factors; 2-year risk of 0.6%), intermediate risk (2-3 factors; 2-year risk of 3.4%), and high risk (4-6 factors; 2-year risk of 10.2%). The influence of the biology of DLBCL on CNS relapse remains a matter of debate. There is still insufficient evidence to demonstrate an influence of B-cell origin (germinal center vs. non-germinal center DLBCL) on CNS dis-

ease. However, many retrospective and recent studies have described a high percentage of CNS involvement in DLBCL cases with MYC rearrangement, particularly when associated with either additional BCL-2 or BCL-6 gene rearrangements: in these patients, the frequency of CNS disease ranges from 9% to 45%. Based on these results Fletcher and Kahl2 recommended that patients with DLBCL and MYC rearrangements be considered at high risk of CNS relapse. In another recent study, Savage et al.29 reported that DLBCL patients and dual expression of MYC (≥40% positivity) and BCL2 (≥50% positivity) determined by immunohistochemistry, had higher risk of CNS relapse (2-year risk of 9.7% vs. 2.2%, P=0.001). This study also showed increased risk for those patients with activated B-cell or non-germinal center B-cell origin, but significance was not retained in the multivariate analysis.

Summary and recommendations for CNS prophylaxis in DLBCL based on the presence of risk factors The authors recommend screening patients for CNS involvement by lumbar puncture and cerebrospinal fluid (CSF) analysis by conventional cytology (CC) and FCM in order to provide prophylaxis in the following situations: • Increased serum LDH and involvement of >1 extranodal site (recommendation 1, level of evidence B)

continued in the previous page

Mitrovic et al. (2012)8

Cao et al. (2012)a Schmitz et al. (2012)9

1197 DLBCL

3.7% CHOP-like 2.1% R-CHOP-like (P=0.049) 315 DLBCL 3.03% CHOP 3.33% R-CHOP 2210 1–13.2% Chemo aggressive B-cell 0–9.7% R-Chemo lymphoma (1809 DLBCL)

Guirguis et al. (2012)b Tomita et al. (2012)20

217 DLBCL 1221 DLBCL

3,7% R-CHOP 6.7% R-CHOP

Kumar et al. (2012)17

989 DLBCL

2% R-CHOP

Deng et al. (2013)c

599 DLBCL

Zhang et al. (2014)1*

4911 DLBCL

6.5% CHOP 4.3% R-CHOP 5.7% Chemo 4.7% R-chemo

-

-

-

-

NA

-

Overall series: >1 extranodal involvement LDH Patients treated with R-chemo: Advanced stage (III-IV) LDH Testicular involvement <60 years Adrenal gland Bone Breast Univariate analysis: IPI (intermediate-high and high)

Stage III/IV IPI>1 PS>1 LDH >1 extranodal involvement BM Testicular involvement

IT MTX (days 1, 15) in first 2 cycles IT MTX and/or HD-MTX NA

High-CHOEP and Mega-CHOEP phase III studies: Upper neck Head, BM Testes High risk patients -

IT prophylaxis (71.8%); BM systemic prophylaxis (28.2%) Other high-risk site >1 extranodal site Higher IPI Higher stage -

-

*Meta-analysis of the first 8 studies of the table. aCao B, et al. Oncol Lett. 2012;4(3):541–5. bGuirguis HR, et al. Br J Haematol. 2012;159(1):39-49. cDeng L, et al. Int J Hematol. 2013;98(6):664–71. CNS: central nervous system; DLBCL: diffuse large B-cell lymphoma; CHOP: cyclophosphamide, doxorubicin, vincristine, and prednisone; R-CHOP: rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone; aaIPI: age-adjusted International Prognostic Index; LDH: lactate dehydrogenase; PS: performance status; BM: bone marrow; IT: intrathecal; CR: complete remission; MTX: methotrexate; HD-MTX: high-dose methotrexate; ECOG PS: Eastern Cooperative Oncology Group Scale Performance Status.

haematologica | 2017; 102(2)

237


F.J. Peñalver et al.

• Extranodal involvement of testis (recommendation 1, level of evidence B) or breast (recommendation 2, level of evidence B) • Extranodal involvement of kidney, adrenal gland (recommendation 2, level of evidence C) or epidural space (recommendation 2, level of evidence D). • High risk CNS-IPI (recommendation 2, level of evidence B) • MYC rearrangements associated to BCL2 or BCL6 rearrangements (recommendation 2, level of evidence C).

Diagnostic screening for CNS disease in DLBCL Definitive diagnosis of central nervous system lymphoma (CNSL) relies on a positive CSF CC.30 However, CSF samples are only obtained in a selected subgroup of DLBCL patients14 due to the low frequency of CNSL, as discussed above in detail.14 Diagnosis based on histopathology of stereotactic biopsy specimens, including ocular biopsy in cases with positive ophthalmological evaluation, is usually limited to a small number of CSFnegative patients, mostly in cases of suspected primary CNS lymphoma (PCNSL). Clinical presentation: Clinical symptoms associated with CNSL are the first indication of CNS disease in many patients. However, DLBCL patients who have CNSL frequently display subtle symptoms, which are either unrecognized or difficult to distinguish from those related to the primary disease or the treatment thereof. Thus, whenever present, neurological symptoms should prompt further CNS imaging and/or CSF analysis, depending on the clinical context of the patient and the results of complementary diagnostic procedures/tests. Imaging techniques: Of the imaging techniques currently available, the most informative is magnetic resonance imaging (MRI), including contrast-enhanced MRI, with a sensitivity of 71% vs. 36% for computerized tomography (CT).31 Thus, evaluation of CNSL in symptomatic patients typically includes cranial MRI,32 except in the few cases in which the procedure is contraindicated and CT is recommended. Most CNSL lesions analyzed by MRI and/or CT are associated with either diffuse or, more frequently, local (contrast) enhancement, which often includes the leptomeninges, cranial nerves, or the periventricular region.33 However, these patterns have relatively low specificity and cannot be usually considered truly diagnostic, even in previously diagnosed DLBCL cases,34 particularly after corticosteroid therapy.35,36 Diagnosis of CNSL based exclusively on imaging techniques (e.g., MRI) thus continues to pose a clinical challenge, underscoring the need for more definitive diagnostic approaches to demonstrate the tumoral nature of the lesions. More recently developed imaging techniques including positron-emission tomography have been proposed to potentially contribute to diagnosis in specific cases. However, due to their limited specificity, additional studies are still necessary to define their precise value in the diagnosis of CNSL.37,38 Histopathology: Histopathological and immunohistochemical analysis of stereotactic biopsy samples is considered a standard procedure for the diagnosis of PCNSL,32,39 but is not a routine procedure in patients who already have been diagnosed with DLBCL. Stereotactic biopsy is an invasive procedure, which is of relatively limited sensi238

tivity (20%–65% in immunocompetent patients), particularly in patients treated with corticosteroids. Moreover, this approach cannot be used in a subset of patients due to the location of the lesions. CSF cytology: While CSF cytology is a highly specific diagnostic approach for CNSL in DLBCL, it is of limited sensitivity, and produces a significant percentage (20%–60%) of false-negative results,40-42 particularly when used to analyze small volumes from single samples, processed with delay, from patients treated with corticosteroids.40,43 Furthermore, the morphological features of inflammatory lymphocytes in CSF can overlap with those of lymphoid tumor cells, leading to false-positive results in some cases.44

Multiparameter flow cytometry analysis of CSF samples Many studies have demonstrated the utility of FCM for detecting CNS disease in DLBCL.10,42,45,46 Early studies analyzing non-stabilized CSF samples by ≤4-color FCM already demonstrated increased sensitivity in between 3% and 20% of cases as compared with cytology (Table 2).42,47– 50 More recent analyses of larger series of CSF-stabilized samples using 4–8 color FCM have confirmed the greater sensitivity of FCM vs. CC with a median proportion of occult CNSL (FCM+/CC- CSF) of 12% (range: 5%–13%) (Table 1).10,45,51 Patients with occult CNSL (i.e., CC- and FCM+ CSF) typically showed lower levels of CSF infiltration (<20% or <1 tumor B cell/μL) than FCM+/CC+ cases,45 further supporting the greater sensitivity of FCM vs. CC. These studies10,45,51 also showed very few false-negative FCM results (range: 0%–<1%), further supporting its greater diagnostic efficiency with respect to CC. Early studies showed that patients with occult CNSL more frequently present neurological symptoms than FCM-/CC- cases (57% and 10%, respectively),45 suggesting a clinical impact of occult CNS disease. More recently, Wilson and colleagues10 confirmed that among DLBCL cases with negative CSF cytology, the presence of occult CNSL as detected by FCM is associated with a significant reduction in CNS-recurrence-free survival (73% vs. 94%) and overall survival (OS at 3 years: 38% vs. 69%) compared with patients without CSF involvement. These results are in agreement with those of an analysis of 174 lymphoma patients, including 125 DLBCL cases.51 However, it should be noted that the prognostic impact of occult CNSL reported by Wilson and colleagues failed to reach significance among DLBCL cases treated with immunochemotherapy regimens.10 Taken together, these results provide sufficient evidence to support the mandatory use of FCM in the diagnostic work-up of CNS involvement in DLBCL. However, particular attention should be paid to the specific FCM approach used. Immediate sample preservation (preferably in TransFix®)52 and the use of standardized sample preparation procedures and validated ≥8-color antibody combinations for simultaneous identification of all cell subsets present in normal/reactive CSF samples, as well as tumor B-cells, is strongly recommended.53 Evaluation for blood contamination should also be considered in cases with peripheral blood involvement by systemic lymphoma, in which CSF infiltration by blood cells (e.g., red cells and neutrophils) is observed.10

Other biochemical and CSF biomarkers Increases in overall protein and LDH levels, the presence of pleocytosis, and decreased glucose levels in CSF have haematologica | 2017; 102(2)


Guidelines for management of CNS involvement in DLBCL

long been associated with CNSL.14,54 However, these parameters are nonspecific and therefore unreliable for routine diagnosis of CNS disease.55,56 Similarly, CSF levels of soluble (s)CD21, sCD22, sCD24, sCD38, sCD44, sCD72, and immunoglobulin (IG) heavy and light chain isotypes are of limited diagnostic utility.57 Similar rates of CSF-positive cases (8%–13% vs. 11%–16%) have been obtained by polymerase chain reaction (PCR) analysis of IG gene sequences and cytomorphology, respectively, with a high frequency of unexplained discrepant cases,58 suggesting that the utility of PCR analysis of IG genes may be limited to selected cases in which CSF cytology and FCM are not informative.56 Furthermore, increased CSF levels of sCD19, sAnti-thrombin III (sATIII), sCD27, b2 microglobulin, IL-6, IL-10, CXCL13, neopterin, osteopontin, and several microRNAs (miRNA19b, miRNA21, and miRNA92a) have emerged as potentially useful biomarkers for CNS lymphoma, particularly in cases of PCNSL.57,59– 63 However, the potential value of these markers has only been investigated in a few studies, which used varying endpoints (usually one per study), and included few DLBCL cases with secondary CNSL.

tion of stabilized CSF in the diagnostic work-up of DLBCL patients at risk of CNS disease for the identification of occult CNSL (CC-/FCM+) (recommendation 1, level of evidence A). • The presence of occult CNSL in high-risk DLBCL may be considered an adverse prognostic factor, although its independent prognostic value has not been definitively established (recommendation 2, level of evidence B). • Despite their potential value, several other CSF biomarkers (e.g., sCD19, sIL-10 and/or sCXCL13, neopterin, and several miRNAs) cannot be currently used for the diagnosis of CNSL in DLBCL (recommendation 2, level of evidence C). • In case of suspected CNSL in DLBCL patients with negative CSF, stereotactic brain biopsy is still not regarded as a useful routine diagnostic test. However, ophthalmological evaluation with ocular and/or brain biopsy may be required in specific cases (recommendation 2, level of evidence C).

Summary and recommendations for diagnosis of CNS disease in DLBCL

CNS relapse in DLBCL mainly occurs within less than one year after diagnosis (median: 6 months).12,64,65 This pattern of early relapse suggests that affected patients probably harbor occult malignant cells in the CNS at diagnosis.16,42,65 Although FCM improves the identification of CNS involvement by 4- to 10-fold as compared with cytology, it identifies only a fraction of patients that are destined to experience CNS relapse.65 These findings sup-

• Include CNS imaging in the diagnostic work-up of DLBCL patients who present with symptoms of suspected CNSL; in such cases, MRI (including contrast enhanced MRI) is preferable (recommendation 1, level of evidence A). • Use standardized and validated >8-color FCM evalua-

Efficacy of chemoprophylaxis in preventing CNS relapse in DLBCL

Table 2. Frequency of cases including diffuse large B-cell lymphoma (DLBCL) patients showing cerebrospinal fluid (CSF) involvement by cytology vs. flow cytometry (FCM).

Study

Finn et al. (1998)47 French et al. (2001)48 Roma et al. (2002)49 Subira et al. (2005)50 Hedge et al. (2005)42 Bromberg et al. (2007)a Di Noto et al. (2008)b Quijano et al. (2009)45 Sancho et al. (2010)46 Cesana et al. (2010)c Schroers et al. (2010)97 Alvarez et al. (2011)d Bommer et al. (2011)e Craig et al. (2011)f Stacchini et al. (2012)g Benevolo et al. (2012)51 Muñiz et al. (2014)57 Wilson et al. (2014)10

No. of samples (cases)

% of CSF cytology+ cases

% of FCM+ cases

No. DLBCL patients

% of CSF cytology+ DLBCL

% of FCM+ DLBCL

% FCM+ / cytology- DLBCL

35 (35) 35 (36) 53 (47) 56 (33) 51 (52) 1054 (219) 42 (46) 123 (122) 105 (105) 110 (227) 37 (41) 114 (113) 70 (73) 153 (77) 62 (48) 174 (174) 113 (113) 326 (326)

26% 17% 23% 20% 2% 9% 10% 6% 6% 15% 19% 1% 29% NS 16% 4% 7% 5%

33% 25% 40% 32% 22% 20% 26% 22% 22% 20% 30% 12% 28% 8% 24% 10% 22% 18%

NS 6 8 0 43 55 25 81 64 73 33 95 40 3 30 125 91 246

NS 20% 25% NA 2% NS 4% 3% 2% NS 15% 0% 33% 0% 13% 4% 6% 4%

NS 25% 38% NA 26% NS 16% 15% 16% 21% 27% 8% 45% 0% 13% 9% 21% 17%

NS 0% 13% NA 24% NS 12% 12% 14% NS 12% 8% 12% 0% 0% 5% 15% 13%

a Bromberg JEC, et al. Neurology. 2007;68(20):1674–9. bDi Noto R, et al. Leuk Res. 2008;32(8):1196–9. cCesana C, et al Leuk Res. 2010;34(8):1027–34. dAlvarez R, et al. Ann Oncol. 2012;23(5):1274-9.eBommer M, et al. Cancer Cytopathol. 2011;119(1):20-6. fCraig FE, et al. Am J Clin Pathol. 2011;135(1):22-34. gStacchini A, et al. Cytometry B Clin Cytom. 2012;82(3):139-44.

haematologica | 2017; 102(2)

239


F.J. Peñalver et al.

port the consensus that any planned prophylactic measures should be adopted early in the treatment course.11

CNS-directed prophylaxis Historically, CNS prophylaxis is most commonly delivered via the intrathecal (IT) route,11,66,67 targeting in particular the leptomeningeal compartment,11 although some authors suggest that IT prophylaxis may be ineffective.2,4 IT methotrexate prophylaxis: The administration of IT methotrexate (MTX) prophylaxis is recommended during each cycle of chemotherapy, with a total of 4 to 8 doses.68 The most common dose used is 12 mg, which achieves therapeutic levels in the CSF (>1 μmol/L) for 24 to 48 hours.68,69 IT MTX doses of 12.5 mg and 15 mg have also been reported.4,13,67,68,70 Of note, studies supporting this approach11,13,70–73 have several limitations, including small sample sizes, absence of a control arm, and co-administration of systemic MTX. In contrast, two large trials4,16 reported no protective benefit of IT MTX prophylaxis. However, these studies were not originally designed to test the efficacy of CNS prophylaxis.11 Moreover, analysis was only possible in the RICOVER-60 trial4 due to a high number of protocol violations (49%). A recent study using the National Comprehensive Cancer Network (NCCN) database for non-Hodgkin lymphoma (NHL) reported no prophylaxisassociated survival benefit,17 although this study was clearly at risk of potential physician bias in selecting patients for IT therapy. Despite all the above, published guidelines11 and clinical trials exploring new treatment options for DLBCL include IT MTX as prophylaxis for high-risk patients. Data reported suggest that several regimens could be active against CNS relapses. Thus, improved outcomes have been suggested for R-DA-EPOCH in low and intermediate IPI patients, and in an ongoing phase 3 study comparing R-DA-EPOCH with R-CHOP that might clarify the potential impact of continuous infusion on CNS relapse rates in the IT MTX settings (IT MTX given for high risk patients as CNS prophylaxis in both protocol arms).74 Improved outcomes and lower CNS relapse rates have been reported in young patients for R-ACVBP vs. R-CHOP associated with IT MTX in both arms, but high dose systemic MTX being administered only in the R-ACVBP arm.75 Other IT drugs: A number of other drugs including liposomal cytarabine (LC) and rituximab can be administered intrathecally. IT LC maintains cytotoxic concentrations in CSF for up to 14 days after a single IT injection,76,77 but is not licensed for prophylactic use.11 The efficacy and toxicity of LC in the prophylaxis of CNS involvement specifically in DLBCL has only been analyzed in two recently published studies.78,79 There are sufficient data to suggest that IT rituximab is efficacious in the treatment of CNS relapse,80,81 but no data support its use in a prophylactic setting. Triple IT: In Spain, triple intrathecal therapy (TIT, methotrexate, cytarabine and hydrocortisone) is the most commonly used schedule for CNS prophylaxis in hematological malignances for the nationwide use of the PETHEMA risk-adapted protocol for lymphoblastic lymphoma and Burkitt lymphoma, which includes TIT for CNS prophylaxis and limits the use of CNS irradiation.66 TIT is also 240

commonly used in DLBCL for CNS prophylaxis, although no studies have compared TIT with IT MTX treatment, and there is no definitive evidence that CNS direct prophylaxis with IT administration improves CNS progression-free survival in patients with parenchymal CNS involvement. Importantly also, IT chemotherapy is not without clinical risk and toxicity, particularly for older and frail patients.

Systemic prophylaxis Data on the potential effectiveness of systemic chemotherapy for CNS prophylaxis in patients with NHL at high risk of CNS relapse are mainly based on information extrapolated from studies of childhood acute lymphoblastic leukemia.11,82 The appropriate intravenous (IV) MTX dose to achieve therapeutic levels in the CNS is controversial. IV MTX doses ≥3 g/m2 appear to produce therapeutic levels in CSF and parenchyma. Three studies conducted in the post-rituximab era examined this method of prophylaxis using high-dose MTX (HD-MTX) doses of 3 g/m2 to 3.5 g/m2, although co-administered drugs, timing, and the number of doses administered varied by protocol.2,75,83,84 Abramson and colleagues reported good outcomes in a retrospective analysis of 65 high-risk patients with DLBCL who received a median of 3 cycles of HD-MTX (3.5 g/m2, range 1–8 cycles) administered on day 15 of alternating cycles of R-CHOP.83 Patients receiving this treatment regimen should have a good baseline condition, and should be closely and carefully monitored for potential toxicity. Adverse effects of MTX include mucositis, myelosuppression, neurotoxicity, and nephrotoxicity. Pre-treatment alkalization of urine and post-treatment leucovorin rescue are considered standard approaches to minimize these toxic effects.68 Systemic prophylaxis with HD-cytarabine in a small sample of DLBCL patients was found to have no clear beneficial role in preventing CNS disease.2,85 New agents like ibrutinib and lenalidomide, which cross the brain barrier, are being explored, and the impact on CNS relapse risk in DLBCL remains to be established.86,87 Systemic prophylaxis with HD-cytarabine in a small sample of DLBCL patients was found to have no clear beneficial role in preventing CNS disease.2,85 New agents like ibrutinib and lenalidomide, which cross the brain barrier, are being explored and the impact on CNS relapse risk in DLBCL remains to be established.86,87

Which prophylactic strategy should be chosen? The question as to the most effective and least toxic route of CNS prophylaxis delivery (IT, parenteral, or a combination thereof) remains largely unanswered, and should be addressed in large scale randomized clinical trials comparing systemic and IT chemoprophylaxis.68 Aviles and colleagues88 analyzed a homogenous group of 3,258 DLBCL patients treated with CHOP or R-CHOP, 1,005 of whom received different CNS prophylaxis schedules (radiotherapy, IT MTX, HD-MTX, or rituximab). No clear differences were observed between the different prophylaxis schedules. Furthermore, rates of CNS relapse were similar in patients who received prophylaxis (6%) and those who did not (5.9%). Cheah and colleagues89 recently performed a retrospective analysis of patients with high-risk DLBCL, comparing three different strategies of CNS-directed therapy: IT haematologica | 2017; 102(2)


Guidelines for management of CNS involvement in DLBCL

Figure 1. Practical algorithm based on the recommendations of the Guidelines. aCSF examination should be also performed in presence of neurological symptoms, in addition to imaging techniques (MRI, CT). bThe use of standardized and validated ≥8-color FCM evaluation of stabilized CSF samples is recommended (with immediate addition of RPMI1640 or Transfix® to CSF samples). cThe group recommends CNS prophylaxis in FMC-/CC- patients with high-risk factors for CNS relapse until future studies are available. IV HD-MTX (≥3 g/m2) alternating with immunochemotherapy or IT MTX administered during primary therapy (12 mg once per cycle, 4– 6 doses), depending on age, performance status, comorbidities and patient and/or physician’s preferences. IV MTX should be given in line with published schedules, and in the context of performance status and renal function. Delay of subsequent cycles of systemic immunochemotherapy should be avoided. Patients with primary testicular lymphoma should receive IT MTX during primary chemotherapy. Triple IT therapy (MTX 15 mg, cytarabine 40 mg, and hydrocortisone 20 mg) is a reasonable option for CNS prophylaxis. dThere is no direct evidence to support the adoption of different treatment decisions in patients with occult leptomeningeal disease (CC/FCM+): HD-MTX and/or IT chemotherapy should be considered for these patients. eIn cases of CNS involvement at the time of DLBCL diagnosis: HD-MTX (associated IT therapy if leptomeningeal disease is demonstrated). In patients for whom HD-MTX is inadequate due to age or comorbidities, IT liposomal cytarabine should be considered. In the case of CNS relapse: salvage therapy (HD-MTX-based induction) followed by ASCT (depending of performance status and age of the patient). Thiotepa and BCNU should be included in the conditioning regimen before ASCT. In the case of refractoriness or early relapse after HD-MTX, consider clinical trial or radiotherapy. DLBCL, diffuse large B-cell lymphoma; CNS, central nervous system; LDH, lactate dehydrogenase; CNS-IPI, central nervous system-International Prognostic Index;28 CC, conventional cytology; FCM, multiparameter flow cytometry; MRI, magnetic resonance imaging; CT, computerized tomography; IV, intravenous; HD-MTX, high dose-MTX; MTX, methotrexate; IT, intrathecal; ASCT, autologous stem cell transplant.

MTX with R-CHOP (group 1); R-CHOP with IT MTX and two cycles of HD-MTX (group 2); and dose-intensive systemic chemotherapy (Hyper-CVAD or CODOXM/IVAC) with IT/IV MTX (group 3). A total of 23 CNS relapses occurred (24%, 8%, and 2.3% in groups 1, 2, and 3, respectively). Although these data are limited by the retrospective nature of the study, the addition of HD-MTX and/or HD-cytarabine appears to be associated with lower incidence of CNS relapse as compared with IT chemotherapy alone.

Primary testicular lymphoma Patients with primary testicular involvement have a particularly high risk of CNS involvement (>15%) when achieving a complete response (CR). Treatment recommendations for these patients differ from those for other forms of extranodal DLBCL. Vitolo and colleagues reported a 6% CNS relapse rate after 5 years in patients treated with a combination of R-CHOP plus four doses of IT MTX and contralateral testis irradiation.72 No data are available on treatment with IV MTX alone in this scenario.

haematologica | 2017; 102(2)

Summary and recommendations for CNS prophylaxis • CNS-directed prophylaxis should be offered to patients at high-risk of CNS relapse (recommendation 1, level of evidence B). • IV MTX is recommended as CNS prophylaxis in highrisk patients (recommendation 2, level of evidence B). • IV MTX as CNS prophylaxis should be administered during primary therapy at a dose of ≥3 g/m2, alternating with immunochemotherapy (recommendation 1, level of evidence B), and should be given according to published treatment schemes and in the context of performance status and renal function. Delay of subsequent cycles of systemic immunochemotherapy should be avoided (recommendation 1, level of evidence B). • IT MTX (recommendation 2, level of evidence C) or triple IT (recommendation 2, level of evidence C) may be reasonable options for prophylaxis, depending on age, performance status, and comorbidities • IT MTX (12–15 mg once per cycle, 4–6 doses) or triple IT (MTX 15 mg, cytarabine 40 mg, hydrocortisone 20 mg) as CNS prophylaxis should be administered during primary therapy (recommendation 1, level of evidence B). • Patients with primary testicular lymphoma should receive IT MTX during primary chemotherapy (recommendation 1, level of evidence B). 241


F.J. Peñalver et al.

Treatment of central nervous system involvement of lymphoma Secondary involvement of CNS in aggressive NHL can occur at presentation or early in the first year, usually associated with or anticipating systemic relapse. Accordingly, both CNS and systemic lymphoma should be considered for the treatment of CNS dissemination.

Whole-brain radiotherapy The usefulness of radiotherapy for the management of CNS lymphoma is limited by its toxicity, especially in older patients. Whole-brain radiotherapy has been used in combination with chemotherapy in PCNSL, but its true impact on outcome remains controversial.90 While reduced-dose radiotherapy may cause less neurotoxicity, there is a paucity of relevant randomized studies. At present, whole-brain radiation is generally reserved for salvage therapy in patients with MTX resistance.91 In secondary CNS lymphoma (SCNSL), radiotherapy could be considered as an adjuvant treatment in patients with large masses or with blockade of CSF flow.92

Systemic chemotherapy Systemic chemotherapy agents that cross the bloodbrain barrier (BBB) become distributed throughout the neural axis, avoiding the need for IT chemotherapy administered via multiple lumbar punctures or ventricular reservoirs. However, toxicity in bone marrow and other organs should be considered.92

High-dose methotrexate IV MTX is active in primary and secondary CNSL although the optimal dosage is yet to be defined. Doses ≥1 g/m2 achieve tumoricidal levels in brain parenchyma, doses of 8 g/m2 produce higher cytotoxic levels in serum and CSF than IT MTX, and doses of 3 g/m2 are sufficient to treat brain and leptomeningeal disease, without associated IT MTX.91 There is no consensus as to the optimal number of cycles needed, although at least 4 cycles of HD-MTX may be necessary. The toxic effects of HD-MTX should be carefully considered, particularly nephropathy. Advanced age, poor performance status, and renal or liver dysfunction should be considered contraindications for HD-MTX.

Polychemotherapy A study of patients with PCNSL by Ferreri and colleagues demonstrated a failure-free survival benefit in patients who received HD-MTX plus HD-cytarabine as induction therapy, followed by radiotherapy as consolidation.93 Other anti-lymphoma agents that cross the BBB such as procarbazine or ifosfamide have been used in combination with HD-MTX, and have showed encouraging activity.91,92,94 Immunochemotherapy consisting of HDMTX, intravenous rituximab, and oral temozolomide may be a feasible option, as demonstrated by Wong and colleagues in a study of PCNSL patients.95

Intensification chemotherapy and autologous hematopoietic stem cell transplantation High-dose chemotherapy consolidation followed by autologous stem cell transplant (ASCT) rescue is a very promising option in patients with recurrent SCNSL, with better outcomes in patients who achieve CR before transplantation.96 242

In a German prospective phase II study, HDMTX, ifosfamide, dexamethasone and IT LC followed by HDcytarabine, thiotepa and IT LC, and, for responding patients, consolidation with BCNU, thiotepa, etoposide, and ASCT rescue, resulted in 50% CR, with a 2-year OS rate of 68% after transplantation.97 In another recent Italian trial, HDMTX and cytarabine, followed by R-HDS (rituximab, cyclophosphamide, cytarabine, and etoposide) supported by ASCT was associated with 63% CR and 5year OS of 68% for transplanted patients.98 Long-term survival in patients who underwent ASCT has also been reported in a retrospective international multicenter study.99 Other published conditioning regimens include other combinations including cyclophosphamide, carmustine, etoposide, busulfan and thiotepa, with or without rituximab.93,100,101 All such studies demonstrate that significant progress has been made toward cure in this difficult condition that was almost systematically fatal a few years ago.102 Hopefully, new molecules that cross the BBB, like ibrutinib or lenalidomide, might further improve the outcome of these patients.86,87 To our understanding, current treatments for this condition should incorporate multifaceted approaches, such as multi-drug regimens with non-cross resistance and CNS activity, rituximab to improve systemic lymphoma control, IT therapy, and treatment intensification with ASCT.96,101

Intrathecal therapy IT MTX, cytarabine, and thiotepa can be administered into the spinal fluid, allowing the drug to reach the spinal cord and brain. However, these agents are rapidly cleared from the CSF, requiring administration two or three times a week. IT LC provides sustained concentrations in CSF for 14 days, allowing a more favorable administration schedule.103,104 The superiority of LC over conventional cytarabine in the treatment of lymphomatous meningitis has been demonstrated in a randomized clinical trial,76 and several studies have shown significant efficacy of LC.104–106 In terms of safety, LC should be administered with concurrent dexamethasone therapy,107,108 maintaining an adequate interval between LC administration and that of other potential neurotoxic cytostatic drugs, especially intravenous HD-MTX and HD-cytarabine.78,107 Intraventricular or IT administration of rituximab may be of value in the treatment of patients with recurrent CD20-positive CNSL.80,81 Intraventricular administration of rituximab (10–25 mg) is feasible, has shown encouraging anti-CNSL activity and clinical benefit, and when combined with intraventricular MTX results in improved responses.81

Therapeutic approach CNS involvement by aggressive lymphoma is an extremely heterogeneous and very complex situation, with many variables determining treatment of choice and outcome, including the B-cell-of-origin subtype.

CNS and systemic involvement at diagnosis Patients with synchronous CNS and systemic aggressive NHL at presentation should receive immunochemotherapy for the systemic disease and CNS-targeted chemotherapy for CNSL. R-CHOP plus HD-MTX followed, in haematologica | 2017; 102(2)


Guidelines for management of CNS involvement in DLBCL

patients with systemic and CNS CR, by etoposide and cytarabine consolidation is one feasible option.91 In cases of lymphomatous meningitis, R-CHOP plus LC is a possible alternative.105

Summary and recommendations for treatment of CNS involvement in DLBCL

cases involving leptomeningeal lymphoma, associated IT LC treatment can be administered (recommendation 1, level of evidence B). • In patients not suitable for HD-MTX treatment due to age or comorbidities, we recommend treatment with IT LC (recommendation 1, level of evidence B). • In patients with relapsed DLBCL with good clinical condition and of appropriate age: HD-MTX-based schemes followed, in responding cases, by consolidation with ASCT (recommendation 1, level of evidence B). • Thiotepa and BCNU should be included in the conditioning regimen prior to ASCT (recommendation 1, level of evidence C). • For patients with occult leptomeningeal lymphoma (CC-/FCM+), there is no direct evidence supporting the value of different therapeutic strategies. In these patients, we recommend considering treatment with HD-MTX and/or IT chemotherapy (particularly in patients for whom HD-MTX is not indicated due to age or comorbidities) (recommendation 2, level of evidence C). • In cases of refractoriness or early relapse after HDMTX: clinical trial or whole brain radiotherapy (recommendation 2, level of evidence C).

• Patients with systemic DLBCL and synchronous CNS parenchymatous and/or leptomeningeal lymphoma at diagnosis should be treated with HD-MTX-containing regimens (recommendation 1, level of evidence B). In

Acknowledgments Adelaida Velasco for her editorial assistance and patience in preparing the manuscript.

CNS relapse High-dose chemotherapy followed by ASCT is feasible and effective for recurrent aggressive CNS lymphoma, and is probably the best currently available curative option.97,99 It is important to determine whether the relapse is "MTX-sensitive" or not. In MTX-sensitive patients, HDMTX administration to achieve maximum cytoreduction is advisable, followed by thiotepa or carmustine-based conditioning regimens and ASCT.91 Patients with MTXresistant lymphoma or those relapsing within 6 months after consolidation schemas may not be candidates for high-dose rescue strategies. These patients should be included in clinical trials or considered for palliative treatment, according to clinical condition and other clinical or laboratory variables.91

References 1. Zhang J, Chen B, Xu X. Impact of rituximab on incidence of and risk factors for central nervous system relapse in patients with diffuse large B-cell lymphoma: a systematic review and meta-analysis. Leuk Lymphoma. 2014;55(3):509–514. 2. Fletcher CD, Kahl BS. Central nervous system involvement in diffuse large B-cell lymphoma: an analysis of risks and prevention strategies in the post-rituximab era. Leuk Lymphoma. 2014;55(10):2228–2240. 3. Feugier P, Virion JM, Tilly H, et al. Incidence and risk factors for central nervous system occurrence in elderly patients with diffuse large-B-cell lymphoma: Influence or rituximab. Ann Oncol. 2004; 15(1):129–133. 4. Boehme V, Schmitz N, Zeynalova S, Loeffler M, Pfreundschuh M. CNS events in elderly patients with aggressive lymphoma treated with modern chemotherapy (CHOP-14) with or without rituximab: An analysis of patients treated in the RICOVER-60 trial of the German High-Grade Non-Hodgkin Lymphoma Study Group (DSHNHL). Blood. 2009; 113(17):3896–3902. 5. Villa D, Connors JM, Shenkier TN, Gascoyne RD, Sehn LH, Savage KJ. Incidence and risk factors for central nervous system relapse in patients with diffuse large B-cell lymphoma: The impact of the addition of rituximab to CHOP chemotherapy. Ann Oncol. 2009;21(5):1046–1052. 6. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008;336(7650):924–926. 7. Shimazu Y, Notohara K, Ueda Y. Diffuse large B-cell lymphoma with central nervous system relapse: Prognosis and risk factors

haematologica | 2017; 102(2)

8.

9.

10.

11.

12.

13.

14.

according to retrospective analysis from a single-center experience. Int J Hematol. 2009;89(5):577–583. Mitrovic Z, Bast M, Bierman PJ, et al. The addition of rituximab reduces the incidence of secondary central nervous system involvement in patients with diffuse large Bcell lymphoma. Br J Haematol. 2012;157 (3):401–403. Schmitz N, Zeynalova S, Glass B, et al. CNS disease in younger patients with aggressive B-cell lymphoma: an analysis of patients treated on the Mabthera International Trial and trials of the German High-Grade NonHodgkin Lymphoma Study Group. Ann Oncol. 2012;23(5): 1267–1273. Wilson WH, Bromberg JEC, StetlerStevenson M, et al. Detection and outcome of occult leptomeningeal disease in diffuse large B-cell lymphoma and Burkitt lymphoma. Haematologica. 2014; 99(7): 1228– 1235. Mcmillan A, Ardeshna KM, Cwynarski K, Lyttelton M, Mckay P, Montoto S. Guideline on the prevention of secondary central nervous system lymphoma: British Committee for Standards in Haematology. Br J Haematol. 2013;163(2):168–181. van Besien K, Ha CS, Murphy S, et al. Risk factors, treatment, and outcome of central nervous system recurrence in adults with intermediate-grade and immunoblastic lymphoma. Blood. 1998;91(4):1178–1184. Haioun C, Besson C, Lepage E, et al. Incidence and risk factors of central nervous system relapse in histologically aggressive non-Hodgkin’s lymphoma uniformly treated and receiving intrathecal central nervous system prophylaxis: A GELA study on 974 patients. Ann Oncol. 2000;11(6):685–690. Hollender A, Kvaloy S, Nome O, Skovlund E, Lote K, Holte H. Central nervous system

15.

16.

17.

18.

19.

20.

21.

involvement following diagnosis of nonHodgkin’s lymphoma: A risk model. Ann Oncol. 2002;13(7):1099–1107. Boehme V, Zeynalova S, Kloess M, et al. Incidence and risk factors of central nervous system recurrence in aggressive lymphoma-a survey of 1693 patients treated in protocols of the German High-Grade NonHodgkin’s Lymphoma Study Group (DSHNHL). Ann Oncol. 2007;18(1): 149– 157. Bernstein SH, Unger JM, Leblanc M, Friedberg J, Miller TP, Fisher RI. Natural history of CNS relapse in patients with aggressive non-Hodgkin’s lymphoma: a 20-year follow-up analysis of SWOG 8516 -- the Southwest Oncology Group. J Clin Oncol. 2009;27(1):114–119. Kumar A, Vanderplas A, LaCasce AS, et al. Lack of benefit of central nervous system prophylaxis for diffuse large B-cell lymphoma in the rituximab era: findings from a large national database. Cancer. 2012; 118(11):2944–2951. Yamamoto W, Tomita N, Watanabe R, et al. Central nervous system involvement in diffuse large B-cell lymphoma. Eur J Haematol. 2010;85(1):6–10. Zucca E, Conconi A, Mughal TI, et al. Patterns of outcome and prognostic factors in primary large-cell lymphoma of the testis in a survey by the International Extranodal Lymphoma Study Group. J Clin Oncol. 2003;21(1):20–27. Tomita N, Yokoyama M, Yamamoto W, et al. Central nervous system event in patients with diffuse large B-cell lymphoma in the rituximab era. Cancer Sci. 2012; 103(2):245– 251. Hosein PJ, Maragulia JC, Salzberg MP, et al. A multicentre study of primary breast diffuse large B-cell lymphoma in the rituximab

243


F.J. Peñalver et al. era. Br J Haematol. 2014; 165(3):358–363. 22. Villa D, Connors JM, Sehn LH, Gascoyne RD, Savage KJ. Diffuse large B-cell lymphoma with involvement of the kidney: Outcome and risk of central nervous system relapse. Haematologica. 2011;96(7): 1002– 1007. 23. Tai WM, Chung J, Tang PL, et al. Central nervous system (CNS) relapse in diffuse large B cell lymphoma (DLBCL): Pre- and post-rituximab. Ann Hematol. 2011; 90(7):809–818. 24. MacKintosh FR, Colby T V, Podolsky WJ, et al. Central nervous system involvement in non-Hodgkin’s lymphoma: an analysis of 105 cases. Cancer. 1982;49(3):586–595. 25. Murawski N, Held G, Ziepert M, et al. The role of radiotherapy and intrathecal CNS prophylaxis in extralymphatic craniofacial aggressive B-cell lymphomas. Blood. 2014; 124(5):720–728. 26. Lee G, Go S, Kim S, et al. Clinical outcome and prognosis of patients with primary sinonasal tract diffuse large B-cell lymphoma treated with rituximab-cyclophosphamide , doxorubicin , vincristine and prednisone chemotherapy: a study by the Consortium for Improving Survival of Lym. Leuk Lymphoma. 2015;56(4):1020–1026. 27. Schmitz N, Zeynalova S, Nickelsen M, Ziepert M, Pfreundschuh M, Glass B LM. A new prognostic model to assess the risk of CNS disease in patients with aggressive Bcell lymphoma. Hematol Oncol. 2013; 31(S1):96–150. 28. Schmitz N, Zeynalova S, Nickelsen M, et al. CNS International Prognostic Index : A Risk Model for CNS Relapse in Patients With Diffuse Large B-Cell Lymphoma Treated With R-CHOP. J Clin Oncol. 2016; 34(26):3150–3156. 29. Savage KJ, Slack GW, Mottok A, et al. The impact of dual expression of MYC and BCL2 by immunohistochemistry on the risk of CNS relapse in DLBCL. Blood. 2016; 127(18):2182–2189. 30. Scott BJ, Douglas VC, Tihan T, Rubenstein JL, Josephson SA. A systematic approach to the diagnosis of suspected central nervous system lymphoma. JAMA Neurol. 2013; 70(3):311–319. 31. Chamberlain MC, Sandy AD, Press GA. Leptomeningeal metastasis: a comparison of gadolinium-enhanced MR and contrastenhanced CT of the brain. Neurology. 1990;40(3 Pt 1):435–438. 32. Abrey LE, Batchelor TT, Ferreri AJM, et al. Report of an international workshop to standardize baseline evaluation and response criteria for primary CNS lymphoma. J Clin Oncol. 2005; 23(22): 5034–5043. 33. Freilich RJ, Krol G, DeAngelis LM. Neuroimaging and cerebrospinal fluid cytology in the diagnosis of leptomeningeal metastasis. Ann Neurol. 1995;38(1):51–57. 34. Pauls S, Fischer AC, Brambs HJ, Fetscher S, Höche W, Bommer M. Use of magnetic resonance imaging to detect neoplastic meningitis: Limited use in leukemia and lymphoma but convincing results in solid tumors. Eur J Radiol. 2012;81(5):974–978. 35. Johnson BA, Fram EK, Johnson PC, Jacobowitz R. The variable MR appearance of primary lymphoma of the central nervous system: Comparison with histopathologic features. Am J Neuroradiol. 1997; 18(3):563– 572. 36. Geppert M, Ostertag CB, Seitz G, Kiessling M. Glucocorticoid therapy obscures the diagnosis of cerebral lymphoma. Acta Neuropathol. 1990;80(6):629–634. 37. Dethy S, Goldman S, Blecic S, Luxen A,

244

38.

39.

40.

41. 42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

Levivier M, Hildebrand J. Carbon-11methionine and fluorine-18-FDG PET study in brain hematoma. J Nucl Med. 1994;35 (7):1162–1166. Tsuyuguchi N, Sunada I, Ohata K, et al. Evaluation of treatment effects in brain abscess with positron emission tomography: comparison of fluorine-18-fluorodeoxyglucose and carbon-11-methionine. Ann Nucl Med. 2003;17(1):47–51. Baraniskin A, Deckert M, SchulteAltedorneburg G, Schlegel U, Schroers R. Current strategies in the diagnosis of diffuse large B-cell lymphoma of the central nervous system. Br J Haematol. 2012; 156(4):421– 432. Balmaceda C, Gaynor JJ, Sun M, Gluck JT, DeAngelis LM. Leptomeningeal tumor in primary central nervous system lymphoma: Recognition, significance, and implications. Ann Neurol. 1995;38(2):202–209. Fine HA, Mayer RJ. Primary central nervous system lymphoma. Ann Intern Med. 1993; 119(11):1093–1104. Hegde U, Filie A, Little RF, et al. High incidence of occult leptomeningeal disease detected by flow cytometry in newly diagnosed aggressive B-cell lymphomas at risk for central nervous system involvement: The role of flow cytometry versus cytology. Blood. 2005;105(2):496–502. Glantz MJ, Cole BF, Glantz LK, et al. Cerebrospinal fluid cytology in patients with cancer: minimizing false-negative results. Cancer. 1998;82(4):733–739. Glass JP, Melamed M, Chernik NL, Posner JB. Malignant cells in cerebrospinal fluid (CSF): the meaning of a positive CSF cytology. Neurology. 1979;29(10):1369–1375. Quijano S, Lopez A, Sancho JM, et al. Identification of leptomeningeal disease in aggressive B-Cell non-Hodgkin’s lymphoma: Improved sensitivity of flow cytometry. J Clin Oncol. 2009; 27(9):1462–1469. Sancho JM, Orfao A, Quijano S, et al. Clinical significance of occult cerebrospinal fluid involvement assessed by flow cytometry in non-Hodgkin’s lymphoma patients at high risk of central nervous system disease in the rituximab era. Eur J Haematol. 2010;85(4):321–328. Finn WG, Peterson LC, James C, Goolsby CL. Enhanced detection of malignant lymphoma in cerebrospinal fluid by multiparameter flow cytometry. Am J Clin Pathol. 1998;110(3):341–346. French CA, Dorfman DM, Shaheen G, Cibas ES. Diagnosing lymphoproliferative disorders involving the cerebrospinal fluid: Increased sensitivity using flow cytometric analysis. Diagn Cytopathol. 2000; 23(6):369–374. Roma AA, Garcia A, Avagnina A, Rescia C, Elsner B. Lymphoid and myeloid neoplasms involving cerebrospinal fluid: Comparison of morphologic examination and immunophenotyping by flow cytometry. Diagn Cytopathol. 2002;27(5):271–275. Subirá D, Górgolas M, Castañón S, et al. Advantages of flow cytometry immunophenotyping for the diagnosis of central nervous system non-Hodgkin’s lymphoma in AIDS patients. HIV Med. 2005;6(1):21–26. Benevolo G, Stacchini A, Spina M, et al. Final results of a multicenter trial addressing role of CSF flow cytometric analysis in NHL patients at high risk for CNS dissemination. Blood. 2012;120(16):3222–3228. de Jongste AH, Kraan J, Van Den Broek PD, et al. Use of TransFixTM cerebrospinal fluid storage tubes prevents cellular loss and enhances flow cytometric detection of

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

malignant hematological cells after 18 hours of storage. Cytom Part B - Clin Cytom. 2014;86(4):272–279. van Dongen JJM, Lhermitte L, Böttcher S, et al. EuroFlow antibody panels for standardized n-dimensional flow cytometric immunophenotyping of normal, reactive and malignant leukocytes. Leukemia. 2012; 26(9):1908–1975. Blakeley J, Laterra J. Neoplastic and Paraneoplastic Disorders. In: Irani DN, editor. Cerebrospinal Fluid in Clinical Practice. Philadelphia: Saunders; 2009. p. 233–247. Chamberlain MC, Glantz M, Groves MD, Wilson WH. Diagnostic Tools for Neoplastic Meningitis: Detecting Disease, Identifying Patient Risk, and Determining Benefit of Treatment. Semin Oncol. 2009; 36(4 Suppl 2):S35–S45. Gleissner B, Siehl J, Korfel A, Reinhardt R, Thiel E. CSF evaluation in primary CNS lymphoma patients by PCR of the CDR III IgH genes. Neurology. 2002;58(3):390–396. Muniz C, Martin-Martin L, Lopez A, et al. Contribution of cerebrospinal fluid sCD19 levels to the detection of CNS lymphoma and its impact on disease outcome. Blood. 2014;123(12):1864–1869. Kiewe P, Fischer L, Martus P, Thiel E, Korfel A. Meningeal dissemination in primary CNS lymphoma: Diagnosis, treatment, and survival in a large monocenter cohort. Neuro Oncol. 2010;12(4):409–417. Roy S, Josephson SA, Fridlyand J, et al. Protein biomarker identification in the CSF of patients with CNS lymphoma. J Clin Oncol. 2008;26(1):96–105. Baraniskin A, Kuhnhenn J, Schlegel U, et al. Identification of microRNAs in the cerebrospinal fluid as marker for primary diffuse large B-cell lymphoma of the central nervous system. Blood. 2011;117(11):3140–3146. Kersten MJ, Evers LM, Dellemijn PL, et al. Elevation of cerebrospinal fluid soluble CD27 levels in patients with meningeal localization of lymphoid malignancies. Blood. 1996;87(5):1985–1989. Fischer L, Korfel A, Pfeiffer S, et al. CXCL13 and CXCL12 in central nervous system lymphoma patients. Clin Cancer Res. 2009; 15(19):5968–5973. Viaccoz A, Ducray F, Tholance Y, et al. CSF neopterin level as a diagnostic marker in primary central nervous system lymphoma. Neuro Oncol. 2015; 17(11):1497–1503. Chihara D, Oki Y, Matsuo K, et al. Incidence and risk factors for central nervous system relapse in patients with diffuse large B-cell lymphoma: analyses with competing risk regression model. Leuk Lymphoma. 2011;52(12):2270–2275. Siegal T, Goldschmidt N. CNS prophylaxis in diffuse large B-cell lymphoma: If, when, how and for whom? Blood Rev. 2012; 26(3):97–106. Sancho JM, Morgades M, Alonso N, et al. Prospective study on the practice of central nervous system prophylaxis and treatment in non-Hodgkin’s lymphoma in Spain. Med Clin (Barc). 2008;131(12):441–446. Cheung CW, Burton C, Smith P, Linch DC, Hoskin PJ, Ardeshna KM. Central nervous system chemoprophylaxis in non-Hodgkin lymphoma: Current practice in the UK. Br J Haematol. 2005;131(2):193–200. Ghose A, Kundu R, Latif T. Prophylactic CNS directed therapy in systemic diffuse large B cell lymphoma. Crit Rev Oncol Hematol. 2014;91(3):292–303. Fleischhack G, Jaehde U, Bode U. Pharmacokinetics following intraventricular administration of chemotherapy in patients

haematologica | 2017; 102(2)


Guidelines for management of CNS involvement in DLBCL

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

with neoplastic meningitis. Clin Pharmacokinet. 2005;44(1):1–31. Arkenau H-T, Chong G, Cunningham D, et al. The role of intrathecal chemotherapy prophylaxis in patients with diffuse large B-cell lymphoma. Ann Oncol. 2007;18(3): 541–545. Tilly H, Lepage E, Coiffier B, et al. Intensive conventional chemotherapy (ACVBP regimen) compared with standard CHOP for poor-prognosis aggressive non-Hodgkin lymphoma. Blood. 2003;102(13): 4284– 4289. Vitolo U, Chiappella A, Ferreri AJM, et al. First-line treatment for primary testicular diffuse large B-cell lymphoma with rituximabCHOP, CNS prophylaxis, and contralateral testis irradiation: Final results of an international phase II trial. J Clin Oncol. 2011;29(20):2766–2772. Tomita N, Kodama F, Kanamori H, Motomura S, Ishigatsubo Y. Prophylactic intrathecal methotrexate and hydrocortisone reduces central nervous system recurrence and improves survival in aggressive non-Hodgkin lymphoma. Cancer. 2002; 95(3):576–580. Wilson WH, Dunleavy K, Pittaluga S, et al. Phase II study of dose-adjusted EPOCH and rituximab in untreated diffuse large B-cell lymphoma with analysis of germinal center and post-germinal center biomarkers. J Clin Oncol. 2008;26(16):2717–2724. Récher C, Coiffier B, Haioun C, et al. Intensified chemotherapy with ACVBP plus rituximab versus standard CHOP plus rituximab for the treatment of diffuse large B-cell lymphoma (LNH03-2B): An open-label randomised phase 3 trial. Lancet. 2011;378(9806):1858–1867. Glantz MJ, LaFollette S, Jaeckle KA, et al. Randomized trial of a slow-release versus a standard formulation of cytarabine for the intrathecal treatment of lymphomatous meningitis. J Clin Oncol. 1999;17(10): 3110– 3116. Corazzelli G, Frigeri F, Russo F, et al. RDCODOX-M/IVAC with rituximab and intrathecal liposomal cytarabine in adult Burkitt lymphoma and “unclassifiable” highly aggressive B-cell lymphoma. Br J Haematol. 2012;156(2):234–244. de la Fuente A, Salar A, Panizo C, et al. Efficacy and Safety of Liposomal Cytarabine as Intrathecal Prophylaxis in Patients with Diffuse Large B Cell Lymphoma at High Risk of CNS Involvement: A Multicentric Study Including 80 Patients in Spain. Blood. 2009; 114(22):1663. González-Barca E, Canales M, Salar A, et al. Central nervous system prophylaxis with intrathecal liposomal cytarabine in a subset of high-risk patients with diffuse large B-cell lymphoma receiving first line systemic therapy in a prospective trial. Ann Hematol. 2016;95(6):893–899. Schulz H, Pels H, Schmidt-Wolf I, Zeelen U, Germing U, Engert A. Intraventricular treatment of relapsed central nervous system lymphoma with the anti-CD20 antibody rituximab. Haematologica. 2004; 89(6):753–754. Rubenstein JL, Li J, Chen L, et al. Multicenter phase 1 trial of intraventricular immunochemotherapy in recurrent CNS

haematologica | 2017; 102(2)

lymphoma. Blood. 2013;121(5):745–751. 82. Vassal G, Valteau D, Bonnay M, Patte C, Aubier F, Lemerle J. Cerebrospinal fluid and plasma methotrexate levels following highdose regimen given as a 3-hour intravenous infusion in children with nonHodgkin’s lymphoma. Pediatr Hematol Oncol. 1990; 7(1):71–77. 83. Abramson JS, Hellmann M, Barnes JA, et al. Intravenous methotrexate as central nervous system (CNS) prophylaxis is associated with a low risk of CNS recurrence in high-risk patients with diffuse large B-cell lymphoma. Cancer. 2010;116(18):4283–4290. 84. Holte H, Leppä S, Björkholm M, et al. Dosedensified chemoimmunotherapy followed by systemic central nervous system prophylaxis for younger high-risk diffuse large Bcell/follicular grade 3 lymphoma patients: Results of a phase II Nordic lymphoma group study. Ann Oncol. 2013; 24(5):1385– 1392. 85. Adde M, Enblad GG, Hagberg H, Sundstrom C, Laurell A, Sundström C. Outcome for young high-risk aggressive B-cell lymphoma patients treated with CHOEP-14 and rituximab (R-CHOEP-14). Med Oncol. 2006;23(2):283–293. 86. Bernard S, Goldwirt L, Amorim S, et al. Activity of ibrutinib in mantle cell lymphoma patients with central nervous system relapse. Blood. 2015; 126(14):1695–1698. 87. Houillier C, Choquet S, Touitou V, et al. Lenalidomide monotherapy as salvage treatment for recurrent primary CNS lymphoma. Neurology. 2015;84(3):325–326. 88. Avilés A, Jesús Nambo M, Neri N. Central nervous system prophylaxis in patients with aggressive diffuse large B cell lymphoma: an analysis of 3,258 patients in a single center. Med Oncol. 2013;30(2):520. 89. Cheah CY, Herbert KE, O’Rourke K, et al. A multicentre retrospective comparison of central nervous system prophylaxis strategies among patients with high-risk diffuse large B-cell lymphoma. Br J Cancer. 2014; 111(6):1072–1079. 90. Graber JJ, Omuro A. Primary central nervous system lymphoma: is there still a role for radiotherapy? Curr Opin Neurol. 2011;24(6):633–640. 91. Rubenstein JL, Gupta NK, Mannis GN, LaMarre AK, Treseler P. How I treat CNS lymphomas. Blood. 2013;122(14): 2318– 2330. 92. Fischer L, Korfel A, Kiewe P, Neumann M, Jahnke K, Thiel E. Systemic high-dose methotrexate plus ifosfamide is highly effective for central nervous system (CNS) involvement of lymphoma. Ann Hematol. 2009;88(2):133–139. 93. Ferreri AJ, Reni M, Foppoli M, et al. Highdose cytarabine plus high-dose methotrexate versus high-dose methotrexate alone in patients with primary CNS lymphoma: a randomised phase 2 trial. Lancet. 2009; 374(9700):1512–1520. 94. Bokstein F, Lossos A, Lossos IS, Siegal T. Central nervous system relapse of systemic non-Hodgkin’s lymphoma: results of treatment based on high-dose methotrexate combination chemotherapy. Leuk Lymphoma. 2002;43(3):587–593. 95. Wong ET, Tishler R, Barron L, Wu JK.

Immunochemotherapy with rituximab and temozolomide for central nervous system lymphomas. Cancer. 2004;101(1):139–145. 96. Bierman P, Giglio P. Diagnosis and treatment of central nervous system involvement in non-Hodgkin’s lymphoma. Hematol Oncol Clin North Am. 2005; 19(4):597–609. 97. Korfel A, Elter T, Thiel E, et al. Phase II study of central nervous system (CNS)-directed chemotherapy including high-dose chemotherapy with autologous stem cell transplantation for CNS relapse of aggressive lymphomas. Haematologica. 2013; 98(3):364–370. 98. Ferreri AJM, Donadoni G, Cabras MG, et al. High doses of antimetabolites followed by high-dose sequential chemoimmunotherapy and autologous stem-cell transplantation in patients with systemic B-cell lymphoma and secondary CNS involvement: Final results of a multicenter phase II trial. J Clin Oncol. 2015; 33(33):3903–3910. 99. Bromberg JE, Doorduijn JK, Illerhaus G, et al. Central nervous system recurrence of systemic lymphoma in the era of stem cell transplantation - An international primary central nervous system lymphoma study group project. Haematologica. 2013; 98(5):808–813. 100. Rubenstein JL, Hsi ED, Johnson JL, et al. Intensive chemotherapy and immunotherapy in patients with newly diagnosed primary CNS lymphoma: CALGB 50202 (Alliance 50202). J Clin Oncol. 2013; 31(25):3061–3068. 101.Tarella C, Zanni M, Magni M, et al. Rituximab improves the efficacy of highdose chemotherapy with autograft for highrisk follicular and diffuse large B-cell lymphoma: A multicenter Gruppo Italiano Terapie Innnovative nei Linfomi survey. J Clin Oncol. 2008;26(19):3166–3175. 102. Schmitz N, Wu HS. Advances in the Treatment of Secondary CNS Lymphoma. J Clin Oncol. 2015;33(33):3851–3853. 103.Kim S, Khatibi S, Howell SB, McCully C, Balis FM, Poplack DG. Prolongation of drug exposure in cerebrospinal fluid by encapsulation into DepoFoam. Cancer Res. 1993; 53(7):1596–1598. 104.Howell SB. Liposomal cytarabine for the treatment of lymphomatous meningitis. Biol Ther Lymphoma. 2003;6:10–14. 105.Garcia-Marco JA, Panizo C, Garcia ES, et al. Efficacy and safety of Liposomal cytarabine in lymphoma Patients with central nervous system involvement from lymphoma. Cancer. 2009;115(9):1892–1898. 106. Gökbuget N, Hartog CM, Bassan R, et al. Liposomal cytarabine is effective and tolerable in the treatment of central nervous system relapse of acute lymphoblastic leukemia and very aggressive lymphoma. Haematologica. 2011;96(2):238–244. 107. DepoCyte: Summary of product characteristics [Internet]. European Medicines Agency. [cited 2016 Apr 1]. Available from: www.ema.europe.eu/ema/ 108. Sanchez-Gonzalez B, Llorente A, Sancho JM, et al. A new modified prophylactic scheme against liposomal cytarabineinduced arachnoiditis in adult patients with lymphoma. Leuk Lymphoma. 2013; 54(4):892–893.

245


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Red Cell Biology & Its Disorders

Ferrata Storti Foundation

ARQ 092, an orally-available, selective AKT inhibitor, attenuates neutrophil-platelet interactions in sickle cell disease

Kyungho Kim,1 Jing Li,1 Andrew Barazia,1 Alan Tseng,1 Seock-Won Youn,1 Giovanni Abbadessa,2 Yi Yu,2 Brian Schwartz,2 Robert K. Andrews,3 Victor R. Gordeuk4,5 and Jaehyung Cho1

Haematologica 2017 Volume 102(2):246-259

Department of Pharmacology, University of Illinois College of Medicine, Chicago, IL, USA; 2ArQule, Inc., Burlington, MA, USA; 3Australian Centre for Blood Diseases, Monash University, Melbourne, Australia; 4Section of Hematology/Oncology, University of Illinois College of Medicine, Chicago, IL, USA and 5Comprehensive Sickle Cell Center, University of Illinois College of Medicine, Chicago, IL, USA 1

ABSTRACT

P

Correspondence: thromres@uic.edu

Received: June 15, 2016. Accepted: October 5, 2016. Pre-published: October 6, 2016. doi:10.3324/haematol.2016.151159

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/246

revious studies identified the Ser/Thr protein kinase, AKT, as a therapeutic target in thrombo-inflammatory diseases. Here we report that specific inhibition of AKT with ARQ 092, an orally-available AKT inhibitor currently in phase Ib clinical trials as an anti-cancer drug, attenuates the adhesive function of neutrophils and platelets from sickle cell disease patients in vitro and cell-cell interactions in a mouse model of sickle cell disease. Studies using neutrophils and platelets isolated from sickle cell disease patients revealed that treatment with 50-500 nM ARQ 092 significantly blocks αMb2 integrin function in neutrophils and reduces P-selectin exposure and glycoprotein Ib/IX/V-mediated agglutination in platelets. Treatment of isolated platelets and neutrophils with ARQ 092 inhibited heterotypic cell-cell aggregation under shear conditions. Intravital microscopic studies demonstrated that short-term oral administration of ARQ 092 or hydroxyurea, a major therapy for sickle cell disease, diminishes heterotypic cell-cell interactions in venules of sickle cell disease mice challenged with tumor necrosis factor-α. Coadministration of hydroxyurea and ARQ 092 further reduced the adhesive function of neutrophils in venules and neutrophil transmigration into alveoli, inhibited expression of E-selectin and intercellular adhesion molecule-1 in cremaster vessels, and improved survival in these mice. Ex vivo studies in sickle cell disease mice suggested that co-administration of hydroxyurea and ARQ 092 efficiently blocks neutrophil and platelet activation and that the beneficial effect of hydroxyurea results from nitric oxide production. Our results provide important evidence that ARQ 092 could be a novel drug for the prevention and treatment of acute vasoocclusive complications in patients with sickle cell disease.

Introduction

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

246

Sickle cell disease (SCD) is an inherited blood disorder caused by a homozygous Glu6Val mutation at position 6 of b-globin, resulting in hemoglobin S (HbS). HbS is polymerized upon deoxygenation, resulting in sickling and hemolysis of red blood cells, endothelial cell activation, and chronic inflammation.1 In addition, there are several heterozygous forms of SCD,2 such as HbS/b0-thalassemia, which is often clinically similar to sickle cell anemia. The many clinical manifestations in SCD patients include recurrent vaso-occlusive episodes mediated by heterotypic cell-cell adhesion/aggregation, which cause pain crises and increase mortality due to organ damage and acute chest syndrome.3,4 Hydroxyurea, an important therapy for SCD, induces production of fetal hemoglobin and also has other beneficial effects, including increasing nitric oxide (NO) species and decreasing the level of soluble vascular haematologica | 2017; 102(2)


Effect of ARQ 092 on acute vaso-occlusion in SCD

cell adhesion molecule 1.5-7 Consistently, in vivo studies showed that intravenous infusion of hydroxyurea increases the level of plasma NO metabolites (NOx) and has beneficial effects on vaso-occlusive events in Berkeley mice, a model of SCD.8,9 However, SCD patients on hydroxyurea therapy often suffer from vaso-occlusive crises, suggesting that a novel or supplemental therapy is required. Intravital microscopy provided strong evidence that neutrophil-platelet interactions on activated endothelial cells can cause microvascular occlusion under thrombo-inflammatory conditions, including SCD and ischemia/reperfusion injury.9-12 Among several receptors and counter-receptors, the neutrophil-platelet association is primarily mediated by the interaction of neutrophil P-selectin glycoprotein ligand-1 (PSGL-1) and αMb2 integrin with platelet Pselectin and glycoprotein Ibα (GPIbα), respectively.13 We have shown that AKT2 positively regulates the function of αMb2 integrin and P-selectin during vascular inflammation12 and that combining hydroxyurea with AKT2 inhibition has immediate benefits in acute vaso-occlusive events and improves survival in SCD mice.9 Although these results suggest that AKT2 inhibition may be a supplemental therapy for SCD patients with vaso-occlusive crises, no AKT2-specific inhibitor is currently available in the clinic. As a Ser/Thr protein kinase, AKT regulates numerous cellular processes, such as cell growth, survival, and metabolism.14 Its activity is controlled by phosphorylation of the Thr308 and Ser473 residues by 3-phosphoinositide-dependent kinase 1 and mammalian target of rapamycin complex 2, respectively.15 Activated AKT then phosphorylates Ser/Thr residues on a variety of substrates.16 Despite 80% sequence homology of the three isoforms, each AKT isoform plays a partially overlapping but distinct role in platelet activation and aggregation.17-19 In neutrophils, which express AKT1 and AKT2, only AKT2 regulates cell migration, NADPH oxidase 2 activation, b2 integrin function, and neutrophil-platelet interactions under inflammatory conditions.12,20 As a major isoform in endothelial cells, AKT1 modulates the activity of endothelial NO synthase and is involved in angiogenesis, acute inflammation, and atherosclerosis.21-23 Human AKT isoforms share around 98% sequence homology with mouse proteins. These studies suggest the importance of each AKT isoform in the pathophysiology of vascular diseases and identify AKT as an attractive therapeutic target. Several AKT inhibitors are being developed as anti-cancer drugs.24,25 ARQ 092 has been reported to be an orallyavailable, highly-selective AKT inhibitor.26,27 Recent studies show that ARQ 092 blocks the activity of AKT1, AKT2, and AKT3 with IC50 values of 5.0, 4.5, and 16 nM, respectively, and has excellent selectivity (>1,000-fold) over other kinases.26 As an allosteric inhibitor, this compound blocks membrane translocation of inactive AKT and even dephosphorylates the membrane-associated active form, thereby perturbing AKT activity.26 Using cells and tissues isolated from patients with Proteus syndrome harboring AKT1E17K mutations, a previous study demonstrated effective inhibition of the mutant AKT1 by ARQ 092.27 This compound is currently in phase Ib clinical studies for the treatment of lymphoma, breast and endometrial cancers, and tumors with AKT or phosphoinositide 3-kinase (PI3K) mutations, and is well tolerated at a continuous daily dose of 60 mg or a dose of 600 mg when administered once a week, for several months.28 In the present study, we demonstrate that ARQ 092 haematologica | 2017; 102(2)

decreases the activation state of neutrophils and platelets isolated from SCD patients, thereby reducing platelet-neutrophil interactions in vitro. Furthermore, in vivo studies revealed that oral administration of hydroxyurea and ARQ 092 efficiently blocks neutrophil-endothelial cell and neutrophil-platelet interactions in venules and impairs neutrophil infiltration into the alveoli, thereby improving survival in SCD mice challenged with tumor necrosis factor (TNF)-α. Our results warrant further study of ARQ 092 in a clinical trial to treat acute vaso-occlusive crises in SCD patients.

Methods Mice Wild-type (C57BL/6, 6-week old, male and female), hemizygous [Tg(Hu-miniLCRα1 GγAγδbS) Hba-/- Hbb+/-] and Berkeley sickle [Tg(Hu-miniLCRα1 GγAγδbS) Hba-/- Hbb-/-] mice were obtained from The Jackson Laboratory (Bar Harbor, ME, USA). SCD mice (20-24 weeks old) were generated by transplantation of bone marrow cells isolated from Berkeley mice into lethally irradiated wildtype mice as described previously.9,29 Three to 4 months after transplantation, polymerase chain reaction and electrophoresis analyses showed that all chimeric mice expressed the transgene (human HbS) (Online Supplementary Figure S1). In contrast, mouse hemoglobin was not detected. These chimeric Berkeley mice are hereafter referred to as SCD mice. Both male and female SCD mice (20-24 weeks old) were used in this study. The University of Illinois Institutional Animal Care and Use Committee approved all animal care and experimental procedures.

Sickle cell disease patients Sixteen homozygous (HbSS) and six HbS/b0-thalassemia patients (20-52 years, 9 men and 13 women) who had not taken aspirin or ibuprofen within 5 days were included in our studies. None of the patients had been treated with hydroxyurea prior to blood donation. No significant differences were observed in the levels of surface markers of resting and stimulated platelets and neutrophils in patients with HbSS or HbS/b0-thalassemia. Blood from all patients was drawn at routine clinic visits without a pain crisis. Multiple experiments were performed using one patient’s blood sample and each experiment was repeated with blood from three or four different patients. All patients enrolled in this study provided informed consent. The collection and use of blood samples for laboratory analysis were approved by the Institutional Review Board of the University of Illinois at Chicago.

ARQ 092 preparation The process used to synthesize ARQ 092 and the compound’s molecular properties are described elsewhere.30

Isolation of neutrophils and platelets Platelets were isolated from mice and SCD patients as we have previously described.11 Platelets were suspended in HEPES-Tyrode buffer (20 mM HEPES, pH 7.3, 136 mM NaCl, 2.7 mM KCl, 12 mM NaHCO3, 1 mM MgCl2, and 5.5 mM glucose without CaCl2 and bovine serum albumin) at a concentration of 3 x 108 platelets/mL. Neutrophils were isolated from SCD patients’ blood and mouse bone marrow as described previously.12 The concentration of neutrophils was adjusted to 1x107 cells/mL in RPMI1640 medium. Human and mouse neutrophils were stimulated for 10 min at 37°C with 0.5 and 10 μM N-formylmethionyl-leucyl-phenylalanine (fMLP), respectively, and platelets were activated with 0.025 U/mL thrombin for 5 min at 37°C, unless otherwise stated. 247


K. Kim et al.

Intravital microscopy Male SCD mice were fasted overnight and treated with saline or 250 mg/kg of hydroxyurea (50 mg/mL) by oral gavage and subsequently with intraperitoneal injection of TNF-α (500 ng), 3 h prior to imaging. Phosphoric acid (0.01 M) or ARQ 092 (100 mg/10 mL/kg) was administered orally 30 min before imaging. Platelets and neutrophils were monitored via infusion of DyLight 488-conjugated anti-CD42c (0.1 μg/g body weight) and Alexa Fluor 647-conjugated anti-Ly-6G antibodies (0.05 μg/g body weight), respectively. Fluorescence and bright-field images were recorded using an Olympus BX61W microscope with a 60 × 1.0 NA water immersion objective and a Hamamatsu C9300 highspeed camera through an intensifier (Video Scope International, Sterling, VA, USA), and data were analyzed using Slidebook v6.0 (Intelligent Imaging Innovations, Denver, CO, USA). Real-time images were captured in the inflamed cremaster venules, which had a diameter of 25-40 μm. The rolling influx of neutrophils (rolling cells/min) and number of adherent neutrophils were determined over a 5-min period (number/field/5 min). Five to six different venules were monitored in each mouse. Since most platelets adhered to the top of adherent neutrophils,12 the kinetics of platelet accumulation was determined by the integrated median fluorescence intensities of the anti-CD42c antibody which were normalized to the number of adherent neutrophils and plotted over time.

Other methods Genotyping and chimerism analysis of the SCD mice, neutrophil-platelet aggregation assays, survival times, plasma NOx (nitrites/nitrates) level, flow cytometry, platelet aggregation/ agglutination assays, immunohistochemistry, reactive oxygen species (ROS) generation, and Ca2+ mobilization are described in detail in the Online Supplementary Methods.

Statistics Data were analyzed using GraphPad Prism 6 software by ANOVA with the Tukey test, Student t-test, and Mantle-Cox logrank test (survival curve). P values less than 0.05 were considered statistically significant.

Results ARQ 092 inhibits activation of neutrophils and platelets isolated from patients with sickle cell disease We previously reported that the basal levels of AKT phosphorylation are significantly higher in neutrophils and platelets isolated from SCD patients under stable conditions than in those from healthy donors.12 We found that AKT in human neutrophils and platelets is maximally phosphorylated 2 min after stimulation with fMLP or thrombin, respectively (Online Supplementary Figure S2A,B). Similar results were obtained with mouse neutrophils and platelets (Online Supplementary Figure S2C,D). To determine the inhibitory effect of ARQ 092 on AKT phosphorylation, neutrophils isolated from SCD patients were pretreated with ARQ 092 and further incubated with or without fMLP for 2 min, followed by immunoblotting. We observed that ARQ 092, at doses of 50 and 500 nM, markedly reduced phosphorylation of AKT but not of PI3K p85α/b or Src, following fMLP treatment (Figure 1A). Treatment of patients’ neutrophils with 50 and 500 nM ARQ 092 significantly inhibited the surface amount of αMb2 integrin following fMLP stimulation (Figure 1B). 248

Furthermore, binding of anti-activated αMb2 antibodies (CBRM1/5) was decreased by treatment with ARQ 092 (Figure 1C). Although 5 nM ARQ 092 caused a moderate and significant inhibition of AKT phosphorylation, no inhibitory effect on αMb2 integrin function was observed (data not shown). Consistent with impaired αMb2 integrin function, 50 or 500 nM ARQ 092 significantly inhibited binding of soluble fibrinogen, a ligand for αMb2 integrin, to fMLP-stimulated neutrophils (Figure 1D). The surface expression of PSGL-1 was not affected by 500 nM ARQ 092 (data not shown). These results suggest that specific AKT inhibition by ARQ 092 attenuates the membrane translocation and ligand-binding function of αMb2 integrin in stimulated neutrophils isolated from SCD patients. We also tested whether ARQ 092 inhibits the function of platelet surface molecules. We found that treatment with 50 and 500 nM ARQ 092 inhibited AKT phosphorylation in thrombin-stimulated platelets from SCD patients, without affecting PI3K or Src phosphorylation (Figure 1E). Treatment of patients’ platelets with 50 and 500 nM ARQ 092 significantly reduced P-selectin exposure, an indicator of α granule secretion, after stimulation with 0.025 U/mL thrombin (Figure 1F). As seen in neutrophils, while 5 nM ARQ 092 caused a significant but not complete inhibition of AKT phosphorylation, minimal inhibitory effects on Pselectin exposure were observed (data not shown). We found that treatment with ARQ 092 did not affect the surface expression of GPIbα but impaired platelet agglutination induced by binding of von Willebrand factor to GPIbα of the GPIb/IX/V complex (Figure 1G,I). This suggests that AKT inhibition impairs the ligand-binding function of GPIbα. In addition, thrombin-induced aggregation of patients’ platelets was also inhibited by 50 and 500 nM ARQ 092 (Figure 1J). However, a higher concentration of thrombin attenuated the inhibitory effect of ARQ 092 (Online Supplementary Figure S3A), implying that the increased concentration of thrombin induces other signaling pathways which are independent of AKT. Similar results were obtained with cross-linked collagen-related peptide, a GPVI-selective ligand (Online Supplementary Figure S3B,C). Our results indicate that specific AKT inhibition by ARQ 092 effectively blocks α-granule secretion and the adhesive function of activated platelets isolated from SCD patients.

ARQ 092 attenuates heterotypic aggregation of sickle cell patients’ neutrophils and platelets under stirring conditions Heterotypic cell-cell interactions can cause vaso-occlusion in SCD patients.13 Since ARQ 092 inhibited the function of surface molecules on neutrophils and platelets isolated from SCD patients, we investigated whether ARQ 092 affects heterotypic cell-cell aggregation in vitro. Neutrophils and platelets isolated from SCD patients aggregated under stirring conditions mimicking venous shear, creating a new cell population (R1 gate) in which most cells were positive for both L-selectin (a leukocyte marker) and CD41a (αIIb, a platelet marker) (Figure 2A). As quantified by the number of cell-cell aggregates (R3) in the R1 gate, pre-treatment of neutrophils or platelets with 500 nM ARQ 092 moderately but significantly decreased neutrophil-platelet aggregation (Figure 2A-B). When both cell types were treated with ARQ 092, the number of aggregates was further reduced. Since aggregated platelets without associated neutrophils are not detected in the R1 gate, haematologica | 2017; 102(2)


Effect of ARQ 092 on acute vaso-occlusion in SCD

A

B

C

D

E

G

F

H

I

J

Figure 1. ARQ 092 inhibits activation of neutrophils and platelets isolated from SCD patients in vitro. (A-D) Neutrophils or (E-J) platelets isolated from SCD patients were pretreated with vehicle (0.1% DMSO), or 50 or 500 nM ARQ 092 and then incubated with or without 0.5 μM fMLP or 0.025 U/mL thrombin for 2 min, respectively. (A) Immunoblotting was performed using equal amounts (50 μg) of neutrophil lysate protein, followed by densitometry (n = 3). (B-D) Flow cytometry was performed using phycoerythin-conjugated control IgG or antibodies against total (ICRF44) or activated αMb2 (CBRM1/5), or DyLight 488-conjugated fibrinogen. The geometric mean fluorescence intensity of antibodies was normalized to that of control IgG, and data are presented as fold increase compared with vehicle-treated, unstimulated cells. (E) Immunoblotting was performed using equal amounts (50 μg) of platelet lysate protein, followed by densitometry (n = 3). (F) Flow cytometry was performed to measure P-selectin exposure. (G-H) Platelets were stimulated with (G) 0.025 or (H) 0.05 U/mL thrombin. The surface amount of GPIbα was measured by flow cytometry. (I) Platelet agglutination was induced by 0.5 μg/mL vWF and 0.1 mg/mL ristocetin. (J) Platelet aggregation was induced by 0.025 U/mL thrombin. The representative agglutination or aggregation trace was obtained from three independent experiments. Data represent the mean ± SD (n = 3-4). *P<0.05, **P< 0.01, ***P<0.001, and ****P<0.0001 versus unstimulated vehicle control (or between two groups), ANOVA and the Tukey test.

haematologica | 2017; 102(2)

249


K. Kim et al.

we further measured the fluorescence intensities of antiCD41a antibodies in the gate. Pretreatment of neutrophils or platelets with ARQ 092 significantly inhibited the antibody signal, and the inhibitory effect was further increased when both cell types were treated with the inhibitor (Figure 2C). As a control, neutrophil-platelet aggregation was not affected by either anti-L-selectin or anti-CD41a antibodies at a concentration used for cell labeling or even at a 10-fold higher concentration (Online Supplementary Figure S4). Thus, these results show that ARQ 092 effectively blocks heterotypic aggregation of patients’ neutrophils and platelets in vitro.

ARQ 092 specifically inhibits AKT phosphorylation in neutrophils and platelets and reduces cell activation in sickle cell disease mice ex vivo Recent studies using mouse xenograft tumor models

demonstrated that oral administration of 75-100 mg/kg of ARQ 092 significantly delays tumor growth.26 Importantly, the authors showed that after oral administration of 100 mg/kg of ARQ 092 to mice, the maximum plasma concentration reached around 2 μM at 30 min and the plasma level remained around 300 nM at 8 h.26 Thus, we sought to determine the ex vivo effect of ARQ 092 on AKT phosphorylation in neutrophils and platelets isolated from SCD mice. Vehicle or 100 mg/kg of ARQ 092 was administered orally to the mice. Blood and bone marrow were collected, 30 min after treatment, to isolate platelets and neutrophils, respectively. No spontaneous bleeding was observed in the abdominal cavity of ARQ 092-treated SCD mice. As a control, ARQ 092 at 100 mg/kg did not alter the number of circulating blood cells in wild-type mice (Table 1) and had no inhibitory effect on splenomegaly in SCD mice (data not shown). We found that phosphorylation of AKT, but not of

Table 1. The number of circulating blood cells in wild-type mice after oral administration of ARQ 092.

Vehicle ARQ 092

WBC (103/μL)

NE (103/μL)

LY (103/μL)

MO (103/μL)

RBC (106/μL)

PLT (103/μL)

MPV (fL)

3.6 ± 0.4 2.7 ± 1.0

0.7 ± 0.2 0.6 ± 0.1

3.7 ± 1.1 2.9 ± 0.7

0.1 ± 0.0 0.1 ± 0.1

7.6 ± 1.5 7.4 ± 1.2

888 ± 116 853 ± 55.9

8.2 ± 1.7 9.2 ± 1.8

Vehicle (0.01 M phosphoric acid) or ARQ 092 (100 mg/kg) was administered orally to wild-type mice. Blood was collected 30 min after treatment and counted using an automated hematology analyzer (ADVIA 120, Siemens AG, Germany). Data represent the mean ± SD (n = 6 mice per group). WBC: white blood cells; NE: neutrophils; LY: lymphocytes; MO: monocytes; RBC: red blood cells; PLT: platelets; MPV: mean platelet volume.

A

B

250

C

Figure 2. ARQ 092 perturbs heterotypic aggregation of neutrophils and platelets from SCD patients under stirring conditions. Neutrophils and platelets isolated from SCD patients were pretreated with vehicle [0.1% DMSO (-)] or 500 nM ARQ 092 and then incubated with FITCconjugated anti-L-selectin and APC-conjugated anti-CD41a antibodies, respectively. Thrombin-activated platelets were mixed with neutrophils under stirring conditions, followed by flow cytometric analysis. R1, leukocyte-platelet aggregates; R2, neutrophils; and R3, the number of cell aggregates in the R1 gate. (B) Cell-cell aggregation was measured by the number of cell-cell aggregates (R3). (C) Heterotypic (neutrophilplatelet) interaction was measured by the fluorescence intensities of anti-CD41a antibodies in the R1 gate. Data represent the mean ± SD (n = 4). *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001 versus vehicle control (or between two groups), ANOVA and the Tukey test.

haematologica | 2017; 102(2)


Effect of ARQ 092 on acute vaso-occlusion in SCD

Src and PI3K, was disrupted in fMLP-stimulated neutrophils isolated from SCD mice treated with ARQ 092, compared to those treated with vehicle (Figure 3A). The membrane translocation of αMb2 integrin and soluble fibrinogen binding were significantly impaired in neutrophils from ARQ 092-treated mice (Figure 3B, C). Consistently, oral administration of ARQ 092 abrogated AKT phosphorylation in thrombin-activated platelets without affecting Src or PI3K phosphorylation (Figure 3D). P-selectin exposure and platelet aggregation were significantly reduced in platelets, activated with either thrombin or cross-linked collagen-related peptide, from ARQ 092-treated SCD mice, (Figure 3E,F). As a control, treatment of isolated neutrophils and platelets with 50 and 500 nM ARQ 092 in vitro specifi-

cally inhibited AKT phosphorylation following agonist stimulation (Online Supplementary Figure S5). In addition, treatment with ARQ 092 also inhibited AKT phosphorylation in TNF-α-stimulated neutrophils in vitro (Online Supplementary Figure S6). Together, these results indicate that oral administration of 100 mg/kg of ARQ 092 significantly attenuates the activation state of neutrophils and platelets in SCD mice by specific inhibition of AKT.

Oral administration of hydroxyurea and ARQ 092 reduces cell-cell interactions in sickle cell disease mice challenged with tumor necrosis factor-α A previous study showed that oral administration of 250 mg/kg of hydroxyurea partially inhibits leukocyte adhesion

B

A

C

E D

F

haematologica | 2017; 102(2)

Figure 3. Oral administration of ARQ 092 blocks AKT phosphorylation and activation of neutrophils and platelets isolated from SCD mice ex vivo. Vehicle (0.01 M phosphoric acid) or ARQ 092, 100 mg/kg, was given orally to SCD mice. Blood and bone marrow were collected 30 min after treatment. Isolated (A-C) neutrophils or (D-F) platelets were incubated with or without 10 μM fMLP or 0.025 U/mL thrombin for 2 min, respectively. (A and D) To determine the phosphorylation levels of Src, PI3K, and AKT, equal amounts (50 μg) of cell lysate protein were immunoblotted, followed by densitometry (n = 3). (B and C) The surface amount of αMb2 and soluble fibrinogen binding were measured by flow cytometry. (E) P-selectin exposure was measured by flow cytometry. (F) Platelet aggregation was induced by 0.025 U/mL thrombin or 0.25 μg/mL cross-liked collagen-felated peptide (CRP). The representative aggregation traces were obtained from two independent experiments. All other data represent the mean ± SD (n = 3). *P<0.05, ***P<0.001, and ****P<0.0001 versus unstimulated vehicle control (or between two groups), ANOVA and Tukey test.

251


K. Kim et al.

to the venules of SCD mice.8 Thus, we investigated whether oral administration of a single dose of ARQ 092 (100 mg/kg) or co-administration with hydroxyurea (250 mg/kg) affects neutrophil-endothelial cell and neutrophil-platelet interactions in the cremaster venules of SCD mice challenged with TNF-α which induces acute vaso-occlusive events in the mice.8,29 Due to the different pharmacokinetics and mechanisms of action,8,26,31 SCD mice were treated with saline or hydroxyurea by oral gavage prior to intraperitoneal injection of TNF-α. For ARQ 092, vehicle or the compound was given orally 2.5 h after the TNF-α injection, followed by surgical procedures (Figure 4A). No spontaneous bleeding was

observed at the surgical site in any of the mice. We found that compared to the vehicle control, treatment with hydroxyurea or ARQ 092 alone significantly decreased the number of adherent neutrophils with a minimal increase in the rolling influx (Figure 4B-D, Online Supplementary Videos 14). No differences were observed between the effects of the two vehicle controls. Compared to hydroxyurea or ARQ 092 treatment, co-administration of both hydroxyurea and ARQ 092 significantly enhanced the rolling influx and decreased the number of adherent neutrophils in the venules (Online Supplementary Video 5). To investigate platelet-neutrophil interactions, we measured the fluorescence intensi-

C A

D

B

E

G

F Figure 4 (A-G). Oral administration of hydroxyurea and ARQ 092 has numerous beneficial effects in TNF-α-challenged SCD mice. Intravital microscopy was performed as described in the Methods. Neutrophils and platelets were labeled by infusion of Alexa Fluor 647-conjugated anti-Ly-6G and DyLight 488-conjugated anti-CD42c antibodies, respectively. (A) Timeline for each treatment and surgery for intravital microscopy (IVM) in SCD mice. S: saline; HU: hydroxyurea; PA: 0.01 M phosphoric acid. (B) Representative images at various time-points. The time “0” was set as the image capture was initiated for each vessel. (C, D) Number of rolling (for 1 min) and adherent neutrophils (for 5 min). Data represent the mean ± SEM (n = 7-8 mice). (E) The integrated median fluorescence intensities of anti-CD42c antibodies (F platelets) were plotted over time. (F) Survival curves. Survival was significantly improved in the groups treated with hydroxyurea alone (P = 0.0028) or both hydroxyurea and ARQ 092 (P=0.0017), compared to each vehicle control. Mantel-Cox log-rank test. (G) Plasma NOx levels were measured as described in the Methods. Data represent the mean ± SD (n = 7-8 mice per group).

252

haematologica | 2017; 102(2)


Effect of ARQ 092 on acute vaso-occlusion in SCD

ties of anti-CD42c antibodies. Treatment with hydroxyurea, ARQ 092 or both abrogated platelet-neutrophil interactions, compared to vehicle controls (Figure 4E). These results suggest that oral administration of both hydroxyurea and ARQ 092 efficiently inhibits acute vaso-occlusive events in TNFα-challenged SCD mice.

Hydroxyurea alone or with ARQ 092, but not ARQ 092 alone, prolongs survival times of sickle cell disease mice challenged with tumor necrosis factor-α We and others reported that a surgical procedure to expose the cremaster muscle after intraperitoneal injection of TNF-α in SCD mice leads to death within several hours

as a result of acute vaso-occlusive events.8,9,29 Compared to the vehicle controls (saline, S; 0.01 M phosphoric acid, PA), treatment with hydroxyurea or both hydroxyurea and ARQ 092 significantly prolonged survival times in TNF-αchallenged SCD mice, whereas treatment with ARQ 092 alone did not improve survival (S versus hydroxyurea, P=0.0028; PA versus hydroxyurea + ARQ 092, P=0.0017; ARQ 092 versus hydroxyurea + ARQ 092, P=0.0087; and hydroxyurea versus hydroxyurea + ARQ 092, P=0.31) (Figure 4F). Fifty percent of SCD mice treated with vehicle (S), hydroxyurea, vehicle (PA), ARQ 092, and both hydroxyurea and ARQ 092 died at 4.0, 5.5, 4.0, 4.1, and >6 hours, respectively, after TNF-α injection.

Figure 4 continued (H-M). Oral administration of hydroxyurea and ARQ 092 has numerous beneficial effects in TNF-α-challenged SCD mice. (H-K) After recording survival times, the cremaster muscle was taken out for immunohistochemistry. Muscle sections were labeled with control IgG or rat monoclonal antibodies against E-selectin or ICAM-1 and then with DyLight 488-conjugated anti-rat IgG antibodies, followed by incubation with APC-conjugated anti-PECAM-1 antibodies and a mounting reagent containing DAPI. (H and J) Representative images. Bar = 10 μm. (I and K) The fluorescence intensities of antibodies. Data represent the mean ± SD (n = 18-22 vessels in 5 mice per group). (L-M) After recording survival times, lungs were taken out for histochemistry. Neutrophils were stained with naphthol AS-D chloroacetate. (L) Representative images. (M) The number of transmigrated neutrophils (arrow heads) was quantified in the field of view (110 mm2). Data represent the mean ± SD (n = 36-42 sections in 5-6 mice per group). *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001 versus each vehicle control (or between two groups), ANOVA and the Tukey test. S: saline; HU: hydroxyurea; PA: phosphoric acid.

J

H

K

I

L

haematologica | 2017; 102(2)

M

253


K. Kim et al.

Hydroxyurea alone, but not with ARQ 092, enhances plasma levels of nitric oxide metabolites in sickle cell disease mice challenged with tumor necrosis factor-α We recently reported that a single intravenous infusion of hydroxyurea significantly enhances plasma NOx levels in hemizygous control (Hbb+/-) and Berkeley (Hbb-/-) mice.9 We further measured the plasma NOx levels in TNF-α-challenged SCD mice after oral administration of hydroxyurea, ARQ 092, or both. The plasma NOx levels were significantly elevated in the mice after oral administration of hydroxyurea, but not ARQ 092, compared to the vehicle control (Figure 4G). Co-administration of hydroxyurea and ARQ 092 increased the plasma NOx level by 1.3-fold relative to the vehicle (PA) control, but this increase was not significantly different from that following administration of vehicle controls or hydroxyurea alone.

Co-administration of hydroxyurea and ARQ 092 downregulates the expression of E-selectin and intercellular adhesion molecule-1 in sickle cell disease mice challenged with tumor necrosis factor-α E-selectin and intercellular adhesion molecule-1 (ICAM1) expressed on activated endothelial cells are required for neutrophil rolling and adhesion, respectively, during vascu-

A

D

lar inflammation.32 Immunohistochemistry of cremaster muscle sections showed that, compared to the vehicle control, treatment with hydroxyurea or both hydroxyurea and ARQ 092, but not ARQ 092 alone, significantly reduced the expression of E-selectin (Figure 4H,I). ICAM-1 expression was significantly decreased in mice treated with hydroxyurea, ARQ 092, or both, compared to each vehicle control (Figure 4J,K). These results imply that, unlike AKT2-specific inhibition which perturbs expression of both E-selectin and ICAM-1,9 inhibition of all AKT isoforms may have additional effects on vascular endothelial cells of TNF-α-challenged SCD mice.

Co-administration of hydroxyurea and ARQ 092 efficiently blocks neutrophil transmigration in sickle cell disease mice Neutrophils rapidly transmigrate from the pulmonary microvasculature and cause lung injury during inflammation. We found that oral administration of hydroxyurea or ARQ 092 significantly blocked neutrophil transmigration into the alveoli of TNF-α-challenged SCD mice (Figure 4L,M). Co-administration of hydroxyurea and ARQ 092, compared to hydroxyurea or ARQ 092 alone, further decreased the number of transmigrated neutrophils. These results suggest that hydroxyurea with AKT inhibition effi-

C

B

E

F

Figure 5. Co-administration of hydroxyurea and ARQ 092 efficiently inhibits αMb2 integrin function and ROS generation in stimulated neutrophils isolated from SCD mice ex vivo. Vehicle (0.01 M phosphoric acid, -), hydroxyurea (HU), ARQ 092, or both were given orally to TNF-α-challenged SCD mice as described in Figure 4A. Bone marrow was collected at 3 h after TNF-α injection. Isolated neutrophils were incubated with or without fMLP. (A, B) The surface level of αMb2 integrin and soluble fibrinogen binding were measured by flow cytometry. (C) Neutrophils were pretreated with Ca2+ dye and incubated with fMLP for 200 s and 2 mM CaCl2 was then added. A representative graph was obtained from three independent experiments. (D) Neutrophils were incubated with DCFH-DA prior to fMLP stimulation. Intracellular ROS generation was measured by the DCF signal in flow cytometry. (E, F) Neutrophils were mixed with Amplex red reaction solution and treated with or without fMLP. Extracellular H2O2 was detected by absorbance (560 nm) as described in the Methods. The quantification graph was obtained at the 15 min mark of the reaction. Data represent the mean ± SD (n = 45). *P<0.05, **P<0.01, and ***P<0.001 versus unstimulated vehicle control (or between two groups), ANOVA and the Tukey test.

254

haematologica | 2017; 102(2)


Effect of ARQ 092 on acute vaso-occlusion in SCD

ciently reduces lung inflammation in TNF-α-challenged SCD mice.

Co-administration of hydroxyurea and ARQ 092 efficiently blocks numerous functions of neutrophils and platelets isolated from sickle cell disease mice challenged with tumor necrosis factor-α ex vivo We further examined which neutrophil and platelet functions are efficiently inhibited by both drugs. Hydroxyurea, ARQ 092, or both were given orally to TNFα-challenged SCD mice as described in Figure 4A without surgery. Blood and bone marrow were collected 3 h after TNF-α treatment to isolate platelets and neutrophils, respectively. We found that following stimulation of neutrophils with fMLP, the surface amount of αMb2 integrin and soluble fibrinogen binding were significantly inhibited by ARQ 092 or both hydroxyurea and ARQ 092, but not hydroxyurea alone, compared with vehicle control (Figure 5A,B). Although studies showed that an elevation in cytosolic Ca2+ levels is critical for neutrophil activation and that deletion or inhibition of neutrophil AKT2 impairs Ca2+ mobilization following fMLP stimulation,33 Ca2+ release and influx in fMLP-stimulated neutrophils from SCD mice were not affected by any treatment (Figure 5C). Since hydroxyurea produces NO species protecting against oxidative stress5,6 and neutrophil AKT2 is important for ROS generation by activating the NADPH oxidase 2 complex,20 we also measured ROS generation. As measured by the DCF signal, intracellular ROS generation was not reduced in fMLP-stimulated neutrophils isolated from TNF-α-challenged SCD mice after oral administration of hydroxyurea or ARQ 092 alone (Figure 5D). When the mice were treated with both drugs, however, the DCF signal was significantly decreased in stimulated neutrophils. Similar results were obtained using Amplex red which detects extracellular H2O2 (Figure 5E,F). We further investigated the combined effect of hydroxyurea and ARQ 092 on platelet function ex vivo. P-selectin exposure was significantly reduced in thrombin-activated platelets isolated from TNF-α-challenged SCD mice treated with hydroxyurea or ARQ 092 (Figure 6A). The inhibitory effect was further enhanced in platelets from the mice treated with both drugs. In platelets, unlike neutrophils, we observed that treatment with hydroxyurea, ARQ 092, or both equally inhibited Ca2+ influx with little effect on Ca2+ release during thrombin activation (Figure 6B-D). We also found that oral administration of hydroxyurea or ARQ 092 did not affect the surface expression of GPIbα but impaired von Willebrand factor-mediated agglutination of platelets isolated from TNF-α-challenged SCD mice (Online Supplementary Figure S7 and Figure 6E). Coadministration of both drugs slightly increased the inhibitory effect on agglutination. Similar results were obtained from platelet aggregation assays (Figure 6F). Moreover, oral administration of hydroxyurea, ARQ 092, or both markedly reduced the generation of intracellular ROS in thrombin-activated platelets isolated from TNF-αchallenged SCD mice (Figure 6G). Extracellular H2O2 generation was weakly decreased by hydroxyurea or ARQ 092 but significantly impaired by both drugs (Figure 6H,I). Although the precise mechanism(s) by which hydroxyurea inhibits platelet functions remains to be determined, our results suggest that the combination therapy of hydroxyurea and ARQ 092 efficiently inhibits numerous platelet functions in TNF-α-challenged SCD mice. haematologica | 2017; 102(2)

The short-term beneficial effect of hydroxyurea results from nitric oxide production Previous studies suggested that the inhibitory effect of hydroxyurea on cell-cell interactions results from NO production in vivo.8 Thus, we sought to determine whether hydroxyurea-generated NO is important for the inhibitory mechanism of the drug. Using water-soluble carboxy-PTIO (a NO scavenger that has no effect on NO synthase activity),34 we repeated our ex vivo studies shown in Figures 5 and 6. We found that PTIO treatment itself did not affect the membrane translocation and ligand binding of αMb2 integrin in fMLP-stimulated neutrophils (Figure 7A, B) or Pselectin exposure in thrombin-activated platelets (Figure 7C). In neutrophils, hydroxyurea alone had a small effect on the function of αMb2 integrin, and PTIO treatment nullified the effect of hydroxyurea, but not ARQ 092, on αMb2 membrane translocation and fibrinogen binding (Figure 7A,B). The potentiated inhibitory effect of both drugs was restored with PTIO treatment to a level similar to the inhibitory effect of ARQ 092 alone. We also found that administration of PTIO reversed the inhibitory effect of hydroxyurea, but not ARQ 092, on P-selectin exposure in activated platelets (Figure 7C). The synergistic effect of both hydroxyurea and ARQ 092 was abolished by PTIO treatment, and the inhibitory effect of both drugs with PTIO treatment was similar to that of ARQ 092 alone. Thus, these results imply that the synergistic effects of both hydroxyurea and ARQ 092 on neutrophil-platelet interactions result from two different signaling pathways: direct NO production by hydroxyurea and AKT inhibition by ARQ 092.

Discussion ARQ 092 is an orally-available, selective AKT inhibitor which is currently in phase Ib clinical trials for the treatment of certain cancers.28 In the present study, we showed that ARQ 092 reduces the adhesive function of neutrophils and platelets isolated from SCD patients in vitro. Importantly, oral administration of a single dose of ARQ 092 abrogated AKT phosphorylation in isolated neutrophils and platelets following agonist stimulation, significantly reduced cell-cell interactions in cremaster venules, and decreased neutrophil transmigration into the alveoli of TNF-α-challenged SCD mice. The inhibitory effects and survival were increased when the mice were pretreated with hydroxyurea. Thus, our studies provide important evidence that a specific inhibitor of AKT may be beneficial to treat acute vaso-occlusive events in SCD patients. Since AKT signaling is crucial for the function of intravascular cells during numerous vascular diseases,12,17-21,23 ARQ 092 is likely to inhibit the activity of all AKT isoforms in intravascular cells and thereby attenuate the process of thrombosis and inflammation in SCD patients. In addition to sickled red blood cells and activated endothelial cells, activation and adhesion of neutrophils and platelets contribute to the vaso-occlusive complications of SCD.35 Our study shows that ARQ 092 reduces the ligand-binding function of αMb2 integrin in stimulated neutrophils isolated from SCD patients. Importantly, we found that ARQ 092 inhibits αMb2 integrin function in neutrophils isolated from SCD mice after oral administration of the inhibitor. Although oral administration of hydroxyurea or ARQ 092 alone did not affect intracellular 255


K. Kim et al. B

A

D

C

E

F

H

G

I

Figure 6. Co-administration of hydroxyurea and ARQ 092 efficiently inhibits numerous functions of activated platelets isolated from SCD mice ex vivo. Vehicle (0.01 M phosphoric acid, -) or drugs were given orally to TNF-α-challenged SCD mice as described in Figure 5. Blood was collected 3 h after TNF-α injection. Isolated platelets were treated with or without thrombin. (A) Flow cytometry was performed to measure P-selectin exposure. (B-D) Platelets were pretreated with Ca2+ dye and incubated with thrombin for 5 min and 2 mM CaCl2 was then added. Ca2+ release (C) and influx (D) were measured and quantified by the AUC (area under the curve, arbitrary units). (E) Platelet agglutination was induced by 10 μg/mL vWF and 10 μg/mL botrocetin. (F) Platelet aggregation was induced by thrombin. The representative agglutination or aggregation trace was obtained from three independent experiments. (G) Platelets were incubated with DCFH-DA prior to thrombin stimulation. Intracellular ROS generation was measured by the DCF signal in flow cytometry. (H-I) Platelets were mixed with Amplex red reaction solution and treated with or without thrombin. Extracellular H2O2 was detected by absorbance (560 nm) as described in the Methods. The quantification graph was obtained at the 30 min mark of the reaction. Data represent the mean ± SD (n = 3-4). *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001 versus unstimulated vehicle control (or between two groups), ANOVA and the Tukey test.

256

haematologica | 2017; 102(2)


Effect of ARQ 092 on acute vaso-occlusion in SCD

A

C

B

Figure 7. Hydroxyurea (HU) reduces neutrophil and platelet activation via NO production. SCD mice were pretreated by intravenous injection of PTIO, a NO scavenger (1 mg/kg) 30 min prior to TNF-α challenge. Vehicle (0.01 M phosphoric acid, -) or drugs were given orally to TNF-α-challenged SCD mice as described in Figure 5. Blood and bone marrow were collected 3 h after TNF-α injection. Isolated (A, B) neutrophils and (C) platelets were treated with or without fMLP (for 10 min) or thrombin (for 5 min), respectively. Flow cytometry was performed to measure (A, B) the surface level of αMb2 integrin and soluble fibrinogen binding in stimulated neutrophils and (C) P-selectin exposure in stimulated platelets. Data represent the mean ± SD (n = 3). *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001 versus unstimulated vehicle control (or between two groups), ANOVA and the Tukey test.

ROS generation in fMLP-stimulated neutrophils isolated from the mice, co-administration of hydroxyurea and ARQ 092 significantly decreased ROS generation, suggesting that combined therapy may efficiently attenuate oxidative stress conditions in SCD patients. Consistent with previous reports showing that the activation and adhesive function of platelets is regulated by AKT,17-19,36 we found that ARQ 092 significantly impairs P-selectin exposure and GPIbα-mediated agglutination in platelets from SCD patients in vitro and in SCD mouse platelets ex vivo. In support of the importance of neutrophil αMb2 integrin and platelet P-selectin and GPIbα for neutrophil-platelet interactions,13 we observed that specific AKT inhibition in both neutrophils and platelets isolated from SCD patients effectively decreases heterotypic cell-cell aggregation. Thus, our results suggest that co-administration of hydroxyurea and ARQ 092 is beneficial to inhibit cell-cell interactions during vaso-occlusion in SCD. Most AKT inhibitors, including ARQ 092, are being developed as anti-cancer drugs.25 Given the fact that some cancer patients are at high risk of thrombogenesis,37 our results also provide indirect evidence that ARQ 092 may be an effective drug in cancer patients with thrombotic complications. It is believed that leukocyte adhesion to endothelial cells initially mediates cell-cell interactions and aggregation in the vessels of SCD patients and induces the inflammatory process.35 Our studies demonstrate that oral administration of a single dose of both hydroxyurea and ARQ 092 efficiently inhibits neutrophil-endothelial cell and neutrophilplatelet interactions in cremaster venules and reduces neutrophil recruitment into the alveoli of TNF-α-challenged haematologica | 2017; 102(2)

SCD mice. Interestingly, survival was improved by hydroxyurea alone or both hydroxyurea and ARQ 092, but not ARQ 092 alone. We previously reported that intravenous infusion of an AKT2 inhibitor decreases E-selectin and ICAM-1 expression on cremaster vessels of TNF-αchallenged SCD mice.9 However, oral administration of ARQ 092 resulted in a significant reduction in the expression of ICAM-1, but not E-selectin. Although we cannot eliminate the possibility of off-target effects of these AKT inhibitors, these results imply that inhibition of endothelial cell AKT1 and/or AKT3 may regulate E-selectin expression during inflammation. Furthermore, our findings that ARQ 092 alone significantly decreases cell-cell interactions but has no benefit on survival in TNF-α-challenged SCD mice may limit usage of this inhibitor as a supplement to hydroxyurea therapy in SCD patients. Oxidative stress results in reduced NO bioavailability in SCD patients and mice, which in turn aggravates inflammatory conditions.38,39 Hydroxyurea has many beneficial effects in SCD patients. Importantly, preclinical and clinical studies demonstrated that hydroxyurea increases the plasma level of NO species and stimulates a cGMP-signaling pathway.5,6,8,9 As seen in SCD mice after intravenous infusion of hydroxyurea,9 plasma NOx levels were also elevated after oral administration of hydroxyurea to SCD mice. Although combination therapy with hydroxyurea and ARQ 092 showed beneficial effects on acute vasoocclusive events and survival in TNF-α-challenged SCD mice, the plasma NOx levels increased by hydroxyurea treatment seemed to be slightly decreased by the dual therapy (hydroxyurea versus hydroxyurea + ARQ 092¸ 257


K. Kim et al.

P=0.067, Student t-test.). Studies showed that endothelial cell AKT1 plays an important role in acute inflammation and angiogenesis, which is associated with phosphorylation of endothelial NO synthase.21,23,40,41 Although ARQ 092 may inhibit endothelial cell AKT1-NO synthase signaling, the plasma NOx level was not changed by ARQ 092 treatment alone in TNF-α-challenged SCD mice, implying that endothelial NO synthase-derived NO production may not be important for maintaining the basal level of plasma NOx in SCD mice. Alternatively, our finding that a NO scavenger, PTIO, does not alter the inhibitory effect of ARQ 092 on platelet and neutrophil function (Figure 7) suggests that ARQ 092 is unlikely to affect cellular NO generation. Since hydroxyurea serves as a NO donor, further studies are required to determine how ARQ 092 influences the production of hydroxyurea-induced NOx in TNF-αchallenged SCD mice. In addition to hydroxyurea, many drugs targeting cell adhesion, inflammation, induction of fetal hemoglobin, coagulation, or platelet activation/aggregation are currently in clinical trials for the treatment of vaso-occlusive crises in SCD patients.35,42 In particular, preclinical and clinical studies with rivipansel (GMI-1070, a pan-selectin inhibitor) revealed that inhibition of leukocyte-endothelial cell interactions attenuates vaso-occlusion and improves survival in SCD mice,29 and reduces time to resolution of vaso-occlusive events and requirement for opioid analgesia in SCD patients.43 Nevertheless, complete inhibition of leukocyte-endothelial cell interactions could cause side effects by disrupting immune responses. We found that oral administration of ARQ 092 significantly,

References 1. Stuart MJ, Nagel RL. Sickle-cell disease. Lancet. 2004;364(9442):1343-1360. 2. Habara A, Steinberg MH. Minireview: genetic basis of heterogeneity and severity in sickle cell disease. Exp Biol Med (Maywood). 2016;241(7):689-696. 3. Manwani D, Frenette PS. Vaso-occlusion in sickle cell disease: pathophysiology and novel targeted therapies. Blood. 2013;122 (24):3892-3898. 4. Potoka KP, Gladwin MT. Vasculopathy and pulmonary hypertension in sickle cell disease. Am J Physiol Lung Cell Mol Physiol. 2015;308(4):L314-324. 5. Cokic VP, Smith RD, Beleslin-Cokic BB, et al. Hydroxyurea induces fetal hemoglobin by the nitric oxide-dependent activation of soluble guanylyl cyclase. J Clin Invest. 2003;111(2):231-239. 6. Nahavandi M, Wyche MQ, Perlin E, Tavakkoli F, Castro O. Nitric oxide metabolites in sickle cell anemia patients after oral administration of hydroxyurea; hemoglobinopathy. Hematology. 2000;5 (4):335-339. 7. Saleh AW, Hillen HF, Duits AJ. Levels of endothelial, neutrophil and platelet-specific factors in sickle cell anemia patients during hydroxyurea therapy. Acta Haematol. 1999;102(1):31-37. 8. Almeida CB, Scheiermann C, Jang JE, et al. Hydroxyurea and a cGMP-amplifying agent have immediate benefits on acute vasoocclusive events in sickle cell disease mice. Blood. 2012;120(14):2879-2888.

258

but not completely, inhibited cell-cell interactions in microvessels of TNF-α-challenged SCD mice. Importantly, when combined with hydroxyurea, ARQ 092 showed synergistic effects on acute vaso-occlusive events and improved survival of the mice. In addition, our studies demonstrated that 50-500 nM ARQ 092 is efficacious in inhibiting platelet and neutrophil functions. In cancer patients, the plasma concentration of ARQ 092 reached 2.6, 6.4, and 8.1 µM after oral administration of 20, 40, and 60 mg per day, respectively, for 15 days (unpublished data). Thus, a minimal oral dose of ARQ 092 may be sufficient to attenuate vaso-occlusive events in SCD patients. We reported that neutrophils and platelets isolated from SCD patients, compared to healthy donor cells, exhibit a significant increase in the basal level of AKT phosphorylation without altering its expression.12 Since the AKT signaling pathway is aberrantly activated in many types of cancers,44 our studies provide strong evidence that in addition to anti-cancer therapy, ARQ 092 has the potential to be developed for the treatment of acute vaso-occlusive episodes in SCD patients, possibly in combination with hydroxyurea. Acknowledgements The authors thank Dr. Lewis Hsu for his helpful comments. This work was in part supported by grants from the National Institutes of Health, National Heart, Lung, and Blood Institute (HL109439 and HL130028, JC), American Society of Hematology Bridge Fund (JC) and Scholar Award (JL), and American Heart Association postdoctoral (KK) and predoctoral fellowship (AT).

9. Barazia A, Li J, Kim K, Shabrani N, Cho J. Hydroxyurea with AKT2 inhibition decreases vaso-occlusive events in sickle cell disease mice. Blood. 2015;126(22):2511-2517. 10. Hidalgo A, Chang J, Jang JE, et al. Heterotypic interactions enabled by polarized neutrophil microdomains mediate thromboinflammatory injury. Nat Med. 2009;15(4):384-391. 11. Kim K, Li J, Tseng A, Andrews RK, Cho J. NOX2 is critical for heterotypic neutrophilplatelet interactions during vascular inflammation. Blood. 2015;126(16):1952-1964. 12. Li J, Kim K, Hahm E, et al. Neutrophil AKT2 regulates heterotypic cell-cell interactions during vascular inflammation. J Clin Invest. 2014;124(4):1483-1496. 13. Li J, Kim K, Barazia A, Tseng A, Cho J. Platelet-neutrophil interactions under thromboinflammatory conditions. Cell Mol Life Sci. 2015;72(14):2627-2643. 14. Hers I, Vincent EE, Tavare JM. Akt signalling in health and disease. Cell Signal. 2011;23(10):1515-1527. 15. Yang WL, Wu CY, Wu J, Lin HK. Regulation of Akt signaling activation by ubiquitination. Cell Cycle. 2010;9(3):487-497. 16. Woulfe DS. Akt signaling in platelets and thrombosis. Expert Rev Hematol. 2010;3 (1):81-91. 17. Chen J, De S, Damron DS, et al. Impaired platelet responses to thrombin and collagen in AKT-1-deficient mice. Blood. 2004;104 (6):1703-1710. 18. Woulfe D, Jiang H, Morgans A, et al. Defects in secretion, aggregation, and thrombus formation in platelets from mice lacking Akt2. J

Clin Invest. 2004;113(3):441-450. 19. O'Brien KA, Stojanovic-Terpo A, Hay N, Du X. An important role for Akt3 in platelet activation and thrombosis. Blood. 2011;118 (15):4215-4223. 20. Chen J, Tang H, Hay N, Xu J, Ye RD. Akt isoforms differentially regulate neutrophil functions. Blood. 2010;115(21):4237-4246. 21. Di Lorenzo A, Fernandez-Hernando C, Cirino G, Sessa WC. Akt1 is critical for acute inflammation and histamine-mediated vascular leakage. Proc Natl Acad Sci USA. 2009;106(34):14552-14557. 22. Fernandez-Hernando C, Ackah E, Yu J, et al. Loss of Akt1 leads to severe atherosclerosis and occlusive coronary artery disease. Cell Metab. 2007;6(6):446-457. 23. Lee MY, Luciano AK, Ackah E, et al. Endothelial Akt1 mediates angiogenesis by phosphorylating multiple angiogenic substrates. Proc Natl Acad Sci USA. 2014;111(35):12865-12870. 24. Pal SK, Reckamp K, Yu H, Figlin RA. Akt inhibitors in clinical development for the treatment of cancer. Expert Opin Investig Drugs. 2010;19(11):1355-1366. 25. Nitulescu GM, Margina D, Juzenas P, et al. Akt inhibitors in cancer treatment: the long journey from drug discovery to clinical use (Review). Int J Oncol. 2016;48(3):869-885. 26. Yu Y, Savage RE, Eathiraj S, et al. Targeting AKT1-E17K and the PI3K/AKT pathway with an allosteric AKT inhibitor, ARQ 092. PLoS One. 2015;10(10):e0140479. 27. Lindhurst MJ, Yourick MR, Yu Y, et al. Repression of AKT signaling by ARQ 092 in cells and tissues from patients with Proteus

haematologica | 2017; 102(2)


Effect of ARQ 092 on acute vaso-occlusion in SCD

syndrome. Sci Rep. 2015;5:17162. 28. Tolcher A, Harb W, Sachdev JC, et al. Results from a phase 1 study of ARQ 092, a novel pan AKT-inhibitor, in subjects with advanced solid tumors or recurrent malignant lymphoma. Eur J Cancer. 2015;51:S66. 29. Chang J, Patton JT, Sarkar A, et al. GMI1070, a novel pan-selectin antagonist, reverses acute vascular occlusions in sickle cell mice. Blood. 2010;116(10):1779-1786. 30. Lapierre JM, Eathiraj S, Vensel D, et al. Discovery of 3-(3-(4-(1-aminocyclobutyl)phenyl)-5-phenyl-3H-imidazo[4,5b]pyridin-2-yl)pyridin -2-amine (ARQ 092): an orally bioavailable, selective, and potent allosteric AKT inhibitor. J Med Chem. 2016;59(13):6455-6469. 31. Gwilt PR, Tracewell WG. Pharmacokinetics and pharmacodynamics of hydroxyurea. Clin Pharmacokinet. 1998;34(5):347-358. 32. Phillipson M, Kubes P. The neutrophil in vascular inflammation. Nat Med. 2011;17(11): 1381-1390. 33. Li J, Kim K, Hahm E, et al. Neutrophil AKT2 regulates heterotypic cell-cell interactions

haematologica | 2017; 102(2)

34.

35.

36.

37. 38.

39.

during vascular inflammation. J Clin Invest. 2014;124(4):1483-1496. Akaike T, Yoshida M, Miyamoto Y, et al. Antagonistic action of imidazolineoxyl Noxides against endothelium-derived relaxing factor/.NO through a radical reaction. Biochemistry. 1993;32(3):827-832. Zhang D, Xu C, Manwani D, Frenette PS. Neutrophils, platelets, and inflammatory pathways at the nexus of sickle cell disease pathophysiology. Blood. 2016;127(7):801809. Yin H, Stojanovic A, Hay N, Du X. The role of Akt in the signaling pathway of the glycoprotein Ib-IX induced platelet activation. Blood. 2008;111(2):658-665. Falanga A, Russo L, Verzeroli C. Mechanisms of thrombosis in cancer. Thromb Res. 2013;131 (Suppl 1):S59-62. Dasgupta T, Fabry ME, Kaul DK. Antisickling property of fetal hemoglobin enhances nitric oxide bioavailability and ameliorates organ oxidative stress in transgenic-knockout sickle mice. Am J Physiol Regul Integr Comp Physiol. 2010;298(2): R394-402. Reiter CD, Wang X, Tanus-Santos JE, et al.

40.

41.

42. 43.

44.

Cell-free hemoglobin limits nitric oxide bioavailability in sickle-cell disease. Nat Med. 2002;8(12):1383-1389. Ackah E, Yu J, Zoellner S, et al. Akt1/protein kinase Balpha is critical for ischemic and VEGF-mediated angiogenesis. J Clin Invest. 2005;115(8):2119-2127. Bauer PM, Fulton D, Boo YC, et al. Compensatory phosphorylation and protein-protein interactions revealed by loss of function and gain of function mutants of multiple serine phosphorylation sites in endothelial nitric-oxide synthase. J Biol Chem. 2003;278(17):14841-14849. Telen MJ. Beyond hydroxyurea: new and old drugs in the pipeline for sickle cell disease. Blood. 2016;127(7):810-819. Telen MJ, Wun T, McCavit TL, et al. Randomized phase 2 study of GMI-1070 in SCD: reduction in time to resolution of vaso-occlusive events and decreased opioid use. Blood. 2015;125(17):2656-2664. Hennessy BT, Smith DL, Ram PT, Lu Y, Mills GB. Exploiting the PI3K/AKT pathway for cancer drug discovery. Nat Rev Drug Discov. 2005;4(12):988-1004.

259


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Iron Metabolism & Its Disorders

Ferrata Storti Foundation

Haematologica 2017 Volume 102(2):260-270

Hemolytic anemia repressed hepcidin level without hepatocyte iron overload: lesson from Günther disease model Sarah Millot,1,2,3,4* Constance Delaby,1,3,5* Boualem Moulouel,1,3,4 Thibaud Lefebvre,1,3,4,6 Nathalie Pilard,1,3 Nicolas Ducrot,1,3,4 Cécile Ged,7 Philippe Lettéron,1,3 Lucia de Franceschi,8 Jean Charles Deybach,1,3,4,5 Carole Beaumont,1,3,4 Laurent Gouya,1,3,4,6 Hubert De Verneuil,6 Saïd Lyoumi,1,4,9 Hervé Puy1,3,4,6 and Zoubida Karim1,3,4

INSERM U1149/ERL CNRS 8252, Centre de Recherche sur l’Inflammation Paris Montmartre, 75018 Paris, France; 2Assistance Publique–Hôpitaux de Paris (AP-HP), Service Odontologie, Hôpital Universitaire, Université de Montpellier, France; 3Université Paris Diderot, Bichat site, Paris, France; 4Laboratory of Excellence, GR-Ex, Paris, France; 5 Institut de Médecine Régénératrice et de Biothérapie-Hôpital Saint Eloi CHU Montpellier, Université de Montpellier, France; 6Assistance Publique–Hôpitaux de Paris (AP-HP), Centre Français des Porphyries, Hôpital Louis Mourier, Colombes, France; 7 INSERM, Biothérapies des Maladies Génétiques et Cancers, U1035, F-33000 Bordeaux, France; 8Department of Clinical and Experimental Medicine, Section of Internal Medicine, University of Verona, Italy and 9Université Versailles Saint Quentin en Yvelines, France 1

*SM, CD, HP and ZK contributed equally to this work

ABSTRACT

H

Correspondence: herve.puy@aphp.fr/zoubida.karim@inserm.fr

Received: July 11, 2016. Accepted: October 28, 2016. Pre-published: November 10, 2016. doi:10.3324/haematol.2016.151621

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/260

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

260

emolysis occurring in hematologic diseases is often associated with an iron loading anemia. This iron overload is the result of a massive outflow of hemoglobin into the bloodstream, but the mechanism of hemoglobin handling has not been fully elucidated. Here, in a congenital erythropoietic porphyria mouse model, we evaluate the impact of hemolysis and regenerative anemia on hepcidin synthesis and iron metabolism. Hemolysis was confirmed by a complete drop in haptoglobin, hemopexin and increased plasma lactate dehydrogenase, an increased red blood cell distribution width and osmotic fragility, a reduced half-life of red blood cells, and increased expression of heme oxygenase 1. The erythropoiesis-induced Fam132b was increased, hepcidin mRNA repressed, and transepithelial iron transport in isolated duodenal loops increased. Iron was mostly accumulated in liver and spleen macrophages but transferrin saturation remained within the normal range. The expression levels of hemoglobin-haptoglobin receptor CD163 and hemopexin receptor CD91 were drastically reduced in both liver and spleen, resulting in heme- and hemoglobin-derived iron elimination in urine. In the kidney, the megalin/cubilin endocytic complex, heme oxygenase 1 and the iron exporter ferroportin were induced, which is reminiscent of significant renal handling of hemoglobin-derived iron. Our results highlight ironbound hemoglobin urinary clearance mechanism and strongly suggest that, in addition to the sequestration of iron in macrophages, kidney may play a major role in protecting hepatocytes from iron overload in chronic hemolysis.

Introduction Iron homeostasis relies on its storage and recycling through tissue macrophages, which contain the largest iron pool [derived from phagocytosis of senescent red cells and subsequent catabolism of hemoglobin (Hb) and heme]. Such recycling provides most of the daily iron requirement (20-30 mg). However, the intestine also takes part in iron homeostasis by providing 1-2 mg of iron per day, which correhaematologica | 2017; 102(2)


Iron and heme metabolisms in hemolytic mouse model

sponds to the daily loss of the metal. The major regulator of iron homeostasis is hepcidin (reviewed by Ganz1), which is directly down-regulated by stimulated activity of erythropoiesis.2 Fam132b [Erythroferrone (ERFE)] has been proposed as a crucial cytokine produced by late erythroblast to repress hepcidin synthesis.3 Since low hepcidin levels favor intestinal iron absorption and mobilization of tissue iron stores, its repression accounts for the paradoxical condition known as iron loading anemia.4,5 In hemolytic anemia, much less is known about hepcidin expression and the pattern of iron loading compared to anemia with ineffective erythropoiesis. Intravascular hemolysis leads to massive red blood cell (RBC)-free Hb and heme, which are chaperoned by haptoglobin (Hp) and hemopexin (Hpx), respectively, and cleared by spleen and liver macrophages via CD163 and CD91, respectively.6 Subsequent cellular endocytosis of these complexes followed by heme oxygenase 1 (HO-1) pathway activation results in heme catabolism and progressive tissue iron accumulation. We generated a mouse model of congenital erythropoietic porphyria (CEP; MIM 2637007), presenting with chronic hemolysis, to study iron and heme metabolisms and hepcidin expression. CEP or Günther’s disease is a rare autosomal recessive disorder8 caused by partial deficiency of Uroporphyrinogen III Synthase (UROS; EC 4.2.1.75), the fourth enzyme of heme metabolism. This deficiency leads to excessive synthesis and accumulation of pathogenic type I isomers of hydrophilic porphyrins (uroporphyrin I and coproporphyrin I) in bone marrow erythroid cells, leading to intravascular hemolysis with massive appearance of these compounds in plasma and urine.9 CEP patients suffer from chronic hemolysis without symptoms of ineffective erythropoiesis10 and from cutaneous photosensitivity with mutilating involvement. Clinical severity of anemia is highly heterogeneous among the patients, suggesting a role of modifier genes in the expression of the disease. Indeed, we identified a gain-of-function mutation in ALAS2, the erythroid isoform of the first enzyme of the heme biosynthetic pathway, in a CEP patient with severe hemolytic phenotype.11 Nevertheless, pathogenesis of hemolysis and iron disturbances resulting from such a chronic hemolysis without an ineffective erythropoiesis model has not yet been characterized. This study aims to identify the precise pathogenesis of iron disturbance occurring in a chronic hemolysis CEP mouse model. We compared this model to Hjv–/– hemochromatosis mice, which exhibited high non-heme iron overload without hemolysis or ineffective erythropoiesis. Our results clearly show that iron metabolism is highly adapted to satisfy the iron needs of bone marrow and spleen for effective erythropoiesis. Hepcidin levels are fully reduced in CEP mice by the regenerative erythropoiesis but do not lead to hepatocyte iron overload. Tissue iron overload derived from heme/Hb is primarily localized in the liver and spleen macrophages rather than hepatocytes. The absence of hepatocyte iron overload is a consequence of both the huge increase in erythroblast production and urinary iron losses. Finally, a tight co-ordination of heme/Hb and iron handling by the liver and kidney, through a dissociation in the tissue expression CD91 and CD163 expression levels limit the toxicity of Hb, heme and iron and allow the recovery of iron needed for erythropoiesis. haematologica | 2017; 102(2)

Methods Animals and biochemical analyses The CEP mouse model was established by knock-in of the human P248Q mutation into the Uros gene and subsequently back-crossed on the BALB/c background.7,12 The hemojuvelin knock-out animals (Hjv-/-) were a kind gift from Nancy Andrews (Duke University, USA). For acute hemolysis and heme-arginate experiments, BALB/c mice were provided by the Janvier laboratory (Janvier, Le Genest Saint Isle, France). Since in male mice the expression of hepcidin has been shown to be repressed by testosterone,13 our experiments were performed on 12- to 14-week-old female mice, unless mentioned otherwise. For hematologic and biochemical analysis, mice were anesthetized by intraperitoneal (i.p.) injection of a xylazine/ketamine mixture before blood puncture in the orbital sinus on heparin- or EDTA-coated tubes. All experimental procedures involving animals were performed with the approval of the Ethics Committee and in compliance with the French and European regulations on Animal Welfare and Public Health Service. Further details are available in the Online Supplementary Appendix.

Cell preparation, RBC turnover and flow cytometry analysis Spleen cells were isolated by mechanical dissociation using a 70 μm cell strainer in the presence of PBS/EDTA/BSA (2 mM/0.5%). Bone marrow cells from femurs and tibias were collected by gentle passage through an 18-gauge needle. Following centrifugation, the cell pellet was washed and resuspended in DMEM (containing 2% FBS and 1 mM HEPES) before flow cytometry analysis. RBC lifespan was assayed by in vivo biotinylation followed by FACS analyses, as previously described.14 Further details are available in the Online Supplementary Appendix.

Iron transport studies in duodenal loop The mice were sacrificed and a segment of approximately 3 cm of duodenum was rapidly isolated and immersed in cold HBSS pH 7.5. Iron transport was monitored by filling the duodenal segment with 100 μL/cm of the radiolabeled transport solution and placing it in a normal HBSS warm bath with 95% O2 / 5% CO2 gas as previously described.15 Then, aliquots from the bath were taken every 5 min for 55Fe-counting by liquid scintillation. Iron transport activity was evaluated as a ratio of counts per duodenum length (cm).

Porphyrin assay, glucose-6-phosphate dehydrogenase activity and erythropoietin assay Porphyrins were extracted and analyzed by high performance liquid chromatography (HPLC).16,17 Glucose-6-phosphate dehydrogenase (G6PD) activity was determined using RANDOX G6PDH kit (RANDOX Laboratories Ltd., Antrim, UK). Serum concentration of erythropoietin (epo) was determined using an assay kit (R&D Systems).

Western blot analysis Crude membrane fractions from mouse tissues were prepared as previously described.18,19 Briefly, 10-20 µg of crude membrane proteins were solubilized in 1× Laemmli buffer, analyzed by SDS/PAGE and transferred onto a PVDF membrane for HO-1, ferroportin, TfR1, H-ferritin and DMT1. For Hpx detection, 1 μL of plasma was loaded on SDS/PAGE and transferred onto nitrocellulose membrane. Primary antibodies used were: anti-ferroportin (a kind gift from D. Haile, San Antonio, USA), anti-H ferritin (kindly provided by Dr P. Arosio, Brescia, Italy), anti-HO1 (Stressgen Biotechnologies), anti-DMT1 (kindly provided by Dr B. Galy, Heidelberg, Germany), anti-cubilin and anti-megalin (kindly pro261


S. Millot et al.

vided by Dr R. Nielsen, Aarhus, Denmark), anti-b-actin (Sigma, Saint Quentin Fallavier, France), and anti-Hpx (kindly provided by Dr E. Tolosano, University of Torino, Italy). The blots were revealed by ECL® (GE Healthcare) after secondary antibody incubations.

Quantitative RT-PCR Total RNA from tissue was isolated using an RNA Plus Extraction Solution, as previously described.20 The results were arbitrarily normalized using the sample with the lowest CT value for S14, actin or GAPDH, as indicated. The relative quantifications were calculated using the comparative CT method. The amplification efficiency of each target was determined using serial 2-fold dilutions of cDNA. Sequences of the primers are available in the Online Supplementary Appendix.

Statistical analysis For biochemical and hematologic parameters, statistical significance was evaluated using the two-tailed Student t-test for comparison between means of two unpaired groups. Two-way ANOVA was used to compare the area under the curves for the duodenal loop experiments. GraphPad Prism software (GraphPad Software, San Diego, CA, USA) was used for statistical analysis.

Results Clinical features of CEP mice Homozygous CEP mice showed growth retardation, reduced fertility in adults of both sexes, and red-colored serum and urine.7 Type I porphyrins (Uro and Copro) were increased in urine and feces (data not shown) and in RBCs of CEP mice (Table 1). Cyanosis was visible on ears and tails of the young CEP mice but no spontaneous photosensitivity lesion was observed under our husbandry conditions.

Severe hemolytic anemia and reduced RBC half-life in CEP mice Congenital erythropoietic porphyria mice showed increased serum bilirubin and LDH levels (Table 1), with almost undetectable levels of Hp and Hpx (Figure 1A and B, respectively), which was strongly indicative of hemolytic anemia. Blood smears revealed anisocytosis and poikilocytosis (Figure 1C). The anemia in CEP mice was severe with significant reduction of Hb levels and RBC number (Table 1), regenerative with marked reticulocytosis (28.8±4.2%) (Table 1), and microcytic and hypochromic with reduced mean cell Hb content (9.95±0.64 pg in CEP vs. 14.5±0.18 in WT mice) (Table 1). The density distribution of red cells indicated the presence of both dense erythrocytes with mean cell hemoglobin concentration (MCHC) more than 37 g/dL and overhydrated cells with MCHC less than 22 g/dL (Figure 1D). The increased RBC distribution width (RDW) was in agreement with the variation in red cell size in CEP, related to the presence of increased reticulocytes and a subpopulation of microcytic cells (Table 1). The reticulocyte CHr and microcytic reticulocytes (MCVr) were diminished in CEP mice (Table 1), which strongly suggests an early onset of iron deficiency during erythropoietic differentiation.21 To further ascertain the hemolytic nature of the anemia, we measured the lifespan of the erythrocytes and revealed that RBC half-life was reduced from 15 days in WT to less than seven days in CEP mice (Figure 1E). However, osmot262

Table 1. Biological and hematologic parameters.

Parameters RBCs Uro I (μmol/L packed RBCs) G6PD(IU/g Hb) Mature RBCs RBCs (x1012/L) Hb (g/dL) Hct (%) MCV (fL) MCHC (g/dL) CH (pg) RDW (%) Reticulocytes Reticulocytes (%) MCVr (fL) MCHCr (g/dL) CHr (pg) Serum Bilirubin (μmol/L) LDH (UI/L) Erythropoietin (ng/L) Iron (μmol/L) Transferrin (g/L)

WT (n=6)

CEP (n=6)

2.44±0.37 22±0.5

794±59* 47±2

10.2±0.3 14.8±0.9 43.2±1.6 50.9±0.5 28.7±0.75 14.5±0.2 13.7±0.3

7.3±0.3* 8.6±0.5* 32.4±0.7* 34.4±1.6* 29.7±0.4 9.95±0.64* 32.2±0.6*

2.3±0.4 53.4±0.6 26.8±0.7 14.1±0.2

28.8±4.2* 44.8±2.4* 25.8±0.20 11.5±0.6*

5.97±1.34 889±207 156±41 38.3±6.5 3.1±0.3

15.1±3.4* 2324±775* 1277±518* 65±0.7* 4.5±0.6*

WT: wild-type (controls); CEP: congenital erythropoietic porphyria; RBC: red blood cells; G6PD: glucose-6-phosphate dehydrogenase; Hb: hemoglobin; Hct: hematocrit; MCV: mean corpuscular volume; MCHC: mean cell hemoglobin concentration; CH: hemoglobin content; RDW: RBC distribution width; MCVr: microcytic reticulocytes; MCHCr: reticulocyte mean corpuscular hemoglobin concentration; CHr: reticulocyte hemoglobin content; LDH: lactate dehydrogenase. *P<0.0025, two-tailed Student t-test. Results are expressed as mean±SD.

ic fragility measurements revealed a higher resistance to lysis of CEP erythrocytes (as compared to WT) in the higher range of NaCl concentrations. Indeed, at NaCl concentration between 5.5 g/L and 9 g/L, 20%-30% more WT erythrocytes were hemolyzed than CEP erythrocytes (Figure 1F). These data suggest that erythrocytes surviving in the circulation are more resistant to in vitro hemolysis: they are likely to correspond to reticulocytes and could explain the increased G6PD activity in CEP erythrocytes (Table 1).

Decreased apoptosis of immature erythroblasts and induction of spleen stress erythropoiesis in CEP mice In mice, bone marrow erythropoiesis is primarily homeostatic whereas “stress erythropoiesis” develops rapidly in the spleen following acute anemia. We analyzed the respective contribution of bone marrow and spleen erythropoiesis in compensating the hemolytic anemia in CEP mice (Figure 2). Using a flow cytometry assay based on cell surface expression of Ter119 and CD71, as previously described,14 we visualized larger Ter119+CD71+ cells [basophilic immature erythroblasts (Ery A)], smaller Ter119+CD71+cells [polychromatic intermediate erythroblasts (Ery B)] and Ter119+CD71– cells [acidophilic late erythroblasts and reticulocytes (Ery C)]. The bone marrow of CEP mice showed an important increase in the number of early erythroblasts (Figure 2A) accompanied by decreased apoptosis (Figure 2B). A moderate but significant increase in intermediate erythroblasts and a decrease in late erythroblasts were also observed haematologica | 2017; 102(2)


Iron and heme metabolisms in hemolytic mouse model

A

B

P=0.0025

C

D

E

F

Figure 1. Red blood cell (RBC) morphology, half-life and osmotic fragility in congenital erythropoietic porphyria (CEP) mice. (A) Measurement of serum Hp in controls (WT) and CEP mice. (B) Western blot of plasma Hpx in WT and CEP mice. (Right) Molecular weights in kDa. (C) Red blood cell smears. Anisocytosis and poikilocytosis are shown by arrow heads and hypochromic red cells by arrows. (D) Plot of mean cell Hb concentration (x-axis) against mean cell volume (y-axis) in controls (WT) and CEP mice. Data were obtained with the ADVIA 120 hematology analyzer (Siemens Healthcare Diagnostics, France). One representative mouse is shown for each condition. Similar results were obtained for the other 5 animals. (E) Red blood cell half-life. Biotin was injected on three consecutive days and a small aliquot of blood was analyzed at the indicated times by streptavidin labeling to detect the decay of biotinylated erythrocytes over time. Remaining biotinylated RBCs are expressed as percentage (%) of the total circulating RBCs. Results are the meanÂąStandard Deviation (SD) of results obtained on n=3 mice for each group. (F) Osmotic fragility of erythrocytes was evaluated in WT (black rectangles) and in CEP (gray triangle) mice. Results are expressed as percentage (%) of hemolysis in distilled water (set to 100%) against the NaCl concentration in the test solution.

(Figure 2A). In addition, the number of late erythroblasts was much lower than the intermediate erythroblasts in CEP mice, suggesting increased maturation rate and cellular exit from the bone marrow. The spleen of CEP animals was grossly enlarged (Online Supplementary Figure S2B), as previously reported,7 resulting from the onset of spleen erythropoiesis. Indeed, in this organ, there was a strong increase in the total number of erythroblasts at all stages of differentiation (Figure 2C). However, in contrast to the bone marrow, there was no significant decrease between intermediate and late erythroblasts (Figure 2C) and we observed decreased rather than increased apoptosis, as usually observed in ineffective erythropoiesis22,23 (Figure 2D). In adult spleen, stress erythropoiesis may be induced during acute anemia and hypoxia by BMP4 (Bone Morphogenic Protein 4) factor.24,25 We thus analyzed BMP4 expression level in CEP mice and show its strong increase in the red pulp of the spleen (Figure 2E), likely contributing to the rapid formation of stress burst-forming unit erythroid progenitors (BFU-Es) as a consequence of the high levels of erythropoietin (Epo) in these mice (Table 1). haematologica | 2017; 102(2)

Therefore, hemolytic anemia in CEP mice activates a compensatory stress erythropoiesis with no sign of ineffective erythropoiesis.

Regenerative anemia represses hepcidin expression Increased Epo levels and regenerative erythropoiesis are known to repress hepcidin expression. We thus investigated their impact on hepcidin synthesis and iron status in CEP mice. As expected, hepcidin was markedly reduced, both in the liver (at the mRNA level) and in the serum (Figure 3A and B). Fam132b mRNA expression in bone marrow cells was significantly increased (30-fold compared to WT mice) (Figure 3C). Using isolated duodenal loops to measure the transepithelial iron transport, we found that CEP mice presented a higher rate of iron absorption than the WT mice, although the differences between the area under the curves did not reach statistical significance (Figure 3D). In addition, we observed an increase of ferroportin protein expression in duodenal enterocytes (Figure 3E and F). Serum iron was also increased in CEP mice, but this did not lead to elevated Tf 263


S. Millot et al. Bone marrow A

C

P=0.0005

Spleen

Bone marrow B

P=0.0003

Spleen D

P=0.0001

E

Figure 2. Erythroblast subpopulations in bone marrow and spleen of congenital erythropoietic porphyria (CEP) mice and spleen BMP4 expression. The number of erythroblasts at different stages of maturation (A and C) and the proportion of Annexin V+ cells in each subset of erythroblast (B and D) was analyzed in bone marrow (A and B) and spleen (C and D). The proportion of erythroblasts at each stage of maturation was determined by FACS and the corresponding number of erythroblasts was calculated based on the observation that a femur contains 20x106 cells and a spleen contains 106 cells/mg wet weight.14 The three stages correspond to early (Ery A, white bars), intermediate (Ery B, gray bars), and late erythroblasts and reticulocytes (Ery C, black bars). Results are expressed as mean±Standard Error of the Mean (SEM) obtained for 3 animals of each genotype. The number of erythroblasts differed significantly between WT and CEP mice at all three stages of differentiation in the bone marrow (A) (*P=0.01, **P<0.05, ***P<0.0001; Student t-test) and in the spleen (C) (P=0.03). (E) Immunohistochemical staining of BMP4 in the spleen of WT and CEP mice (original magnifications 10x and 20x). n.s.: not significant.

saturation because Tf was also significantly increased, which is reminiscent of iron deficiency anemia and facilitates iron delivery to a larger number of erythroblasts (Table 1 and Figure 3G).

Tissue iron overload is strongly held in macrophages As expected, serum ferritin was significantly increased in CEP mice (Figure 3H). This increase was associated with higher iron content of the liver (1589±395 µg/g tissue in CEP vs. 443±573 μg/g tissue in WT; P=0.004). In the spleen, the concentration of iron was only moderately increased in CEP mice (1158±205 µg/g tissue in WT vs. 1739±332 μg/g tissue in CEP; P=0.004); however, due to a sharp increase in size (Online Supplementary Figure S2B), the total amount of iron was grossly increased from 145±35 in WT to 2157±740 g in CEP (P<0.0001). Furthermore, Perl’s staining in CEP mice shows that heavy iron deposits in the liver were restricted to Kupffer cells while only a diffuse, fainter staining was observed in hepatocytes (Figure 4A). This pattern of iron accumulation differs from the primary hemochromatosis mouse models (including Hjv–/– mice)26-28 where Tf saturation is high19 and 264

iron is mostly accumulated in hepatocytes (Figure 4A). Iron was also detected in the macrophages of the red pulp of CEP mice, while almost no iron deposit was observed in the spleen of Hjv–/– mice (Figure 4B), confirming that Hb and “free” heme are the likely source of macrophage iron accumulation. Therefore, these data suggest that, in chronic hemolysis, release of Hb in plasma contributes to macrophage iron overload preferentially and that Tf-bound iron that is massively used to meet the high demand of regenerative erythropoiesis in bone marrow and spleen contributes only moderately to tissue iron loading. However, one cannot exclude the possibility that hepatocyte iron overload could appear in older animals. The Hp plasma concentration was very low in CEP mice (Figure 1A), probably because of an increased rate of its endocytosis and subsequent lysosomal degradation. Interestingly, in both liver and spleen, the expression of the Hb-Hp receptor (CD163)29 was found to be fully suppressed at both the mRNA and protein levels (Figure 4C and D and Online Supplementary Figure S1B), suggesting a slowdown of Hb uptake in macrophages which may prevent excess iron overload. Heme released from Hb is haematologica | 2017; 102(2)


Iron and heme metabolisms in hemolytic mouse model

A

B

C

D P=0.002

P=0.002

E

P=0.01

F

G P=0.05

H P=0.0003

Figure 3. Iron parameters in congenital erythropoietic porphyria (CEP) mice. Wild-type (WT, control) and CEP mice were explored as follows. (A) Quantitative PCR analysis of Hamp1 (hepcidin) mRNA expression in the liver, normalized by S14 mRNA. (B) Serum hepcidin level measured by LC-MS/MS. (C) Quantitative PCR analysis of Fam132b mRNA expression in the bone marrow, normalized by Gapdh mRNA. (D) Transepithelial iron transport in isolated duodenal loops evaluated as described in the Methods section. (E) Western blot analysis of ferroportin (Fpn) in duodenal enterocytes. b-actin is shown as a loading control. The Fpn antibody is known to give a non-specific 55 KDa band.14,41 (Left) Molecular weight markers are shown in kDa (F) Quantitative analysis of Fpn protein expression evaluated as ratio to actin abundance. (G) Calculation of transferrin saturation (%) based on the measurement of serum iron and transferrin. (H) Serum ferritin level (μg/L). All results are mean±Standard Error of the Mean (SEM) of at least 4 independent mice. n.s.: not significant.

bound by Hpx and the heme-Hpx complex is mostly taken up by hepatocyte CD91 receptor to be degraded in lysosomes.30 The mRNA level of CD91 was significantly reduced in the liver and was fully suppressed in the spleen (Online Supplementary Figure S1A and B). Thus, like Hb, heme uptake appears to be decreased to protect hepatocytes from iron overload and heme pro-inflammatory effects. Moreover, HO-1 was highly expressed in the liver of CEP compared to WT mice (Figure 4E), confirming that residual heme uptake is rapidly degraded in the liver. In addition, despite an increase in ferritin expression, ferroportin expression was also induced in the liver of CEP mice (Figure 4E), suggesting an increase of iron release in the circulation to satisfy the high iron demand. In the spleen, where the steady state protein levels of HO1 and ferroportin were already high, we observed no difference between control and mutant mice (Online Supplementary Figure S2A).

Urinary iron losses contribute to the iron-restricted anemia Since the CD163 expression was fully suppressed both in liver and spleen, we analyzed Hb excretion in urine. Total urinary iron measured using Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES) was found at very high levels in CEP mice (from 9.9±5.5 μmol/L in WT to 223.6±113 µmol/L in CEP animals; P<0.001) (Figure 5A). The determination of urinary nonheme iron by the method normally used for serum iron haematologica | 2017; 102(2)

gave values of 2.2 μmol/L in WT mice and 56 µmol/L in CEP mice, indicating that approximately 75% of the urinary iron is organic. Indeed, using affinity-purified antimouse Hb, the histological analysis of kidney sections confirmed the appearance of significant Hb labeling at the apical membrane of CEP proximal tubules (Figure 5B). Interestingly, in contrast with the preserved mRNA levels (data not shown), the immunostaining of the endocytic receptor megalin/cubilin complexes revealed evidence of increased cubilin protein expression (Figure 5B) with no significant changes in megalin protein abundance (data not shown). Given that megalin/cubilin receptors are responsible for the uptake of glomerular filtrate proteins by the proximal tubules, we measured proteinuria in both WT and CEP mice. The results showed reduced amount of total urinary proteins and urinary albumin in CEP compared to WT mice (Figure 5C and D), suggesting induced function of the megalin/cubilin endocytic receptor in these mice. The toxicity of Hb on renal function was evaluated by the measurement of both serum urea levels and creatinine clearance calculations, and by the histological examination of kidney sections. We could not detect any evidence of renal injury in the CEP mice compared to WT mice, at least until the age of 14 weeks (Figure 5E and F and Online Supplementary Figure S3). Next, we explored the intrinsic heme and iron handling by the kidney in CEP mice. As expected, Perl’s staining of CEP kidneys showed significant accumulation of iron in the renal cortical part, particularly in the proximal tubules (Figure 6A). Ferritin 265


S. Millot et al.

A

B

C

D

E

P<0.0001

Figure 4. Hemoglobin, heme and iron processing in liver and spleen of congenital erythropoietic porphyria (CEP) mice. Perl’s staining of iron loading in the liver (A) and spleen (B) of wild-type (WT, control), CEP and Hjv–/– mice. In CEP mice, non-heme iron deposits were detected in liver Kupffer cells and in the spleen red pulp macrophages in CEP mice, whereas in Hjv–/– mice, non-heme iron was accumulated in hepatocytes (magnification 10X and 20X). (C) Western blot of the CD163 in the liver of WT and CEP mice. b-actin is shown as a loading control. (D) Quantitative PCR analysis of CD163 mRNA expression in the liver of WT and CEP mice, normalized by S14 mRNA. Results are mean±Standard Error of the Mean (SEM) of 6 independent mice. (E) Western blot analysis of HO1, H ferritin and ferroportin in the liver of of WT and CEP mice. b-actin is shown as a loading control.

protein level was significantly increased in the cortex but not in the medulla of CEP mice, confirming exclusive iron handing in the proximal renal tubules in these mice (Figure 6B); no change in the mRNA-expression of ferritin was observed (Figure 6B). However, the mRNA and protein expression of both TFR1 and DMT1 responsible for iron entry into renal cells were slightly decreased, although not reaching statistical significance (Online Supplementary Figure S4). Interestingly, the mRNA expression levels of HO-1 and ferroportin, which are both induced transcriptionally by free heme, were enhanced significantly in the cortex, but not in the medulla, of CEP mice, resulting in large increases in their protein abundance (Figure 6C). We also measured the portion of free heme in the urine of these mice using a hemin assay kit and found that only small traces of urinary free heme were detectable in WT mice and these were only slightly increased in the urine of CEP mice (0.68±0.2 µmol/L in WT to 6.7±0.2 μmol/L in CEP animals; P<0.03). To test whether this portion of heme results from Hb catabolism or from free heme waste, we studied heme and iron metabolism in PHZtreated mice (used as a model of acute hemolysis) and in mice that were (i.p.)-injected daily with heme arginate (HA) for three weeks. The control mice were injected with 266

excipients. Both increased urinary iron and heme levels and induced cortical HO-1 and ferroportin mRNA levels were observed solely in PHZ-mice but not in non-heme (NH) mice (Online Supplementary Table S1 and Figure S5). In addition, by Perl’s staining, iron overload in NH mice was localized primarily in the spleen, and to a lesser extent in the liver, but not in the kidney, indicating that the kidney may contribute to Hb rather than free heme handling in conditions of hemolysis (Online Supplementary Figure S5). Altogether, our results suggest that the kidney, by specifically handling Hb and recovering iron through ferroportin, likely limits deleterious effects of hemolytic anemia in CEP mice.

Discussion Our CEP mouse model exhibiting induced chronic intravascular hemolysis highlighted new insights into iron disorders related to heme and Hb catabolisms. The phenotype of this iron disorder, although associated with downregulation of hepcidin expression, differs strongly from the iron-loading anemia that is associated to ineffective erythropoiesis. Indeed, CEP mice mounted an efficient haematologica | 2017; 102(2)


Iron and heme metabolisms in hemolytic mouse model

A

B P<0.001

P<0.03

C

D

E

F

P<0.003 P<0.02

Figure 5. Hb clearance in the kidney of congenital erythropoietic porphyria (CEP) mice. (A) Non-heme iron (gray bars) and total iron (striped bars) was measured in the urine of wild-type (WT, control) and CEP mice. The results are meanÂąStandard Error of the Mean (SEM) of 6 independent mice. (B) Immunofluorescence staining of hemoglobin (Hb) and cubilin in the kidney of WT and CEP mice. G: glomerous; PT: proximal tubule (original magnifications 40X). (C-F) Urinary total protein, albumin, creatinine clearance and serum urea, respectively, measured in the urine of WT and CEP mice. Results are meanÂąStandard Error of the Mean (SEM) of 6 independent mice.

erythroid response and normal transferrin saturation, which, together with urinary iron-bound Hb losses, limited excess iron deposition in hepatocytes. The increased RDW of CEP red cells most likely contributed to hemolysis. The inability of RBCs to control their volume is known to impair their function, including their ability to undergo membrane deformation during circulation in the vasculature.31 The observed anemia was highly regenerative but was also microcytic, with microcytosis already present at the level of reticulocytes. This finding suggests that heme synthesis was reduced during erythroblast maturation most likely because of reduced protoporphyrin IX (PPIX) synthesis. In addition, the CEP mice showed decreased CHr, a known marker of true iron deficiency.21,32 Therefore, iron supply to the developing erythroblast could also be a rate-limiting factor in our CEP mice because of both a highly regenerative erythropoiesis and iron losses associated with hemolysis. Indeed, hepcidin expression was reduced and led to increased intestinal iron absorption, as shown by our experiments on isolated duodenal loops. Furthermore, ferroportin, which exports iron from tissues back to the plasma, was highly induced in CEP mice, suggesting that tissue iron stores can be efficiently mobilized. Transferrin saturation is still within the normal range, suggesting that increased intestinal iron absorption and macrophages heme-iron recycling are sufficient to compensate for the urinary iron losshaematologica | 2017; 102(2)

es. We also showed that Fam 132b (ERFE), a factor produced by developing erythroblasts and implicated in Hamp gene repression,3 was highly up-regulated in bone marrow and erythropoietic spleen in conditions of chronic hemolysis and is likely to contribute to hepcidin repression. Furthermore, it has been shown that this erythropoietic signaling pathway can override hepcidin regulation by iron, as shown by the paradoxical association of low serum hepcidin levels and tissue iron overload both in mouse models22,33,34 and patients5,35 with b thalassemia, congenital dyserythropoietic anemia,36 or myelodysplastic syndrome.4 The normal Tf saturation seemed to limit iron loading of hepatocytes, on the contrary to what is observed in hemochromatosis with normal erythropoiesis or in ironloading anemia with ineffective erythropoiesis, where hepcidin is also fully suppressed but leads to heavy hepatocyte iron overload.22,33,34 In hemolytic conditions, hemeand Hb-derived iron contributes to tissue iron loading independently of Tf-bound iron. Hb makes stable complexes with Hp before being taken up by macrophages through binding and internalization by CD163.29 Interestingly, Hp was almost undetectable in the plasma of the CEP mice and the CD163 mRNA expression was fully suppressed both in liver and spleen, suggesting that both tissues are able to deploy intrinsic mechanisms to reduce free Hb management, and thus to be protected from its 267


S. Millot et al. A

B

C P<0.0159 P<0.0317

Figure 6. Heme and iron processing in the kidney of congenital erythropoietic porphyria (CEP) mice. (A) Perl’s staining of iron loading in the kidney of wild-type (WT, control) and CEP mice. Non-heme iron deposits were detected in the proximal tubules. G: glomerous; PT: proximal tubule (magnification 10x and 20x). (B and C) Quantitative PCR and western-blot analysis of H-ferritin (B), HO1, and Fpn (C) in the cortex and medulla of WT and CEP kidneys. For quantitative PCR analysis, mRNA expression is normalized by actin mRNA. Results are mean±Standard Error of the Mean (SEM) of 6 independent mice. For western blot analysis, b-actin is shown as a loading control. n.s.: not significant.

cytotoxicity. Both free heme and Hb have been shown to down-regulate the mRNA expression of the CD163 in cultures of human monocyte-derived macrophages.37 In addition, when the buffering capacity of Hp is overwhelmed, Hb is oxidized into methemoglobin which liberates its heme rapidly.30 Heme is then bound by Hpx and cleared by internalization of the complex by CD91.38 Expression of this receptor is ubiquitous but the presence of iron deposits in the CEP and HA mice, mostly in macrophages of the spleen and the liver, is indicative of dominant heme uptake by macrophages. Furthermore, Hpx was almost undetectable in the plasma of the CEP mice, suggesting that some heme remained either “free” or loosely bound to albumin and was probably taken up by liver macrophages, as previously demonstrated.30 Perl’s staining of the liver showed that iron accumulated predominantly in macrophages, suggesting that this heme uptake pathway was operative in macrophages, similarly to what was seen in the superoxide dismutase 1 knockout mouse model also characterized by chronic hemolysis, and the animals were analyzed at one year of age.39 Our results also highlight the involvement of the kidney in Hb and iron in the context of hemolytic anemia. We 268

have previously shown that the kidney exhibits a cell specificity of iron handling in the kidney, which depends on the pathological origin of the iron overload. In hemochromatosis models, transferrin-bound iron was specifically handled in the thick ascending limb;19 however, as shown in our hemolytic model, Hb-bound iron was specifically taken up by the proximal tubule, the endocytic megalin/cubilin complex was stimulated, and HO-1 and ferroportin were induced to generate and return iron to the bloodstream, although the urinary iron-bound Hb losses remained significant and certainly contribute to the hypochromic anemia. The involvement of renal megalin/cubilin receptors for the binding and uptake of Hb has been previously demonstrated by in vitro studies and mouse models.40 In addition, the upregulation of cubilin in the CEP model illustrates the physiological importance of this receptor in the renal clearance of Hb. Altogether, we show that chronic intravascular hemolysis in CEP mice is associated with an efficient erythroid response in bone marrow and spleen (Figure 7). Microcytic anemia persists despite repression of hepcidin, increased intestinal iron absorption, renal iron recovery, and regenerative erythropoiesis. This hepcidin repression does not haematologica | 2017; 102(2)


Iron and heme metabolisms in hemolytic mouse model

Figure 7. Schematic representation of iron balance in congenital erythropoietic porphyria (CEP) mice. Regenerative anemia associated with hemolysis represses hepcidin expression, thereby increasing intestinal iron absorption and allowing iron recycling by macrophages. Transferrin (Tf) is increased to favor iron delivery to developing erythroblasts and Tf saturation remains normal. Plasma heme and hemoglobin (Hb) contribute to macrophage iron overload, mostly in the liver. However, the reduction of CD163 expression should prevent the uptake of free Hb in the liver, but favor its wasting in the urine. Kidney recovers part of this Hb to recycle iron into the circulation. Combined with urinary Hb-iron losses, normal Tf saturation prevents hepatocyte iron overload.

lead to a significant hepatocyte iron overload. Genes such as CD163 and CD91, involved in iron redistribution following hemolysis, could play a role as a modifier of disease severity in human CEP patients as well as in other chronic hemolytic disorders. Moreover, our results highlight the crucial involvement of kidney in eliminating the extra toxic Hb and in the recovery of iron to satisfy iron demand for regenerative purposes in the context of hemolytic anemia. Acknowledgments The authors are very grateful to Catherine Vernimmen, Olivier Thibaudeau, Margarita Hurtado-Nedelec and Valérie Andrieu for their help with the animal work and FACS analysis, and to

References 1. Ganz T. Hepcidin and iron regulation, 10 years later. Blood. 2011;117(17):4425-4433. 2. De Falco L, Silvestri L, Kannengiesser C, et al. Functional and clinical impact of novel TMPRSS6 variants in iron-refractory irondeficiency anemia patients and genotypephenotype studies. Hum Mutat. 2014;35(11):1321-1329. 3. Kautz L, Jung G, Valore EV, et al. Identification of erythroferrone as an ery-

haematologica | 2017; 102(2)

Joel Poupon (Laboratoire de Toxicologie Biologique, Hôpital Lariboisière, Paris) for urinary iron determinations by atomic absorption spectrometry. Funding INSERM and the Université Paris Diderot, France supported this work. Part of this work is funded by the labex GR-Ex, reference ANR-11-LABX-0051, by the program “Investissements d’avenir” of the French National Research Agency, reference ANR-11-IDEX-0005-02, and by the program “University of Sorbonne Paris Cité, Excellence Initiative, IDEX”, reference Hemir. SM was supported by the Université Paris Diderot and by the Société Française d'Hématologie. BM was supported by a grant from the Fondation pour la Recherche Médicale Française.

throid regulator of iron metabolism. Nat Genet. 2014;46(7):678-684. 4. Santini V, Girelli D, Sanna A, et al. Hepcidin levels and their determinants in different types of myelodysplastic syndromes. PLoS One. 2012;6(8):e23109. 5. Tanno T, Bhanu NV, Oneal PA, et al. High levels of GDF15 in thalassemia suppress expression of the iron regulatory protein hepcidin. Nat Med. 2007;13(9):1096-1101. 6. Nielsen MJ, Moller HJ, Moestrup SK. Hemoglobin and heme scavenger recep-

tors. Antioxid Redox Signal. 2009;12(2): 261-273. 7. Ged C, Mendez M, Robert E, et al. A knock-in mouse model of congenital erythropoietic porphyria. Genomics. 2006;87(1):84-92. 8. Ged C, Moreau-Gaudry F, Richard E, Robert-Richard E, de Verneuil H. Congenital erythropoietic porphyria: mutation update and correlations between genotype and phenotype. Cell Mol Biol (Noisyle-grand). 2009;55(1):53-60.

269


S. Millot et al. 9. Puy H, Gouya L, Deybach JC. Porphyrias. Lancet. 2011;375(9718):924-937. 10. Katugampola RP, Anstey AV, Finlay AY, et al. A management algorithm for congenital erythropoietic porphyria derived from a study of 29 cases. Br J Dermatol. 2012;167(4):888-900. 11. To-Figueras J, Ducamp S, Clayton J, et al. ALAS2 acts as a modifier gene in patients with congenital erythropoietic porphyria. Blood. 2011;118(6):1443-1451. 12. Robert-Richard E, Moreau-Gaudry F, Lalanne M, et al. Effective gene therapy of mice with congenital erythropoietic porphyria is facilitated by a survival advantage of corrected erythroid cells. Am J Hum Genet. 2008;82(1):113-124. 13. Latour C, Kautz L, Besson-Fournier C, et al. Testosterone perturbs systemic iron balance through activation of epidermal growth factor receptor signaling in the liver and repression of hepcidin. Hepatology. 2014;59(2):683-94. 14. Millot S, Andrieu V, Letteron P, et al. Erythropoietin stimulates spleen BMP4dependent stress erythropoiesis and partially corrects anemia in a mouse model of generalized inflammation. Blood. 2011; 116(26):6072-6081. 15. Brasse-Lagnel C, Karim Z, Letteron P, Bekri S, Bado A, Beaumont C. Intestinal DMT1 cotransporter is down-regulated by hepcidin via proteasome internalization and degradation. Gastroenterology. 2011; 140(4):1261-1271 e1. 16. Lockwood WH, Poulos V, Rossi E, Curnow DH. Rapid procedure for fecal porphyrin assay. Clin Chem. 1985;31(7):1163-1167. 17. Lim CK, Peters TJ. Urine and faecal porphyrin profiles by reversed-phase high-performance liquid chromatography in the porphyrias. Clin Chim Acta. 1984; 139(1):55-63. 18. Delaby C, Pilard N, Goncalves AS, Beaumont C, Canonne-Hergaux F. Presence of the iron exporter ferroportin at the plasma membrane of macrophages is enhanced by iron loading and down-regulated by hepcidin. Blood. 2005;106(12):3979-3984. 19. Moulouel B, Houamel D, Delaby C, et al. Hepcidin regulates intrarenal iron handling

270

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

at the distal nephron. Kidney Int. 2013; 84(4):756-766. Lyoumi S, Abitbol M, Andrieu V, et al. Increased plasma transferrin, altered body iron distribution, and microcytic hypochromic anemia in ferrochelatasedeficient mice. Blood. 2007;109(2):811-818. Brugnara C, Laufer MR, Friedman AJ, Bridges K, Platt O. Reticulocyte hemoglobin content (CHr): early indicator of iron deficiency and response to therapy. Blood. 1994;83(10):3100-3101. De Franceschi L, Daraio F, Filippini A, et al. Liver expression of hepcidin and other iron genes in two mouse models of beta-thalassemia. Haematologica. 2006; 91(10):1336-1342. Li H, Rybicki AC, Suzuka SM, et al. Transferrin therapy ameliorates disease in beta-thalassemic mice. Nat Med. 2010; 16(2):177-182. Perry JM, Harandi OF, Paulson RF. BMP4, SCF, and hypoxia cooperatively regulate the expansion of murine stress erythroid progenitors. Blood. 2007;109(10):44944502. Paulson RF, Shi L, Wu DC. Stress erythropoiesis: new signals and new stress progenitor cells. Curr Opin Hematol. 2011;18(3): 139-145. Huang FW, Pinkus JL, Pinkus GS, Fleming MD, Andrews NC. A mouse model of juvenile hemochromatosis. J Clin Invest. 2005; 115(8):2187-2191. Meynard D, Kautz L, Darnaud V, et al. Lack of the bone morphogenetic protein BMP6 induces massive iron overload. Nat Genet. 2009;41(4):478-481. Nicolas G, Bennoun M, Devaux I, et al. Lack of hepcidin gene expression and severe tissue iron overload in upstream stimulatory factor 2 (USF2) knockout mice. Proc Natl Acad Sci USA. 2001;98(15):8780-8785. Kristiansen M, Graversen JH, Jacobsen C, et al. Identification of the haemoglobin scavenger receptor. Nature. 2001; 409(6817): 198-201. Tolosano E, Fagoonee S, Morello N, Vinchi F, Fiorito V. Heme scavenging and the other facets of hemopexin. Antioxid Redox Signal. 2011;12(2):305-320.

31. An X, Mohandas N. Disorders of red cell membrane. Br J Haematol. 2008;141(3): 367-375. 32. Brugnara C. Reticulocyte cellular indices: a new approach in the diagnosis of anemias and monitoring of erythropoietic function. Crit Rev Clin Lab Sci. 2000;37(2):93-130. 33. Gardenghi S, Marongiu MF, Ramos P, et al. Ineffective erythropoiesis in beta-thalassemia is characterized by increased iron absorption mediated by down-regulation of hepcidin and up-regulation of ferroportin. Blood. 2007;109(11):5027-5035. 34. Ramos P, Melchiori L, Gardenghi S, et al. Iron metabolism and ineffective erythropoiesis in beta-thalassemia mouse models. Ann N Y Acad Sci. 2010;1202:24-30. 35. Origa R, Galanello R, Ganz T, et al. Liver iron concentrations and urinary hepcidin in beta-thalassemia. Haematologica. 2007; 92(5):583-588. 36. Tamary H, Shalev H, Perez-Avraham G, et al. Elevated growth differentiation factor 15 expression in patients with congenital dyserythropoietic anemia type I. Blood. 2008;112(13):5241-5244. 37. Schaer CA, Vallelian F, Imhof A, Schoedon G, Schaer DJ. Heme carrier protein (HCP-1) spatially interacts with the CD163 hemoglobin uptake pathway and is a target of inflammatory macrophage activation. J Leukoc Biol. 2008;83(2):325-333. 38. Hvidberg V, Maniecki MB, Jacobsen C, et al. Identification of the receptor scavenging hemopexin-heme complexes. Blood. 2005; 106(7):2572-2579. 39. Starzynski RR, Canonne-Hergaux F, Willemetz A, et al. Haemolytic anaemia and alterations in hepatic iron metabolism in aged mice lacking Cu,Zn-superoxide dismutase. Biochem J. 2009;420(3):383-390. 40. Gburek J, Verroust PJ, Willnow TE, et al. Megalin and cubilin are endocytic receptors involved in renal clearance of hemoglobin. J Am Soc Nephrol. 2002;13(2):423-430. 41. Lasocki S, Millot S, Andrieu V, et al. Phlebotomies or erythropoietin injections allow mobilization of iron stores in a mouse model mimicking intensive care anemia. Crit Care Med. 2008;36(8):23882394.

haematologica | 2017; 102(2)


ARTICLE

Coagulation & Its Disorders

The C1 and C2 domains of blood coagulation factor VIII mediate its endocytosis by dendritic cells

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Bagirath Gangadharan,1,2,3* Mathieu Ing,1,2,3* Sandrine Delignat,1,2,3 Ivan Peyron,1,2,3 Maud Teyssandier,1,2,3 Srinivas V. Kaveri1,2,3 and Sébastien Lacroix-Desmazes1,2,3

Sorbonne Universités, UPMC Université Paris 06, UMR S 1138; 2INSERM, UMR S 1138, and 3Université Paris Descartes, Sorbonne Paris Cité, UMR S 1138, Centre de Recherche des Cordeliers, F-75006, Paris, France 1

Haematologica 2017 Volume 102(2):271-281

ABSTRACT

T

he development of inhibitory antibodies to therapeutic factor VIII is the major complication of replacement therapy in patients with hemophilia A. The first step in the initiation of the anti-factor VIII immune response is factor VIII interaction with receptor(s) on antigenpresenting cells, followed by endocytosis and presentation to naïve CD4+ T cells. Recent studies indicate a role for the C1 domain in factor VIII uptake. We investigated whether charged residues in the C2 domain participate in immunogenic factor VIII uptake. Co-incubation of factor VIII with BO2C11, a monoclonal C2-specific immunoglobulin G, reduced factor VIII endocytosis by dendritic cells and presentation to CD4+ T cells, and diminished factor VIII immunogenicity in factor VIII-deficient mice. The mutation of basic residues within the BO2C11 epitope of C2 replicated reduced in vitro immunogenic uptake, but failed to prevent factor VIII immunogenicity in mice. BO2C11 prevents factor VIII binding to von Willebrand factor, thus potentially biasing factor VIII immunogenicity by perturbing its half-life. Interestingly, a factor VIIIY1680C mutant, that does not bind von Willebrand factor, demonstrated unaltered endocytosis by dendritic cells as well as immunogenicity in factor VIII-deficient mice. Co-incubation of factor VIIIY1680C with BO2C11, however, resulted in decreased factor VIII immunogenicity in vivo. In addition, a previously described triple C1 mutant showed decreased uptake in vitro, and reduced immunogenicity in vivo, but only in the absence of endogenous von Willebrand factor. Taken together, the results indicate that residues in the C1 and/or C2 domains of factor VIII are implicated in immunogenic factor VIII uptake, at least in vitro. Conversely, in vivo, the binding to endogenous von Willebrand factor masks the reducing effect of mutations in the C domains on factor VIII immunogenicity.

Correspondence: sebastien.lacroix-desmazes@crc.jussieu.fr

Received: April 27, 2016. Accepted: October 3, 2016. Pre-published: October 6, 2016. doi:10.3324/haematol.2016.148502

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/271

Introduction Hemophilia A is a monogenic disorder associated with mutations causing reductions in functional levels of coagulation factor VIII (FVIII). FVIII consists of a heavy chain (A1-a1-A2-a2-B domain) and a light chain (a3-A3-C1-C2) held together by non-covalent interactions.1 It rapidly associates with von Willebrand factor (VWF) in blood, and this interaction is necessary for maintaining its circulatory half-life.2 Current treatment for FVIII deficiency requires prophylactic infusions of plasmaderived or recombinant FVIII. However, up to 30% of patients with severe hemophilia A develop an anti-FVIII immune response, thus rendering treatment ineffective.3 The development of anti-FVIII antibody responses is dependent on T helper cells, requiring antigen uptake and presentation by antigen-presenting cells (APCs).4 haematologica | 2017; 102(2)

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

271


B. Gangadharan et al.

Hence, understanding the initial steps involved in FVIII capture by APCs may provide novel strategies to prevent the onset of the immune response. Several groups, including ours, have investigated the endocytic pathways involved in FVIII uptake. Candidate receptors such as macrophage mannose receptor (MMR, CD206), low-density lipoprotein receptor-related protein (LRP, CD91), or other receptor-associated protein (RAP)sensitive receptors have been proposed.5-10 Equally, the nature of the residues within FVIII domain(s) that contribute to these interactions is also an area of active investigation. Despite these efforts, the in vivo relevance of these receptors and the nature of the FVIII residues involved in FVIII uptake remain unclear. Recently, Herczenik et al.10 demonstrated that KM33, a human C1 domain-specific monoclonal immunoglobulin G (IgG), inhibits FVIII endocytosis by monocyte-derived dendritic cells (MoDCs) or mouse bone marrow-derived dendritic cells (BMDCs). KM33 engages K2092, F2093 and R2090 residues, involved in the interactions with phospholipid membrane surfaces.11 Additionally, KM33 inhibits interactions of the C1 domain with membrane surfaces, VWF and LRP.12 FVIII uptake by LRP in MoDCs, used as model APCs, has been ruled out,8 while a role for CD206 has been controversially documented.7,9,10 This suggests that FVIII uptake by APCs may involve other endocytic receptor(s). Importantly, FVIII mutants containing alanine substitutions of the K2092, F2093 and R2090 C1 residues, exhibit diminished uptake in vitro and reduced immunogenicity in a mouse model of severe hemophilia A.11 Together, these results point to the significance of membrane-binding residues within the C1 domain for FVIII uptake both in vitro and in vivo. Similar to the C1 domain, the C2 domain of FVIII interacts with phospholipid membrane surface, a binding that involves several basic residues.13 In the study herein, we investigated whether membrane-interacting residues within the C2 domain of FVIII are involved in FVIII uptake. We first show that BO2C11, a well-characterized human monoclonal IgG that engages membrane-binding residues in the C2 domain,14 inhibits FVIII uptake and presentation in vitro, and reduces FVIII immunogenicity in vivo. We also demonstrate that this reduced immunogenicity is independent of the ability of FVIII to interact with endogenous VWF. Additionally, by site-directed mutagenesis, we demonstrate that the R2215 residue, which is part of the BO2C11 epitope, is implicated in the cellular uptake of FVIII by APCs in vitro. Together with the published data, our results suggest a potential synergy between membrane-binding residues in both the C1 and C2 domains of FVIII in mediating FVIII recognition or uptake by APCs in vitro. Furthermore, we also demonstrate that FVIII C domain mutations exhibit diminished immunogenicity in vivo only in the absence of endogenous VWF.

Methods Reagents Recombinant human FVIII (Refacto) came from Pfizer (New York, NY, USA). Murine granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin-4 (IL-4) came from Cellgenix Technology Transfer (Freiburg, Germany). The monoclonal mouse anti-FVIII (mAb6), the human anti-A2 (BO2BII) and anti-C2 (BO2C11) domains antibodies were kind gifts from Drs 272

JM Saint-Remy and M Jacquemin (KUL, Leuven, Belgium). The monoclonal human-derived anti-C1 (KM33) was a gift from Dr J Voorberg (Sanquin, Amsterdam, The Netherlands). The mouse monoclonal anti-A2 domain (GMA-8015) and anti-C2 domain (ESH8) antibodies were purchased from Green Mountain Antibodies (Burlington, VT, USA) and Sekisui Diagnostics (Kings Hill, Kent, UK), respectively. Biotin-labeled GMA-8015 was prepared upon incubation with a 20-fold molar excess of EZ-Link Sulfo-NHS-LC-Biotin (Thermo Fisher Scientific, Courtaboeuf, France) for 30 min at room temperature, and the removal of excess biotin by diafiltration was carried out using 30 kDa Amicon Ultra15 centrifugal filter units (Merck Millipore, Saint-Quentin-enYvelines, France). BO2C11 fragment antigen binding (Fab) or F(ab’)2 fragments were digested by papain or pepsin following the manufacturer’s instructions (Thermo Fisher Scientific, Courtaboeuf, France).

Production and purification of recombinant mutated or wild-type FVIII Complementary DNA (cDNA) encoding human B-domain deleted (BDD) FVIII (FVIIIHSQ), containing the 14-amino acid segment SFSQNPPVLKRHQR in place of the B domain, cloned in the ReNeo mammalian expression plasmid with a geneticin resistance, has been described previously.15 The cDNA encoding FVIIIHSQ was used as a template to generate the R2215A, R2220A, R2215A-R2220A, R2090A-K2092A-F2093A and Y1680C FVIII mutants by splicing by overlap extension mutagenesis as described in the Online Supplementary Methods. The presence of the mutations was confirmed by standard sequencing analysis. The stable expression of wild-type and mutated FVIII by baby hamster kidney-derived cells, and FVIII purification, were performed as described in the Online Supplementary Methods. The concentration of purified wild-type and mutated FVIII was calculated by absorbance at 280 nm using a molar extinction coefficient of 256,300 M-1cm-1 and a molecular weight of 165,300 Da. Specific activities were estimated by a one-stage clotting assay and ranged between 4800-9000 IU/mg. The sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) migration profiles of the different purified recombinant FVIII are shown in the Online Supplementary Figure S1. In parallel, the ability of the different FVIII molecules to generate activated factor X was assessed in a chromogenic assay (Siemens Healthcare Diagnostics, Marburg, Germany).

FVIII binding to VWF and monoclonal antibodies Ninety-six-well ELISA plates (Maxisorp, Nunc, Roskilde, Denmark) were coated with human plasma-derived VWF (Wilfactin, LFB, Les Ulis, France), BO2C11, BO2BII, ESH8 or KM33 at 1 µg/ml in bicarbonate buffer, pH 9.5, for 1 hr at 37°C. Wells were blocked with 20 mM HEPES, 0.15 M NaCl, 0.05% Tween 20 and 5% bovine serum albumin (BSA), pH 7.4. Wild-type and mutated FVIII were then diluted in blocking buffer and incubated in the coated wells for 1 hr at 37°C. Bound FVIII was revealed using either biotinylated GMA-8015 (1 µg/ml), followed by streptavidin conjugated to horseradish peroxidase (R&D systems, Lille, France) or, in the case of BO2BII-bound FVIII, with ESH8 followed by a horseradish peroxidase conjugated polyclonal goat antimouse IgG antibody (Southern Biotech, Anaheim, CA, USA) and the o-phenylenediamine dihydrochloride (OPD) substrate (SigmaAldrich, Saint-Louis, MO, USA). Absorbance was read at 492 nm.

Generation of immature human MoDCs and mouse BMDCs Monocytes were isolated from the blood of healthy donors using anti-CD14+ magnetic microbeads (Miltenyi Biotec, Paris, haematologica | 2017; 102(2)


Immunogenic factor VIII uptake

A

B

C

Figure 1. The anti-C2 antibody BO2C11 inhibits FVIII uptake in vitro and modulates FVIII immunogenicity in vivo. Panel A. B domain-deleted FVIII (20 nM) was preincubated alone or with equimolar concentrations of ESH8, BO2C11 or with a 2 molar excess BO2C11 Fab fragments. Uptake by human MoDCs was analyzed by fluorescence-activated cell sorting (FACS). Results are expressed as the percentage of median fluorescence intensity (MFI), whereby 100% corresponds to MFI obtained with wild-type FVIII (FVIIIHSQ) incubated alone. The graph is representative of 4 independent donors (mean±SEM). Panel B. Immature MoDCs were cultured for 24 hr in 96 round bottom plates with the FVIII-specific HLA-DRB1*0101-restricted CD4+ T-cell hybridoma (ratio 1:10, clone 1G8-A2) in the presence of B domain-deleted FVIII alone or pre-incubated with 2 molar excess of BO2C11. Activation of T cells was measured by IL-2 secretion in the supernatant by ELISA (BD Biosciences). The graph is representative of 4 experiments (mean±SEM). Panel C. Hemophilic FVIII exon 16 knock-out mice (n=6/group) were injected once a week for 4 weeks with 0.2 µg of B domain-deleted FVIII pre-incubated with 6 μM F(ab’)2 of BO2C11 (closed square) or a human IgG4 isotype control (open square). After 4 weeks, the antiFVIII antibody response was measured. Anti-FVIII IgG titers are defined as arbitrary units using the mouse monoclonal anti-FVIII IgG mAb6 as a standard. Data are represented as serum dilution versus mean±SEM of absorbance (492 nm). FVIII: factor VIII; MoDCs: monocyte-derived dendritic cells; A.U.: arbitrary unit; IL-2: interleukin-2; IgG: immunoglobulin G; Ab: antibody.

France). Ethics committee approval was obtained for the use of buffy bags from healthy donors. Monocytes (0.5.106 cells/ml) were cultured in RPMI-1640 (Lonza, Verviers, Belgium) with 10% fetal calf serum (Life Technologies, Saint-Aubin, France), supplemented with GM-CSF (1000 IU/106 cells) and IL-4 (500 IU/106 cells) (Miltenyi Biotec) for 5 days to generate immature MoDCs. After 5 days, the differentiation of MoDCs (> 90% purity) was confirmed by flow cytometry upon loss of CD14 staining (M5E2 clone, BD Pharmingen, San Jose, CA, USA), expression of major histocompatibility complex (MHC) class II and CD1a (HI149 and G46.6 clones, respectively, BD Pharmingen). Acquisition was performed on a LSR II cytometer with FACSDiva software (version 6.1, BD Biosciences, Le Pont au Claix, France). Murine BMDCs were generated as described previously.16 Briefly, bone marrow cells were extracted from FVIII exon 16 knock-out C57BL/6 mice and cultured in Petri dishes (2.106 cells/10 ml/plate) for 10 days in RPMI1640 supplemented with 10% fetal calf serum, 50 mM 2-mercaptoethanol and 200 U/ml murine recombinant GM-CSF (Cellgenix Technology Transfer, Freiburg, Germany). Supplemented medium was replaced at days 3, 6 and 8. BMDCs purity and phenotype were validated by the expression of CD11c (HL3 clone, BD Pharmingen).

permeabilization buffer (eBiosciences), and FVIII was stained using biotinylated GMA-8015 (2 µg/ml), followed by streptavidinPE (1 µg/ml, BD Biosciences) for 30 min at room temperature. Cells were analyzed by flow cytometry. The uptake was quantified as the difference in median fluorescence intensities between 37°C and 4°C (ΔMFI37°C-4°C), to exclude the signal generated by the binding of FVIII to the cell surface.

In vitro activation of a FVIII-specific HLA-DRB1*0101restricted mouse CD4+ T cell hybridoma Activation of the HLA-DRB1*0101-restricted mouse CD4+ T cell hybridoma specific for human FVIII (1G8-A2), was assessed as previously described.17 FVIII (10 nM) pre-incubated or not with 2 molar excess of BO2C11 was incubated with 10,000 MoDCs or 200,000 mitomycin C-treated mouse splenocytes from SUREL1 mice and co-cultured with 100,000 T cells in X-VIVO15 medium (Life Technologies) for 18 hr at 37°C. Controls included T cells incubated alone, or incubated with MoDCs/BMDCs in the presence of concanavalin A (2 μg/ml, Sigma-Aldrich), or in the absence of FVIII. Levels of interleukin-2 (IL-2) secreted in the supernatant by T cells were assessed using the BD OptEIA™ mouse IL-2 ELISA set (BD Biosciences). Of note, MoDCs do not produce IL-2 when incubated with FVIII alone.18

In vitro FVIII uptake by immature MoDCs and BMDCs B domain-deleted FVIII (20 nM) was pre-incubated with equimolar concentrations of ESH8, BO2C11 or with a 2 molar excess BO2C11 Fab fragments for 30 min at 37°C. Samples were then incubated with 5-day-old immature MoDCs or with 9-dayold immature BMDCs (0.2.106 cells/100 μl) in Iscove's Modified Dulbecco's Medium for 30 min at 37°C or 4°C. Cells were washed with ice-cold phosphate buffered saline (PBS) and fixed with BD CytofixTM Fixation buffer (BD Biosciences) for 20 min at 4°C. Cells were then permeabilized for 30 min at room temperature with haematologica | 2017; 102(2)

Animals and administration of FVIII Eight to 12 week-old FVIII exon 16 knock-out C57Bl/6 mice and VWF/FVIII exon 16 double knock-out C57Bl/6 mice (Professor H.H. Kazazian, University of Pennsylvania School of Medicine, Philadelphia, PA, USA, and Drs. C Denis and O Christophe, INSERM U770, Le Kremlin-Bicêtre, France, respectively), were injected intravenously once a week for 4 weeks with either: i) Bdomain deleted FVIII (0.2 μg, Refacto, Pfizer) pre-incubated with equimolar amounts of F(ab’)2 fragments of an IgG4 isotype control 273


B. Gangadharan et al. or BO2C11, ii) B-domain deleted FVIIIY1680C (0.4 μg) alone or preincubated with Fab fragments of BO2C11, or iii) B-domain deleted FVIIIHSQ, FVIIIC1 or FVIIIR2215-20A (0.5 or 1 μg). The presence of endotoxins in the different recombinant FVIII was evaluated using the ToxinSensor™ Chromogenic LAL Endotoxin Assay Kit (Genscript, Piscataway, NJ, USA). The measured values were below the accepted threshold (i.e., <0.01 ng endotoxin/20 g mouse weight). Blood was collected on heparinized capillaries by retro-

orbital bleeding 4 days after the fourth administration of FVIII. Plasma was collected and kept at -80°C until use. Mice were handled in agreement with French ethical authorities (authorization #23BA53).

Titration of anti-FVIII IgG and FVIII inhibitors

ELISA plates were coated with FVIII (1 μg/ml, Recombinate®, Baxter, Maurepas, France), in bicarbonate buffer pH 9.5, overnight

A

B

C

D

E

F

Figure 2. Characterization of FVIII containing alanine substitutions of the residues predicted to interact with BO2C11. Panel A. B domain-deleted wild-type FVIII (FVIIIHSQ), FVIIIR2215A or FVIIIR2220A were serially diluted 2-fold starting at 1 nM. Factor Xa generation was measured using a FVIII chromogenic assay. The reaction was stopped at 10 min using 20% acetic acid and the final absorbance measured at 405 nm. The data are represented as factor Xa generated in arbitrary units and are equivalent to the absorbance at 405 nm. VWF (Panel B), BO2C11 (human anti-C2 domain antibody, Panel C), ESH8 (mouse anti-C2 domain antibody that does not compete with BO2C11, Panel D), BO2BII (human anti-A2 domain antibody, Panel E) and KM33 (human anti-C1 domain antibody, Panel F) were immobilized on microtiter plates. After blocking, FVIIIHSQ, FVIIIR2215A, FVIIIR2220A, FVIIIR2215-20A or FVIIIC1 were serially diluted across the plates, and bound FVIII was revealed using either the biotinylated anti-A2 antibody, GMA-8015, or, in the case of BO2BII-bound FVIII, ESH8, followed by an anti-mouse-HRP antibody. The graphs are represented as absorbance measured at 492 nm (mean±SEM). FXa: factor Xa; VWF: von Willebrand factor; A.U.: arbitrary unit; FVIII: factor VIII.

274

haematologica | 2017; 102(2)


Immunogenic factor VIII uptake

at 4°C. After blocking with PBS, 0.05% Tween 20 and 2% BSA, the mouse plasma was incubated for 1 hr at 37°C. Serial dilutions of the samples were incubated for 1 hr at 37°C, and bound IgG were revealed using a horseradish peroxidase conjugated polyclonal goat anti-mouse IgG antibody (Southern Biotech) and the OPD substrate. Absorbance was read at 492 nm. The monoclonal mouse anti-human FVIII IgG mAb6 was used as a standard. Titers are expressed in arbitrary units (A.U.). Inhibitory titers were estimated by incubating heat-inactivated mouse plasma with human standard plasma (Siemens Healthcare Diagnostics) for 2 hr at 37°C. The residual FVIII procoagulant activity was measured using a chromogenic assay (Siemens Healthcare Diagnostics). Results are expressed in Bethesda units (BU)/ml that correspond to the reciprocal dilution of the mouse plasma that yielded 50% residual FVIII activity.

In vivo clearance of wild-type and mutated FVIII FVIII exon 16 knock-out mice and VWF/FVIII exon 16 double knock-out C57Bl/6 mice were injected intravenously with wildtype and mutated FVIII (10 nM in 100 μl). Blood was collected in 0.129 M sodium citrate at different time points after FVIII administration. FVIII residual levels in mouse plasma were determined by sandwich ELISA using ESH8 and biotinylated GMA-8015 as capture and detection antibodies, as described above. Data are plotted as a percentage of the initial FVIII level, measured 5 min-

A

B

C

D

utes after FVIII infusion, versus time (mean±SEM). Values at 5 min post-injection did not differ between the groups of mice treated with the different FVIII. Experimental data was fitted with a onephase exponential decay equation using GraphPad Prism software (version 6.01).

Results FVIII bound to BO2C11 induces diminished immune responses in vivo We first evaluated whether BO2C11, a human C2-specific anti-FVIII IgG, inhibits FVIII uptake by MoDCs or BMDCs in vitro. We hypothesized that the uptake is restricted within the BO2C11 binding region14 and thus, as a control, employed another C2-specific antibody, ESH8. ESH8 does not bind to the BO2C11 epitope and, unlike BO2C11, does not inhibit FVIII binding to VWF or phosphatidylserine membrane surfaces.19 We incubated 20 nM FVIII or FVIII pre-incubated with equimolar concentrations of anti-C2 antibodies prior to incubation with MoDCs or BMDCs. Additionally, we evaluated the inhibition caused by pre-incubating with a 2-fold molar excess of the Fab fragments of BO2C11. Following the addition of BO2C11 or the corresponding Fab fragments, we

E

Figure 3. domain mutations alter FVIII endocytosis and presentation by APCs, but do not alter FVIII immunogenicity in vivo. B domain-deleted wild-type FVIII (FVIIIHSQ) or the mutants FVIIIR2215A, FVIIIR2215-20A or FVIIIC1 were added at 20 nM to MoDCs (Panel A) or BMDCs (Panel C) for 30 min. Internalized FVIII was detected as described in Methods. Results are expressed as the percentage of MFI, whereby 100% corresponds to MFI obtained with FVIIIHSQ. Panels B and D represent activation of a FVIII-specific HLA-DRB1*0101-restricted T-cell hybridoma by B domain-deleted FVIII-loaded (10 nM) HLA-matched human MoDCs or splenocytes from SURE-L1 mice, respectively. Supernatant was collected after 24 hr and the IL-2 produced by the activated T cells was measured. Representative of three experiments (mean±SEM). Panel E. FVIII-deficient mice were injected intravenously once weekly for 4 weeks with 1 μg of B domain-deleted FVIIIHSQ, FVIIIC1 or FVIIIR2215-20A. One week after the last injection, blood samples were collected. Anti-FVIII IgG titers are defined as arbitrary units based on standard curves generated using mAb6. Statistical significances were assessed using the two-tailed nonparametric Mann-Whitney U test. ns: not significant; FVIII: factor VIII; MoDCs: monocyte-derived dendritic cells; A.U.: arbitrary unit; IL-2: interleukin-2; IgG: immunoglobulin G; MFI: median fluorescence intensity; BMDCs: bone marrow-derived dendritic cells.

haematologica | 2017; 102(2)

275


B. Gangadharan et al.

observed a more than 70% reduction in FVIII internalization by MoDCs (Figure 1A), and a reduction of about 30% in the case of BMDCs (Online Supplementary Figure 2A). The addition of ESH8 did not reduce FVIII uptake by MoDCs. Similarly, BO2C11 or Fab of BO2C11 inhibited FVIII presentation to a HLA-DR1-restricted CD4+ T-cell hybridoma by more than 80% in the case of both MoDCs (Figure 1B) and splenocytes purified from HLA-DR1 Tg SURE-L1 mice (Online Supplementary Figure S2B), used as sources of APCs. Together, our results implicate a role for C2 membrane-binding residues in FVIII uptake by APCs. We next investigated the effect of BO2C11 on FVIII immunogenicity in vivo. We administered FVIII pre-incubated with F(ab’)2 of an isotype control or of BO2C11 for 4 weeks at weekly intervals. After 4 weeks, BO2C11bound FVIII exhibited diminished immunogenicity compared to the isotype control-treated mice (Figure 1C).

A

B

C

D

Characterization of FVIII variants mutated in the BO2C11 epitope We generated FVIII variants, wherein the two arginine residues located at position 2215 and 2220 that belong to the BO2C11 epitope, are mutated to alanine residues. The purified FVIII mutants exhibited specific activities between 4800-8000 IU/mg (Figure 2A) similar to that of non-mutated FVIII (FVIIIHSQ). Substitutions at either R2215 or R2220 did not alter binding to VWF, while substitution of both R2215 and R2220 lead to a marginally reduced binding to VWF (Figure 2B). Substitutions at R2215 or R2220 resulted in diminished binding to BO2C11 (Figure 2C), with substitution at R2220 resulting in a near complete inhibition of FVIII interaction with BO2C11. Double mutation of these residues (FVIIIR2215-20A) provided little additional benefit, confirming that R2220 contributes to most of the binding to BO2C11. Importantly, these FVIII

E

Figure 4. FVIII binding to VWF does not alter endocytosis in vitro or immunogenicity in vivo. Panel A. The binding of B domain-deleted FVIIIHSQ and FVIIIY1680C to VWF was studied using an ELISA as described in the Methods. Results (mean±SEM) are expressed in arbitrary units. Panel B. FVIIIHSQ (10 nM, open squares) or FVIIIY1680C (closed squares) in 100 μl were administered into FVIII-deficient mice and the residual FVIII levels were measured at different time points (n=3 mice per time point) by ELISA (FVIII:Ag). The data is plotted as a percentage of the initial FVIII level (measured 5 minutes after administration, mean±SEM) versus time. Panel C. FVIIIHSQ or mutants FVIIIY1680C were added at increasing concentrations (10, 20 and 40 nM) to immature MoDCs (n=3 different donors) for 30 min at 37°C in serum-free medium and in the absence of VWF. Cells were subsequently washed, fixed, permeabilized and stained for FVIII using a FITC-labeled anti-A2 domain-specific antibody followed by fluorescence-activated cell sorting (FACS). Data are expressed as the percentage of MFI (mean±SEM), whereby 100% corresponds to the MFI observed for 10 nM of FVIIIHSQ. Panels D and E. FVIII-deficient mice were administered with FVIIIHSQ (0.5 µg, open squares) or FVIIIY1680C (0.5 μg, closed squares), intravenously once a week for 4 weeks. Anti-FVIII IgG titers were measured using an ELISA as described above (D). Inhibitory titers towards FVIII were assessed using a modified Bethesda assay (E). Horizontal bars depict medians. The statistical significance was assessed using the two-tailed non-parametric Mann-Whitney U test. ns: not significant; FVIII: factor VIII; VWF: von Willebrand factor; A.U.: arbitrary unit; MFI: median fluorescence intensity; IgG: immunoglobulin G; Ag: antigen.

276

haematologica | 2017; 102(2)


Immunogenic factor VIII uptake

variants retained binding to antibodies targeting other domains of FVIII. In particular, the mutations did not have significant effects on the ability of FVIII to interact with ESH8 (mouse anti-C2 IgG), BO2BII (human anti-A2 IgG) or KM33 (human anti-C1 IgG, Figure 2D-F). Recently, the three residues R2090A/F2092A/K2093A in the FVIII C1 domain that belong to the KM33 epitope, were implicated in FVIII uptake.11 Hence, as a control, we generated the R2090A/F2092A/K2093A FVIII mutant, referred to as “FVIIIC1”. As expected, FVIIIC1 showed drastically reduced binding to KM33 (Figure 2F), and unaltered binding to BO2C11 and ESH8 (Figure 2C-D).

utes (Figure 4B). We also confirmed that the Y1680C mutation does not alter the in vitro endocytosis of FVIII by MoDCs (Figure 4C). Next, we performed comparative immunogenicity studies of FVIIIHSQ and FVIIIY1680C using FVIII-deficient mice. Remarkably, FVIIIY1680C was as immunogenic as FVIIIHSQ (Figure 4D). The inhibitor titers were also not significantly different (Figure 4E). In addition, the immunogenicity of FVIIIY1680C in FVIII-deficient mice was similar to that of FVIIIHSQ in double FVIII/VWFdeficient mice (data not shown).

C2 domain residues are implicated in FVIII uptake by APCs

Having demonstrated that the ability of FVIII to interact with endogenous VWF does not modulate the magnitude of the anti-FVIII immune response, we investigated the immune protective effect of BO2C11 using FVIIIY1680C. FVIIIY1680C alone or pre-incubated with a 2 molar excess of BO2C11 Fab fragments was administered to FVIII-deficient mice at weekly intervals for 4 weeks. Five days after the last FVIII administration, the anti-FVIII immune response was measured. We observed a significant reduction both in the anti-FVIII antibody response (Figure 5A), and FVIII-specific T-cell proliferation (Online Supplementary Figure S4), when FVIIIY1680C was pre-incubated with BO2C11 Fab fragments. In contrast, in double FVIII/VWFdeficient mice, the immunogenicity of FVIIIR2215-20A (1300±303 μg/ml, 90±17 BU/ml, mean±SEM) was not statistically different from that of FVIIIHSQ (2258±596 μg/ml, 180±62 BU/ml, Figure 5C,D, P>0.2), despite an increased residence time in the circulation of FVIIIR2215-20A (Figure 5B, 41 min [95% confidence interval (CI): 39-51]) as compared to FVIIIHSQ (17 min [13-23]). Conversely, FVIIIC1 showed a statistically significant reduction in immunogenicity in double FVIII/VWF-deficient mice as compared to FVIIIHSQ, both in terms of levels of anti-FVIII IgG (460±144 and 2258±596 μg/ml, respectively; P<0.001) and inhibitory titers (38±17 and 180±62 BU/ml, respectively; P<0.05). Similar observations were obtained with a FVIII mutant that combined the triple C1 mutation and a replacement of the R2215 residue with a serine (Online Supplementary Figure S5).

We next evaluated the role of C2 residues in FVIII uptake and presentation. Human MoDCs and murine BMDCs were incubated with 20 nM FVIII, fixed and permeabilized, and FVIII was revealed with the anti-A2 antibody GMA-8015. FVIII uptake was dose-dependent (Online Supplementary Figure S3) and significantly reduced for FVIIIR2215A and FVIIIR2215-20A (Figure 3A and 3C). We confirmed our observation in an antigen presentation assay to a HLA-DR1-restricted CD4+ T-cell hybridoma using MoDCs derived from a HLA-DR1 donor, and using splenocytes from HLA-DR1 transgenic SURE-L1 mice (Figure 3B and 3D). Importantly, the epitope of CD4+ T-cell hybridoma is located in the 2004-2031 amino acid stretch at the A3-C1 junction of FVIII (data not shown); it is therefore distant from the mutated residues, thus ruling out the fact that the lack of T-cell activation is consecutive to a disruption of the T-cell epitope. We observed a significant decrease in FVIII presentation by human and mouse dendritic cells (DCs), as measured by reduced IL-2 secretion in the case of both FVIIIR2215A and FVIIIR2215A-R20A. Our present data indicate that the C2 domain participates in FVIII uptake and involves at least R2215. To further evaluate the importance of the C2 mutants in in vivo immunogenic uptake, we compared the immunogenicity of the different FVIII variants in FVIII-deficient mice. Cage-matched siblings were administered intravenously 4 times weekly with 1 μg of FVIIIHSQ, FVIIIR2215-20A or FVIIIC1, used as a control. In contrast to previous reports,11 FVIII-deficient mice generated an antibody response not only to FVIIIHSQ but also to FVIIIC1 (Figure 3E). In addition, FVIIIR2215-20A also presented with the same immunogenicity as FVIIIHSQ. The data suggest that mutating residues in the C1 and C2 domains of FVIII, at least in our hands, has no effect on FVIII immunogenicity in FVIIIdeficient mice in vivo.

The immunogenicity of FVIII is independent from its ability to interact with endogenous VWF Because BO2C11 inhibits FVIII interaction with VWF, we investigated the possibility that FVIII immunogenicity in the presence of BO2C11 (Figure 1C) was diminished due to its faster clearance. To address this potential bias, we generated a FVIII variant, FVIIIY1680C, which cannot associate with VWF. The specific activities of FVIIIHSQ and FVIIIY1680C were similar as assessed by one-stage and chromogenic assay (7300±165 and 6300±185 IU/mg, respectively, mean±SD). FVIIIY1680C exhibited a significant reduction in its ability to interact with VWF by ELISA (Figure 4A). Accordingly, FVIIIY1680C exhibited a rapid clearance in FVIII-deficient mice with a t1/2 of approximately 20 minhaematologica | 2017; 102(2)

C domain mutations exhibit reduced immunogenicity only in the absence of VWF

Discussion The endocytosis of an antigen by APCs is not sufficient for the induction of a primary CD4+ T-cell-dependent immune response. It is however a prerequisite to its presentation to naïve antigen-specific CD4+ T cells. Over the years, several receptors have been identified for FVIII5,6,20,21 that have mostly been implicated in FVIII catabolism.8,22 To date, the only receptor incriminated in FVIII internalization by APCs is the C-type lectin receptor CD206, that binds mannose-ending glycans linked to Asn239 in the A1 domain of FVIII, and Asn2118 in the C1 domain.7 The importance of the C1 domain of FVIII in binding to phospholipids and participating in FVIII endocytosis has recently been proposed. Thus, Lys 2092 and Phe 2093 in C1 were shown to participate in binding to phosphatidylserine and to platelets.23,24 Interestingly, co-incubation of FVIII with KM33, a monoclonal anti-C1 IgG that targets the R2090, K2092 and F2093 residues, prior to injection into FVIII-deficient mice, was associated with a decrease in FVIII immunogenicity.10 Additionally, a triple 277


B. Gangadharan et al.

R2090A/K2092A/F2093A FVIII mutant demonstrated reduced endocytosis by MoDCs and macrophages, and reduced immunogenicity in FVIII-deficient mice.11 Because the C2 domain of FVIII shares structural homology with the C1 domain,25 and is a major membrane-binding motif26,27 containing several charged/basic residues, we investigated whether it also participates in the endocytic process by APCs, that leads to FVIII processing and presentation to CD4+ T cells. Our results demonstrate that masking part of the C2 domain with BO2C11, a human

A

P<0.01

C

monoclonal IgG that interacts with solvent-exposed basic and hydrophobic side chains on the C2 membrane-binding loops,14 reduces FVIII endocytosis by MoDCs and presentation to T cells in vitro, as well as immunogenicity in mice. The protective effect of BO2C11 was epitopespecific, since ESH8, a non-overlapping mouse monoclonal anti-C2 IgG, failed to block FVIII uptake in vitro. ESH8 in fact increased the uptake process. Since the binding of ESH8 to FVIII was proposed to prevent a conformational change in the C2 domain of the FVIII light chain,28 our

B

D

P=0.021

P<0.01

Figure 5. Contribution of the C1 and C2 domains of FVIII to FVIII immunogenicity in the absence of binding to VWF. Panel A. FVIII-deficient mice were administered with B domain-deleted FVIIIY1680C (0.4 μg, open squares) or FVIIIY1680C pre-incubated with a 2-fold molar excess of BO2C11 Fab (closed squares) intravenously once a week for 4 weeks. One week after the last injection, blood samples were collected. Anti-FVIII IgG titers are defined as arbitrary units based on standard curves generated using mAb6. The graph depicts a pool of two independent experiments. Panel B. B domain-deleted FVIIIHSQ, FVIIIC1 or FVIIIR2215-20A (10 nM in 100 μl) were administered to FVIII-deficient mice and the residual FVIII levels were measured at different time points (n=3 mice per time point) by ELISA. The data are plotted as a percentage of the initial FVIII levels (measured 5 minutes after administration) versus time (mean±SEM) and are representative of 2 independent experiments. Experimental data were fitted using a one-phase decay curve to determine the in vivo half-lives. Panels C and D. Double FVIII/VWF-deficient mice were injected intravenously once a week for 4 weeks with 0.5 μg of FVIIIHSQ, FVIIIC1 or FVIIIR2215-20A. One week after the last injection, blood samples were collected. Anti-FVIII IgG titers were measured as described above (C). Inhibitory titers towards FVIII were assessed using a modified Bethesda assay (D). Horizontal bars depict medians. Statistical significances were assessed using the two-tailed non-parametric Mann-Whitney U test. ns: not significant: FVIII: factor VIII; IgG: immunoglobulin G; A.U.: arbitrary unit; Fab: fragment antigen binding; VWF: von Willebrand factor; Ag: antigen.

278

haematologica | 2017; 102(2)


Immunogenic factor VIII uptake

observation may partly relate to a particular conformation of ESH8-bound FVIII, that orients FVIII to facilitate the uptake process. Interestingly, the mutation of key charged residues on C2 that belong to the BO2C11 epitope, resumed the effect of BO2C11 on FVIII endocytosis and presentation in vitro. Our results are thus reminiscent of the aforementioned findings on the role played by the FVIII C1 domain in immunogenic FVIII uptake. Together, the data demonstrate that both the C1 and C2 domains of FVIII play a role in FVIII endocytosis by MoDCs. Since lactadherin, a potent competitor for phosphatidylserine binding, does not inhibit FVIII endocytosis by MoDCs,11 FVIII uptake is probably independent of binding to phospholipids. Rather, we speculate that FVIII uptake in the absence of VWF in vitro involves charged interactions with a putative highly negatively charged receptor, or a potential role for co-receptors, such as heparan sulfate proteoglycans. The respective contribution, either redundant, additive or synergistic, to FVIII internalization of residues in the C1 and C2 domains, and of mannose-ending Nlinked glycans in the A1 and C1 domains, as well as the sequence of events that allows FVIII binding at the cell surface and internalization, remain to be deciphered. While co-incubation of FVIII prior to injection into FVIIIdeficient mice, with BO2C11 or with KM3310 reduced the anti-FVIII IgG response, neither the FVIIIR2215-20A C2 mutant nor the triple FVIIIC1 mutant demonstrated reduced immunogenicity in FVIII-deficient mice in our hands. These observations may be explained either by a steric hindrance of the C domains of FVIII by the antigen-binding domains of BO2C11/KM33, or by the implication in FVIII internalization of additional amino acids within the BO2C11 epitope on C2, or within the KM33 epitope on C1. Alternatively, the findings may be accounted for by an interfering effect of endogenous VWF. Indeed, the T2086-S2095 region of the C1 domain was recently shown to be buried upon association with VWF,29 and the triple FVIIIC1 mutant retains the ability to interact with VWF. In support of this, in the absence of endogenous VWF, the triple FVIIIC1 mutant showed reduced immunogenicity as compared to FVIIIHSQ in vivo. In contrast, the FVIIIR2215-20A C2 mutant, that also retains most of VWF binding and demonstrates reduced endocytosis in vitro, was as immunogenic as FVIIIHSQ in mice lacking VWF. Of note, the lack of reduced immunogenicity of the triple FVIIIC1 mutant in FVIII-deficient mice is at odds with the report by Wroblewska et al.;11 it remains unclear whether the observed difference may relate to differences in levels of VWF in different strains of mice.30 Although the C2 domain has been identified as an essential membrane interactive motif, with the C1 domain providing additional membrane binding affinity,23,26,27 the results suggest that C2 residues do not play a predominant role in immunogenic FVIII in the absence of endogenous VWF. Instead, we speculate that the in vivo role in endocytosis of the targeted R2215 and R2220 basic residues in the C2 domain is only secondary to that of other membrane accessible residues within the BO2C11 epitope. This is supported by the observation that the pre-incubation of the FVIIIY1680C mutant, which does not bind VWF, with BO2C11, reduced its immunogenicity in FVIII-deficient mice. Additional mutational analysis of membrane accessible hydrophobic residues within the BO2C11 epitope shall shed light on the specific residues of the C2 domain of FVIII that are implicated in the in vivo immunogenic uptake process. haematologica | 2017; 102(2)

VWF has controversially been proposed to reduce the immunogenicity of therapeutic FVIII.31 Unexpectedly, the present results indicate that the presence of endogenous VWF, or the capacity of exogenous FVIII to bind endogenous VWF, does not alter the immunogenicity of FVIII in mice. Indeed, FVIII presented with the same degree of immunogenicity in FVIII-deficient mice and in double FVIII/VWF-deficient mice. Likewise, FVIIIY1680C with impaired binding to VWF induced similar anti-FVIII IgG levels as native FVIII in FVIII-deficient mice. The potential immunomodulatory role of VWF towards FVIII has been suggested by a suspected reduced prevalence of FVIII inhibitors in hemophilia A patients treated with exogenous VWF-containing plasma FVIII concentrates, as compared to patients receiving recombinant or highly purified products.32,33 Controversial results have, however, been obtained upon studying large retrospective and prospective patient cohorts.34,35 In parallel, studies performed in pre-clinical mouse models of hemophilia A have indicated that pre-incubation of recombinant FVIII with exogenous VWF leads to a reduction in the levels of inhibitory anti-FVIII antibodies following administration to mice,36-39 although contradictory results have occasionally been generated.40 In vitro, a protective role for VWF on FVIII endocytosis by human MoDCs was clearly shown,10,41 although presentation of processed FVIIIderived peptides could still be detected, at least in vitro.42 In addition, under shear stress, VWF fails to block FVIII internalization by macrophages, and both VWF and FVIII co-localize within the same cells.43 This is in agreement with the observation that exogenous FVIII and exogenous VWF may be co-detected in splenic macrophages after injection into double FVIII/VWF-deficient mice.44 The internalization of FVIII and VWF was, however, inhibited by the LRP antagonist receptor-associated protein (RAP) in the case of macrophages under shear stress, but not in the case of MoDCs in static conditions,8 suggesting that this process is more relevant to FVIII catabolism than immunogenicity. Interestingly, the intricate role played by VWF in FVIII immunogenicity is reminiscent of its complex role towards FVIII catabolism. Indeed, binding of FVIII to VWF dictates FVIII residence time in the circulation, as shown in patients with type 3 von Willebrand disease.45 Conversely, VWF-binding may mediate, at least in part, the catabolism of therapeutic FVIII,20,46 which is further suggested by the fact that the half-life of modified Fc-fused or PEGylated FVIII barely exceeds that of VWF.47,48 Taken together, the results indicate that residues in the C1 and/or C2 domains of FVIII are implicated in immunogenic FVIII uptake, at least in vitro. Conversely, in vivo, the binding to endogenous VWF masks the reducing effect of mutations in the C domains on FVIII immunogenicity. Acknowledgments KM33 and VWF-deficient mice were kind gifts from Dr Jan Voorberg (Department of Plasma Proteins, Sanquin-AMC Landsteiner Laboratory and Van Creveld Laboratory, Amsterdam, The Netherlands) and Dr Olivier Christophe (Institut National de la Santé et de la Recherche Médicale U770, Le Kremlin-Bicêtre, France), respectively. FVIIIHSQ in the ReNeo plasmid and BHK-M cells were kind gifts from Prof Pete Lollar (Aflac Cancer and Blood Disorders Center, Department of Pediatrics, Emory University, Atlanta, GA, USA). The monoclonal mouse and human anti-FVIII antibodies, mAb6, BO2BII 279


B. Gangadharan et al.

and BO2C11 were kind gifts from Drs JM Saint-Remy and M Jacquemin (KUL, Leuven, Belgium). We would like to extend our thanks to the Centre d’Explorations Fonctionnelles (Centre de Recherche des Cordeliers, Paris, France) for assistance. Funding This work was supported by Institut national de la santé et de

References 1. Fay PJ. Factor VIII structure and function. Int J Hematol. 2006;83(2):103-108. 2. Lenting PJ, van Mourik JA, Mertens K. The life cycle of coagulation factor VIII in view of its structure and function. Blood. 1998;92(11):3983-3996. 3. Ehrenforth S, Kreuz W, Scharrer I, et al. Incidence of development of factor VIII and factor IX inhibitors in haemophiliacs. Lancet. 1992;339(8793):594-598. 4. Lacroix-Desmazes S, Navarrete AM, Andre S, Bayry J, Kaveri SV, Dasgupta S. Dynamics of factor VIII interactions determine its immunologic fate in hemophilia A. Blood. 2008;112(2):240-249. 5. Lenting P, Neels J, van den Berg B, et al. The light chain of factor VIII comprises a binding site for low density lipoprotein receptor-related protein. J Biol Chem. 1999; 274(34):23734-23739. 6. Saenko E, Yakhyaev A, Mikhailenko I, Strickland D, Sarafanov A. Role of the low density lipoprotein-related protein receptor in mediation of factor VIII catabolism. J Biol Chem. 1999;274:37685-37692. 7. Dasgupta S, Navarrete AM, Bayry J, et al. A role for exposed mannosylations in presentation of human therapeutic self-proteins to CD4+ T lymphocytes. Proc Natl Acad Sci U S A. 2007;104(21):8965-8970. 8. Dasgupta S, Navarrete AM, Andre S, et al. Factor VIII bypasses CD91/LRP for endocytosis by dendritic cells leading to T-cell activation. Haematologica. 2008;93(1):83-89. 9. Repesse Y, Dasgupta S, Navarrete AM, Delignat S, Kaveri SV, Lacroix-Desmazes S. Mannose-sensitive receptors mediate the uptake of factor VIII therapeutics by human dendritic cells. J Allergy Clin Immunol. 2012;129(4):1172-1173. 10. Herczenik E, van Haren SD, Wroblewska A, et al. Uptake of blood coagulation factor VIII by dendritic cells is mediated via its C1 domain. J Allergy Clin Immunol. 2012; 129(2):501-509, 9 e1-5. 11. Wroblewska A, van Haren SD, Herczenik E, et al. Modification of an exposed loop in the C1 domain reduces immune responses to factor VIII in hemophilia A mice. Blood. 2012;119(22):5294-5300. 12. Meems H, van den Biggelaar M, Rondaij M, van der Zwaan C, Mertens K, Meijer AB. C1 domain residues Lys 2092 and Phe 2093 are of major importance for the endocytic uptake of coagulation factor VIII. Int J Biochem Cell Biol. 2011;43(8):1114-1121. 13. Lu J, Pipe SW, Miao H, Jacquemin M, Gilbert GE. A membrane-interactive surface on the factor VIII C1 domain cooperates with the C2 domain for cofactor function. Blood. 2011;117(11):3181-3189. 14. Spiegel PC, Jr., Jacquemin M, Saint-Remy JM, Stoddard BL, Pratt KP. Structure of a factor VIII C2 domain-immunoglobulin G4kappa Fab complex: identification of an

280

15.

16.

17. 18.

19.

20.

21.

22.

23.

24.

25.

26.

la recherche médicale (INSERM), Centre National de la Recherche Scientifique (CNRS), and Université Pierre et Marie Curie (UPMC) Paris 6 (France), and by an unrestricted grant from Grifols (Investigator-Sponsored Research Program, Barcelona, Spain). MI and IP were the recipients of fellowships from Ministère de l’enseignement supérieur et de la recherche (France).

inhibitory antibody epitope on the surface of factor VIII. Blood. 2001;98(1):13-19. Healey JF, Barrow RT, Tamim HM, et al. Residues Glu2181-Val2243 contain a major determinant of the inhibitory epitope in the C2 domain of human factor VIII. Blood. 1998;92(10):3701-3709. Lutz MB, Kukutsch N, Ogilvie AL, et al. An advanced culture method for generating large quantities of highly pure dendritic cells from mouse bone marrow. J Immunol Methods. 1999;223(1):77-92. Delignat S, Repesse Y, Gilardin L, et al. Predictive immunogenicity of Refacto AF. Haemophilia. 2014;20(4):486-492. Pfistershammer K, Stockl J, Siekmann J, Turecek PL, Schwarz HP, Reipert BM. Recombinant factor VIII and factor VIIIvon Willebrand factor complex do not present danger signals for human dendritic cells. Thromb Haemost. 2006;96(3):309316. Meeks SL, Healey JF, Parker ET, Barrow RT, Lollar P. Antihuman factor VIII C2 domain antibodies in hemophilia A mice recognize a functionally complex continuous spectrum of epitopes dominated by inhibitors of factor VIII activation. Blood. 2007;110(13):4234-4242. Sarafanov A, Ananyeva N, Shima M, Saenko E. Cell surface heparan sulfate proteoglycans participate in factor viii catabolism mediated by low density lipoprotein receptor-related protein. J Biol Chem. 2001;276(15):11970-11979. Bovenschen N, Rijken DC, Havekes LM, Vlijmen BJ, Mertens K. The B domain of coagulation factor VIII interacts with the asialoglycoprotein receptor. J Thromb Haemost. 2005;3(6):1257-1265. Navarrete AM, Dasgupta S, Teyssandier M, et al. Endocytic receptor for pro-coagulant factor VIII: relevance to inhibitor formation. Thromb Haemost. 2010;104(6):10931098. Meems H, Meijer AB, Cullinan DB, Mertens K, Gilbert GE. Factor VIII C1 domain residues Lys 2092 and Phe 2093 contribute to membrane binding and cofactor activity. Blood. 2009;114(18):3938-3946. Bloem E, van den Biggelaar M, Wroblewska A, et al. Factor VIII C1 domain spikes 20922093 and 2158-2159 comprise regions that modulate cofactor function and cellular uptake. J Biol Chem. 2013;288(41):2967029679. Ngo JC, Huang M, Roth DA, Furie BC, Furie B. Crystal structure of human factor VIII: implications for the formation of the factor IXa-factor VIIIa complex. Structure. 2008;16(4):597-606. Scandella D, Gilbert GE, Shima M, et al. Some Factor VIII inhibitor antibodies recognize a common epitope corresponding to C2 domain amino acids 2248 through 2312, which overlap a phospholipid-binding site. Blood. 1995;86(5):1811-1819.

27. Gilbert GE, Kaufman RJ, Arena AA, Miao H, Pipe SW. Four hydrophobic amino acids of the factor VIII C2 domain are constituents of both the membrane-binding and von Willebrand factor-binding motifs. J Biol Chem. 2002;277(8):6374-6381. 28. Saenko EL, Shima M, Gilbert GE, Scandella D. Slowed release of thrombin-cleaved factor VIII from von Willebrand factor by a monoclonal and a human antibody is a novel mechanism for FVIII inhibition. J Biol Chem. 1996;271(44):27424-27431. 29. Chiu PL, Bou-Assaf GM, Chhabra ES, et al. Mapping the interaction between factor VIII and von Willebrand factor by electron microscopy and mass spectrometry. Blood. 2015;126(8):935-938. 30. Shavit JA, Manichaikul A, Lemmerhirt HL, Broman KW, Ginsburg D. Modifiers of von Willebrand factor identified by natural variation in inbred strains of mice. Blood. 2009; 114(26):5368-5374. 31. Oldenburg J, Lacroix-Desmazes S, Lillicrap D. Alloantibodies to therapeutic factor VIII in hemophilia A: the role of von Willebrand factor in regulating factor VIII immunogenicity. Haematologica. 2015;100(2):149156. 32. Goudemand J, Rothschild C, Demiguel V, et al. Influence of the type of factor VIII concentrate on the incidence of factor VIII inhibitors in previously untreated patients with severe hemophilia A. Blood. 2006; 107(1):46-51. 33. Calvez T, Laurian Y, Goudemand J. Inhibitor incidence with recombinant vs. plasma-derived FVIII in previously untreated patients with severe hemophilia A: homogeneous results from four published observational studies. J Thromb Haemost. 2008;6(2):390-392. 34. Gouw SC, van der Bom JG, Auerswald G, Escuriola Ettinghausen C, Tedgard U, van den Berg HM. Recombinant versus plasmaderived factor VIII products and the development of inhibitors in previously untreated patients with severe hemophilia A: the CANAL cohort study. Blood. 2007; 109(11):4693-4697. 35. Gouw SC, van der Bom JG, Ljung R, et al. Factor VIII products and inhibitor development in severe hemophilia A. N Engl J Med. 2013;368(3):231-239. 36. Behrmann M, Pasi J, Saint-Remy JM, Kotitschke R, Kloft M. Von Willebrand factor modulates factor VIII immunogenicity: comparative study of different factor VIII concentrates in a haemophilia A mouse model. Thromb Haemost. 2002;88(2):221229. 37. Delignat S, Dasgupta S, Andre S, et al. Comparison of the immunogenicity of different therapeutic preparations of human factor VIII in the murine model of hemophilia A. Haematologica. 2007;92(10):14231426. 38. Kallas A, Kuuse S, Maimets T, Pooga M.

haematologica | 2017; 102(2)


Immunogenic factor VIII uptake

von Willebrand factor and transforming growth factor-beta modulate immune response against coagulation factor VIII in FVIII-deficient mice. Thromb Res. 2007; 120(6):911-919. 39. Delignat S, Repesse Y, Navarrete AM, et al. Immunoprotective effect of von Willebrand factor towards therapeutic factor VIII in experimental haemophilia A. Haemophilia. 2012;18(2):248-254. 40. Qadura M, Waters B, Burnett E, et al. Recombinant and plasma-derived factor VIII products induce distinct splenic cytokine microenvironments in hemophilia A mice. Blood. 2009 ;114(4):871-880. 41. Dasgupta S, Repesse Y, Bayry J, et al. VWF protects FVIII from endocytosis by dendritic cells and subsequent presentation to

haematologica | 2017; 102(2)

42.

43.

44.

45.

immune effectors. Blood. 2007;109(2):610612. Sorvillo N, Hartholt RB, Bloem E, et al. von Willebrand factor binds to the surface of dendritic cells and modulates peptide presentation of factor VIII. Haematologica. 2016;101(3):309-318. Castro-Nunez L, Dienava-Verdoold I, Herczenik E, Mertens K, Meijer AB. Shear stress is required for the endocytic uptake of the factor VIII-von Willebrand factor complex by macrophages. J Thromb Haemost. 2012;10(9):1929-1937. van Schooten CJ, Shahbazi S, Groot E, et al. Macrophages contribute to the cellular uptake of von Willebrand factor and factor VIII in vivo. Blood. 2008;112(5):1704-1712. Goodeve AC, Eikenboom JC, Ginsburg D,

et al. A standard nomenclature for von Willebrand factor gene mutations and polymorphisms. On behalf of the ISTH SSC Subcommittee on von Willebrand factor. Thromb Haemost. 2001;85(5):929-931. 46. Pegon JN, Kurdi M, Casari C, et al. Factor VIII and von Willebrand factor are ligands for the carbohydrate-receptor Siglec-5. Haematologica. 2012;97(12):1855-1863. 47. Mahlangu J, Powell JS, Ragni MV, et al. Phase 3 study of recombinant factor VIII Fc fusion protein in severe hemophilia A. Blood. 2014;123(3):317-325. 48. Konkle BA, Stasyshyn O, Chowdary P, et al. Pegylated, full-length, recombinant factor VIII for prophylactic and on-demand treatment of severe hemophilia A. Blood. 2015;126(9):1078-1085.

281


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Platelet Biology & Its Disorders

Ferrata Storti Foundation

Haematologica 2017 Volume 102(2):282-294

Germline variants in ETV6 underlie reduced platelet formation, platelet dysfunction and increased levels of circulating CD34+ progenitors

Marjorie Poggi,1,* Matthias Canault,1,* Marie Favier,1,2,* Ernest Turro,3,4,* Paul Saultier,1 Dorsaf Ghalloussi,1 Veronique Baccini,1 Lea Vidal,1 Anna Mezzapesa,1 Nadjim Chelghoum,5 Badreddine Mohand-Oumoussa,5 Céline Falaise,6 Rémi Favier,7 Willem H. Ouwehand,3,8 Mathieu Fiore,6,9 Franck Peiretti, 1 Pierre Emmanuel Morange,1,6 Noémie Saut,1,6 Denis Bernot,1 Andreas Greinacher,10 NIHR BioResource,11 Alan T. Nurden,12 Paquita Nurden,6,12 Kathleen Freson,13,* David-Alexandre Trégouët,14,15,16,* Hana Raslova2* and Marie-Christine Alessi1,6,*

Aix Marseille Univ, INSERM, INRA, NORT, Marseille, France; 2Inserm U1170, Gustave Roussy, University Paris Sud, Equipe labellisée Ligue contre le Cancer 94805 Villejuif, France; 3Department of Haematology and National Health Service Blood & Transplant, Cambridge University, UK; 4MRC Biostatistics Unit, Cambridge, UK; 5Post-Genomic Platform of Pitié-Salpêtrière (P3S), Pierre and Marie Curie University, F-75013 Paris, France; 6French Reference-Center on Inherited Platelet Disorders, Marseille, France; 7Assistance PubliqueHôpitaux de Paris, Hôpital Armand Trousseau, Paris, France; 8Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK; 9Laboratoire d’hématologie, CHU de Bordeaux, Pessac, France; 10Institute for Immunology and Transfusion Medicine, University Medicine Greifswald, Germany; 11NIHR BioResource - Rare Diseases, Cambridge University Hospitals, Cambridge Biomedical Campus, UK; 12LIRYC, Plateforme Technologique et d’Innovation Biomédicale, Hôpital Xavier Arnozan, Pessac, France; 13 Department of Cardiovascular Sciences, Center for Molecular and Vascular Biology, KU Leuven, Belgium; 14ICAN Institute of Cardiometabolism and Nutrition, F-75013 Paris, France; 15Inserm, UMR_S 1166, Team Genomics and Pathophysiology of Cardiovascular Diseases, F-75013 Paris, France and 16Sorbonne Universités, Université Pierre et Marie Curie (UPMC Univ Paris 06), UMR_S 1166, F-75013 Paris, France 1

*MP, MC, MF, ET, KF, D-A T, HR and M-C A contributed equally to this work.

ABSTRACT

Correspondence: marie-christine.alessi@univ-amu.fr

Received: April 18, 2016. Accepted: September 22, 2016. Pre-published: September 23, 2016. doi:10.3324/haematol.2016.147694

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/282

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

282

V

ariants in ETV6, which encodes a transcription repressor of the E26 transformation-specific family, have recently been reported to be responsible for inherited thrombocytopenia and hematologic malignancy. We sequenced the DNA from cases with unexplained dominant thrombocytopenia and identified six likely pathogenic variants in ETV6, of which five are novel. We observed low repressive activity of all tested ETV6 variants, and variants located in the E26 transformation-specific binding domain (encoding p.A377T, p.Y401N) led to reduced binding to corepressors. We also observed a large expansion of megakaryocyte colony-forming units derived from variant carriers and reduced proplatelet formation with abnormal cytoskeletal organization. The defect in proplatelet formation was also observed in control CD34+ cell-derived megakaryocytes transduced with lentiviral particles encoding mutant ETV6. Reduced expression levels of key regulators of the actin cytoskeleton CDC42 and RHOA were measured. Moreover, changes in the actin structures are typically accompanied by a rounder platelet shape with a highly heterogeneous size, decreased platelet arachidonic response, and spreading and retarded clot retraction in ETV6 deficient platelets. Elevated numbers of circulating CD34+ cells were found in p.P214L and p.Y401N carriers, and two patients from different families suffered from refractory anemia with excess blasts, while one patient from a third family was successfully treated for acute myeloid leukemia. Overall, our study provides novel insights into the role of ETV6 as a driver of cytoskeletal regulatory gene expression during platelet production, and the impact of variants resulting in platelets with altered size, shape and function and potentially also in changes in circulating progenitor levels. haematologica | 2017; 102(2)


ETV6 and platelet formation

Introduction The genetic determinants of non syndromic autosomal dominant (AD) thrombocytopenia with normal platelet size remain largely unknown, yet it is important to identify such variants because they may predispose carriers to hematological malignancy. Germline variants in RUNX1 cause a familial platelet disorder with an increased risk of acute myeloid leukemia (FPD/AML), while variants in the 5′ untranslated region (UTR) of ANKRD26 have also been shown to predispose individuals to hematologic malignancies. Recently, germline variants in ETV6 (TEL) have been reported to underlie AD thrombocytopenia with predisposition to leukemia.1-3 ETV6, which was initially identified as encoding a tumor suppressor in humans, is often found fused with partner genes in samples from human leukemia of myeloid and lymphoid origin.4 Somatic ETV6 variants have also been found in solid tumors, T-cell leukemias and myelodysplastic syndromes, hence the widespread interest in this gene.5,6 ETV6 encodes an E26 transformation-specific (Ets) family transcriptional repressor. It can bind DNA via a highly conserved Ets DNA-binding consensus site located at the C-terminus. The N-terminal domain (pointed domain) is necessary for homotypic dimerization and interaction with the Ets family protein FLI.7,8 The central region is involved in repressive complex recruitment (including SMRT, Sin3A and NCOR)9 and autoinhibitory activity.10 ETV6 plays an important role in hematopoiesis. In mice, ETV6 is essential for hematopoietic transition from the fetal liver to the bone marrow (BM).11 Conditional disruption of the ETV6 gene has shown that ETV6 plays a unique, non redundant role in megakaryocytopoiesis. Data concerning ETV6 involvement in megakaryocytopoiesis in humans remains scarce, however, a recent study has shown that patients expressing a mutated form of ETV6 displayed abnormal megakaryocyte (MK) development with a likely impact on platelet production.1 We have assessed the biological impact of six likely pathogenic variants in ETV6, of which five are novel. We describe in detail how variants in ETV6 lead to increased megakaryocyte proliferation and various cytoskeletonrelated platelet defects that include altered platelet shape, reduced Rho GTPase expression in platelets, decreased proplatelet (PPT) formation and reduced platelet spreading. Additionally, we show that patients exhibit elevated levels of circulating CD34+ progenitors and a predisposition to myelodysplastic syndrome and leukemia.

Methods

The platelet survival assay was based on the method of Thakur and colleagues.13

High-throughput and Sanger sequencing DNA samples from 957 patients enrolled in the BRIDGE-BPD project were subjected to whole-genome or whole-exome sequencing, and the results were used for variant calling as described previously.14,15 DNA samples from eight patients in the French cohort were subjected to whole-exome sequencing or Sanger sequencing at the ETV6 lotus. Sequence analysis was carried out using Chromas X software, and aligned using Multalin.16

Site-directed mutagenesis and luciferase assays ETV6 cDNA was ligated into a pcDNA3 expression vector, and mutagenesis was performed using the GENEART® Site-directed Mutagenesis System kit (Life technologies).17 Transcriptional regulatory properties of wild-type (WT) and mutant ETV6 (mutETV6) as well as ETV6 corepressor binding,18 were determined by using the luciferase reporter systems in transfected GripTite™ 293 macrophage scavenger receptor (MSR) cells.

Immunoassays Immunoblots were performed with antibodies directed against human ETV6; SMRT, RHOA (Santa Cruz Biotechnology), CDC42, RAC1 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Millipore) and MYH10 (Cell Signaling Technology) antibodies. Chemiluminescence signals were detected and quantified (CCD camera-based ImageQuant LAS 4000, GE Healthcare). Levels of thrombopoietin (TPO) and stromal cell-derived factor 1α (SDF1α) were quantified via ELISA (Abcam). For the co-immunoprecipitation assays, whole cell extracts were prepared in NP-40 buffer and pre-cleared with protein A/G magnetic beads (Millipore). Immunoprecipitation of the cell extracts with antiETV6-coated beads was carried out overnight.

Megakaryocyte differentiation and quantification of proplatelet-bearing megakaryocytes CD34+ cells were grown in serum-free medium supplemented with TPO and stem cell factor (SCF) (Life Technologies).19 At culture day 10, we assessed ploidy in the Hoechst+CD41+CD42a+ cell population20 (Navios, BD Biosciences). Proplatelets were quantified between day 11 and 15. Microtubule and F-actin organization was determined in megakaryocytes adhering to fibrinogen with fluorescently labeled polyclonal rabbit anti-tubulin antibody (Sigma-Aldrich) and phalloidin (Life Technologies).

Lentiviral particle production and CD34+ cell transduction Lentiviral particles were prepared as previously described.21,22 CD34+ cells were infected twice. After 8 hours, the cells were washed and cultured in serum-free medium.

Platelets and circulating CD34+-cells analysis Blood samples were collected after informed written consent, in accordance with our local Institutional Review Boards and the Declaration of Helsinki. Platelet-rich plasma (PRP), washed platelets and circulating CD34+ cells were prepared according to standard procedures. For electron microscopy (EM), platelets were fixed in glutaraldehyde and processed as previously described.12 For platelet spreading, fibronectin-adherent platelets were stained with Alexa Fluor 488 phalloidin (filamentous (F)-actin) and Alexa Fluor 594 DNAse I (globular (G)-actin). Filopodia and lamellipodia were manually quantified. For clot retraction, coagulation of PRP was triggered using thrombin, and clots were allowed to retract. Images were recorded using a CoolSNAP CCD camera and analyzed to evaluate the reduction of the initial clot surface (ImageJ). haematologica | 2017; 102(2)

Clonogenic progenitor assays CD34+ cells were plated in human methylcellulose medium H4434 (STEMCELL Technologies), supplemented with erythropoietin (EPO), interleukin-3 (IL-3), SCF, granulocyte colony-stimulating factor (G-CSF), interleukin-6 (IL-6) and TPO to quantify erythroid (burst forming unit-erythroid (BFU-E), colony-forming unit-erythroid (CFU-E), granulocytic/macrophage (CFU-GM), mixed (CFUGEMM) and megakaryocyte (CFU-MK)) progenitors at day 12.23

Statistical analyses Analyses were performed using GraphPad Prism software. Statistical significance was determined via a two-tailed MannWhitney test. P<0.05 was considered statistically significant. 283


M. Poggi et al.

Results Identification of affected families Screening of patients with thrombocytopenia for rare non-synonymous variants in ETV6 revealed six families with patients carrying one of six possibly pathogenic variants. The variants encode p.P214L, which has been previ-

A

ously reported,3 and the novel substitutions p.I358M, p.A377T, p.R396G, p.Y401N and p.Y401H (Figure 1A). Family studies by Sanger sequencing showed segregation between the ETV6 variant and thrombocytopenia in all cases for which DNA samples were available (Figure 1B). Henceforth, we refer to the likely pathogenic variants described above as mutETV6.

B

C

D

E

F

Figure 1. Identification of variants in ETV6 underlying AD thrombocytopenia, megakaryocyte and platelet characteristics. (A) Schematic representation of the different domains of the ETV6 protein. The N-terminal domain (PNT), central domain and C-terminal domain containing a DNA-binding domain (ETS) are depicted. Arrows indicate the location of the ETV6 variants and the corresponding family is mentioned in brackets. (B) Pedigrees for the affected families. Squares denote males, circles denote females and slashes represent deceased family members. Black filled symbols represent thrombocytopenic family members and dotted line symbols represent non-tested members. The families F1, F2, F3, F4, F5 and F6 carried the ETV6 p.P214L, p.A377T, p.Y401N, p.I358M, p.R396G and p.Y401H variants, respectively, which segregated with thrombocytopenia. See Table 1 for blood cell count values. (C) Sex-stratified histograms of platelet count and mean platelet volume measurements, obtained using a Coulter hematology analyzer, from 480,001 UK Biobank volunteers, after adjustment for technical artifacts. The red arrows superimposed upon the histograms indicate the sex and values for patients with a deleterious variant in ETV6. The green arrows indicate the sex and values for relatives homozygous for the corresponding wild-type allele. (D) BM smears (May-GrĂźnwald-Giemsa staining) from family F1 propositus (F1-IV3). Left: a relatively immature MK with reduced cytoplasm. Middle: a micromegakaryocyte without granules, with immature cytoplasm (basophilic) and nucleus. Signs of impaired proplatelet formation can be observed. Right: a mature MK of reduced size with a hypolobulated nucleus. Table 1 indicates the % of MKs at each stage of maturation in the BM samples from family F1 proposita (F1-IVI3) and a healthy control. (E) Ultrastructural aspects of platelets from patients F3-I2 and F3-II4, F4-I2 and F4-II3 and unrelated healthy controls. Upper panel: aspect of healthy platelets; middle panel: series of mostly rounder platelets from patients F4-I2 and F4-II3, lower panel: a series of platelets emphasizing anisocytosis in patient F3-114 and a platelet from patient F3-I2 with abnormal membrane complex (MC). Note the heterogeneous presence of Îą-granules with an occasional granule of increased size. (F) The platelet area and roundness was quantified. Perfect round platelets would have a value of 1. Values are the means and SD as quantified for 50 randomly selected platelets per subject using two-tailed unpaired t-test with Welchâ&#x20AC;&#x2122;s correction. ***P<0.0001. WT: wild-type; MK: megakaryocyte; PNT: pointed; ETS: ET6 transformation-specific.

284

haematologica | 2017; 102(2)


ETV6 and platelet formation

Description of the families Our study included six families with mutETV6 variants. The proband of the first family (F1-IV3) was a 7-year-old girl who was admitted for emergency care due to the suspicion of acute leukemia with asthenia, weakness, paleness, severe thrombocytopenia (44 x 109/L) and anemia (hemoglobin: 50 g/L). BM examination unequivocally dismissed a diagnosis of leukemia, and the anemia was attributed to an iron deficiency subsequent to repeated episodes of severe epistaxis. The patient underwent a red blood cell transfusion, and the anemia was progressively corrected via iron supplementation. However, the platelet

count remained low (50 x 109/L). Clinical examination of the parents and two siblings did not reveal any particular bleeding tendency. However, the patient's mother (F1-III3) had undergone a splenectomy at the age of 17 because of chronic thrombocytopenia, and she exhibited subnormal platelet counts (116-210 x 109/L) at the time of examination. To gain further insight into the possibility of inherited thrombocytopenia, we screened the extended family for platelet counts. An AD form of thrombocytopenia was evidenced (Figure 1B), with normal mean platelet volume (MPV) compared with a large population of blood donors (Figure 1C). Of note, fourteen out of twenty-three carriers

Table 1. Hematological parameters in the six studied family members.

Family- Sex ETV6 Current Red cell Hyperdense Hemoglobin Mean Mean Platelet Mean Absolute Absolute Absolute Individuals genotype age count Red blood corpuscular corpuscular count platelet neutrophil lymphocyte monocyte cells hemoglobin volume volume count count count concentration Normal range 4.0-5.0x 0.0-2.5 115-160 310-350 80-100 150-400x 7.0-9.0 2.0-7.5x 1.5-4.0x 0.2-2.0x 1012/L % g/L g/L fL 109/L fL 109/L 109/L 109/L F1-II1 F1-II2 F1-II4 F1-III1 F1-III3 F1-III5 F1-III6 F1-III7 F1-III8 F1-IV1 F1-IV2 F1-IV3 F2-II2 F2-II3 F2-III1 F2-III2 F3-I1 F3-I2 F3-II1 F3-II2 F3-II3 F3-II4 F4-I1 F4-I2 F4-II1* F4-II2 F4-II3 F5-I2 F5-II1 F5-II2 F6-I1 F6-II1 F6-II2

F M F M F M M M M F M F F M M F F M F M F F F M M F F M F F F F F

WT P214L P214L WT P214L P214L P214L P214L P214L P214L WT P214L A377T A377T ND A377T ND Y401N ND Y401N ND Y401N WT I358M I358M I358M I358M R396G R396G R396G WT Y401H ND

70 69 69 53 43 27 18 43 27 13 11 8 28 24 7 2 54 56 29 21 22 16 55 56 24 30 27 59 26 20 ND ND ND

5.0 3.3 4.5 4.6 4.0 4.3 4.8 4.5 5.0 4.2 5.0 4.3 4.9 4.8 4.4 4.6 5.2 5.0 4.8 4.7 4.9 4.4 3.9 4.6 3.0 4.3 4.1 ND ND ND 4.4 3.9 ND

1.1 4.3 1.1 1.5 0.6 5.0 6.4 ND 2.6 2.0 1.7 5.7 ND ND ND ND 1.2 2.0 1.1 1.3 0.5 1.3 ND ND ND ND ND ND ND ND ND ND ND

155 117 144 130 125 140 152 147 169 130 138 131 162 167 125 122 131 160 130 147 139 145 120 149 125 142 142 140 133 109 137 132 ND

343 350 342 340 323 359 353 ND 354 337 338 354 342 343 330 325 347 360 347 350 352 350 348 353 349 351 351 ND ND ND ND ND ND

89 101 94 84 95 92 89 94 92 91 81 83 97 102 87 81 88 86 79 88 80 87 88 91 92 95 98 ND ND ND ND 100 ND

242 44 64 224 116-210 55 51 58 38 85 184 50 84 60 85 111 280 125 285 112 389 80 209 57 29 134 121 58 76 75 199 77 159

8.4 7.2 8.0 10.2 12.4 8.8 8.4 10.6 7.9 8.7 8.5 10.2 9.0 7.9 7.7 8.7 9.3 9.8 10.4 10.0 9.6 9.4 9 11.7 10.7 10.5 9.8 10.8 11.2 9.9 ND ND ND

2.4 1.3 3.6 4.3 2.9 3.9 3.1 4.6 2.7 2.3 2.0 1.6 4.0 3.2 1.6 1.0 3.2 4.4 7.5 2.1 3.1 2.5 3.1 2.6 1.5 2.5 2.8 ND ND ND ND ND ND

1.2 1.0 0.8 1.6 1.5 1.0 1.4 0.8 1.8 1.7 2.1 1.5 2.0 1.2 2.8 4.1 1.5 2.3 5.0 1.8 1.6 1.5 4.9 1.6 2.1 1.5 1.9 3.5 4.0 3.9 ND ND ND

0.3 0.2 0.3 0.3 0.5 0.5 0.5 0.6 0.5 0.2 0.2 0.2 0.4 0.5 0.5 0.8 0.3 0.5 0.6 0.3 0.4 0.3 0.2 0.5 0.5 0.7 0.7 ND ND ND ND ND ND

*in remission (2 years after chemotherapy). Bold values are outside the normal range. F: female; M: male; ND: not done; WT: wild-type.

haematologica | 2017; 102(2)

285


M. Poggi et al.

exhibit MPV >9 fL (Table 1). Plasma TPO levels were decreased in affected F1 members (n=4): 160 ± 9 pg/mL vs. controls (n=8): 296 ± 39 pg/mL, P=0.02. May-GrünwaldGiemsa staining of BM smears of patient F1-IV3 showed that megakaryocytes were present, although a high pro-

portion were of medium size, in the early stages of maturation and tended to be hypolobulated (Figure 1D). The 7year-old patient’s grandfather (F1-II2) was diagnosed with refractory anemia with excess blasts type 2 (RAEB-2) at the age of 70. Individuals from the second and third fami-

A

B

C

D

E

Figure 2. Effect of the variants on repressive activity and corepressor recruitment. (A) Western blot analysis of ETV6 expression in platelets of 6 affected F1 members and 7 external controls. GAPDH was used as a protein loading control. (B-C) GripTite™ 293 MSR cells were co-transfected with the luciferase reporter plasmid containing 3 tandem copies of the Ets Binding Site (EBS) upstream of HSV-Tk (E743tk80Luc), pCDNA3.1 expression vector (empty, WT or mutETV6) or pGL473 Renilla luciferase control vector. (B) Western blot analysis of ETV6 expression in whole cell lysates of GripTite™ 293 MSR transfected with WT ETV6 or mutETV6 expression vectors. GAPDH was used as a protein loading control. The data are representative of 4 to 8 independent experiments. (C) The firefly to renilla luminescence ratios (Fluc/Rluc) were calculated to compensate for transfection efficiency. The data represent the mean ± SEM of 4 to 8 independent experiments, student’s t-test ***P<0.001 (each condition was compared with WT). (D) Effects of the ETV6 variants on corepressor recruitment. Mammalian two-hybrid analysis of the protein interactions between WT NCOR, SMRT or Sin3A (expressed using the GAL4 DNA-binding domain (DBD) plasmid) and WT ETV6 or mutETV6 (expressed using the GAL4-VP16 activation domain vector). The results are expressed as mean ± SEM of 3 to 8 independent experiments, student’s t-test *P<0.05, **P<0.01, ***P<0.001. (E) Immunoprecipitation of endogenous corepressor SMRT and ETV6 from GripTite™ 293 MSR cells transfected with WT and mutETV6. Immunoprecipitation was performed on cell lysates with ETV6 antibody. The total cell lysates (lower panel) and immunoprecipitates (upper panel) were analyzed via immunoblotting with anti-SMRT antibody. Quantification of band intensity for SMRT and SMRT-extended (SMRTe) is shown below the western blot. The results are expressed as mean ± SEM, student’s t-test, *P<0.05 vs. WT. GAPDH: glyceraldehyde-3-phosphate dehydrogenase; WT: wild-type; A.U: arbitrary unit; IP: immunoprecipitation; IB: immunoblot.

286

haematologica | 2017; 102(2)


ETV6 and platelet formation

lies had platelet counts between 60 and 125 x 109/L (Table 1). BM from F2-II3 displayed a delay in granulocyte maturation and dyserythropoiesis (data not shown). Peripheral blood smears revealed platelet anisocytosis (data not shown), as confirmed by EM (Figure 1E), which further highlighted the presence of occasional hypogranular platelets with a poorly organized open canalicular system. Patient F2-I1 presented with refractory anemia with excess blasts (RAEB) and required BM transplantation. The propositus (F4-II1) from pedigree 4 was referred with AML type M0 at the age of 8 years, and suffered from epistaxis, ecchymosis and infections. After 2 years of chemotherapy treatment, the BM showed no blasts and the peripheral blood counts normalized, except for a persistent low platelet count (Table 1). BM studies showed the presence of many hypolobulated small megakaryocytes (data not shown). Thrombocytopenia was also present in his father and two sisters without any bleeding problems (Table 1). EM investigation of platelets from affected members F4-I2 and F4-II3 showed the presence of both larger and smaller platelets that, significantly, were of round shape rather than of discoid shape (Figure 1E,F; P<0.0001). These platelets had normal dense and α-granules numbers, but some α-granules were elongated (data not shown). Pedigree 5 was referred for genetic testing of AD thrombocytopenia in a father with very mild bleeding problems (propositus F5-I2), and his 2 asymptomatic daughters (Figure 1B). BM investigation in F5-II2 showed the presence of dysmegakaryopoiesis with almost no mature megakaryocytes (data not shown). A mother (F6-I1) and daughter (propositus F6-II1) from pedigree 6 (Figure 1B) were diagnosed with platelet dense storage pool deficiency (SPD), with platelet aggregation defects and abnormal dense granules. Thrombocytopenia was only recorded for the daughter, who suffered from severe menorrhagia and had an increased bleeding tendency with bruising and nosebleeds. The mother had a normal platelet count and did not carry the ETV6 variant. No clinical information or DNA was available from the father. Therefore, the ETV6 variant in F6-II1 could be present as a de novo or somatic variant. SPD in the mother and daughter was likely to be caused by another additional genetic factor. Indeed, in contrast to the obvious platelet aggregation and

A

secretion defects for these two patients, such abnormalities were not present in the other five families, except for a decreased aggregation response to arachidonic acid as the only consistent finding in every family (Online Supplementary Table S1). Consistent with normal dense granules found by EM (Figure 1E), adenosine triphosphate (ATP) secretion and mepacrine uptake and release were normal (Online Supplementary Table S1). Flow cytometry analysis of key platelet surface receptors (αIIbb3, glycoprotein (GP) Ibα, GPIa, GPIV, CD63 and CD62P) was also normal (Online Supplementary Table S2). A platelet survival assay was performed on patient F3II4 (Online Supplementary Table S3), and revealed decreased platelet lifespan (4.6 days) without significant splenic or hepatic sequestration. Notably, this patient had not undergone platelet transfusion. Patient F1-III3 underwent a 111In-oxine platelet survival assessment (autologous transfusion) in 1981 prior to a splenectomy, which revealed short platelet half-life (24 h vs. 3.5 days in the control) and hepatic and splenic platelet sequestration, with predominant sequestration in the liver (data not shown). Patient F1-III3 was assessed for anti-human leukocyte antigen (HLA) antibodies on several occasions (National Center of Blood Transfusion, Marseille, France), but all results were negative (data not shown).

Variants in ETV6 lead to a functional defect in transcriptional activity Western blot analysis showed that ETV6 protein expression was not reduced in platelets from the patients, nor in GripTite™ 293 MSR cells transfected with the ETV6 variants (Figure 2A,B). To investigate the transcriptional regulatory properties of mutETV6 compared with WT ETV6, we analyzed repressive activity. Cotransfection of the reporter plasmid along with expression of a plasmid encoding WT ETV6 resulted in an almost 90% inhibition of luciferase activity. The substitution of WT ETV6 with any of the mutETV6 variants led to a significant reduction in repressive activity (85% to 100%) (Figure 2C). To evaluate whether this reduction in repressive activity perhaps resulted from variations in nuclear corepressor complex recruitment, we investigated the interaction of ETV6 with NCOR, SMRT and Sin3A using a mammalian

B

Figure 3. Increased numbers of circulating CD34 positive cells in variant carriers. Flow cytometry analysis of CD34+ cells. (A) Representative CD34+/CD38+ dot plot of cells from 2 controls and 1 patient (F1-III7). (B) Histograms show the percentage of CD34+ cells in 8 controls and 5 affected family members (F1-III3, F1-III7, F1III8, F1-IV1, F1-IV3) (mean ± SEM, student’s t-test, **P<0.01).

haematologica | 2017; 102(2)

287


M. Poggi et al.

two-hybrid assay. p.P214L ETV6 interacted with NCOR, SMRT and Sin3A, whereas p.A377T and p.Y401N ETV6 did not (Figure 2D). Immunoprecipitation assays showed that the p.A377T and p.Y401N variants reduced ETV6 binding to SMRT and SMRTe (Figure 2E)

Increased numbers of circulating CD34 positive cells in affected family members F1 carriers (F1-III3, F1-III7, F1-III8, F1-IV1 and F1-IV3) exhibited a 4- to 6-fold increase in circulating CD34+/CD38+ cells compared with healthy donors (Figure 3A,B). Similarly, F3-I2 and F3-II4 exhibited a 5- and 3-fold increase in circulating CD34+ cells compared with controls (0.16% and 0.09%, respectively, vs. 0.035%). The expression levels of immature cell markers CD133 and CD117 did not differ between F1 members and controls.

A

D

Expression of the myeloid lineage marker CD33 contrasted with the absence of megakaryocyte lineage markers CD123, CD41, CD61 and CD42b (data not shown). Additionally, plasma levels of SDF1α did not vary between patients (F1: 1966 ± 95 pg/mL; n=8) and controls (2068 ± 75 pg/mL; n=9).

Variants in ETV6 cause megakaryocyte hyperplasia but reduced proplatelet formation in vitro The percentage of CD41+CD42a+ megakaryocytes derived from CD34+ (gated on Hoechst+ cells) was significantly higher in mutETV6 carriers (Figure 4A,B). No significant difference in mean ploidy was detected between patients and healthy donors (Figure 4C). Accordingly, the number of CD34+-derived CFU-GM/G/M colonies was higher in patients compared with controls (Figure 4D).

B

C

E

F

Figure 4. Megakaryocyte differentiation and colony-forming cell potential. (A-C) In vitro megakaryocyte (MK) differentiation in control or patient peripheral blood CD34+ cells, the cells were analyzed at culture day 10. (A) The data show a representative dot plot of CD41 and CD42a expression in Hoechst+ cells from a control individual and F1-III6. The gate represents mature MKs. (B) The histogram represents the MK (CD41+CD42a+Hoechst+) numbers (nb) in the affected family members (n=9) expressed as fold increase over healthy controls (n=10), student’s t-test, **P<0.01. (C) The ploidy level (N) was analyzed for CD41+CD42a+ MKs, and mean ploidy was calculated using the percentage of cells with 2N, 4N, 8N, 16N and 32N. (D) Methylcellulose assay. The histograms present the number of erythroid (BFUE), granulo-monocyte (CFU-G/M/GM) and mixed (CFU-GEMM) progenitors from two patients of family F3 with the p.Y401N variant (F3-I2 and F3-II4) and two independent controls. Mean ± SEM, student’s t-test, *P<0.05. (E) Fibrin clot culture. The histograms present the number of MK progenitors (CFU-MK) from two independent controls and two patients (F3-I2 and F3-II4). The CFU-MKs are divided into four categories: <5 MKs per colony, 5-10 MKs per colony, 10-50 MKs per colony or >50 MKs per colony. Error bars represent ± SD of triplicate experiments. (F) Representative pictures of CFU-MKs after CD41 immunostaining. Control 1 and Control 2 represent 2 independent controls, and F3-I2 and F3-II4 are two affected patients. CFU-GEMM: colony-forming unit–granulocyte, erythrocyte, monocyte, megakaryocyte; BFU-E: burst forming unit-erythroid; CFU-G/M/GM: colony-forming unit-granulocytes, macrophages, granulocyte-macrophages; CFU-MK: colony-forming unitmegakaryocyte.

288

haematologica | 2017; 102(2)


ETV6 and platelet formation

The number of megakaryocyte progenitors (CFU-MK) from patients F3-I2 and F3-II4 did not differ from controls, although the size of CFU-MKs was significantly increased in the two patients (Figure 4E,F), thereby suggesting an increased proliferation of megakaryocyte precursors in the presence of mutETV6. Proplatelet-bearing megakaryocytes derived from controls showed multiple branched thin extensions, swellings and tips. In contrast, megakaryocytes from patients formed very few proplatelets with a reduced number of thicker extensions. Although there was no swelling, tips were of increased size (Figure 5A). A 2- to 15-fold decrease in the percentage of proplatelet-bearing megakaryocytes

A

was observed in carriers of the p.P214L (F1-III3, F1-III7 and F1-IV3) and p.Y401N (F3-II2 and F3-II4) variants (Figure 5B). b-tubulin and F-actin staining confirmed the megakaryocyte-proplatelet extension defect together with a reduced concentration of both actin filaments and microtubules in the residual larger megakaryocyte cell body (Figure 5C). Additionally, b-tubulin failed to accumulate normally in the few extension tips observed in mutETV6 megakaryocytes compared with controls. To confirm that mutETV6 leads to a defect in proplatelet formation, CD34+ cells from healthy donors were transduced with lentivirus containing ETV6 sequences encoding the WT or the p.P214L mutant. Non-transduced cells

B

C

D

Figure 5. ETV6 variants lead to defective proplatelet formation. (A-B) In vitro MK differentiation induced from control or patient peripheral blood CD34+ progenitors in the presence of TPO and SCF. (A) Representative microscopic images of PPT formation in control (n=2) and patient (F1-III7, F1-III3) MKs after 11 or 13 days of culture. (B) The histograms show the percentage of PPT-bearing MKs from members of 2 families (F1-III3, F1-IV3, F1-III7, F3-I2, F3-II4) and 5 independent controls evaluated (3 to 5 evaluations) between culture days 10 to 15. The percentage of PPT-forming MKs was estimated by counting MKs exhibiting ≥1 cytoplasmic processes with areas of constriction. Double-blinded researchers quantified a total of 300-500 cells. The results are expressed as mean ± SEM, student’s t-test **P<0.01 and ***P<0.001. (C) F-actin and b-tubulin staining on PPT-forming MKs from F1-III3 and a control individual, adhering to fibrinogen. Confocal images were acquired at day 12 of culture (x60). (D) In vitro MK differentiation was induced from control peripheral blood CD34+ progenitors transduced with WT or mutETV6 (family F1, c.641C>T, p.P214L) lentiviral particles in the presence of TPO and SCF. Microscopic images of PPT formation were acquired at days 13 and 15 of culture. PPT: proplatelet; MKs: megakaryocytes; F-actin: filamentous actin.

haematologica | 2017; 102(2)

289


M. Poggi et al.

were also included as control. After 13 and 15 days of culture in the presence of TPO and SCF, cells transduced with mutETV6 lentivirus did not form proplatelets, in contrast to non-transduced cells or those transduced with WT ETV6 (Figure 5D).

A

mutETV6 does not alter MYH10 expression but was associated with decreased expression and activity of the key regulators of the actin cytoskeleton CDC42 and RHOA To assess potential cooperation between the ETV6, RUNX1 and FLI1 pathways we examined MYH10 protein

B

C

D

Figure 6. Rho GTPase expression analysis. (A) Western blot analysis and quantification of CDC42, RAC1 and RHOA expression in platelet lysates from healthy controls (n=7 for CDC42, n=4 for RAC1, n=4 for RHOA) and affected members from F1 (n=6 for CDC42, n=4 for RAC1 and n=5 for RHOA). GAPDH was used as a protein loading control. The results are expressed as mean ± SEM, student’s t-test *P<0.05 and ***P<0.001. (B) Quantification of CDC42, RAC1 and RHOA mRNA levels in CD34+-derived megakaryocytes (MKs) from healthy controls (n=10) and affected family members from F1 (n=6) and F3 (n=2). mRNA expression levels were measured via reverse transcription polymerase chain reaction (RT-PCR), and expression levels were normalized to housekeeping 36b4 RNA. The results are expressed as mean ± SEM, student’s t-test, **P<0.005. (C) Western blot analysis and quantification of CDC42, RHOA and GAPDH expression in platelets from two affected members of the F3 family (F3-I2 and F3-II4) and a healthy control. (D) In vitro MK differentiation was induced from F1-III7 CD34+ progenitors transduced with control or CDC42 lentiviral particles in the presence of TPO and SCF. Microscopic images of proplatelet (PPT) formation were acquired at days 14 and 16 of culture. The arrows indicate thinner PPT extensions and swellings in the presence of CDC42. Extensions were enlarged in the control. mRNA: messenger RNA; GAPDH: glyceraldehyde-3-phosphate dehydrogenase; A.U: arbitrary unit.

290

haematologica | 2017; 102(2)


ETV6 and platelet formation

expression levels in patient platelets. We did not detect increased MYH10 levels in platelets from ETV6 patients (F1-III6 and F1-III8), which contrasted with RUNX1 and FLI1 defects24 (Online Supplementary Figures S1A and S1B). Proplatelet formation is dependent on massive reorganization of the actin cytoskeleton. Rho GTPase family members (e.g., CDC42, RAC1 and RHOA) are key regulators of actin cytoskeleton dynamics in platelets25 and megakaryocytes. p.P214L (n=5) and p.Y401N (n=2) variants led to significantly reduced platelet expression levels of CDC42 and RHOA, without affecting RAC1 expression (Figure 6A). Likewise, CDC42 and RHOA messenger ribonucleic acid (mRNA) levels were decreased in

megakaryocytes from F1 and F3, while RAC1 mRNA levels remained unaffected (Figure 6B). Notably, patient F3I2, with 112 x 109 platelets/L, exhibited only slightly decreased levels of CDC42 and RHOA in platelets (Figure 6C) compared with other affected members. CDC42 and RHOA levels significantly correlated with platelet count (n=6 from F1 and F3) (P=0.03 and r=0.84 for CDC42; P=0.008 and r=0.92 for RHOA). To confirm the specificity of this effect, we quantified CDC42 protein levels in FLI1 deficient patients with thrombocytopenia (n=2) (122 and 131 x 109/L). None of these patients exhibited reduced levels of CDC42 (Online Supplementary Figure S1C). Overexpression of CDC42 in CD34+-derived megakary-

A

B

C

Figure 7. Platelet spreading and clot retraction. (A) Left: Representative images of unstimulated platelets spread over immobilized fibronectin. Middle: filopodia formation was quantified according to the number of extensions per unstimulated platelet derived from affected individuals (F1-III3, F1-III7) and healthy controls (n=2). Right: Quantification of lamellipodia-forming cells, at resting and ADP-stimulated conditions, from affected members (F1-III7, F1-III8) and healthy controls (n=2). The data are expressed as mean ± SEM of 5 different view fields. Student’s t-test, ***P<0.001. (B) Actin polymerization quantification in spread unstimulated platelets. Left: representative images of G-actin, F-actin and the G-actin/F-actin ratio in control platelets and the rare spread platelets detected in F1-III7. Platelets were spread over fibronectin and stimulated with ADP. Right: quantification of the area with the high G-actin/F-actin ratio. Quantification of the ratio was performed according to the look-up table as the percentage of the platelet surface (n=20 different cells; mean ± SEM. Student’s t-test *P<0.05). (C) Clot retraction. Left: representative images at 0, 30 and 50 minutes. Right: quantification of the extent of clot retraction expressed as percentage of the initial clot (mean ± SEM, n=2 for F1-III7 and n=4 for controls. Two-way ANOVA, ***P<0.001). F-actin: filamentous actin; G-actin: globular-actin; ADP: adenosine diphosphate; LUT: look-up table.

haematologica | 2017; 102(2)

291


M. Poggi et al.

ocytes from patients (F1-III3, F1-III7, F1-III8) did not fully reverse the phenotype, although it did improve the proplatelet-bearing megakaryocyte phenotype. Transduced cells produced thinner extensions and swellings, which were not observed in control transduced cells (Figure 6D).

mutETV6 alters platelet spreading The reduced expression of CDC42 and RHOA suggests that ETV6 is involved in cytoskeletal reorganization, and thus does not only have an important role in platelet shape (EM showed more round platelets) and proplatelet formation, but also in regulating platelet spreading. We assessed whether the p.P214L transition in ETV6 affects the spreading of platelets over immobilized fibronectin and clot retraction. Platelets showed a reduced capacity to form filopodia and lamellipodia, under unstimulated and adenosine diphosphate (ADP)-stimulated conditions, respectively (Figure 7A). The G/F actin ratio was significantly higher in the rare patient platelets that spread (Figure 7B). Furthermore, reduced clot retraction velocity was noticeable in the mutETV6 (F1-lll7) (Figure 7C).

Discussion Herein, we presented six families with AD thrombocytopenia associated with germline variants in ETV6. CD34+-derived megakaryocytes from mutETV6 carriers showed a reduced ability to form proplatelets. The variants in the Ets domain impaired interaction with the corepressors NCOR, SMRT and Sin3A. Patient platelets were more round and had a reduced capacity to form filopodia and lamellipodia, which was associated with reduced expression levels of cytoskeletal regulators CDC42 and RHOA. Additionally, mutETV6 carriers displayed increased numbers of circulating CD34+ progenitor cells, which may contribute to a predisposition to hematologic malignancy. Loss of ETV6 function has been reported to contribute to leukemia, predominantly due to somatic variants and fusion transcripts.3 Four amino acid substitutions described in this study are listed in the Catalog Of Somatic Mutations In Cancer (COSMIC). The p.P214L, p.R396G and p.A377T variants were present in digestive tract tumors,26 while p.Y401C has been associated with AML.27 More recently, germline ETV6 variants have also been found to predispose to cancer. Eleven patients carrying mutETV6 (p.P214L, p.R399C, p.R369Q, p.L349P, p.N385fs or p.W380R) developed acute lymphocytic leukemia or myelodysplastic syndrome.1-3,28 Two affected members of the families F1 and F2 had myelodysplasia with RAEB and one member of the F4 family was successfully treated with chemotherapy for AML-M0. Variants reduced the repressive activity of ETV6 without altering ETV6 protein expression levels in platelets. The alteration of ETV6 repressive activity can be explained by the modification of ETV6 cellular localization, as p.P214L and four other variants affecting the ETS domain lead to ETV6 sequestration in the cytoplasm in both HeLa transfected cells and cultured megakaryocytes.1-3 However, three of these variants only partially prevented nuclear localization, thereby indicating other possible mechanisms. The p.A377T and p.Y401N variants prevented corepressor complex recruitment. These substitutions are located in the ETV6 Ets second and third 292

Îą-helix, contiguous to amino acids involved in key hydrophobic contacts with the H5 helix of the C-terminal inhibitory domain (aa 426-436),29 thus possibly affecting ETV6 DNA-binding ability. In immunoprecipitation assays, overexpression of WT or p.P214L ETV6 did not modify the interaction between SMRT and ETV6, while p.A377T and p.Y401N ETV6 significantly reduced this interaction. Overall, this suggests that variants in the Ets DNA-binding domain exert a dominant negative effect. ETV6 has been shown to drive megakaryocyte differentiation of hematopoietic stem cells.30 From the literature, and supported by BM studies in F4-II1 and F5-II2, ETV6 defects seem to result in an increased percentage of small megakaryocytes.1 Our megakaryocyte colony assays confirmed an increased proliferation of early megakaryocyte progenitors, characterized by an increased production of CD41+CD42a+ megakaryocytes compared to control conditions. This may explain the reduced TPO levels observed in the affected members of family F1. Accordingly, the loss of ETV6 in the erythro-megakaryocytic lineage in mice also results in large, highly proliferative early megakaryocytes and mild thrombocytopenia. We cannot exclude that ETV6-driven deregulation of megakaryocyte proliferation may take place in hematopoietic progenitors, thereby promoting oncogenic transformation. Altogether, these data do not support the concept that signaling between the ETV6 and RUNX1/FLI1/ANKRD26 pathways is involved in the underlying mechanism, as variants in these genes were associated with a decreased or normal megakaryocyte colony formation.19,31,32 Furthermore, MYH10 expression levels remained low in patients with mutETV6, which indicates unaltered RUNX1 and FLI1 function.24 Despite the increased early megakaryocyte proliferation potential, CD34+-derived megakaryocytes from patients with mutETV6 showed a reduced capacity to form proplatelets. These altered megakaryocyte features suggest that a defect in cytoskeletal reorganization during proplatelet formation likely causes thrombocytopenia in patients. Sequencing of platelet RNA from patients with p.P214L ETV6 revealed a considerable reduction in the levels of several cytoskeletal transcripts.1 Furthermore, Palmi et al.33 showed that the ETV6-RUNX1 fusion protein, which is associated with a loss of ETV6 repressive activity,34,35 alters the expression of genes regulating cytoskeletal organization. In particular, the ETV6-RUNX1 fusion protein led to reduced expression of CDC42. The mechanism by which the loss of ETV6 repressive activity results in reduced CDC42 and RHOA expression remains to be resolved. CDC42 is an important mediator of platelet and megakaryocyte cytoskeleton reorganization.36 Therefore, we hypothesize that ETV6 repressive activity is a key regulator of megakaryocyte cytoskeleton remodeling, driven via Rho GTPases in mutETV6 carriers. mutETV6 was associated with a decrease in CDC42 and RHOA expression levels in platelets without affecting RAC1 expression. Additionally, mutETV6 platelets showed defects in functions classically associated with CDC42 (i.e., filopodia formation) and RHOA (i.e., lamellipodia formation and clot retraction).36 EM also showed platelets of variable sizes and having a more circular instead of discoid shape. RNA sequencing previously performed on mutETV6 transfected cells, patient platelets and leukemia cells did not reveal any modification in Rho GTPase mRNA levels,1,3 which haematologica | 2017; 102(2)


ETV6 and platelet formation

may be due to variations in the models applied. Indeed, Rho GTPase mRNA levels were evaluated in CD34+derived megakaryocytes, and the reduced mRNA levels were confirmed at the protein level in patient platelets. Moreover, we observed a correlation between platelet count and CDC42 and RHOA expression levels, thereby suggesting a relationship between thrombocytopenia severity and Rho GTPase levels. Key regulators of the actin cytoskeleton CDC42 and RHOA have already been shown to be associated with thrombocytopenia, due to defects in cytoskeleton organization.37-41 In affected individuals, we found abnormal tubulin organization in proplatelet-forming megakaryocytes and altered actin polymerization in platelets. Rescue experiments with CDC42 lentiviral particles were not able to fully reverse the phenotype, although the cells produced thinner extensions and swellings, which were barely observed in control cells. In mice, Cdc42 or RhoA deficiency causes increased platelet clearance.37,38 Such an observation was noted in two patients: one young girl who never received platelets (F3-II4), and a patient for whom splenectomy improved the platelet count (F1-III3). This suggests that ETV6 mutations are linked to several defects with reduced platelet formation and survival, although this latter mechanism requires further confirmation. Individuals carrying a germline ETV6 variant showed increased numbers of circulating CD34+/CD133+ cells. The phenotype of these stem cells did not differ between patients and controls. This increase has to be considered as a helpful marker of the ETV6-related thrombocytopenia. It may not be attributed to excessive proliferation, as Zhang et al. showed reduced proliferation of CD34+ cells expressing WT or mutETV6.3 Interestingly, the defect in CDC42 expression may also account for increased hematopoietic progenitor mobilization, as chemical inhibition of CDC42 in mice efficiently improved progenitor

References 1. Noetzli L, Lo RW, Lee-Sherick AB, et al. Germline mutations in ETV6 are associated with thrombocytopenia, red cell macrocytosis and predisposition to lymphoblastic leukemia. Nat Genet. 2015;47(5):535-538. 2. Topka S, Vijai J, Walsh MF, et al. Germline ETV6 mutations confer susceptibility to acute lymphoblastic leukemia and thrombocytopenia. PLoS Genet. 2015; 11(6):e1005262. 3. Zhang MY, Churpek JE, Keel SB, et al. Germline ETV6 mutations in familial thrombocytopenia and hematologic malignancy. Nat Genet. 2015;47(2):180-185. 4. Rowley JD. The critical role of chromosome translocations in human leukemias. Annu Rev Genet. 1998;32:495-519. 5. Bejar R, Stevenson K, Abdel-Wahab O, et al. Clinical effect of point mutations in myelodysplastic syndromes. N Engl J Med. 2011;364(26):2496-2506. 6. Van Vlierberghe P, Ambesi-Impiombato A, Perez-Garcia A, et al. ETV6 mutations in early immature human T cell leukemias. J

haematologica | 2017; 102(2)

recruitment in the peripheral blood. Altered interaction between mutated progenitors and the BM microenvironment, as reported in the case of the ETV6-RUNX1 fusion protein, may also be involved.33 Further investigations are required to more precisely delineate the role that ETV6 plays in stem cell progenitor mobilization. In conclusion, we identified six variants in ETV6, of which five are novel, associated with dominant thrombocytopenia. Our study provides novel insights into the role that ETV6 plays in platelet function, morphology and formation that seem to be driven by changes in the cytoskeleton and potentially also in circulating CD34+ progenitor levels. Funding Bioinformatics analyzes benefit from the C2BIG computing centre funded by the Région Ile de France and UPMC. This work was partially supported by the ICAN Institute of Cardiometabolism and Nutrition (ANR-10-IAHU-05), the Ligue nationale contre le cancer (Labeled team H Raslova), and the “Fondation pour la Recherche Médicale FRM” (grant to PS FDM20150633607). We thank Dr. J. Ghysdael and Dr. F. Guidez for providing the plasmid constructs, Laboratory of Pr. D. Raoult (URMITE), microscopy unit (P. Weber), Dr Paola Ballerini (Hopital Trousseau) for genotyping; Dr. JC. Bordet for transmission electron microscopy; Dr C Chomiene and C Dosquet for the platelet survival assay; M Crest for experimental help, the French Reference Center on Hereditary Platelet Disorders (CRPP) for patients’ clinical exploration. For the F2-F6 families, study makes use of whole genome sequencing data and analysis approaches generated by the NIHR BioResource - Rare Disease BRIDGE Consortium. The NIHR BioResource - Rare Diseases is funded by the National Institute for Health Research of England (NIHR; award number RG65966)). KF is supported by the Fund for Scientific Research-Flanders (FWO-Vlaanderen, Belgium, G.0B17.13N] and by the Research Council of the University of Leuven (BOF KU Leuven‚ Belgium, OT/14/098].

Exp Med. 2011;208(13):2571-2579. 7. Kwiatkowski BA, Bastian LS, Bauer TR, Jr., Tsai S, Zielinska-Kwiatkowska AG, Hickstein DD. The ets family member Tel binds to the Fli-1 oncoprotein and inhibits its transcriptional activity. J Biol Chem. 1998;273(28):17525-17530. 8. Chakrabarti SR, Sood R, Ganguly S, Bohlander S, Shen Z, Nucifora G. Modulation of TEL transcription activity by interaction with the ubiquitin-conjugating enzyme UBC9. Proc Natl Acad Sci U S A. 1999;96(13):7467-7472. 9. Wang L, Hiebert SW. TEL contacts multiple co-repressors and specifically associates with histone deacetylase-3. Oncogene. 2001;20(28):3716-3725. 10. Green SM, Coyne HJ, 3rd, McIntosh LP, Graves BJ. DNA binding by the ETS protein TEL (ETV6) is regulated by autoinhibition and self-association. J Biol Chem. 2010;285(24):18496-18504. 11. Wang LC, Swat W, Fujiwara Y, et al. The TEL/ETV6 gene is required specifically for hematopoiesis in the bone marrow. Genes Dev. 1998;12(15):2392-2402.

12. Nurden P, Debili N, Vainchenker W, et al. Impaired megakaryocytopoiesis in type 2B von Willebrand disease with severe thrombocytopenia. Blood. 2006;108(8):25872595. 13. Thakur ML, Walsh L, Malech HL, Gottschalk A. Indium-111-labeled human platelets: improved method, efficacy, and evaluation. J Nucl Med. 1981;22(4):381385. 14. Westbury SK, Turro E, Greene D, et al. Human phenotype ontology annotation and cluster analysis to unravel genetic defects in 707 cases with unexplained bleeding and platelet disorders. Genome Med. 2015;7(1):36. 15. Turro E, Greene D, Wijgaerts A, et al. A dominant gain-of-function mutation in universal tyrosine kinase SRC causes thrombocytopenia, myelofibrosis, bleeding, and bone pathologies. Sci Transl Med. 2016;8(328):328ra330. 16. Barton GJ, Sternberg MJ. A strategy for the rapid multiple alignment of protein sequences. Confidence levels from tertiary structure comparisons. J Mol Biol.

293


M. Poggi et al. 1987;198(2):327-337. 17. Lopez RG, Carron C, Oury C, Gardellin P, Bernard O, Ghysdael J. TEL is a sequencespecific transcriptional repressor. J Biol Chem. 1999;274(42):30132-30138. 18. Guidez F, Petrie K, Ford AM, et al. Recruitment of the nuclear receptor corepressor N-CoR by the TEL moiety of the childhood leukemia-associated TEL-AML1 oncoprotein. Blood. 2000;96(7):2557-2561. 19. Bluteau D, Balduini A, Balayn N, et al. Thrombocytopenia-associated mutations in the ANKRD26 regulatory region induce MAPK hyperactivation. J Clin Invest. 2014;124(2):580-591. 20. Lordier L, Bluteau D, Jalil A, et al. RUNX1induced silencing of non-muscle myosin heavy chain IIB contributes to megakaryocyte polyploidization. Nat Commun. 2012;3:717. 21. Naldini L, Blomer U, Gallay P, et al. In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science. 1996;272(5259):263-267. 22. Raslova H, Komura E, Le Couedic JP, et al. FLI1 monoallelic expression combined with its hemizygous loss underlies ParisTrousseau/Jacobsen thrombopenia. J Clin Invest. 2004;114(1):77-84. 23. Klimchenko O, Mori M, Distefano A, et al. A common bipotent progenitor generates the erythroid and megakaryocyte lineages in embryonic stem cell-derived primitive hematopoiesis. Blood. 2009;114(8):15061517. 24. Antony-Debre I, Bluteau D, Itzykson R, et al. MYH10 protein expression in platelets as a biomarker of RUNX1 and FLI1 alterations. Blood. 2012;120(13):2719-2722. 25. Goggs R, Williams CM, Mellor H, Poole

294

26. 27.

28.

29.

30.

31.

32.

33.

AW. Platelet Rho GTPases-a focus on novel players, roles and relationships. Biochem J. 2015;466(3):431-442. Seshagiri S, Stawiski EW, Durinck S, et al. Recurrent R-spondin fusions in colon cancer. Nature. 2012;488(7413):660-664. Dolnik A, Engelmann JC, ScharfenbergerSchmeer M, et al. Commonly altered genomic regions in acute myeloid leukemia are enriched for somatic mutations involved in chromatin remodeling and splicing. Blood. 2012;120(18):e83-92. Melazzini F, Palombo F, Balduini A, et al. Clinical and pathogenetic features of ETV6 related thrombocytopenia with predisposition to acute lymphoblastic leukemia. Haematologica. 2016. Coyne HJ, 3rd, De S, Okon M, et al. Autoinhibition of ETV6 (TEL) DNA binding: appended helices sterically block the ETS domain. J Mol Biol. 2012;421(1):67-84. Sakurai T, Yamada T, Kihara-Negishi F, et al. Effects of overexpression of the Ets family transcription factor TEL on cell growth and differentiation of K562 cells. Int J Oncol. 2003;22(6):1327-1333. Bluteau D, Glembotsky AC, Raimbault A, et al. Dysmegakaryopoiesis of FPD/AML pedigrees with constitutional RUNX1 mutations is linked to myosin II deregulated expression. Blood. 2012;120(13):2708-2718. Breton-Gorius J, Favier R, Guichard J, et al. A new congenital dysmegakaryopoietic thrombocytopenia (Paris-Trousseau) associated with giant platelet alpha-granules and chromosome 11 deletion at 11q23. Blood. 1995;85(7):1805-1814. Palmi C, Fazio G, Savino AM, et al. Cytoskeletal regulatory gene expression and migratory properties of B cell progeni-

34.

35.

36. 37.

38.

39.

40.

41.

tors are affected by the ETV6-RUNX1 rearrangement. Mol Cancer Res. 2014; 12:1796-1806. Song H, Kim JH, Rho JK, Park SY, Kim CG, Choe SY. Functional characterization of TEL/AML1 fusion protein in the regulation of human CR1 gene promoter. Mol Cells. 1999;9(5):560-563. Rho JK, Kim JH, Yu J, Choe SY. Correlation between cellular localization of TEL/AML1 fusion protein and repression of AML1mediated transactivation of CR1 gene. Biochem Biophys Res Commun. 2002;297(1):91-95. Aslan JE, McCarty OJ. Rho GTPases in platelet function. J Thromb Haemost. 2013;11(1):35-46. Pleines I, Eckly A, Elvers M, et al. Multiple alterations of platelet functions dominated by increased secretion in mice lacking Cdc42 in platelets. Blood. 2010;115(16): 3364-3373. Pleines I, Hagedorn I, Gupta S, et al. Megakaryocyte-specific RhoA deficiency causes macrothrombocytopenia and defective platelet activation in hemostasis and thrombosis. Blood. 2012;119(4):1054-1063. Suzuki A, Shin JW, Wang Y, et al. RhoA is essential for maintaining normal megakaryocyte ploidy and platelet generation. PLoS One. 2013;8(7):e69315. Pleines I, Dutting S, Cherpokova D, et al. Defective tubulin organization and proplatelet formation in murine megakaryocytes lacking Rac1 and Cdc42. Blood. 2013;122(18):3178-3187. Chen C, Song X, Ma S, et al. Cdc42 inhibitor ML141 enhances G-CSF-induced hematopoietic stem and progenitor cell mobilization. Int J Hematol. 2015;101(1):5-12.

haematologica | 2017; 102(2)


ARTICLE

Bone Marrow Failure

Necroptosis in spontaneously-mutated hematopoietic cells induces autoimmune bone marrow failure in mice

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Junping Xin,1,2,3 Peter Breslin,1,4,5 Wei Wei,1 Jing Li,6 Rafael Gutierrez,1 Joseph Cannova,1 Allen Ni,1 Grace Ng,1 Rachel Schmidt,1 Haiyan Chen,7 Vamsi Parini,7 Paul C. Kuo,1 Ameet R. Kini,7 Patrick Stiff,1 Jiang Zhu8 and Jiwang Zhang1,7

Oncology Institute, Cardinal Bernardin Cancer Center, Loyola University Medical Center, Maywood, IL, USA; 2Research and Development Service, Hines VA Hospital, Hines, IL, USA; 3 Department of Molecular Pharmacology and Therapeutics, Loyola University Medical Center, Maywood, IL, USA; 4Department of Biology, Loyola University Chicago, IL, USA; 5 Department of Molecular/Cellular Physiology, Loyola University Medical Center, Maywood, IL, USA; 6Department of Biology, College of Life and Environment Science, Shanghai Normal University, P.R. of China; 7Department of Pathology, Loyola University Medical Center, Maywood, IL, USA and 8State Key Laboratory for Medical Genomics and Shanghai Institute of Hematology and Collaborative Innovation Center of Hematology, Rui-Jin Hospital; Shanghai Jiao-Tong University School of Medicine, P.R. of China

1

Haematologica 2017 Volume 102(2):295-307

ABSTRACT

A

cquired aplastic anemia is an autoimmune-mediated bone marrow failure syndrome. The mechanism by which such an autoimmune reaction is initiated is unknown. Whether and how the genetic lesions detected in patients cause autoimmune bone marrow failure have not yet been determined. We found that mice with spontaneous deletion of the TGFb-activated kinase-1 gene in a small subset of hematopoietic cells developed bone marrow failure which resembled the clinical manifestations of acquired aplastic anemia patients. Bone marrow failure in such mice could be reversed by depletion of CD4+ T lymphocytes or blocked by knockout of interferon-γ, suggesting a Th1-cell-mediated autoimmune mechanism. The onset and progression of bone marrow failure in such mice were significantly accelerated by the inactivation of tumor necrosis factor-α signaling. Tumor necrosis factor-α restricts autoimmune bone marrow failure by inhibiting type-1 T-cell responses and maintaining the function of myeloid-derived suppressor cells. Furthermore, we determined that necroptosis among a small subset of mutant hematopoietic cells is the cause of autoimmune bone marrow failure because such bone marrow failure can be prevented by deletion of receptor interacting protein kinase-3. Our study suggests a novel mechanism to explain the pathogenesis of autoimmune bone marrow failure.

Correspondence: neuroimmune@gmail.com/jzhang@luc.edu

Received: June 27, 2016. Accepted: September 12, 2016. Pre-published: September 15, 2016. doi:10.3324/haematol.2016.151514

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/295

Introduction Acquired aplastic anemia (AAA) is a common type of bone marrow failure (BMF) syndrome characterized by significant reductions in bone marrow (BM) cellularity and peripheral blood (PB) pancytopenia.1-3 Activated Th1 and Tc1 lymphocyte-mediated autoimmune responses are known to be the major reasons underlying hematopoietic repression since elimination of such cells results in the successful recovery of hematopoiesis in more than 70% of patients.4-9 Cryptic clonal somatic mutations are detected in BM samples from most AAA patients.5,10-16 As a consequence, AAA patients show an increased risk for the development of paroxysmal nocturnal hemoglobinuria, myelodysplastic syndrome and acute myeloid leukemia. In addition, autoimmune-related inflammatory reactions were detected haematologica | 2017; 102(2)

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

295


J. Xin et al.

in almost all AAA patients, as demonstrated by increased levels of both tumor necrosis factor-α (TNFα) and interferon-γ (IFNγ) in patient samples.17-20 However, the mechanism of activation of this autoimmune reaction is still not known. The relationship between the cryptic clonal somatic mutations and the autoimmune pathogenesis of AAA has not been identified, and the roles of TNFα and IFNγ in the development and progression of AAA have not been fully determined. TGF1b-activated kinase-1 (Tak1) is a member of the MAP3K family which mediates TNFα, IL1b and TLR-stimulated NF-kB, JNK and p38 signaling.21 It is required for cell survival and prevention of inflammatory reactions in many tissues.22 We previously reported that Tak1 is required to protect hematopoietic stem progenitor cells (HSPCs) from inflammatory cytokine-induced apoptosis and necroptosis. Necroptosis is a type of programmed necrosis which is regulated by the Rip1-Rip3Mlkl pathway. As distinct from apoptotic cells, which are eliminated by macrophages and do not induce inflammatory reactions, necroptotic cells secrete factors which do stimulate immune/inflammatory reactions.23,24 Mx1Cre+Tak1fx/fx mice (Tak1mut) develop severe BMF within 2-3 days of polyI:C injection to induce Tak1 deletion in up to 100% of BM hematopoietic cells.25,26 Interestingly, we found that without polyI:C injection, spontaneous deletion of the Tak1 gene in a low percentage of hematopoietic cells (1%-3%) occurs due to the leakage of Mx1Cre,27 causing a chronic autoimmune BMF in mice after longterm observation. TNFα is thought to be a key mediator of autoimmune BMF.28,29 However, inactivation of TNFα signaling surprisingly accelerates the progression of BMF in Tak1mut mice, an observation which runs counter to generally accepted theory. The BMF phenotypes of Tak1mut and Tak1mutTnfr–/– mice (deletion of both TNF receptor 1 and 2) resemble most of the clinical manifestations seen in chronic and severe AAA patients, respectively.4 Thus these experimental animals provide a better model system by which to study the autoimmune pathogenesis of AAA. Importantly, although necroptosis has been implicated in BMF in hematopoietic-specific Tak1 and receptor interacting protein kinase-1 (Rip1)-knockout mice, these studies suggested that both Tak1 and Rip1 are required for the survival of HSPCs by repressing apoptosis and necroptosis.25,30,31 However, due to the rapid death of the knockout mice, none of these studies was able to detect the activation of immune cells. In addition, although it has been proposed that necroptotic cells can be immunogenic,23,24,32 to our knowledge, we are the first to experimentally demonstrate that necroptotic cells actually do initiate autoimmune diseases. It was known that some hematopoietic parameters (such as HSPC numbers, red blood counts and hemoglobin concentration), as well as immune responses, differ between male and female mice.33,34 We were concerned that the phenotypic differences in autoimmune reactivity and hematopoiesis between male and female mice might influence our data interpretation. Therefore, in this study, only male mice were used for analysis. In addition, we and others found that Tnfr–/–, Ifnγ–/– and Rip3–/– mice are phenotypically normal and their hematopoietic parameters are comparable to those of WT mice. Thus, to make our data easier to comprehend, we include only data from single-gene knockout mice as controls for studies of compound-gene knockout mice. 296

Methods Mice and genotyping All mice were maintained in a C57BL6/J background and housed under a 12-h light/dark cycle in micro-isolator cages contained within a laminar flow ventilation system. All procedures were conducted in accordance with the National Institutes of Health guidelines for the care and use of laboratory animals for research purposes and were approved by Loyola University Chicago’s Institutional Animal Care and Use Committee (IACUC) (AU#513380). Ifnγ–/– and Tnfr–/– mice (B6.129STnfrsf1aTnfrsf1b, knockout of both Tnfr 1 and 2) were purchased from the Jackson Laboratory. Rip3–/– mice were kindly provided by Dr. Vishva Dixit (Genentech Inc., South San Francisco, CA, USA).35 Mx1cre+Tak1fx/fxTnfr–/– mice (Tak1mutTnfr–/–) are maintained from breeders of Mx1cre+Tak1fx/+Tnfr–/– and Tak1fx/fxTnfr–/– mice. Mx1cre+Tak1fx/+Tnfr–/– and Tak1fx/fxTnfr–/– mice were further crossed with Ifnγ–/– and Rip3–/– mice, respectively, to produce Mx1cre+Tak1fx/fxTnfr–/– Ifnγ–/– (Tak1mutTnfr–/–Ifnγ–/–) and Mx1cre+Tak1fx/fxTnfr–/–Rip3–/– (Tak1mutTnfr–/–Rip3–/–) mice, as well as the heterozygous littermates including Tak1mutTnfr–/– Ifnγ–/+ and Tak1mutTnfr–/–Rip3–/+. The gross phenotypes of these mice are summarized in Table 1. The PCR primers used for genotyping of experimental mice are listed in the Online Supplementary Appendix.

Bone marrow failure monitoring, blood analysis and mouse survival All mice were monitored for BMF development by dynamic examination of peripheral blood cell counts and observation for symptoms such as hunched body, growth arrest and/or significant weight loss. The death of mice specifically from BMF was confirmed by hematologic analysis using a Hemavet 950 Hematology System (Drew Scientific Inc., FL, USA). Spleens, livers and BM were collected upon animal sacrifice for further validation. All mice were monitored up to two years of age and analyzed by Kaplan-Meier survival graphing (GraphPad Prism v.5.04).

Histological analysis Bones were decalcified and fixed with CalforTM decalcifying solution (Cancer Diagnostic Inc., Durham, NC, USA) according to the manufacturer’s instructions. Spleens were fixed in 10% zinc-formalin. Tissues were sectioned and stained with H&E. Photographs were taken using an Olympus BX50 microscope equipped with a digital camera system (DP21).

Flow cytometric analysis For intracellular staining, cells were pre-treated with PMA (50 ng/mL; St. Louis, MO, USA) for six hours in the presence of 10 μg/mL Brefeldin-A; St. Louis, MO, USA) for the final four hours, followed by a 5-minute fixation in 4% paraformaldehyde, and were permeabilized with saponin (0.1%; St. Louis, MO, USA). Cells were suspended in FACS buffer (1xPBS supplemented with 2% FBS) at a concentration of 1x107 cells mL-1 and aliquotted into flow cytometry tubes for antibody staining. Surface staining was performed without fixation or permeabilization. Stained cells were subjected to multi-color analysis using a BD LSRFortessaTM flow cytometer. Data were analyzed using Flowjo software. During the analysis, cells were first gated on live cells, then further analyzed for specific staining. The antibody resource and clone information are listed in the Online Supplementary Appendix. Further details on the methods used can be found in the Online Supplementary Appendix. haematologica | 2017; 102(2)


Necroptotic cells initiate autoimmune bone marrow failure

Statistical analysis Data are expressed as means±SD. Two-way ANOVA (multiple groups) and Student’s t-test (two groups) were performed to determine the statistical significance of differences among and between experimental groups. P<0.05 was considered significant.

Results Development of chronic BMF in Tak1mut mice Due to endogenous Ifnα expression, even without polyI:C injection, Mx1Cre induces the spontaneous dele-

tion of target genes in BM hematopoietic cells which is sufficient to induce the development of T-cell leukemia in Ptenfx/fx mice.27 Using GFP-reporter mice, we found Mx1Creinduced GFP expression was detected in 1%-3% and 5%7% of BM hematopoietic cells when analyzed in 3- and 12-month old mice, respectively (Online Supplementary Figure S1). To determine whether deletion of Tak1 in this small subset of hematopoietic cells influences normal hematopoietic homeostasis, we maintained a cohort of Tak1mut mice for long-term observation to dynamically examine the PB cell counts. Mx1Cre– littermates and Cre+ littermates (including Mx1Cre+Tak1fx/+ and Mx1Cre+Tak1+/+)

A

C

B

D

E

F H

G

Figure 1. Spontaneous Tak1 deletion in a small subset of hematopoietic stem progenitor cells (HSPCs) results in chronic bone marrow failure (BMF). Peripheral blood (PB) and bone marrow (BM) were collected from Tak1mut mice and their WT and Cre+ littermates at age 14 months. (A) White blood cells (WBC), red blood cells (RBC), hemoglobin (Hb) and platelets (plt) were analyzed using Hemavet 950 Hematology System. (B) H&E-stained bone marrow section (tibia) after decalcification. (C) Number of total nucleated cells (TNCs) in BM from two hind limbs (four bones pooled: 2 tibias and 2 femurs from each mouse) were compared. (D) Representative flow cytometric plots for analysis of BM hematopoietic stem cells (HSCs) and hematopoietic progenitor cells (HPCs). BM cells were first gated on Lin– population and then analyzed for LK, LSK and LS populations. (E) Percentages of LK, LSK and LS populations in TNCs from BM. (F) Proliferation of Lin-c-kit+ HSPCs was compared among WT, Cre+ and Tak1mut mice (14 months old) by BrdU pulse-labeling and flow cytometric analysis. (G) Death of Lin-c-kit+ HSPCs was compared among WT, Cre+ and Tak1mut mice at indicated ages by Annexin-V staining followed by flow cytometric analysis. (I) Percentages of cells with ΔTak1 (Tak1 deletion) in Lin- HSPCs, Lin+ differentiated BM cells and CD3+ T lymphocytes from Tak1mut mice were determined by quantitative PCR. Lin– HSPCs from WT mouse BM and Tak1–/– leukemic cells (LCs) were used as negative and positive controls. *P<0.05 compared to WT and Cre+ mice.

haematologica | 2017; 102(2)

297


J. Xin et al. were studied in parallel as controls. We found that Tak1mut mice were generally normal up to eight months of age, with no detectable pathological phenotype. The body size and weight of Tak1mut mice are comparable to their littermate controls (Table 1 and Online Supplementary Figure S2A and B). However, most Tak1mut mice start to develop chronic BMF after eight months, as demonstrated by reduction of white blood cell counts (WBC), red blood cell counts (RBC), hemoglobin concentration (Hb), and platelet numbers (plt) in PB, as well as a reduction in cell counts in BM (Figure 1A-C and Online Supplementary Figure S2H). In addition, all Tak1mut mice developed more pronounced thymic degeneration as indicated by decreased size and cell counts (Online Supplementary Figure S2C and D). Interestingly, despite the splenomegaly observed (increased spleen size and weight) (Online Supplementary Figure S2E and F), significantly lower numbers of cells can be collected from the spleens of Tak1mut mice (Online Supplementary Figure S2G) due to increased fibrosis. These phenotypic changes in Tak1mut mice resemble the clinical manifestations seen in chronic AAA patients.3,4 Further analysis of HSPCs showed that, compared to WT and Cre+ controls, the BM of Tak1mut mice showed a significantly increased percentage of Sca1v cells, including Lin–Sca1+c-kit+ (LSK) and Lin–Sca1+c-kit– (LS) cells, in Lineage– (Lin– ) (Figure 1D) or total nucleated cell (TNC)

A

populations (Figure 1E). However, the absolute number of LSK cells was reduced due to the reduction in TNC (Online Supplementary Figure S2I). Such alterations in HSPCs in Tak1mut mice are similar to the HSPC changes observed in mice with chronically enhanced Ifnγ.20,36 HSCs and MPPs in such mice cannot be reliably analyzed using any of the current standard panels of surface markers such as CD150, CD34, or FLT3 (data not shown).

Enhanced Th1 and Tc1-immune responses in Tak1mut mice Theoretically, increased elimination of 1%-3% of HSPCs in BM is insufficient to induce BMF in animals because it has been shown that approximately 5% of normal HSPCs are sufficient to maintain normal hematopoietic homeostasis in mice. One of the possible mechanisms of BMF development in TAK1mut mice is exhaustion of HSPCs due to the constant elimination of the TAK1mut HSPCs. Using Annexin-V staining assays, we found that the percentage of death in HSPCs is comparable between TAK1mut mice and the littermate controls before BMF developed. Reduced proliferation and increased death of HSPCs were only observed in TAK1mut mice after BMF had developed (Figure 1F and G), which did not support such a hypothesis. Another possible explanation is the accumulation of Takmut HSPCs, which have compromised

B

C E

D

F

Figure 2. Increased activation of Th1 and Tc1 cells in Takmut mice. Bone marrow (BM) samples were collected from Tak1mut mice and their WT and Cre+ littermates at age 14 months. (A) Flow cytometric analysis of BM CD3+ T cells and B220+ B cells (gated on lymphocytes). (B) Flow cytometric analysis of naïve T cells (CD62L+ CD44–) and effector T cells (CD62LlowCD44hi) in BM (gated on CD3+ cells). (C) Percentages of Th1, Tnfa+, Th17, Th2, and Treg cells in BM (gated on CD4+ T cells). (D) Percentages of Tc1, Tnfa+, Tc17, Tc2, and Treg cells in BM (gated on CD8+ T cells). Ifnγ and Tnfα levels in BM CD4+ (E) and CD8+ (F) T cells were analyzed by mean fluorescence intensity (MFI) of intracellular antibody staining. Data are presented as means±SD. *P<0.05 compared to WT and Cre+ mice.

298

haematologica | 2017; 102(2)


Necroptotic cells initiate autoimmune bone marrow failure

hematopoietic reconstitutive capacity. We excluded such a possibility since fewer than 0.5% of HSPCs with Tak1 deletion (ΔTak1) could be detected in Tak1mut mice of any age before or after the development of BMF (Figure 1H); most Takmut HSPCs died of apoptosis or necroptosis shortly after Tak1 was deleted (Online Supplementary Figure S3).

A

Interestingly, WT mice whose BM cells were chimerized with Mx1Cre+Tak1fx/fx hematopoietic cells but not WT hematopoietic cells also developed a pancytopenic phenotype 1-2 months after 3 polyI:C injections (Online Supplementary Figure S4B-D). Low percentages of Mx1Cre+Tak1fx/fx hematopoietic cells were detected in these

B

C

E

G

I

K

D

F

H

J

L

Figure 3. Deficiency of Tnfr accelerates bone marrow failure (BMF) and enhances Th1 cell responses in Tak1mut mice. (A) Survival of Tak1mutTnfr–/– and Tak1mut and Tnfr–/– mice were recorded and compared (*P<0.05, compared to Tnfr–/– mice; #P<0.05 compared to Tak1mut mice). Peripheral blood (PB) and bone marrow (BM) were collected from 4-month old Tak1mutTnfr–/– mice and age-matched Tak1mut and Tnfr–/– mice. (B) White blood cells, (WBC), red blood cells (RBC), hemoglobin (Hb) and platelets (plt) were analyzed using the Hemavet 950 Hematology System. (C) H&E stained bone marrow section (tibia) after decalcification. (D) Number of total nucleated cells (TNCs) in BM from two hind limbs were counted and compared. (E) Representative flow cytometric plots for analysis of BM hematopoietic stem cells (HSCs) and hematopoietic progenitor cells (HPCs). BM cells were first gated on the Lin– population and then analyzed for LK, LSK and LS populations. (F) Percentages of LK, LSK and LS populations in total nucleated cells (TNCs) from BM. (G) Flow analysis of BM CD3+ T cells and B220+ B cells (gated on lymphocytes). (H) Flow cytometric analysis of naïve T cells (CD62L+ CD44–) and activated T cells (CD62LlowCD44hi) in BM (gated on CD3+ cells). (I) Percentages of Th1, Tnfα+, Th17, Th2, and Treg cells in BM (gated on CD4+ T cells). Ifnγ and Tnfα levels in BM CD4+ (J) and CD8+ (I) T cells were analyzed by mean fluorescence intensity (MFI) of intracellular antibody staining. (K) Percentages of Tc1, Tnfα+, Tc17, Tc2, and Treg cells in BM (gated on CD8+ T cells). Data are presented as means±SD. *P<0.05 compared to Tnfr--/-; #P<0.05 compared to Tak1mut.

haematologica | 2017; 102(2)

299


J. Xin et al.

recipient mice after pancytopenia developed (Online Supplementary Figure S4A), suggesting that the factors released by dying Takmut hematopoietic cells might be inducing BMF. The detection of cryptic clonal genetic mutations in AAA patients suggests that a small subset of mutant cells might be the cause of T-cell mediated autoimmune reactions and hematopoietic repression.5,10-15 Thus, we speculated that BMF in Tak1mut mice might be induced by Tak1mut cells through the stimulation of autoimmune responses. To test this hypothesis, we examined T and B lymphocytes in BM, PB, lymph nodes and spleens of Tak1mut mice. While the frequencies of CD3+ T lymphocytes were significantly increased in all tested tissues, B220+ B lymphocytes were remarkably reduced in Tak1mut mice compared to control mice (Figure 2A; only data from BM are shown), suggesting B lymphopoiesis is repressed. In addition, T-cell activation was increased in Tak1mut mice as shown by a decrease in naïve T cells (CD62L+CD44–)

A

and a simultaneous increase in CD62LlowCD44high effector T cells (Figure 2B; only data from BM are shown). Furthermore, we used intracellular staining and FACS analysis to examine the subtypes of the CD4+ and CD8+ Tcell subsets in the BM (Online Supplementary Figure S5). Compared to control mice, Tak1mut mice showed a significant increase in Th1 and Tc1 responses, as indicated by increased frequencies of Tnfα/Ifnγ-expressing Th1 (CD4+) and Tc1 (CD8+) cells (Online Supplementary Figure S2C and D), as well as by an increase in Tnfα and Ifnγ expression (Figure 2E and F). Of note, compared to WT and Cre+ controls, Tak1mut mice displayed a decrease in Foxp3+ Treg cells but no significant difference in the frequencies of IL-17expressing Th17/Tc17 or IL-4-expressing Th2/Tc2 cells. Taken together, these data suggested that Th1/Tc1 cellmediated autoimmune responses might be the cause of BMF in Tak1mut mice. TCR repertoire analysis suggested an expansion of oligoclone Th1 cells in Tak1mut BMF mice as demonstrated by a skewed spectratype profile with 1-2

B

C

D

E

F

* * G

H

* *

300

Figure 4. Enhanced function of Tnfr–/– APCs and decreased function of Tnfr–/– MDSCs. (A and B) T cells from wild-type (WT) spleens were first isolated using a pan-T-cell isolation kit, then separated with anti-CD62L. Naïve T cells (CD62L+) were further cultured with WT and Tnfr–/– APCs for seven days under three conditions: 1) neutral condition (N); 2) Th1-priming condition (C1, in the presence of IL-12 and anti-IL-4); and 3) Th2-priming condition (C2, in the presence of IL-4 and anti-IL-12). Cells were collected and subjected to intracellular staining for cytokines (A). The mean fluorescence intensity (MFI) of Ifnγ in all of the groups was calculated and is presented in bar graphs (B). (C-E) CFSE-labeled WT T cells were co-cultured with either WT or Tnfr–/– APCs in the presence of anti-CD3 and anti-CD28. After three days of culturing, cells were collected and analyzed for the percentage of T cells with more than 3 divisions (C). WT and Tnfr–/– APCs were stained for intracellular IL-12 (D). MFI of IL-12 in the two groups in (D) was calculated and presented in bar graphs (E). (F) CD11b+Gr1+ cells were sorted from WT and Tnfr–/– bone marrow and added to anti-CD3activated, CFSE-labeled WT T cells at a ratio of T:MDSC=1:2 and co-cultured for six days. The CFSE signals were analyzed on day 6. The results were analyzed to show the cells with a low number of divisions (2-6 divisions). (G) T cells from WT spleens were isolated using a pan-T-cell isolation kit and were cultured with WT and Tak1mutTnfr–/–APCs for seven days under Th1-priming condition. Cells were collected and subjected to intracellular staining for Ifnγ. (I) CD11b+Gr1+ cells were sorted from WT and Tak1mutTnfr–/– BM, added to anti-CD3activated, CFSE-labeled WT T cells at a ratio of T:MDSC=1:2 and co-cultured for six days. The CFSE signals were analyzed on day 6. The results were analyzed to show the cells with a low number of divisions. The samples were analyzed in triplicate and experiments were repeated independently twice. *P<0.05 compared to WT.

haematologica | 2017; 102(2)


Necroptotic cells initiate autoimmune bone marrow failure

dominant peaks (Online Supplementary Figure S6). We excluded the possibility that accumulation of Tak1 deletion in T cells is the direct cause of T-cell overactivation in Tak1mut mice because: 1) Tak1 is required for the survival of peripheral T cells;37-39 and 2) ΔTak1 is barely detected in T cells in Tak1mut mice at any age before or after BMF development (Figure 1H).

Tnfα signaling inactivation accelerated disease onset and progression of BMF

Increased levels of TNFα have been detected in many AAA patients, and such levels have been suggested to be contributory to the pathogenesis of AAA.17-19,28,29 We have reported that inactivation of Tnfα signaling by knockout of both Tnfr1 and 2 (Tnfr–/–) can partially prevent the death of Tak1-null HSPCs.25 Thus, we expected that inactivation of Tnfα signaling may attenuate the progression of BMF in Tak1mut mice. However, to our surprise, we found that the progression of BMF in Tak1mut Tnfr–/– mice without polyI:C induction was dramatically exacerbated and much more severe than what was observed in Tak1mut mice. All Tak1mut Tnfr–/– mice were infertile and started to show growth retardation at two months of age compared to Tak1mut and Tnfr–/– mice (Online Supplementary Figure S7A and B). All Tak1mut Tnfr–/– mice developed more pronounced thymic degeneration as indicated by decreased thymic size and cell counts (Online Supplementary Figure S7C and D), as well as splenomegaly but with reduced cellularity (Online

Supplementary Figure S7E-G). BMF could be detected in Tak1mut Tnfr–/– mice as early as two months of age. Almost all Tak1mut Tnfr–/– mice become morbid and died by approximately 3-7 months of age (Table 1 and Figure 3A). In contrast, Tak1mut mice of the same age grow normally and breed with no signs of BMF. These mice develop severe pancytopenia by four months of age, characterized by significant reductions in red blood cell (RBC), hemoglobin (Hb), platelet (plt), white blood cell (WBC) and BM cell counts (Figure 3B-D and Online Supplementary Figure S7H). Reduced proliferation and increased death of HSPCs could only be detected after BMF had developed, suggesting they are not the initiators of BMF (Online Supplementary Figure S7K and I). The hematopoietic phenotype in 4month old Tak1mutTnfr–/– mice resembles the clinical manifestations of severe AAA,3,4,40 and is much more severe than that of 14-month old Tak1mut mice. Analysis of HSPCs showed that among Lin- BM cells, almost all c-Kit+ cells expressed Sca1 (Figure 3E and F and Online Supplementary Figure S7I), a phenomenon commonly observed in mice with increased Ifnγ activity.41-44 These results suggested that Tnfα signaling restricts the development of BMF. Although inactivation of Tnfα signaling can partially prevent necroptosis of Tak1mut HSPCs,25 most Tnfr–/–Tak1mut HSPCs eventually die of apoptosis. This explains why inactivation of Tnf signaling only slightly extends the lifespan of Tak1–/– mice upon polyI:C injection (maximum of 1 week).25 Consistent with this observation, we did not

A

B

D

C

E

Figure 5. Depletion of CD4+ T cells restores normal hematopoiesis to Tak1mutTnfr–/– mice. (A) Bone marrow (BM) samples were collected from wild-type (WT) mice and Tak1mutTnfr–/– mice treated with IgG and anti-CD4 antibody. Flow cytometric analyses of the frequencies of CD4+ and CD8+ T cells are shown. (B) White blood cells (WBC), red blood cells (RBC), hemoglobin (Hb) and platelets (plt) were analyzed using the Hemavet 950 Hematology System. (C) Number of total nucleated cells (TNCs) in BM from two hind limbs were counted and compared. (D) Representative flow cytometric plots for analysis of BM hematopoietic stem cells (HSCs) and hematopoietic progenitor cells (HPCs). BM cells were first gated on the Lin– population and then analyzed for LK, LSK and LS populations. (E) Percentages of LK, LSK and LS populations in TNCs of BM. Data are presented as means±SD. *P<0.05 compared to WT; #P<0.05 compared to IgG treatment.

haematologica | 2017; 102(2)

301


J. Xin et al. detect any accumulation of HSPCs in ΔTak1 in Tak1mut Tnfr–/– mice without polyI:C injection (Online Supplementary Figure S7J). We also failed to detect increased HSPC death in Tak1mut Tnfr–/– mice before BMF development when compared to WT and Tak1mut mice

(Online Supplementary Figure S7K). Increased apoptosis and reduced proliferation of HSPCs were detected only after BMF had developed (Online Supplementary Figure S7J and K). This suggested that inactivation of Tnfα signaling did not directly alter the survival or proliferation of HSPCs.

A

C

E

B

D

F

G

H

I

J

Figure 6. Ifnγ knockout prevents BMF in Tak1mutTnfr–/– mice. Peripheral blood (PB) and bone marrow (BM) samples were collected from Ifnγ–/–, Tak1mutTnfr–/–Ifnγ+/– and Tak1mutTnfr–/–Ifnγ–/– mice at the age of ten months (n=5/group). (A) White blood cells (WBC), red blood cells (RBC), hemoglobin (Hb) and platelets (plt) were analyzed using the Hemavet 950 Hematology System. (B) Number of total nucleated cells (TNCs) in BM from two hind limbs were counted and compared. (C) Representative flow cytometric plots for analysis of BM hematopoietic stem cells (HSCs) and hematopoietic progenitor cells (HPCs). BM cells were first gated on the Lin- population and then analyzed for LK, LSK and LS populations. (D) Percentages of LK, LSK and LS populations in total nucleated cells (TNCs) of BM. (E) Flow analysis of BM CD3+ T cells and B220+ B cells (gated on lymphocytes). (F) Flow cytometric analysis of naïve T cells (CD62L+ CD44–) and activated T cells (CD62LlowCD44hi) in BM (gated on CD3+ cells). (G) Percentages of Th1, Tnfα+, Th17, Th2, and Treg cells in BM (gated on CD4+ T cells). (I) Percentages of Tc1, Tnfa+, Tc17, Tc2, and Treg cells in BM (gated on CD8+ T cells). (H and J) Ifnγ and Tnfα levels in BM CD4+ (H) and CD8+ (J) T cells were analyzed by mean fluorescence intensity (MFI) of intracellular antibody staining. Data are presented as means±SD. *P<0.05 compared to Tnfr-/-; #P<0.05 compared to Tak1mutTnfr–/–Ifnγ+/–.

302

haematologica | 2017; 102(2)


Necroptotic cells initiate autoimmune bone marrow failure

Deficiency in Tnfα signaling enhances Th1- and Tc1-cell responses

mice, age-matched Tak1mut mice, and Tnfr–/– and WT control mice. At four months of age, when all the parameters analyzed in Tak1mut mice were still comparable to those of Tnfr–/– and WT mice (data not shown), a consistent decrease in B cells and an increase in T-cell frequency, as well as elevated T-cell activation, were observed in the spleens, PB and BM of Tak1mutTnfr–/– mice (Figure 3G and H; only data

To determine whether the acceleration of BMF in the absence of Tnfα signaling is due to an enhancement of autoimmune responses, we compared the frequency of lymphocytes and activation of T lymphocytes in hematopoietic and lymphoid tissues among Tak1mutTnfr–/–

A

C

E

B

D

F

G

H

I

J

Figure 7. Rip3 knockout prevents BMF in Tak1mutTnfr–/– mice. Peripheral blood (PB) and bone marrow (BM) samples were collected from 4-month old Rip3-/–, Tak1mutTnfr–/–Rip3+/– and Tak1mutTnfr–/–Rip3-/– mice. (A) White blood cells (WBC), red blood cells (RBC), hemoglobin (Hb) and platelets (plt) were analyzed using the Hemavet 950 Hematology System. (B) Number of total nucleated cells (TNCs) in BM from two hind limbs were counted and compared. (C) Representative flow cytometric plots for analysis of BM hematopoietic stem cells (HSCs) and hematopoietic progenitor cells (HPCs). BM cells were first gated on the Lin– population and then analyzed for LK, LSK and LS populations. (D) Percentages of LK, LSK and LS populations in TNCs of BM. (E) Flow analysis of BM CD3+ T cells and B220+ B cells (gated on lymphocytes). (F) Flow cytometric analysis of naïve T cells (CD62L+ CD44–) and activated T cells (CD62LlowCD44hi) in BM (gated on CD3+ cells). (G) Percentages of Th1, Tnfα+, Th17, Th2, and Treg cells in BM (gated on CD4+ T cells). (I) Percentages of Tc1, Tnfa+, Tc17, Tc2, and Treg cells in BM (gated on CD8+ T cells). Ifnγ and Tnfα levels in BM CD4+ (H) and CD8+ (J) T cells were analyzed by mean fluorescence intensity (MFI) of intracellular antibody staining. Data are presented as means±SD. *P<0.05 compared to Rip3–/–; #P<0.05 compared to Tak1mutTnfr–/–Rip3+/–.

haematologica | 2017; 102(2)

303


J. Xin et al.

from BM were shown). Intracellular staining and FACS analysis (Online Supplementary Figure S8) revealed a significant increase in Th1 and Tc1 cells and a decrease in Th17, Th2 and Treg cells in Tak1mutTnfr–/– mice (Figure 3I and K). The expression of Ifnγ and Tnfα in Th1 cells and Tc1 cells was also significantly elevated (Figure 3J and L). These data were similar to those obtained for Tak1mut BMF mice but to a much greater extent. For example, in BM of Tak1mutTnfr–/– mice, over 80% of T cells displayed an activated phenotype and over 70% of CD4+ T cells were Th1 cells and 45% of CD8+ T cells were Tc1 cells. These data indicate that a Tnfr deficiency prompts Th1/Tc-cell responses, which in turn accelerates BMF. Again, no accumulation of ΔTak1 in T cells was detected (Online Supplementary Figure S7J), suggesting that ΔTak1 in T cells is not the direct cause of T-cell overactivation in Tak1mutTnfr–/– mice.

Deficiency in Tnfα signaling enhances the ability of APCs to prime Type 1 T-cell development and reduces the ability of MDSCs to suppress T-cell proliferation

There are three potential mechanisms which could explain why a Tnfα signaling deficiency enhances Th1/Tc1-cell responses: 1) Tnfr–/– T cells are more prone to develop into Th1 cells than are WT T cells; 2) Tnfr–/– antigen-presenting cells (APCs) have an enhanced ability to support Th1/Tc1 responses compared to WT APCs; 3) a reduced number and/or capacity of Tnfr–/– myeloid-derived suppressor cells (MDSCs). To discriminate among these possibilities, we compared the frequencies of Ifnγ-expressing cells in both CD62Lhigh naïve and CD62Llow effector T cells from WT and Tnfr–/– mice. We found Ifnγ-expressing cells in both populations of T cells from Tnfr–/– mice were reduced compared to those from WT mice (Online Supplementary Figure S9A), suggesting that T-cell responses in Tnfr–/– mice are normally maintained at lower levels in the absence of pathogenic stimuli. To determine whether there are functional changes in APCs in Tnfr–/– mice compared to WT mice, we compared the expression of Ifnγ in WT T cells following co-culturing with either WT or Tnfr–/– APCs under three different conditions: neutral condition (N), Th1-inducing condition (C1), and Th2-inducing condition (C2). Following seven days of co-culturing, the frequency of Ifnγ-expressing T cells is consistently higher when cultured with Tnfr–/– APCs than with WT APCs under all three conditions (Figure 4A). In the presence of Tnfr–/– APCs, the frequency of Ifnγ-expressing T cells increases when compared to co-culturing with WT APCs (Figure 4A). In addition, T cells cultured with Tnfr–/– APCs showed higher levels of Ifnγ expression, as demonstrated by increased mean fluorescence intensity (Figure 4B). The frequency of Ifnγ-expressing T cells and expression levels of Ifnγ are consistent with the three conditions, as C1>N>C2 for both WT and Tnfr–/– groups. Tnfα and IL-17 expression were not obviously different between WT and Tnfr–/– groups (Online Supplementary Figure S9B). Furthermore, we measured the proliferation of T cells cultured with WT or Tnfr–/– APCs in the presence of anti-CD3 and anti-CD28. After three days in culture, the portion of T cells with more than 3 divisions was higher in the Tnfr–/– group than in the WT group (Figure 4C). As IL-12 is the critical cytokine that determines the differentiation of Th1 cells, we measured the expression of IL-12 in WT and Tnfr–/– APCs. We found that both the frequency of IL-12expressing cells and the levels of IL-12 were higher in 304

Tnfr–/– APCs than in WT APCs (Figure 4D and E). Using the same co-culture system, we determined that the activity of APCs in Tak1mutTnfr–/– mice was also increased compared to APCs isolated from WT littermate controls (Figure 4G). To test whether the number and/or function of MDSCs was altered in Tnfr–/– mice, we measured the frequency of CD11b+Gr1lowLy6Chi (containing granulocytic MDSCs) or CD11b+Gr1highLy6Clow (containing monocytic MDSCs) cells in the BM and conducted an inhibition assay. The frequencies of both populations in the BM were not significantly different when comparing WT and Tnfr–/– mice (Online Supplementary Figure S9C). However, Tnfr–/– MDSCs showed a reduced ability to suppress T-cell proliferation compared to WT MDSCs (Figure 4F and Online Supplementary Figure S9D). We also determined that the MDSCs isolated from Tak1mutTnfr–/– mice showed reduced ability to suppress T-cell proliferation compared to MDSCs isolated from WT littermate controls (Figure 4H). These data suggest that enhanced functions of APCs and reduced functions of MDSCs may concurrently contribute to the T-cell overactivation and high Ifnγ levels seen in Tak1mutTnfr–/– mice.

CD4+ lymphocyte depletion restores normal hematopoiesis to Tak1mutTnfr–/– mice T-cell depletion by anti-thymocyte globulin (ATG) has been demonstrated to be the most effective treatment for AAA.4-9 Our data indicate that CD4+ Th1 cells are the major type of cell that is increased in the BM of BMF mice. Therefore, we assessed whether depletion of CD4+ T cells would restore normal hematopoiesis to BMF mice. After the onset of BMF in Tak1mutTnfr–/– mice at the age of two months, we treated them with anti-CD4 antibody (i.p. injection of 250 µg/mouse, weekly). We found that antiCD4 antibody treatment efficiently depleted CD4+ T cells from BM (Figure 5A) and PB (Online Supplementary Figure 10A). Of note, following the depletion of CD4+ T cells, the frequency of CD8+ T cells also decreased to a normal level (Figure 5A), suggesting that the increased number of activated CD8+ T cells in BMF mice is secondary to the activation of CD4+ T cells. Strikingly, after anti-CD4 treatment for two months (9 injections), all BMF mice regained body weight and became grossly normal (Online Supplementary Figure S10B). In addition, anti-CD4 treatment prevented both degeneration of the thymus and splenomegaly (Online Supplementary Figure S10D-H). While all IgG-treated control mice died by six months of age, anti-CD4 treated mice (weekly treatment) survived for over ten months (Table 1). At the conclusion of the experiments, we examined the hematopoietic tissues of the experimental animals. We found that anti-CD4-treated mice showed normal thymic and splenic sizes and normal cell counts (Online Supplementary Figure S10D-F). RBC/Hb, WBC and plt in PB and BM cellularity values were comparable to WT control mice (Figure 5B and C and Online Supplementary Figure S10). Analysis of HSPCs showed that the expression of Sca1 in Lin– BM cells was also comparable in anti-CD4-treated mice and WT mice (Figure 5D and E). Also, Tnfα and Ifnγ levels in PB returned to normal levels (Online Supplementary Figure S15). Thus, we concluded that the BMF observed in Tak1mutTnfr–/– mice is caused by autoreactive CD4+ T lymphocytes. In addition, the reversible hematopoiesis in Tak1mutTnfr–/– BMF mice suggested that HSPCs were not exhausted in such mice. Th1-mediated hematopoietic repression is the key haematologica | 2017; 102(2)


Necroptotic cells initiate autoimmune bone marrow failure

Necroptosis

CD4+ T-cell activation Th1 cells

Figure 8. A working model for necroptosis-initiated autoimmune bone marrow failure (BMF). Antigens released from cells undergoing necroptosis owing to genetic mutation of Tak1 in a small subset of hematopoietic stem and progenitor cells (HSPC) stimulate CD4+ T-cell activation and promote the development of Th1/Tc1 cells, which in turn produce a large amount of IFNγ, resulting in bone marrow (BM) exhaustion. Tnf restricts the onset and progression of BMF by repressing Th1 and Tc1 activation. BMF in the current model can be prevented/cured by targeting the key components of the necropotosis-Th1 cell-Ifnγ axis.

pathogenic factor for BMF in such mice. Th1 and Tc1 cells did not seem to directly attack the HPSCs (Online Supplementary Figure S11), but induced apoptosis and repression of proliferation (Figure 1F and G), as well as induction of differentiation of HSPCs (Online Supplementary Figure S11) by secreting Ifnγ.

Ifnγ knockout prevents BMF development in Tak1mutTnfr-/- mice

Increased levels of IFNγ are believed to be important in the pathogenesis of AAA.20,45 Under appropriately clean housing conditions, Ifnγ–/– mice are grossly healthy without detectable hematopoietic defects. To determine whether Ifnγ mediates the hematopoietic repressive activity of activated CD4+ T cells in Tak1mutTnfr–/– mice, we studied whether Ifnγ deletion can prevent the development of AAA. As expected, Ifnγ deletion completely prevented development of the disease in these mice. Tak1mutTnfr–/–Ifnγ–/+ mice showed normal growth and development of both the thymus and bone marrow, bred normally and had the same lifespan as WT controls (Table 1 and Online Supplementary Figure S12A-G). All Tak1mutTnfr–/–Ifnγ–/– mice were comparable to WT and Ifnγ–/– control mice with respect to hematopoietic parameters in PB and BM (Figure 6A-F); B/T cell numbers and T-cell activation (Figure 6G-J) were likewise comparable. In addition, Tak1mutTnfr–/–Ifnγ–/+ mice (Ifnγ heterozygous littermates) showed an attenuated phenotype compared to Tak1mutTnfr–/– mice. BMF in Tak1mutTnfr–/–Ifnγ+/– mice was less severe than that seen in Tak1mutTnfr–/– mice but more severe than that seen in Tak1mut mice based on studies of hematopoietic parameters in PB, BM, spleen and thymus (Online Supplementary Figure S12C-G). Tak1mutTnfr–/–Ifnγ–/+ mice could survive up to ten months, a significant extension in lifespan compared to Tak1mutTnfr–/– mice (Table 1). These mice also showed a reduced Th1 responsiveness and Ifnγ expression than did Tak1mutTnfr–/–Ifnγ+/+ mice (Online Supplementary Figure S12I and J). It can be concluded that Ifnγ derived from Th1 cells is responsible for the development of BMF in a gene dose-dependent manner.

Rip3 deletion prevented BMF development in Tak1mutTnfr–/– mice HSPCs with Tak1 deletion died from either apoptosis or necroptosis.25,26 Apoptotic cells are usually removed by macrophages, which normally do not induce immune haematologica | 2017; 102(2)

reactions. However, necroptotic cells release intracellular components called DAMPs (damage-associated molecular patterns) which can induce immune responses.23,24,46-48 Thus we speculated that the Th1-mediated autoimmune BMF in Tak1mut mice might be induced by factors released from cells undergoing necroptosis. Rip3 is the key mediator of necroptosis.49,50 To test such a hypothesis, we crossed Rip3–/– mice with Cre+Tak1f/+Tnfr–/– mice to produce mice with Tak1mutTnfr–/–Rip3+/– and Tak1mutTnfr–/– Rip3–/– genotypes. Heterozygous deletion of Rip3 did not influence the disease phenotype of Tak1mutTnfr–/– mice. Interestingly, homozygous Rip3 deletion completely prevented BMF development. Tak1mutTnfr–/–Rip3–/– mice were grossly normal in growth and reproductive capability (Table 1 and Online Supplementary Figure S13A and B). RBC, Hb, plt and WBC in PB (Figure 6A), cellularity in BM, thymuses and spleens (Figure 6B and Online Supplementary Figure S13CG), BM HSPCs (Figure 6C and D), as well as B/T cell numbers and T-cell activation (Figure 6E-J) of Tak1mutTnfr–/–Rip3–/– mice were comparable to WT and Rip3–/– controls at the age of four months, a time point at which all Tak1mutTnfr–/– mice show severe BMF. Again, no accumulation of ΔTak1 HSPCs was detected in Tak1mutTnfr–/–Rip3–/– mice (Online Supplementary Figure S14), suggesting: 1) BMF in Tak1mutTnfr–/– mice is not due to an increase in ΔTak1 HSPCs; 2) inactivation of Rip3 did not prevent the elimination of ΔTak1 HSPCs. To support this notion further, we found that ΔTak1 HSPCs died of apoptosis when Rip3 was inactivated (Online Supplementary Figure S3).

Discussion Although the autoimmune feature of AAA has been well documented, the pathogenesis of this life-threatening disease is still not clear owing to the lack of an adequately representative animal model which can be used to test hypotheses. Most previous studies used an allogeneic Tcell-transfer model which induces BMF by donor T cells from a different genetic background.51,52 Such a model best resembles a graft-versus-host disease dynamic which does not sufficiently reproduce the disease onset and progression observed with AAA patients. By using geneticallymutant animal models, we developed an autoimmune BMF model. In this model, the Tak1 gene is spontaneously 305


J. Xin et al.

mutated in a small subset of HSPCs, which scenario might mimic the cryptic genetic lesions (clonal somatic mutations in a small subset of HSPCs) which are typically detected in AAA patients.5,10-15 Using such a model, we demonstrated that mutant HSPCs undergoing Rip3-mediated necroptosis might be the cause of autoimmune BMF. The mutant HSPCs release factors such as DNA and intracellular proteins which may serve as autoimmunogenic antigens to induce Th1-related autoimmune responses in mice. Our data suggest that necroptosis-Th1 response-Ifnγ is the axis of development of this disease (Figure 8). It might also explain the pathogenesis of some viral infection-related BMF because viral DNA products also induce necroptosis.53 Elevated levels of both IFNγ and TNFα are commonly detected in AAA patients’ blood. It has been proposed that both IFNγ and TNFα contribute to hematopoietic repression in the pathogenesis of AAA.17-19,28,29 The elevated serum levels of both IFNγ and TNFα are consistent with the occurrence of BMF in our animal model (Online Supplementary Figure S15). However, we determined that these two inflammatory cytokines function in opposite ways in the pathogenesis of AAA. Ifnγ is the key effector of the autoimmune response which represses BM hematopoiesis,20,45 whereas Tnfα is a factor which restricts autoimmune responsiveness and inhibits the progression of AAA. This observation is contrary to the currently accepted dogma for the role of TNFα in the pathogenesis of autoimmune diseases.28,29 Based on the data we have presented in the current studies, we suggest that anti-TNF treatment should be avoided in treating patients with autoimmune-related BMF such as AAA. In fact, TNF blockade-related autoimmune BMF has been reported consistently.54,55 Many potential mechanisms could explain how Tnfα restricts the development of BMF. After experimentally ruling out other mechanisms, we concluded that Tnf signaling restricts autoimmune reactions by: 1) repressing the Th1-stimulating activity of APC by suppressing IL-12 expression; and 2) promoting the T-cell-repressing functions of MDSCs.56 Treg cell changes in our model are more likely to be a result rather than the cause of abnormal immune responsiveness because such changes are not always correlated to disease onset, progression and severity. The clonal evolution of mutant HSPCs and associated subsequent leukemic transformation is one of the critical problems for AAA patients. Inhibition of apoptosis is

References 1. Young NS, Scheinberg P, Calado RT. Aplastic anemia. Curr Opin Hematol. 2008 (3):162-168. 2. Brodsky RA, Jones RJ. Aplastic anaemia. Lancet. 2005;365(9471):1647-1656. 3. Alter BP. Diagnosis, genetics, and management of inherited bone marrow failure syndromes. Hematology Am Soc Hematol Educ Program. 2007:29-39. 4. Scheinberg P, Young NS. How I treat acquired aplastic anemia. Blood. 2012; 120(6):1185-1196. 5. Young NS, Bacigalupo A, Marsh JC.

306

always correlated to increased risk of tumor development. However, the role of necroptosis in carcinogenesis and drug resistance has not been well explored. One of the concerns for the inhibition of necroptosis in BMF patients is whether it also increases the development of leukemia by promoting the clonal expansion of mutant cells. Myeloid-specific Tak–/– mice develop chronic myelomonocytic leukemia.57 We predicted that Tak1mut in Tak1mutRip3–/– mice might be more susceptible to the development of leukemia as they age. However, we did not detect an agerelated accumulation of Tak1mut HSPCs in BM of Tak1mutRip3–/– mice (Online Supplementary Figure S14). Such mice did not show any sign of hematopoietic abnormalities for up to one year of age. We found that Tak1mut HSPCs died of apoptosis when necroptosis is inhibited because Rip3-mediated necroptotic signaling and caspase 8-mediated apoptotic signaling are mutually repressed and interconvertible. In fact, in another of our studies, we found that Rip3 signaling in required for maintaining the undifferentiated state of leukemic cells by repressing calpainmediated STAT3α and SOCS1 degradation (J Xin, 2016, unpublished data). Thus, we believe that inhibition of Rip3 signaling might be a reasonable explanation for the repression of the clonal evolution of mutant HSPCs by promoting apoptosis or differentiation in BMF patients. Acknowledgments The authors thank the staff of the Department of Comparative Medicine of Loyola University Medical Center for their excellent animal care services, as well as Drs. Manuel Diaz, Nancy Zeleznik-Le, Andrew Dingwall and Wei Qiu for their ongoing professional collaboration and scientific suggestions and discussions, which improved the scientific quality of the present studies. We appreciate laboratory support from Patience Oladeinde, Danielle Howard and Emma Yao, and FACS sorting and analysis assistance from Patricia Simms, Ashley Hess, Shwetha Ravichandran and Veronica Volgina. We also appreciate Drs. Xiaoping Du and Aleksandra Stojanovic-Terpo (University of Illinois Chicago) for their assistance with CBC analysis. Funding This work was supported by US Department of Defense grant (BM120072) and NIH grants (R01HL95896 and R21CA181970) to JZ through Loyola University Chicago, and also by the Leukemia Research Foundation New Investigator Award (8th Annual George Richard Memorial Grant) to JX. JX was also supported in part by a grant from the Muscular Dystrophy Association (MDA202906).

Aplastic anemia: pathophysiology and treatment. Biol Blood Marrow Transplant. 2010;(1 Suppl):S119-125. 6. Young NS. Pathophysiologic mechanisms in acquired aplastic anemia. Hematology Am Soc Hematol Educ Program. 2006:7277. 7. Scheinberg P, Nunez O, Weinstein B, Scheinberg P, Biancotto A, Wu CO, et al. Horse versus rabbit antithymocyte globulin in acquired aplastic anemia. N Engl J Med. 2011;365(5):430-438. 8. Kordasti S, Marsh J, Al-Khan S, et al. Functional characterization of CD4+ T cells in aplastic anemia. Blood. 2011;119(9): 2033-2043.

9. Risitano AM, Maciejewski JP, Green S, et al. In-vivo dominant immune responses in aplastic anaemia: molecular tracking of putatively pathogenetic T-cell clones by TCR beta-CDR3 sequencing. Lancet. 2004; 364(9431):355-364. 10. Afable MG 2nd, Wlodarski M, Makishima H, et al. SNP array-based karyotyping: differences and similarities between aplastic anemia and hypocellular myelodysplastic syndromes. Blood. 2011;117(25):6876-6884. 11. Tiu R, Gondek L, O'Keefe C, Maciejewski JP. Clonality of the stem cell compartment during evolution of myelodysplastic syndromes and other bone marrow failure syndromes. Leukemia. 2007;21(8):1648-1657.

haematologica | 2017; 102(2)


Necroptotic cells initiate autoimmune bone marrow failure

12. Bagby GC, Fleischman AG. The stem cell fitness landscape and pathways of molecular leukemogenesis. Front Biosci (Schol Ed). 2011;3:487-500. 13. Li Y, Li X, Ge M, et al. Long-term follow-up of clonal evolutions in 802 aplastic anemia patients: a single-center experience. Ann Hematol. 2011;90(5):529-537. 14. Yoshizato T, Dumitriu B, Hosokawa K, et al. Somatic Mutations and Clonal Hematopoiesis in Aplastic Anemia. N Engl J Med. 2015;373(1):35-47. 15. Babushok DV, Olson TS, Bessler M. Somatic Mutations and Clonal Hematopoiesis in Aplastic Anemia. N Engl J Med. 2015;373(17):1673. 16. Kulasekararaj AG, Jiang J, Smith AE, et al. Somatic mutations identify a subgroup of aplastic anemia patients who progress to myelodysplastic syndrome. Blood. 2014;124(17):2698-2704. 17. Sloand E, Kim S, Maciejewski JP, et al. Intracellular interferon-gamma in circulating and marrow T cells detected by flow cytometry and the response to immunosuppressive therapy in patients with aplastic anemia. Blood. 2002;100(4):1185-1191. 18. Dufour C, Corcione A, Svahn J, et al. TNFalpha and IFN-gamma are overexpressed in the bone marrow of Fanconi anemia patients and TNF-alpha suppresses erythropoiesis in vitro. Blood. 2003;102(6): 20532059. 19. Dybedal I, Bryder D, Fossum A, et al. Tumor necrosis factor (TNF)-mediated activation of the p55 TNF receptor negatively regulates maintenance of cycling reconstituting human hematopoietic stem cells. Blood. 2001;98(6):1782-1791. 20. de Bruin AM, Voermans C, Nolte MA. Impact of interferon-gamma on hematopoiesis. Blood. 2014;124(16):24792486. 21. Adhikari A, Xu M, Chen ZJ. Ubiquitinmediated activation of TAK1 and IKK. Oncogene. 2007;26(22):3214-3226. 22. Mihaly SR, Ninomiya-Tsuji J, Morioka S. TAK1 control of cell death. Cell Death Differ. 2014;21(11):1667-1676. 23. Pasparakis M, Vandenabeele P. Necroptosis and its role in inflammation. Nature. 2015; 517(7534):311-320. 24. Linkermann A, Green DR. Necroptosis. N Engl J Med. 2014;370(5):455-465. 25. Xiao Y, Li H, Zhang J, et al. TNF-alpha/FasRIP-1-induced cell death signaling separates murine hematopoietic stem cells/progenitors into 2 distinct populations. Blood. 2011;118(23):6057-6067. 26. Tang M, Wei X, Guo Y, et al. TAK1 is required for the survival of hematopoietic cells and hepatocytes in mice. J Exp Med. 2008;205(7):1611-1619. 27. Zhang J, Xiao Y, Guo Y, et al. Differential requirements for c-Myc in chronic hematopoietic hyperplasia and acute

haematologica | 2017; 102(2)

28.

29.

30.

31.

32.

33.

34. 35.

36.

37.

38.

39.

40.

41.

hematopoietic malignancies in Pten-null mice. Leukemia. 2011;25(12):1857-1868. Chen Y, Zou Z, Wu Z, et al. TNF-alphainduced programmed cell death in the pathogenesis of acquired aplastic anemia. Expert Rev Hematol. 2015;8(4):515-526. Dufour C, Ferretti E, Bagnasco F, et al. Changes in cytokine profile pre- and postimmunosuppression in acquired aplastic anemia. Haematologica. 2009;94(12):17431747. Roderick JE, Hermance N, Zelic M, et al. Hematopoietic RIPK1 deficiency results in bone marrow failure caused by apoptosis and RIPK3-mediated necroptosis. Proc Natl Acad Sci USA. 2014;111(40):14436-14441. Dillon CP, Weinlich R, Rodriguez DA, et al. RIPK1 blocks early postnatal lethality mediated by caspase-8 and RIPK3. Cell. 2014;157(5):1189-1202. Silke J, Rickard JA, Gerlic M. The diverse role of RIP kinases in necroptosis and inflammation. Nat Immunol. 2015;16(7): 689-697. Nakada D, Oguro H, Levi BP, et al. Oestrogen increases haematopoietic stemcell self-renewal in females and during pregnancy. Nature. 2014;505(7484):555558. Kovats S. Estrogen receptors regulate innate immune cells and signaling pathways. Cell Immunol. 2015;294(2):63-69. Newton K, Sun X, Dixit VM. Kinase RIP3 is dispensable for normal NF-kappa Bs, signaling by the B-cell and T-cell receptors, tumor necrosis factor receptor 1, and Tolllike receptors 2 and 4. Mol Cell Biol. 2004;24(4):1464-1469. Libregts SF, Gutierrez L, de Bruin AM, et al. Chronic IFN-gamma production in mice induces anemia by reducing erythrocyte life span and inhibiting erythropoiesis through an IRF-1/PU.1 axis. Blood. 2011;118(9):2578-2588. Wan YY, Chi H, Xie M, et al. The kinase TAK1 integrates antigen and cytokine receptor signaling for T cell development, survival and function. Nat Immunol. 2006;7(8):851-858. Sato S, Sanjo H, Tsujimura T, et al. TAK1 is indispensable for development of T cells and prevention of colitis by the generation of regulatory T cells. Int Immunol. 2006;18(10):1405-1411. Liu HH, Xie M, Schneider MD, et al. Essential role of TAK1 in thymocyte development and activation. Proc Natl Acad Sci USA. 2006;103(31):11677-11682. Young NS. Current concepts in the pathophysiology and treatment of aplastic anemia. Hematology Am Soc Hematol Educ Program. 2013;2013:76-81. Baldridge MT, King KY, Goodell MA. Inflammatory signals regulate hematopoietic stem cells. Trends Immunol. 2010;32(2):57-65.

42. MacNamara KC, Jones M, Martin O, Winslow GM. Transient activation of hematopoietic stem and progenitor cells by IFNgamma during acute bacterial infection. PLoS One. 2011;6(12):e28669. 43. Baldridge MT, King KY, Boles NC, et al. Quiescent haematopoietic stem cells are activated by IFN-gamma in response to chronic infection. Nature. 2010; 465(7299):793-797. 44. Sato T, Onai N, Yoshihara H, et al. Interferon regulatory factor-2 protects quiescent hematopoietic stem cells from type I interferon-dependent exhaustion. Nat Med. 2009;15(6):696-700. 45. Lin FC, Karwan M, Saleh B, et al. IFNgamma causes aplastic anemia by altering hematopoietic stem/progenitor cell composition and disrupting lineage differentiation. Blood. 2014;124(25):3699-3708. 46. Kaczmarek A, Vandenabeele P, Krysko DV. Necroptosis: the release of damage-associated molecular patterns and its physiological relevance. Immunity. 2013;38(2):209223. 47. Han J, Zhong CQ, Zhang DW. Programmed necrosis: backup to and competitor with apoptosis in the immune system. Nat Immunol. 2011;12(12):1143-1149. 48. Zhou W, Yuan J. Necroptosis in health and diseases. Semin Cell Dev Biol. 2014;35:1423. 49. Moriwaki K, Chan FK. RIP3: a molecular switch for necrosis and inflammation. Genes Dev. 2013;27(15):1640-1649. 50. Zhang DW, Shao J, Lin J, et al. RIP3, an energy metabolism regulator that switches TNF-induced cell death from apoptosis to necrosis. Science. 2009;325(5938):332-336. 51. Bloom ML, Wolk AG, Simon-Stoos KL, et al. A mouse model of lymphocyte infusioninduced bone marrow failure. Exp Hematol. 2004;32(12):1163-1172. 52. Chen J, Desierto MJ, Feng X, Biancotto A, Young NS. Immune-mediated bone marrow failure in C57BL/6 mice. Exp Hematol. 2015;43(4):256-267. 53. Kaiser WJ, Sridharan H, Huang C, et al. Toll-like receptor 3-mediated necrosis via TRIF, RIP3, and MLKL. J Biol Chem. 2013; 288(43):31268-31279. 54. Kuruvilla J, Leitch HA, Vickars LM, et al. Aplastic anemia following administration of a tumor necrosis factor-alpha inhibitor. Eur J Haematol. 2003;71(5):396-398. 55. Kozak N, Friedman J, Schattner A. Etanercept-associated transient bone marrow aplasia: a review of the literature and pathogenetic mechanisms. Drugs R D. 2014;14(2):155-158. 56. Sade-Feldman M, Kanterman J, Ish-Shalom E, et al. Tumor necrosis factor-alpha blocks differentiation and enhances suppressive activity of immature myeloid cells during chronic inflammation. Immunity. 2013; 38(3):541-554.

307


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Myelodysplastic Syndrome

Ferrata Storti Foundation

Haematologica 2017 Volume 102(2):308-319

Immunophenotypic analysis of erythroid dysplasia in myelodysplastic syndromes. A report from the IMDSFlow working group

Theresia M. Westers,1 Eline M.P. Cremers,1 Uta Oelschlaegel,2 Ulrika Johansson,3 Peter Bettelheim,4 Sergio Matarraz,5 Alberto Orfao,5 Bijan Moshaver,6 Lisa Eidenschink Brodersen,7 Michael R. Loken,7 Denise A. Wells,7 Dolores Subirá,8 Matthew Cullen,9 Jeroen G. te Marvelde,10 Vincent H.J. van der Velden,10 Frank W.M.B. Preijers,11 Sung-Chao Chu,12 Jean Feuillard,13 Estelle Guérin,13 Katherina Psarra,14 Anna Porwit,15,16 Leonie Saft,16 Robin Ireland,17 Timothy Milne,17 Marie C. Béné,18 Birgit I. Witte,19 Matteo G. Della Porta,20 Wolfgang Kern21 and Arjan A. van de Loosdrecht,1 on behalf of the IMDSFlow Working Group

Department of Hematology, VU University Medical Center, Cancer Center Amsterdam, The Netherlands; 2Department of Internal Medicine, Universitätsklinikum “Carl-GustavCarus”, Dresden, Germany; 3Department of Haematology, University Hospitals NHS Foundation Trust, Bristol, UK; 4First Medical Department, Elisabethinen Hospital, Linz, Austria; 5Servicio Central de Citometría (NUCLEUS) and Department of Medicine, Centro de Investigación del Cáncer, Instituto de Biologia Celular y Molecular del Cáncer, (CSIC/USAL and IBSAL), Universidad de Salamanca, Spain; 6Isala Clinics, Zwolle, The Netherlands; 7HematoLogics, Inc., Seattle, WA, USA; 8Department of Hematology, Hospital Universitario de Guadalajara, Spain; 9HMDS, St. James’s University Hospital, Leeds, UK; 10Department of Immunology, Erasmus MC, University Medical Center Rotterdam, The Netherlands; 11Department of Laboratory Medicine – Laboratory for Hematology, Radboud University Medical Center, Nijmegen, The Netherlands; 12 Department of Hematology and Oncology, Buddhist Tzu Chi General Hospital, Tzu Chi University, Hualien, Taiwan; 13Laboratoire d’Hématologie, CHU Dupuytren, Limoges, France; 14Department of Immunology-Histocompatibility, Evangelismos Hospital, Athens, Greece; 15Department of Pathobiology and Laboratory Medicine, University of Toronto, University Health Network, Toronto General Hospital, ON, Canada; 16Department of Pathology, Karolinska University Hospital, Stockholm, Sweden; 17King’s College Hospital, London, UK; 18Laboratoire d'Hématologie, CHU de Nantes, France; 19Department of Epidemiology and Biostatistics, VU University Medical Center, Amsterdam, The Netherlands; 20Department of Hematology and Oncology, Fondazione IRCCS Policlinico San Matteo, and University of Pavia, Italy and 21MLL Munich Leukemia Laboratory, Germany 1

Correspondence: tm.westers@VUmc.nl

Received: April 15, 2016. Accepted: September 22, 2016. Pre-published: October 6, 2016. doi:10.3324/haematol.2016.147835

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/308

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

308

ABSTRACT

C

urrent recommendations for diagnosing myelodysplastic syndromes endorse flow cytometry as an informative tool. Most flow cytometry protocols focus on the analysis of progenitor cells and the evaluation of the maturing myelomonocytic lineage. However, one of the most frequently observed features of myelodysplastic syndromes is anemia, which may be associated with dyserythropoiesis. Therefore, analysis of changes in flow cytometry features of nucleated erythroid cells may complement current flow cytometry tools. The multicenter study within the IMDSFlow Working Group, reported herein, focused on defining flow cytometry parameters that enable discrimination of dyserythropoiesis associated with myelodysplastic syndromes from non-clonal cytopenias. Data from a learning cohort were compared between myelodysplasia and controls, and results were validated in a separate cohort. The learning cohort comprised 245 myelodysplasia cases, 290 pathological, and 142 normal controls; the validation cohort comprised 129 myelodysplasia cases, 153 pathological, and 49 normal controls. Multivariate logistic regression analysis performed in the learning cohort revealed that analysis of expression of CD36 and CD71 (expressed as coefficient of variation), in combination with CD71 fluorescence intensity and the percentage of CD117+ erythroid progenitors provided the best discrimination between myelodysplastic syndromes and non-clonal cytopenias (specificity 90%; 95% confidence interval: 84–94%). The high specificity of this marker set was confirmed in the validation cohort (92%; 95% confidence interval: 86–97%). This erythroid flow cytometry marker combination may improve the evaluation of cytopenic cases with suspected myelodysplasia, particularly when combined with flow cytometry assessment of the myelomonocytic lineage. haematologica | 2017; 102(2)


Flow cytometric analysis of erythrodysplasia in MDS

Introduction Discriminating between cytopenia due to myelodysplastic syndromes (MDS) and cytopenia due to other (non-clonal) causes can be challenging, especially when dysplasia as assessed by cytomorphology does not fulfill the diagnostic criteria of MDS according to WHO1, and when other MDS-associated features are absent (e.g., >15% ring sideroblasts (RS) and/or cytogenetic aberrations). Current recommendations for the diagnosis of MDS endorse flow cytometry (FC) as a valuable additional diagnostic tool. In this respect, it has been recommended to follow the guidelines set down by the International/European LeukemiaNet Working Group for FC in MDS (IMDSFlow).2-4 Despite the fact that FC for MDS correlates with cytomorphology, the sensitivity of current validated FC scores for diagnosing MDS requires improvement.5-8 So far, most of the designed FC scores have comprised the analysis of the (im)mature myelomonocytic lineage with a median sensitivity of 75% for identifying MDS (median specificity, 94%; see Westers et al.4). Since anemia is frequently observed in MDS, often accompanied by erythroid dysplasia, analysis of immunophenotypic changes of nucleated erythroid cells (NEC) may complement current FC analysis.9,10 Thus far, this has not been studied in great detail. Integration of results from analysis of the erythroid lineage to the primarily myelomonocytic and progenitor cell-based FC scores may improve sensitivity of FC analysis in MDS.8;11-13 Incorporating erythroid markers in FC protocols requires knowledge of normal erythroid differentiation, and of potential aberrancies and pitfalls. The characteristic morphological stages of normal erythroid differentiation are reflected by their light scatter properties and by their differential expression of CD45, CD117, CD105, CD36, CD71 and/or CD235a (Figure 1).14-17 Some of the FC aberrancies that have been reported to reflect MDSrelated dyserythropoiesis are: a) an increased number of NEC within total nucleated cells; b) an altered proportion of consecutive erythroid differentiation stages, such as an increased number of immature erythroid cells (CD117+ and/or CD105+) or, by contrast, a decrease in erythroid progenitors; c) an abnormal pattern of CD71 versus CD235a; d) a reduced expression of CD71 and/or CD36; and e) an overexpression of CD105. Most of these aberrancies are present in 70–80% of MDS cases.8,11-13,15,18-22 However, a number of features may be shared across the spectrum of non-clonal cytopenias.23-25 The multicenter study reported herein focused on defining the erythroid FC parameters that enable distinction of dyserythropoiesis associated with MDS from nonclonal cytopenias. Hereto, data from a learning cohort were compared between MDS patients and controls, and the results were validated in a separate cohort.

Methods MDS patients and controls Nineteen centers (members of the IMDSFlow group) collected FC data on the erythroid lineage in low grade MDS cases (<5% blasts) and controls. Data were acquired from bone marrow samples taken from 1008 patients and healthy controls after informed consent in accordance with the Declaration of haematologica | 2017; 102(2)

Helsinki; where required, local ethics committee approval was obtained. The learning cohort comprised 677 cases (18 centers, data collected between October 2012 and September 2013), and the validation cohort comprised 331 cases (9 centers, data collected between December 2013 and April 2014). Inclusion criterion for pathological controls was cytopenia not associated with MDS. In total, data on 374 MDS cases, 443 pathological and 191 normal controls were collected (specified in Tables 1 and 2). Information regarding age, gender, cytomorphology and cytogenetics was requested. One center with limited access to cytomorphology results only included MDS patients with typical features of MDS as the presence of more than 15% RS and/or MDS-associated cytogenetic anomalies. In 325/374 MDS cases, sub-classification according to the WHO-20081 was provided. The median contribution per center to the total study cohort was 47 cases (range 6–100); the median number of erythroid FC markers analyzed per case was 7 (range 1–9 of 10 proposed markers).

Sample preparation and antibody combinations Flow cytometric analysis in MDS requires the removal of mature, enucleated erythrocytes through the use of lysis protocols. The vast majority of centers used ammonium chloridebased solutions, either home-made or commercial (e.g., PharmLyse; BD Biosciences, San Jose, CA); two centers used FACSLyse (BD Biosciences), and one other used VersaLyse (Beckman Coulter, Miami, FL). FACSLyse contains a fixative, whereas VersaLyse is recommended for use with a fixative when the sample contains anticoagulants other than EDTA. The duration of lysis and temperature varied among centers (5–25 minutes and 4–37°C, respectively), but most lysed for 10 minutes (n=10) at room temperature (n=16). Two centers reported the use of an additional fixative in their staining protocols, both in combination with an ammonium chloride-based lysing solution. Detailed information can be found in Online Supplementary Information. Most centers used the IMDSFlow-recommended stain-lyse-wash procedure; five centers performed stain-lysewash. Antibody combinations were similar between centers, but clones and fluorochromes differed. Most centers used a backbone of CD45 and CD34 and/or CD117, and added antibodies such as CD235a, CD71, CD36, and CD105. Examples of antibody combinations and panels have been described previously.3,26,27 Nuclear dyes were not routinely included in the panels, and only one center applied the live/dead stain 7-AAD. The flow cytometers used included: FACSCalibur (BD Biosciences; n=3); FACS CANTO-II (BD Biosciences; n=10); a combination of FACSCalibur and FACS CANTO-II (both BD Biosciences; n=2); and Navios (Beckman Coulter; n=4). Panels comprised 4-, 5-, 6, 8- and/or 10-color FC; WinList7.0 (Verity Software House, Topsham, ME), Kaluza (Beckman Coulter), CellQuestPro, FACSDIVA (both BD Biosciences), and/or Infinicyt (Cytognos, Salamanca, Spain) software packages were used for data analysis.

Gating strategy and data collection The gating strategy was discussed during the IMDSFlow meeting in 2011 and re-evaluated in 2012. All participants performed FC analysis of the erythroid lineage defined as CD45dimto-negative and SSClow-to-intermediate. It is noteworthy that the initially proposed gating strategy (erythroid lineage defined by CD45 negativity) was altered to include early erythroid precursors that are within the CD45dim cell population.3,4,13 Six or more color panels enabled the inclusion of a myeloid-defining marker such as CD13 or CD33, and a more accurate separation of myeloid and erythroid progenitors. Moreover, to exclude platelets and 309


T.M. Westers et al.

platelet aggregates, a combination of scatter properties and CD36high/CD71- was suggested. The final gating strategy was distributed among all centers (detailed information in Online Supplementary Information). The following parameters were collected: the percentage of NEC within all nucleated cells; the expression pattern of CD71 versus CD235a; the percentage of CD71dimCD235a+ cells within the CD71/CD235a pattern; CD71 and CD36 expression levels; the percentage of CD117+ cells in the erythroid compartment; CD105 expression level and the percentage of CD105+ cells in the erythroid compartment. Recent knowledge, such as the finding that CD71 and CD36 expression represented as CV is statistically more significant than when represented as mean fluorescence intensity (MFI)28, led to adjustments in the initially proposed protocol and, hence, reanalysis of the list mode data files by the individual centers. Gating strategies and analyzed parameters are shown in the Online Supplementary Information with FC plots of MDS in comparison to normal subjects.

performed to determine the erythroid markers that discriminate between pathological controls and MDS; data were analyzed dichotomously. All variables that displayed a univariate difference of P<0.1 were included in a backward selection procedure based on the Likelihood Ratio score. Regression coefficients of the variables in the final model were used to define the weight of these markers in a descriptive score for dyserythropoiesis. Cut-off level of the score indicating MDS-associated erythroid aberrancies was determined based on the total weight of these variables and a specificity of at least 90%. The sensitivity and specificity of the marker combination were calculated to illustrate predictive accuracy. The data were analyzed using SPSS 20.0 (IBM Corp, Armonk, NY), and GraphPad 6.0 software (La Jolla, CA). P-values <0.05 were considered significant.

Table 2. Subcategories of MDS and pathological controls in the learning and validation cohort.

Statistical analyses Due to differences in sample processing, instrument settings, clones, and fluorochromes between centers, the expression levels of CD71, CD36, and CD105 varied. Therefore, the median expression levels of CD36, CD71, and CD105 in the subset of normal bone marrow samples were calculated for each individual center. Expression levels were then normalized against the median value for that particular marker for each center separately. Patient and control groups were compared using the Kruskal Wallis test for continuous data, and the Chi-square or the Fisher’s exact test for dichotomous data. Correlations between certain markers, and between markers and age, were analyzed using Spearman’s rank correlation coefficients. Comparing single parameters between MDS and control groups demonstrated that results overlapped (results section figure 2); hence, receiver-operator-curve (ROC) analyses did not yield applicable cut-offs. Therefore, cut-off values for aberrancies were based on the 10th and/or 90th percentile of the data of pathological controls in the learning cohort. Multivariate logistic regression analyses were

Table 1. Characteristics of MDS patients and controls in the learning and validation cohorts.

Learning cohort Normal n age male:female Pathological controls n age male:female MDS n age male:female Comparison of age normal vs. pathological controls pathological controls vs. MDS Comparison of gender

142 60 (20-86)* 1.4:1

Validation cohort 49 64 (24-86)* 2.3:1

290 62 (18-92)* 1:1

153 72 (20-98)* 1.2:1

245 72 (23-94)* 1.1:1

129 75 (40-95)* 1.6:1

P = 0.007 P < 0.001 n.s.

P = 0.193 P = 0.013 n.s.

*Data are expressed as the median and range; abbreviations: MDS: myelodysplastic syndromes; n: number; n.s.: not significant.

310

MDS subcategories RCUD RARS RARS-t RCMD del(5q) MDS-U other not specified Subcategories of pathological controls iron deficiency anemia anemia in chronic disease† vitamin B12/folic acid deficiencies anemia in auto-immune diseases† anemia due to renal failure anemia other cytopenia associated with marrow infiltration cytopenia induced by chemotherapy or medication or post-SCT ITP or neutropenia or auto immune cytopenia NOS reactive conditions or cytopenia induced by infections normal bone marrow, peripheral cytopenia other than defined subcategories inconclusive non clonal cytopenia NOS ET, PV, primary myelofibrosis PNH AA

Learning cohort

Validation cohort

23 (9.4%) 16 (6.5%) 2 (0.8%) 155 (63%) 14 (5.7%) 3 (1.2%) 32 (13%)

14 (11%) 12 (9.3%) 1 (0.8%) 75 (58%) 5 (3.9%) 3 (2.3%) 2 (1.6%)* 17 (13%)**

22 (7.6%) 42 (14.5%) 11 (3.8%) 13 (4.5%) 6 (2.1%) 9 (3.1%) 20 (6.9%)

13 (8.5%) 8 (5.2%) 11 (7.2%) 7 (4.6%) 5 (3.3%) 9 (5.9%) 7 (4.6%)

27 (9.3%)

8 (5.2%)

30 (10.3%)

14 (9.2%)

32 (11%) 6 (2.1%) 29 (10%) 11 (3.8%) 25 (8.6%) 5 (1.7%) 1 (0.3%) 1 (0.3%)

10 (6.5%) 0 23 (15%) 8 (5.2%) 19 (12,4%) 7 (4.6%) 0 4 (2.6%)

Values between brackets represent the relative distribution within MDS or pathological control subgroups; *other concerns one case of hypoplastic MDS and one case of MDS with fibrosis; **not specified, but the presence of MDS-associated cytogenetic aberrations indicated; AA: aplastic anemia, ET: essential thrombocythemia; MDS: myelodysplastic syndromes, MDS-U: MDS-unclassifiable; NOS: not otherwise specified, PNH: paroxysmal nocturnal hemoglobinuria, PV: polycythemia vera; RARS: refractory anemia with ring sideroblasts; RARS-t: RARS and thrombocytosis; RCMD: refractory cytopenia with multilineage dysplasia, RCUD: refractory cytopenia with unilineage dysplasia, del(5q): MDS with isolated del(5q). Comparison of the distribution of MDS subsets among the learning and validation cohorts did not differ (c2 test; P=0.511); the distribution of subsets of pathological controls did (P<0.001).The subcategories† ”anemia in chronic disease” comprises iron incorporation disorders, bowel diseases, diabetes, etc., “anemia in auto-immune diseases” comprises AIHA, AITP, Rheumatoid Arthritis, SLE, etc.; “anemia other” comprises, among others, cases of normocytic anemia, anemia unexplained.

haematologica | 2017; 102(2)


Flow cytometric analysis of erythrodysplasia in MDS

A

B

C

D

F

E

G

Figure 1. Flow cytometric profiles of normal erythroid differentiation. Early erythroid precursors are defined as CD45dim/SSCint/CD34+/CD117+/CD105+/CD235a-, basophilic erythroblasts as CD45dim/–/ proerythroblasts as CD45dim/–/SSCint/CD34–/CD117+/CD105+/CD36+/CD71+/CD235a+, SSCint/CD117–/CD105+/CD36++/CD71+/CD235a+, polychromatic erythroblasts as CD45–/SSClow/FSCint/CD105–/CD36+/CD71+/CD235a+ and orthochromatic erythroblasts CD45–/ SSClow/FSClow/CD36+/-/CD71+/CD235a+. Indicated colors reflecting erythroid subsets are not visible in the CD71 vs. CD235a plot (Fig 1F). Herein, pink colored cells represent the total erythroid lineage in this plot. Mature erythrocytes (CD45–/CD36–/CD71–/CD235a++) can be seen in improperly lysed cell preparations (Figure 1F). Reticulocytes are not covered in these graphs, but they may appear as CD71dim-to-negative in non-lysed cell preparations. Myeloid progenitors are CD34+/CD117+/HLA-DR+/CD105– (Figure 1C and D.); these cells have slightly higher CD45 expression than erythroid precursors; moreover, in contrast to myeloid progenitors erythroid cells do not express HLA-DR (adapted from references 14-16).

Results Flow cytometric analysis of the erythroid lineage in normal bone marrow samples: comparison of results from participating centers Discrepancies in erythroid analysis between centers (and samples) can occur at several levels: a) sample quality (e.g., hemodilution); b) sample preparation (e.g., lysing procedure); c) data acquisition (e.g., acquisition rate and threshold of forward light scatter); and d) degree of adherence to the proposed gating strategy. Therefore, we first compared the FC results for normal bone marrow samples (learning cohort) between centers in terms of each defined marker. The percentage of NEC was highly diverse among centers; yet, it seemed to be independent of the lysing method applied (Online Supplementary Figure S1). Similarly, the percentage of CD71dim cells differed largely between centers. haematologica | 2017; 102(2)

Furthermore, two centers reported higher percentages of CD117+ erythroid progenitors (up to 50% within the NEC) than the other centers (<15%). Results for one center could be explained by their stringent lysing procedures (i.e., 15 minutes at 37°C) which removed more mature (orthochromatic and polychromatic) erythroblasts resulting in a relative increase in early progenitors (data not shown). To circumvent the issues regarding differences in percentages of erythroid (sub)populations between centers, the percentages of NEC, CD117+ and CD105+ erythroid progenitors were also normalized as described for antigen expression levels (see Material and Methods statistics section); these are further referred to as relative percentages, i.e., relative to the median percentage in normal bone marrow samples (Online Supplementary Figure S2). CD71dim cells were rarely seen in normal controls; therefore, results for the percentage CD71dim could not be normalized. 311


T.M. Westers et al.

Erythroid aberrancies that may discriminate between MDS and pathological controls Next, all FC-erythroid parameters in the learning cohort were compared between MDS and controls. The results from the learning cohort are summarized in Figure 2 (P-values in Online Supplementary Table S1). The relative percentage of NEC within the total nucleated cell population was significantly higher in MDS than in the pathological and normal controls (P<0.001). Similarly, the CD71CD235a differentiation pattern was more frequently considered aberrant in MDS (65%, 109/167 cases) than in pathological and normal controls (18% (44/254) and 3.7% (5/134) of cases, respectively (P<0.001)). To objectify the evaluation of this pattern, we analyzed its components separately. CD71 expression was analyzed in terms of MFI, CV, and the presence of a subpopulation with reduced

CD71 expression (CD71dim). The relative CD71 MFI was significantly reduced in MDS, whereas the relative CV for CD71 and the percentage of CD71dim cells were significantly higher than those in both control groups (P<0.001); no significant differences were observed between the pathological controls and normal bone marrow samples. The expression of CD235a largely depends on the success of removing mature erythrocytes from a sample. Moreover, membrane fragments of lysed erythrocytes may stick to other cells in the analysis sample, mimicking positivity. Hence, this parameter was considered too unreliable for evaluation when considered individually. The percentage of immature erythroid progenitors can also affect the appearance of the differentiation pattern. Analysis of the relative percentage of CD117+ (and CD105+) erythroid progenitor cells revealed a broader

Figure 2. Distribution of erythroid markers analyzed by flow cytometry among MDS patients and controls within the learning cohort. Results of the analysis of indicated markers of the erythroid lineage are plotted along the X-axes: relative (rel.) percentages of nucleated erythroid cells (NEC); rel. mean fluorescence intensity (MFI) for CD36, CD71, and CD105; rel. coefficient of variation (CV) of CD36 and CD71; and rel. percentages of CD117+ and CD105+ erythroid progenitors. Relative frequencies (as percentage of the MDS or control cohort for a particular marker) are depicted along the Y-axes. Dotted lines represent results for normal bone marrow (NBM) samples, dashed lines pathological controls (PC) and solid lines MDS cases. P-values of comparison between groups are depicted: **: <0.001, *: <0.05, ns: not significant (Kruskal Wallis test). Grey boxes indicate reference ranges for the analyzed markers as defined by 10th and 90th percentiles of pathological controls. Scatterplots of results for the markers (depicted here as frequency histograms) that were selected as FC-markers for erythroid dysplasia from the multivariate analysis are depicted in Online Supplementary Figure S3.

312

haematologica | 2017; 102(2)


Flow cytometric analysis of erythrodysplasia in MDS

range of these cells in MDS, although not significantly different from the control groups. Relative expression of CD105 was either increased or decreased in MDS. Nonetheless, CD105 expression did not discriminate between MDS and pathological controls. Similar to that for CD71, the relative MFI of CD36 was significantly lower and the relative CV for CD36 was significantly higher in MDS than in the control groups. To summarize, the markers that showed a significantly different distribution in MDS as compared to controls were: the relative percentage of NEC and the percentage of CD71dim cells (increased in MDS); the relative MFI of CD71 and CD36 (decreased in MDS); and the relative CV for CD71 and CD36 (increased in MDS); Figure 2.

Selection of a combination of erythroid FC aberrancies that distinguish MDS from pathological controls To be applicable in FC analysis of a single patient in daily practice, cut-offs for the identification of MDS-associated changes of all potential aberrancies were defined (10th and 90th percentiles of the pathological controls, Online Supplementary Table S2A) and compared between MDS and controls (Table 3). All parameters differed between groups at a P-value of <0.1, and thus could have been considered for the multivariate logistic regression analysis. However, due to large differences between centers, regarding the percentages of NEC and CD71dim cells (likely due to technical variation as shown for normal controls), these parameters were not entered in the multivariate analysis. Besides, irrespective of the finding that data for CD105 significantly discriminated between subgroups (Table 3), this marker was not included. Entering CD105 data would have reduced the power of the (multicenter) analysis, since data on CD105 were only available in a limited number of centers (5/18) and cases. Multivariate logistic regression analysis was based on 119 MDS cases and 153 pathological controls that had available data for all parameters entered in the test. CD36 CV was identified as the best discriminator between MDS and pathological controls in combination with the CV of CD71, the MFI of CD71 and the percentage of CD117+ erythroid cells (Table

4). These four markers were used to define a FC-erythroid dysplasia score in which aberrancies were considered in a weighted manner: four points for increase in CD36 CV; three points for increase in CD71 CV; two points for decreased CD71 MFI; and two points in the case of decreased or increased percentage of CD117+ erythroid cells (reference ranges are summarized in Table 5). A cutoff of ≥5 points resulted in the identification of MDS-associated erythroid aberrancies by FC at a specificity of 90% (95% CI: 84–94%) and a sensitivity of 33% (95% CI: 24– 42). Results for the selected markers and the application of the FC-erythroid dysplasia score in the learning cohort are displayed in Table 6 and Figure 3A. In daily practice, a numerical way of counting aberrancies would be more convenient. This involves the definition of a new cut-off; i.e., ≥2 aberrant markers (Figure 3C). Note that, the exception to this numerical score is that the combination of aberrancies in CD71 MFI and percentage of CD117+ alone is not sufficient to conclude dyserythropoiesis by FC (<5 points in the weighted score). The latter was seen in only one pathological control and three MDS cases.

Correlation between erythroid markers and age The incidence of MDS increases with age; hence, erythroid markers that are significantly correlated with age may be less suitable for discriminating between MDS and controls. Since we observed significant differences between the groups regarding age (Table 1); correlations between FC results for erythroid markers and age were evaluated for normal bone marrow samples. Only CD105 MFI, a variable that was not included in the multivariate analysis, demonstrated a moderate-to-good inverse correlation with age, i.e., CD105 expression decreased with increasing age (Spearman’s Rho -0.55, P<0.001, n=47, Online Supplementary Table S4 and Online Supplementary Figure S4).

Validation of FC aberrancies in the erythroid lineage in MDS and pathological controls The value of the defined antigenic combinations was tested in an independent cohort. Nine centers provided

Table 3. Aberrancies in FC markers of the erythroid lineage between MDS and controls within the learning cohort.

rel. %NEC pattern CD71 vs. CD235a %CD71dim rel. MFI of CD71 rel.CV of CD71 rel. MFI of CD36 rel. CV of CD36 rel. %CD117 progenitors rel. %CD105 progenitors rel. MFI of CD105

NBM

PC

MDS

P MDS vs. PC

P MDS vs. NBM

P PC vs. NBM

2.9 3.7 0.8 4.0 4.5 0.8 0.0 8.2 6.7 29.8

10.1 17.3 10.0 10.0 10.3 10.3 10.2 19.4 20.8 22.5

32.1 64.9 31.5 27.8 45.5 25.8 30.1 33.8 48.4 58.7

<0.001 <0.001 <0.001 <0.001 <0.001 0.001 <0.001 0.005 0.001 <0.001

<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.004

0.013 <0.001 <0.001 0.045 0.152 <0.001 0.001 0.008 0.092 0.395

After applying cut-offs as defined in the set of pathological controls, the results were expressed as ‘0’ and ‘1’ for within and beyond reference range(s), respectively (ranges as displayed in Online Supplementary Table S2A). Percentages of subjects with aberrancy are displayed for normal bone marrow (NBM), pathological controls (PC) and MDS cases. Results were compared among subgroups using Fisher’s Exact test; P-values are depicted (P). CV: coefficient of variation; dim: diminished; MFI: mean fluorescence intensity; NEC: nucleated erythroid cells; P: P-value; rel.: relative.

haematologica | 2017; 102(2)

313


T.M. Westers et al.

data for this validation cohort, the results are depicted in Figure 4 and Online Supplementary Table S1. Similar to results in the learning cohort, the relative CVs of CD36 and CD71 were significantly increased in MDS as compared to controls, whereas CD36 MFI was significantly decreased. Since the distribution of subcategories was similar in the MDS learning and validation cohorts, we compared FC results between the two MDS cohorts. This

revealed that the increase in CD71 CV and the decrease in CD71 MFI were significantly less evident in the MDS validation cohort than in the learning cohort (t-test, P<0.001). Results for CD36 CV and the percentage of CD117+ erythroid cells did not differ between both MDS cohorts (P=0.134 and 0.116, respectively). Reference ranges, as defined in the learning cohort, were applied to evaluate the data from the validation cohort,

Table 4. Results of multivariate logistic regression analysis in learning cohort.

Parameter rel. CD36 CV rel. CD71 CV rel. CD71 MFI rel. %CD117

odds ratio

95% CI

P

3.7 3.2 2.2 1.7

1.6 – 8.5 1.6 – 6.4 1.1 – 4.5 0.92 – 3.2

0.003 0.001 0.033 0.084

Markers entered in the analysis were relative CD36 MFI, CD36 CV, CD71 MFI and CD71 CV, and the relative percentage of CD117+ erythroid cells (%CD117). 272/535 cases were available for analysis of which 153 pathological controls and 119 MDS cases in the learning cohort; P<0.001). CI: confidence interval; CV: coefficient of variation; MFI: mean fluorescence intensity; P: P-value; rel.: relative.

A

C

314

B

D

Figure 3. FC-erythroid dysplasia score in learning and validation cohorts. The weighted score consists of four parameters: increase in CD36 CV (4 points) and CD71 CV (3 points); decrease in CD71 MFI (2 points); and decrease or increase of CD117+ erythroid progenitors (2 points). A maximum score of 11 points can be reached. Data are grouped as normal bone marrow (NBM), pathological controls (PC) and MDS, relative distribution of the results for the score is displayed along the Y-axes. Panel A. represents the learning cohort consisting of 79 normal bone marrow samples (NBM), 153 pathological controls (PC) and 119 MDS cases. The FC-erythroid dysplasia score could only be calculated in the cases with data on all four defined parameters (351/670 cases). Panel B. represents the results in the validation cohort consisting of 42 NBM samples, 106 pathological controls and 93 MDS cases (241/320 cases). Clonal disorders as aplastic anemia and those within the category of essential thrombocythemia, polycythemia vera and primary myelofibrosis were excluded from both cohorts (two and nine cases for learning and validation cohorts, respectively). A cut-off of ≥5 points resulted in a specificity of 90% (95% CI: 84–94%) and a sensitivity of 33% (95% CI: 24–42) in the learning cohort; in the validation cohort, specificity was 92% (95% CI: 86–97%) and sensitivity 24% (95% CI: 15–34%). The numerical score, depicted in panels C and D, consists of four parameters: increase in CD36 CV and CD71 CV ; decrease in CD71 MFI; and decrease or increase of CD117+ erythroid progenitors. A maximum score of 4 points can be reached. A cut-off of ≥2 points resulted in a specificity of 90% (95% CI: 84–94%) and a sensitivity of 35% (95% CI: 27–45) in the learning cohort; in the validation cohort, specificity and sensitivity were 92% (95% CI: 86– 97%) and 25% (95% CI: 16–35%), respectively.

haematologica | 2017; 102(2)


Flow cytometric analysis of erythrodysplasia in MDS

and then to calculate the weighted FC-erythroid dysplasia score. This resulted in a specificity of 92% (95% CI: 86– 97%) and a sensitivity of 24% (95% CI: 15–34%) for identifying MDS-associated erythroid aberrancies by FC (Figure 3B and Online Supplementary Table S5). In most

cases, the numerical score could have been applied (cut-off ≥2 aberrancies; Figure 3D) with the same result regarding presence of erythroid dysplasia; only one MDS had decreased CD71 MFI in combination with an altered CD117+ percentage.

Table 5. Reference ranges of FC parameters incorporated in the FC-erythroid dysplasia score (learning cohort).

relative CV of CD36 relative CV of CD71 relative MFI of CD71 relative %CD117+ erythroid cells

Reference Ranges

# of PC cases*

# of NBM cases†

<145% <133% >46% 37–212%

175 177 250 182

92 86 126 122

Reference ranges were determined in the learning cohort. Reference ranges represent values relative to median values for the analyzed markers in the erythroid compartment of normal bone marrow (NBM) subjects (learning cohort). These values represent 10th and/or 90th percentiles as determined in the set of pathological controls (PC) within the learning cohort. *number of PC cases that were available to calculate cut-off values (10th and 90th percentiles); †number of NBM cases that were available to calculate median values. CV: coefficient of variation; MFI: mean fluorescence intensity.

Figure 4. Distribution of erythroid markers analyzed by FC within subgroups of MDS patients and controls within the validation cohort. Results of the analysis of selected markers of the erythroid lineage in the validation cohort are plotted along the X-axes: relative coefficient of variation (CV) of CD36 and CD71, and relative percentages of CD117+ erythroid progenitors. Normalization was performed against results for the normal bone marrow (NBM) samples of the validation cohort per each individual center. Relative frequencies are depicted along the Y-axes. Dotted lines represent results for NBM samples, dashed lines pathological controls (PC), and solid lines MDS cases. Grey boxes indicate 10th and/or 90th percentiles of pathological controls defined in the learning cohort that were applied for evaluating aberrancies.

haematologica | 2017; 102(2)

315


T.M. Westers et al.

Discussion Analysis of erythroid dysplasia is rarely included in current FC protocols for MDS, since the significance of FC data from the erythroid lineage is, to a large extent, still under debate.4 Here, we reported the results of a multicenter study within the IMDSFlow group, which focused on defining erythroid parameters that enable discrimination of dyserythropoiesis associated with MDS from non-clonal cytopenia. The majority of erythroid FC markers that are recommended for evaluation of dysplasia according to ELNet guidelines4 were significantly different between MDS and controls. Analysis of the presence of aberrancies in the erythroid markers CD71 and CD36 (expressed as the CV), together with the MFI of CD71 and an abnormal percentage of CD117+ erythroid progenitor cells, provided the best discrimination between MDS and non-clonal cytopenia. A weighted score based on these four parameters yielded a specificity of 90% (95% CI: 84–94%) in the learning cohort and 92% (95% CI: 86–97%) in the validation cohort. Sensitivity of the weighted score was 33% (95% CI: 24–42%) and 24% (95% CI: 15–34%) in the

learning and validation cohorts, respectively. The latter lower sensitivity could be explained by a less evidently increased CD71 CV and decreased CD71 MFI in the MDS validation cohort compared to the learning cohort. Hence, fewer MDS cases scored CD71 CV and/or CD71 MFI as aberrant (Online Supplementary Table S3). Notably, these scores only reflect the presence of FC-erythroid dysplasia, not the likelihood of an MDS diagnosis. Tenth and 90th percentiles in the validation cohort’s control cases slightly differed from the learning cohort (Online Supplementary Table S2B). Yet, application of these cut-offs in the validation cohort resulted in comparable specificity: 91% (95% CI: 84–96%; sensitivity: 27% (95% CI: 18– 37%). In general, cut-off values are most reliable when defined by standardized analyses of control samples in a single center. Notably, no consensus has been reached as to whether percentiles, standard deviations or log differences should be applied as cut-offs for any of the MDSassociated aberrancies. Specificity of the defined markers for identification of MDS-associated erythroid changes is considered to be more important than their general diagnostic value for

Table 6. Results of FC aberrancies in the erythroid lineage and the FC-erythroid dysplasia score among MDS cases and controls within the learning cohort.

Increased CV Increased CV Decreased De/increased FC-erythroid # of cases of CD36 of CD71 relative % of CD117+ dysplasia score in flow MFI CD71 erythroid ≥5 score* progenitors Categories normal controls pathological controls MDS MDS subcategories RCUD RARS(-t) RCMD del(5q) MDS NOS Pathological control subcategories iron deficiency anemia anemia in chronic disease† vitamin B12/folic acid deficiencies anemia in auto-immune diseases† anemia due to renal failure anemia other† cytopenia associated with marrow infiltration cytopenia induced by chemotherapy or medication or post-SCT ITP or neutropenia or auto immune cytopenia NOS reactive conditions or cytopenia induced by infections normal bone marrow (peripheral cytopenia NOS) other than defined subcategories NOS inconclusive

0 10 30

5 10 46

4 10 28

8 19 34

3 10 33

2/79 15/153 39/119

25 60 28 10 -

31 64 44 30 80

26 33 25 38 40

41 31 32 63 -

33 57 30 13 -

5/15 8/14 24/79 1/8 -

11 7 0 0 25 0 0 10 17

0 10 17 0 0 22 20 0 9

0 3 0 20 0 11 16 24 12

11 23 25 0 60 0 40 9 9

6 10 0 0 0 0 14

1/18 3/29 0/6 0/5 0/5 0/7 3/21

19

39

24

19

28

5/18

0

20

33

67

20

1/5

5 0

0 0

4 0

16 18

0 0

0/20 0/11

Displayed numbers correspond to the percentage of cases per subgroup that were beyond the reference ranges (Table 5). The coefficient of variation (CV) for CD71 and CD36 were tested against the 90th percentile; the expression level (mean fluorescence intensity: MFI) of CD71 was against the 10th percentile; and the percentage of CD117 erythroid progenitor cells was tested against both the 10th and 90th percentiles. Only data of subsets with five or more cases are depicted. Diagonally marked cells represent data not available or reliable (i.e., missing or only small data sets (<5 cases)) Note: *number of cases with a FC-erythroid dysplasia score of ≥2 per total number of cases in which all parameters were available that comprise the score; the subcategories† ”anemia in chronic disease” comprises iron incorporation disorders, bowel diseases, diabetes, etc., “anemia in auto-immune diseases” comprises AIHA, AITP, Rheumatoid Arthritis, SLE, etc. and “anemia other” comprises, among others, cases of normocytic anemia, anemia unexplained; NOS: not otherwise specified.

316

haematologica | 2017; 102(2)


Flow cytometric analysis of erythrodysplasia in MDS

MDS. The specificity may be optimized by increasing the cut-off from ≥5 to ≥6 points (specificity 96% in both cohorts), at the cost, however, of a decrease in sensitivity (24% and 14%, in the learning and validation cohorts, respectively). To simplify interpretation of results from erythroid analysis, a numerical way of counting aberrancies was tested; a cut-off of ≥2 aberrant markers led to comparable specificity and sensitivity as for the weighted score. However, it must be taken into account that the sole combination of CD71 MFI and percentage of CD117+ erythroid progenitors is not sufficient to indicate MDS-associated changes in the erythroid lineage. Cremers et al. compared the analysis of the set of four FC-parameters to erythroid dysplasia as assessed by cytomorphology.29 They demonstrated that FC correlated well with cytomorphology, albeit at a lower sensitivity (low/int-1 risk MDS, 64% vs. 84%, respectively); controls showed 11% and 10% of dysplasia by FC and cytomorphology, respectively. The findings presented herein confirmed results from a recent study28 that reported a significant increase in CD71 CV and CD36 CV to be highly suggestive for MDS. Yet, discrimination between MDS and controls based on CV values was less clear in the current dataset. Mathis et al.28 stated that the difference in CD71 CV between MDS and controls was less pronounced after erythrocyte lysis; however, this was not the case for CD36 CV. Since all data in the present study were obtained after erythrocyte lysis, it might explain the observed differences. It may seem paradoxical to use erythrocyte lysing procedures when the focus is on analysis of the erythroid lineage; but lyse-stainwash is the recommended protocol for processing samples for FC in MDS.3 Despite IMDSFlow recommendations, methodological variation between centers may have led to differences in results as demonstrated in normal controls. Harmonization, or even standardization, of methods may narrow differences and improve validity of conclusions from multicenter studies, as has been demonstrated within the Euroflow consortium.30 Hence, grouping of data per technical procedure could have been informative from a practical perspective; yet, the power of the analyses within and between numerous subgroups of centers would have been strongly limited by sample sizes. Notably, in daily practice, FC results in subjects suspected of MDS should preferably be compared with a center’s own cohort of control samples. Despite these technical considerations, our data confirm the robustness of the evaluation of an increase in CD71 and CD36 CV on erythroid cells.28 Another discriminatory marker was the percentage of erythroid progenitors defined as CD117+. A potential marker for future inclusion in erythroid data analysis by FC is CD105. It has been demonstrated (in normal and pathological controls) that CD105 is lost before carbonic anhydrase is expressed, which suggests that the majority of CD105+ erythroid progenitor cells are not subject to ammonium chloride-based lysing protocols.16 This confirms the robustness of the percentage of erythroid progenitors (CD117+ and/or CD105+) as a marker for erythroid dysplasia. Notably, hemodilution impacts the analysis of the erythroid compartment as it may result in a lack or paucity of erythroid progenitors, similar to what is seen in the myelomonocytic compartments. In heavily hemodiluted samples, the erythroid lineage should be considered non-evaluable. Application of CD105 may overcome the potential error haematologica | 2017; 102(2)

of assigning CD117+ myeloid progenitors as erythroid progenitors, especially when combined with CD117 and an additional myeloid marker such as CD33. Furthermore, CD105 overexpression was confirmed in some cases of MDS in our dataset.18,20 However, we also observed a decreased expression in MDS. Notably, CD105 expression was negatively correlated with age in normal controls. Future studies in larger data sets may elucidate whether CD105 is truly valuable in analysis of erythroid dysplasia in MDS. New insights may improve the impact of FC in the diagnosis of MDS. A recent report showed that increased expression of CD44 on all maturational stages of erythroid cells was associated with MDS, irrespective of presence or absence of morphologic dyserythropoiesis.31 In addition, decreased expression of the major coxsackie-adenovirus receptor (CAR) was demonstrated in dysplastic CD105+ erythroid progenitors.32,33 The diagnostic value of the herein presented parameter combination is limited. Yet, ultimately, the analysis of the erythroid and myelomonocytic lineages and hematogones should be combined. Further validation should reveal the power of the herein defined erythroid markers. Results from a prospective clinical trial in low/int-1 risk MDS demonstrated that the addition of proposed erythroid FC-markers to the more widely acknowledged analysis of the myelomonocytic lineage increased the sensitivity of MDS-FC from 68 to 80% (specificity only slightly decreased from 98% to 95%).29 In addition, it would be relevant to elucidate the value of the combination of myeloid and erythroid FC markers in indeterminate cases according to cytomorphology. MDSFC of the myelomonocytic lineage has shown to be of negative predictive value in these cases.34,35 In view of emerging knowledge from next generation sequencing analysis, future research may also concern the analysis of the relation between gene modifications/mutations and FC findings. Data comparing cytogenetic aberrations and FC have already demonstrated distinct profiles.36,37 Parallel mutational data in the current cohort are not available, but it would be of interest to compare, for instance, the immunophenotypic profile of nucleated erythroid cells in relation to the presence of a SF3B1 mutation since there is a relation between this mutation and the occurrence of ring sideroblasts.38 Cytomorphology reports dysplastic features in erythropoiesis in non-MDS cases, such as reactive conditions.23-25 Moreover, patients with cytopenia due to marrow infiltration may demonstrate FC aberrancies associated with MDS. MDS may even coincide with the other malignancy in these patients.39,40 A subset of patients with reactive marrows or marrow infiltration in our dataset indeed showed multiple erythroid aberrancies (5/23 and 2/9, respectively; Online Supplementary Table S5). Follow-up analysis after several months may exclude or confirm MDS in these cases.34,35 This stresses that FC analysis in suspected MDS, though proven very specific, should always be part of an integrated diagnostic approach rather than a solitary diagnostic tool.41 In line with this, FC may attribute to the diagnostic work-up in cases that show clonal hematopoiesis of indeterminate potential (CHIP), particularly when patients present with cytopenia and have indeterminate cytomorphology and/or non-informative cytogenetics. To summarize, we identified significant aberrancies with respect to the FC markers recommended by 317


T.M. Westers et al.

IMDSFlow for analysis of the erythroid lineage in MDS. The best indicators of dysplastic changes associated with MDS were an increased CV of CD36 and CD71, a decreased MFI of CD71 in combination with decreased or increased percentages of erythroid progenitors (CD117+). Application of the defined marker set demonstrated high specificity. Future studies should assess the contribution of the selected erythroid markers to the evaluation of the myeloid progenitors, the maturing myelomonocytic lineage, and hematogones in current FC protocols in MDS. This will be implemented in an upcoming multicenter data collection exercise within IMDSFlow.

References 11. 1. Brunning R, Orazi A, Germing U, et al. Myelodysplastic syndromes/neoplasms. In: Swerdlow et al., ed. WHO classification of Tumours and Haematopoietic and Lymphoid Tissues. Lyon: IARC; 2008 2. Malcovati L, Hellstrom-Lindberg E, Bowen D, et al. Diagnosis and treatment of primary myelodysplastic syndromes in adults: recommendations from the European LeukemiaNet. Blood. 2013;122(17):29432964. 3. van de Loosdrecht AA, Alhan C, Bene MC, et al. Standardization of flow cytometry in myelodysplastic syndromes: report from the first European LeukemiaNet working conference on flow cytometry in myelodysplastic syndromes. Haematologica. 2009;94(8):1124-1134. 4. Westers TM, Ireland R, Kern W, et al. Standardization of flow cytometry in myelodysplastic syndromes: a report from an international consortium and the European LeukemiaNet Working Group. Leukemia. 2012;26(7):1730-1741. 5. Wells DA, Benesch M, Loken MR, et al. Myeloid and monocytic dyspoiesis as determined by flow cytometric scoring in myelodysplastic syndrome correlates with the IPSS and with outcome after hematopoietic stem cell transplantation. Blood. 2003;102(1):394-403. 6. van de Loosdrecht AA, Westers TM, Westra AH, et al. Identification of distinct prognostic subgroups in low- and intermediate-1-risk myelodysplastic syndromes by flow cytometry. Blood. 2008; 111(3):1067-1077. 7. Chu SC, Wang TF, Li CC, et al. Flow cytometric scoring system as a diagnostic and prognostic tool in myelodysplastic syndromes. Leuk Res. 2011;35(7):868-873. 8. Kern W, Haferlach C, Schnittger S, Haferlach T. Clinical utility of multiparameter flow cytometry in the diagnosis of 1013 patients with suspected myelodysplastic syndrome: correlation to cytomorphology, cytogenetics, and clinical data. Cancer. 2010;116(19):4549-4563. 9. Maassen A, Strupp C, Giagounidis A, et al. Validation and proposals for a refinement of the WHO 2008 classification of myelodysplastic syndromes without excess of blasts. Leuk Res. 2013;37(1):64-70. 10. Germing U, Strupp C, Giagounidis A, et al. Evaluation of dysplasia through detailed cytomorphology in 3156 patients from the Dusseldorf Registry on

318

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

Acknowledgments The authors would like to thank Claudia Cali and Kelly Schouten (VU University Medical Center Amsterdam, The Netherlands), Jeroen Lauf (Elisabethinen Hospital, Linz, Austria), Frauke Bellos (MLL Munich Leukemia Laboratory, Munich, Germany), Hans Veenstra (Radboud University Medical Center, Nijmegen, The Netherlands), Rik Brooimans and Andre Bijkerk (Erasmus Medical Center, Rotterdam, The Netherlands) for their contribution to data collection, although they did not take part in the IMDSFlow working group conferences. Contributors from The Netherlands are all members of the â&#x20AC;&#x153;Flow Cytometry in MDSâ&#x20AC;? working group within the Dutch Society of Cytometry.

myelodysplastic syndromes. Leuk Res. 2012;36(6):727-734. Stetler-Stevenson M, Arthur DC, Jabbour N, et al. Diagnostic utility of flow cytometric immunophenotyping in myelodysplastic syndrome. Blood. 2001;98(4):979-987. Malcovati L, Della Porta MG, Lunghi M, et al. Flow cytometry evaluation of erythroid and myeloid dysplasia in patients with myelodysplastic syndrome. Leukemia. 2005;19(5):776-783. Matarraz S, Lopez A, Barrena S, et al. Bone marrow cells from myelodysplastic syndromes show altered immunophenotypic profiles that may contribute to the diagnosis and prognostic stratification of the disease: a pilot study on A series of 56 patients. Cytometry B Clin Cytom. 2010; 78(3):154-168. Machherndl-Spandl S, Suessner S, Danzer M, et al. Molecular pathways of early CD105-positive erythroid cells as compared with CD34-positive common precursor cells by flow cytometric cell-sorting and gene expression profiling. Blood Cancer J. 2013;3:e100. Fajtova M, Kovarikova A, Svec P, Kankuri E, Sedlak J. Immunophenotypic profile of nucleated erythroid progenitors during maturation in regenerating bone marrow. Leuk Lymphoma. 2013;54(11):2523-2530. Wangen JR, Eidenschink Brodersen L, Stolk TT, Wells DA, Loken MR. Assessment of normal erythropoiesis by flow cytometry: important considerations for specimen preparation. Int J Lab Hematol. 2014; 36(2):184-196. McGrath KE, Bushnell TP, Palis J. Multispectral imaging of hematopoietic cells: where flow meets morphology. J Immunol Methods. 2008;336(2):91-97. Della Porta MG, Malcovati L, Invernizzi R, et al. Flow cytometry evaluation of erythroid dysplasia in patients with myelodysplastic syndrome. Leukemia. 2006; 20(4):549-555. Lorand-Metze I, Ribeiro E, Lima CS, Batista LS, Metze K. Detection of hematopoietic maturation abnormalities by flow cytometry in myelodysplastic syndromes and its utility for the differential diagnosis with non-clonal disorders. Leuk Res. 2007; 31(2):147-155. Xu F, Wu L, He Q, et al.Immunophenotypic analysis of erythroid dysplasia and its diagnostic application in myelodysplastic syndromes. Intern Med J. 2012;42(4):401-411. Hu J, Liu J, Xue F, et al. Isolation and functional characterization of human erythroblasts at distinct stages: implications for

22.

23. 24.

25.

26.

27.

28.

29.

30.

31.

32.

understanding of normal and disordered erythropoiesis in vivo. Blood. 2013; 121(16):3246-3253. Eidenschink Brodersen L, Menssen AJ, Wangen JR, et al. Assessment of Erythroid Dysplasia by "Difference from Normal"in Routine Clinical Flow Cytometry Work-up. Cytometry B Clin Cytom. 2015;88(2):125135. Bain BJ. The bone marrow aspirate of healthy subjects. Br J Haematol. 1996; 94(1):206-209. Parmentier S, Schetelig J, Lorenz K et al. Assessment of dysplastic hematopoiesis: lessons from healthy bone marrow donors. Haematologica. 2012;97(5):723-730. Della Porta MG, Travaglino E, Boveri E, et al. Minimal morphological criteria for defining bone marrow dysplasia: a basis for clinical implementation of WHO classification of myelodysplastic syndromes. Leukemia. 2015;29(1):66-75. Westers TM, van der Velden, VHJ, Alhan C, et al. Implementation of flow cytometry in the diagnostic work-up of myelodysplastic syndromes in a multicenter approach: report from the Dutch Working Party on Flow Cytometry in MDS. Leuk Res. 2012; 36(4):422-430. van Dongen JJ, Lhermitte L, Bottcher S, et al. EuroFlow antibody panels for standardized n-dimensional flow cytometric immunophenotyping of normal, reactive and malignant leukocytes. Leukemia. 2012;26(9):1908-1975. Mathis S, Chapuis N, Debord C, et al. Flow cytometric detection of dyserythropoiesis: a sensitive and powerful diagnostic tool for myelodysplastic syndromes. Leukemia. 2013;27(10):1981-1987. Cremers, EMP, Westers, TM, Alhan C, et al. Implementation of erythroid analysis by flow cytometry in diagnostic models for myelodysplastic syndromes. Haematologica. 2016;102(2):320-326. Kalina T, Flores-Montero J, Lecrevisse Q, et al. Quality assessment program for EuroFlow protocols: Summary results of four-year (2010-2013) quality assurance rounds. Cytometry A. 2015;87(2):145-156. Laranjeira P, Rodrigues R, Carvalheiro T, et al. Expression of CD44 and CD35 during normal and myelodysplastic erythropoiesis. Leuk Res. 2014;39(3):361-370. Macchherndl-Spandl S, Suessner S, Danzer M, et al. Abnormal Expression of the Major Coxsackie-Adenovirus Receptor CAR on Immature Dysplastic CD105+ Erythroid Progenitor Cells in Patients with MDS and Related Bone Marrow Neoplasms. Blood.

haematologica | 2017; 102(2)


Flow cytometric analysis of erythrodysplasia in MDS

2013;122(21):2818. 33. Bauer K, Machherndl-Spandl S, Suessner S, et al. Identification of CAR as Novel Mediator of Erythroid Differentiation and Migration That is Specifically Downregulated in Erythroid Progenitor Cells in Patients with MDS. Blood. 2014; 124(21):1570. 34. Kern W, Haferlach C, Schnittger S, Alpermann T, Haferlach T. Serial assessment of suspected myelodysplastic syndromes: significance of flow cytometric findings validated by cytomorphology, cytogenetics, and molecular genetics. Haematologica. 2013(2);98:201-207. 35. Cremers EMP, Westers TM, Alhan C, et al. Multiparameter flow cytometry is instrumental to distinguish myelodysplastic syndromes from non-neoplastic cytopenias.

haematologica | 2017; 102(2)

Eur J Cancer. 2016;54(2):49-56. 36. Cutler JA, Wells DA, van de Loosdrecht AA, et al. Phenotypic abnormalities strongly reflect genotype in patients with unexplained cytopenias. Cytometry B Clin Cytom. 2011;80(3):150-157. 37. Oelschlaegel U, Westers TM, Mohr B, et al. Myelodysplastic syndromes with a deletion 5q display a characteristic immunophenotypic profile suitable for diagnostics and response monitoring. Haematologica. 2015;100(3):e93-6. 38. Papaemmanuil E, Cazzola M, Boultwood J, et al. Somatic SF3B1 mutation in myelodysplasia with ring sideroblasts. N Engl J Med. 2011;365(15):1384-1395. 39. Matarraz S, Paiva B, DĂ­ez -Campelo M, et al. Immunophenotypic alterations of bone marrow myeloid cell compartments in mul-

tiple myeloma patients predict for myelodysplasia-associated cytogenetic alterations. Leukemia. 2014;28(8):17471750. 40. Rodriguez-Caballero A, Henriques A, Criado I, et al. Subjects with chronic lymphocytic leukaemia-like B-cell clones with stereotyped B-cell receptors frequently show MDS-associated phenotypes on myeloid cells. Br J Haematol. 2015; 168(2):258-267. 41. Porwit A, van de Loosdrecht AA, Bettelheim P, et al. Revisiting guidelines for integration of flow cytometry results in the WHO classification of myelodysplastic syndromes-proposal from the International/European LeukemiaNet Working Group for Flow Cytometry in MDS. Leukemia. 2014;28(9):1793-1798.

319


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Myelodysplastic Syndromes

Ferrata Storti Foundation

Haematologica 2017 Volume 102(2):320-326

Implementation of erythroid lineage analysis by flow cytometry in diagnostic models for myelodysplastic syndromes Eline M.P. Cremers,1 Theresia M. Westers,1 Canan Alhan,1 Claudia Cali,1 Heleen A. Visser-Wisselaar,2 Dana A. Chitu,2 Vincent H.J. van der Velden,3 Jeroen G. te Marvelde,3 Saskia K. Klein,4 Petra Muus,5 Edo Vellenga,6 Georgina E. de Greef,7 Marie-Cecile C.J.C. Legdeur,8 Pierre W. Wijermans,9 Marian J.P.L. Stevens-Kroef,10 Pedro da Silva-Coelho,11 Joop H. Jansen,11 Gert J. Ossenkoppele1 and Arjan A. van de Loosdrecht1 A study on behalf of the HOVON89 study group

Department of Hematology, VU University Medical Center, Cancer Center Amsterdam; HOVON Data Center, Erasmus MC Cancer Institute, Clinical Trial Center, Rotterdam; Department of Immunology, Erasmus University Medical Center, Rotterdam; 4 Department of Internal Medicine, Meander Medical Center, Amersfoort; 5Department of Hematology, Radboud University Medical Centre, Nijmegen; 6Department of Hematology, University Medical Center Groningen; 7Department of Hematology Erasmus MC Cancer Institute Rotterdam; 8Department of Internal Medicine, Medisch Spectrum Twente, Enschede; 9Department of Internal Medicine, Haga Ziekenhuis, The Hague; 10 Department of Human Genetics, Radboud University Medical Centre, Nijmegen and 11 Laboratory of Hematology, Radboud University Medical Centre, Nijmegen, The Netherlands

1

2

3

ABSTRACT

F

Correspondence: a.vandeloosdrecht@vumc.nl

Received: April 15, 2016. Accepted: September 14, 2016. Pre-published: September 22, 2016. doi:10.3324/haematol.2016.147843

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/320

Š2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

320

low cytometric analysis is a recommended tool in the diagnosis of myelodysplastic syndromes. Current flow cytometric approaches evaluate the (im)mature myelo-/monocytic lineage with a median sensitivity and specificity of ~71% and ~93%, respectively. We hypothesized that the addition of erythroid lineage analysis could increase the sensitivity of flow cytometry. Hereto, we validated the analysis of erythroid lineage parameters recommended by the International/European LeukemiaNet Working Group for Flow Cytometry in Myelodysplastic Syndromes, and incorporated this evaluation in currently applied flow cytometric models. One hundred and sixty-seven bone marrow aspirates were analyzed; 106 patients with myelodysplastic syndromes, and 61 cytopenic controls. There was a strong correlation between presence of erythroid aberrancies assessed by flow cytometry and the diagnosis of myelodysplastic syndromes when validating the previously described erythroid evaluation. Furthermore, addition of erythroid aberrancies to two different flow cytometric models led to an increased sensitivity in detecting myelodysplastic syndromes: from 74% to 86% for the addition to the diagnostic score designed by Ogata and colleagues, and from 69% to 80% for the addition to the integrated flow cytometric score for myelodysplastic syndromes, designed by our group. In both models the specificity was unaffected. The high sensitivity and specificity of flow cytometry in the detection of myelodysplastic syndromes illustrates the important value of flow cytometry in a standardized diagnostic approach. The trial is registered at www.trialregister.nl as NTR1825; EudraCT n.: 2008-002195-10

Introduction Myelodysplastic syndromes (MDS) are a heterogeneous group of clonal hematopoietic disorders characterized by cytopenia(s) and risk of leukemic transformation.1 Multi-parameter flow cytometric (FC) analysis is a recommended tool haematologica | 2017; 102(2)


Erythroid lineage analysis by FC in diagnosis MDS

to support the diagnosis of MDS, which is based on dysplastic features by cytomorphology and typical cytogenetInternational/European ic abnormalities.2 The LeukemiaNet Working Group for Flow Cytometry in MDS (IMDS-flow) provided recommendations on how to process and analyze bone marrow aspirates of patients with unexplained cytopenias suspected of MDS.3,4 Analytic methods have been developed and validated for characterization and quantification of dysplasia and enable accurate diagnosis of MDS.5–12 The most straightforward is a four-parametric diagnostic score that integrates percentage of CD34-positive myeloid progenitors, percentage of B-cell progenitors within the CD34-positive compartment, CD45 expression level of CD34-positive myeloid progenitors (related to CD45 expression level on lymphocytes), and sideward light scatter peak channel value (SSC) of granulocytic cells (related to SSC of lymphocytes). This diagnostic score has a sensitivity and specificity of 69% and 92%, respectively, in low-intermediate risk MDS.13,14 More elaborate scores can reach specificities of up to 100%; this, however, is accompanied by lower sensitivities.15 In accordance with recommendations issued by the IMDS-flow, our group designed and validated an integrated MDS-FC score (iFS).16 The iFS comprises the diagnostic score and evaluation of frequently described aberrant expression levels of lineage defining markers and presence of lineage infidelity markers on (im)mature myelo-/monocytic cells. Sensitivity and specificity of the iFS within a large cohort of patients with persistent cytopenias of unknown origin were 63% and 98%, respectively.17 The lower sensitivity in this and other reports can be explained by the fact that most MDS-FC approaches only evaluate the myeloid cell compartment. Since dyserythropoiesis is the most prevalent feature by cytomorphology in MDS, the addition of in-depth evaluation of the erythroid compartment is expected to improve sensitivity.18 MDS patients with erythroid dysplasia, but without dysmyelopoiesis, may then be identified by FC. For evaluation of the erythroid compartment, different antibody combinations of CD45, CD235a, CD71, CD36, CD105, and intracellular markers such as cytosolic H-ferritin, cytosolic L-ferritin and mitochondrial ferritin have been described.19–22 The IMDS-flow group recently proposed guidelines for erythroid evaluation, advising the evaluation of CD36 coefficient of variation (CV), CD71 CV and mean fluorescence intensity (MFI), and percentage of progenitors (CD117 positive within CD45 negativediminished cell fraction) within the erythroid compartment. Sensitivity and specificity of this marker combination for the detection of MDS-associated erythroid aberrancies were 35% and 90%, respectively. The current study aimed to validate these erythroid parameters in an independent cohort of patients diagnosed with MDS treated within a prospective clinical study and in a reference group of patients with proven non-clonal cytopenias. Furthermore, the additive value of erythroid evaluation to currently applied MDS-FC diagnostic approaches was explored.

Methods Patients A well-defined MDS group and cytopenic control group were assembled between May 2009 and July 2014 (Table 1). The MDS haematologica | 2017; 102(2)

group consisted of patients enrolled in the HOVON89 study. Bone marrow aspirates for FC analysis were taken prior to inclusion, and MDS was diagnosed in accordance with the minimum diagnostic criteria established by the WHO 2001 criteria.23 The definition of non-clonal cytopenias was based on clinical characteristics, cytomorphology, cytogenetic and biochemical indicators. The median age of the MDS group was 71 (range 38-85). The median age of the control group was 65 (range 23-91). The research program was approved by the local ethics committee, and all patientrelated research strictly abided by the Declaration of Helsinki.

Sample preparation, antibody combinations, and cell acquisition Sample processing was performed according to ELN guidelines for FC within 24 hours.15 A 4-color analysis was performed from 2009-2012, and an 8-color analysis from 2012-2014. The staining panels are outlined in the Online Supplementary Table S1. At least 100,000 CD45-positive events were acquired using a FACSCaliburTM or FACSCantoIITM (BD Biosciences, San Jose, CA, USA). Cells were analyzed using Cell QuestPro (BD Biosciences) or Infinicyt software (Cytognos, Salamanca, Spain), respectively. Gating was performed as previously described.15,24

MDS-FC scores For evaluation of the erythroid compartment, guidelines as described by the IMDS-flow were applied. Erythroid evaluation included analysis of CD71 (CV and MFI), CD36 (CV), and CD117 (percentage within the CD45-negative-diminished cell fraction). Cut-off values were assessed as described in the tandem-paper (see also the mathematical examples in the Online Supplementary Files of the paper). Examples are provided in Online Supplementary Figure S1. Following the simplified recommendations, an increased CV of CD71, a decreased MFI of CD71, an increased CV of CD36, and a decreased or increased percentage of CD117 were each assigned one point. A score of ≥2 points was defined as MDSassociated erythroid aberrancies. The four-parameter diagnostic score was calculated according to guidelines as previously described, using the defined cut-offs.13 The iFS was established as described previously.25 The diagnostic score, the iFS, and the erythroid score are described in Table 2A.

Models for incorporation of erythroid analysis Tables 2B-2C describe the two models designed to add erythroid FC analysis to validated MDS-FC approaches. Patients with MDS-associated erythroid aberrancies received one extra point in comparison with the original diagnostic score; a total of ≥2 points was labeled as MDS. The second model added erythroid evaluation to the iFS. Patients with iFS results B with erythroid aberrancies by FC were labeled compatible with MDS.

Statistics Results from MDS-FC were compared between the MDS and control group. Absolute numbers and relative percentages described the data. To test the concordance between presence of MDS-associated erythroid aberrancies and patient group, a chisquare test was performed. To compare the results of different techniques the McNemar test was used. P-values <0.05 were statistically significant. Specificity and sensitivity, and 95%-confidence intervals, were calculated for each MDS-FC model using a two-by-two model. Inter-observer analysis of MDS-FC aberrancies and the diagnostic score was tested by an independent MDSexpert center: the Department of Immunology of the Erasmus University Medical Center, Rotterdam, The Netherlands. Analyses were performed using PASW Statistics version 20.0 (SPSS, Chicago, IL, USA). 321


E.M.P. Cremers et al.

Results Evaluation of erythroid markers In accordance with the IMDS-flow recommendations, we analyzed CD36 (CV), CD71 (CV and MFI), and CD117 (percentage within the CD45 negative-diminished cell fraction). Table 1 lists the analyzed erythroid markers per group. An increased CV of CD71 was the most sensitive marker for MDS as it was positive in 66% of MDS patients, followed by an increased/decreased percentage of CD117 (64%). An increased CV of CD36 was the most

specific marker as only 3% of controls were positive for this marker. Within the MDS group, 64% patients showed multiple erythroid aberrancies (≥2 points), compared with 11% of patients within the control group. The presence of multiple erythroid aberrancies was significantly correlated with the diagnosis of MDS (P<0.001).

Correlation between patient group and cytomorphology Since we found a significant correlation between patient group (MDS or control) and the presence of erythroid aberrancies, the next step was to evaluate the relation

Table 1. Erythroid markers that comprise the IMDS-Flow erythroid FC score and the cumulative score, stratified by patient group. Control group Alcohol abuse Aplastic anemia Auto-immune cytopenia Chronic disease Eosinophilia* Iron deficiency Iron incorporation disorder Medication caused cytopenia Vitamin B12 deficiency MDS group RCUD RARS RCMD RCMD-RS RAEB-1 Isolated del(5q) MDS-U CMML

CV CD71

%

14/61 0/2 2/3 3/10 0/3 0/1 6/26 2/10 0/3 1/3 70/106 2/4 19/20 12/23 19/27 11/14 3/12 1/2 3/4

23 0 67 30 0 0 23 20 0 33 66 50 95 52 70 79 25 50 75

MFI CD71 % 3/61 1/2 0/3 0/10 0/3 0/1 1/26 0/10 0/3 1/3 22/106 0/4 6/20 4/23 7/27 2/14 1/12 0/2 2/4

5 50 0 0 0 0 4 0 0 33 21 0 30 17 26 14 8 0 50

CV CD36

%

% CD117

%

≥2 points

%

2/61 0/2 0/3 1/10 0/3 0/1 1/26 0/10 0/3 0/3 36/106 2/4 9/20 8/23 8/27 7/14 2/12 0/2 0/4

3 0 0 10 0 0 4 0 0 0 34 50 45 35 30 50 17 0 0

35/61 0/2 0/3 8/10 3/3 1/1 15/26 4/10 3/3 1/3 64/106 3/4 8/20 15/23 20/27 6/14 10/12 1/21/4

57 0 0 80 100 100 58 40 100 33 60 75 40 65 74 43 83 50 25

7/61 0/2 0/3 3/10 0/3 0/1 4/26 0/10 0/3 0/3 68/106 3/4 17/20 13/23 19/27 10/14 4/12 0/2 2/4

11 0 0 30 0 0 15 0 0 0 64 75 85 57 70 71 33 0 50

*Normal bone marrow by cytomorphological assessment. The CD36 CV is the most specific parameter (2/61 control patients; 3%), and CD71 CV is the most sensitive parameter (70/106 MDS patients; 66%). In summary: 11% of the controls and 64% of the MDS patients show MDS-associated erythroid aberrancies, as defined by ≥2 points.

Figure 1 MDS-FC results in the MDS and control group. The diagnostic score and the integrated MDS-FC score in patients within the MDS group and control group. The arrows demonstrate the patients changing groups after addition of erythroid evaluation as recommended by the IMDS-flow group. *Flow cytometric results showed minimal dysplastic features, not enough for MDS.

322

haematologica | 2017; 102(2)


Erythroid lineage analysis by FC in diagnosis MDS

between erythroid evaluation by cytomorphology and FC in more detail. As controls might have minimal dyserythropoiesis by cytomorphology, FC might also detect erythroid aberrancies in controls.26,27 Information about erythroid features by cytomorphology was available in 92% of patients in the MDS group, and in 98% patients within the control group. Table 3 provides an overview of the results. Although the positive test results (dyserythropoiesis by cytomorphology and erythroid aberrancies by FC) seem equally distributed between the MDS and the controls, FC identified more dysplastic cases than cyto-

morphology (MDS-FC-positive cases within the cytopenic controls based on morphology). Therefore, the McNemar test, which focuses on the differences between two correlated proportions, was not significant (P=0.01).

Addition of erythroid markers to current MDS-FC scoring systems - diagnostic score The original diagnostic score was indicative for MDS in 78/106 MDS patients, and negative for MDS in 53/61 of the control patients (Figure 1). Hence, sensitivity and specificity of this diagnostic score were 74% (95% CI:

Table 2A. The parameters that describe the original integrated MDS-FC score, the erythroid score and the diagnostic score.

Diagnostic Score Two of the following: Increased percentage of myeloid progenitor cells

Myeloid progenitors

Granulocytes**

Monocytes**

Erythrocytes

>5% myeloid progenitors

Two of the following: Decreased SSC Abnormal CD11b/CD13 Abnormal CD16/CD13 Expression of HLA-DR Lack of CD33 expression Asynchronous shift to the left Abnormal expression of CD15

Two of the following: Abnormal CD45/SSC Decreased/increased number as compared to lymphocytes Abnormal CD11b Abnormal HLA-DR Abnormal CD11b/HLA-DR Abnormal expression of CD14 Abnormal expression of CD13 Loss of CD16 Abnormal expression of CD33

Two of the following***: Increased CD36 coefficient of variation Increased CD71 coefficient of variation

OR: <5% myeloid progenitors with one of the following: Lymphoid markers present Decreased SSC on granulocytes (CD2, CD5, CD19, CD25, CD56) Abnormal expression of CD45 on myeloid progenitor cells

Decreased percentage of B-cell progenitor cells

OR: <5% myeloid progenitors with two of the following: Decrease in CD45 expression Abnormal expression of CD34 Abnormal expression of CD117 Abnormal expression of CD13 Abnormal expression of CD33 Abnormal expression of HLA-DR Expression of CD11b Expression of CD15*

OR: Presence of lymphoid markers

Decreased expression of CD71 Decreased / increased percentage of CD117 positive within nucleated erythroid cells

OR: Presence of lymphoid markers

OR: Presence of CD34 on mature myeloid cells

OR: Presence of CD34 on OR: Myeloid/Lymphoid ratio < 1

mature monocytic cells

If a cell compartment is considered abnormal, a ‘+’ is assigned in Tables 2B-2C.*Note that normal myeloid progenitors might also express CD15. **The granulocytic and monocytic cell compartments were integrated into one compartment in Table 2C (the iFS). ***in case of aberrant CD71 percentage and CD117 percentage one extra abnormality is mandatory. This figure is adapted from Wells et al., scores adjusted as by Cutler et al., and Cremers et al.7,17,25

Table 2B. The addition of the erythroid evaluation to the diagnostic score.13,14 Diagnostic score 0 0 Aberrant erythroid + MDS according to FC No No

1 No

Table 2C. The addition of the erythroid evaluation to the integrated MDS-FC score (iFS).16

Diagnostic score Aberrant myeloid progenitors Aberrant neutrophils (≥2 other aberrancies) Aberrant monocytes (CD56 / ≥2 aberrancies) Original iFS* Aberrant erythroid (≥2 aberrancies) New iFS* Labeled MDS

<2 abnormalities -

-

+

A A No

A + B No

A/B B No

+ +

+ -

A/B A/B + C B Yes No

≥2 Yes

1 + Yes

+ A/B + C Yes

+ +

+

C C - + C C Yes Yes

≥2 + Yes

≥2 abnormalities

-

-

+

+ +

+ + - -

A/B A/B B/C B/C B/C B/C - + - + - + A/B C C C C C No Yes Yes Yes Yes Yes

+ +

+ +

C C - + C C Yes Yes

The four-parameter diagnostic score as described by Della Porta et al.,14 Aberrant myeloid markers, neutrophils and monocytes based on the modified FCSS score. Aberrant myeloid markers as describes in table 2A; more than 2 points per lineage. Aberrant erythroid markers as recommended by the ELNet iMDS-flow, described in Table 2A and the tandem-paper. *Category A ‘no MDS-related features’, B ‘limited number of changes associated with MDS’, or C ‘features consistent with MDS’. Choice for A or B and B or C depends on the kind and number of aberrancies that are encountered. Note that patients with ≥2 points in the diagnostic score can still be labeled as no MDS by the iFS when there are no other abnormalities.

haematologica | 2017; 102(2)

323


E.M.P. Cremers et al.

64%-82%) and 87% (95% CI: 76%-94%), respectively. By erythroid evaluation, 64% of MDS patients and 11% of controls revealed erythroid aberrancies by FC (Table 1). Erythroid results were added to the diagnostic score as illustrated in Table 2B. This led to an upgrade in MDS-FC category in 13 MDS patients and 2 controls. Consequently, the sensitivity and specificity for the diagnostic score including erythroid evaluation were 86% (95% CI: 78%-92%) and 84% (95% CI: 72%-92%), respectively.

Addition of erythroid markers to current MDS-FC scoring systems - integrated MDS-FC score With the addition of erythroid analysis, two extra RARS patients, five RCMD patients, four RCMD-RS patients, and one del(5q) patient were subsequently recognized as MDS. The addition of erythroid analysis did not alter the results for the RAEB-1, MDS-U and CMML patients (Figure 2 and Table S2). Results of the original iFS were C ‘compatible with MDS’ in 73/106, B ‘minor MDS related aberrancies’ in 21/106, and A ‘not compatible with MDS’ in 12/106 MDS patients. Interestingly, each MDS patient not recognized by the original iFS showed only dyserythropoiesis with or without dysmegakaryopoiesis by cytomorphological assessment. In the control group, results were A in 40/61 patients, B in 20/61 patients, and C in only 1/61 patients. The calculated sensitivity and specificity of the iFS were 69% (95% CI: 59% to 78%) and 98% (95% CI: 91%-100%), respectively. In the MDS group, 33 patients were not assigned to MDS by the original iFS (Figure 1; category A and B). After addition of erythroid evaluation, 12 MDS patients

changed from B to C (now allocated MDS), and 5 patients in category A were changed to B (limited changes but still no MDS). In total, 21 patients were not assigned to MDS; 7 in category A, 14 in category B. In the control group, one patient was incorrectly identified as MDS (category C). After addition of erythroid evaluation, two extra patients in category B were upgraded to C and thus allocated as MDS (Figure 1). Overall, the sensitivity of the iFS increased to 80% (95% CI: 71%-87%), and the specificity showed only a minor decline to 95% (95% CI: 86%-99%) In summary, the sensitivity for both the diagnostic score and the iFS increased significantly after addition of erythroid evaluation. For the diagnostic score, sensitivity increased from 74% to 86%, and the iFS sensitivity increased from 69% to 80%. For both strategies, specificity was only marginally affected: 87% to 84% for the diagnostic score; and 98% to 95% for the iFS. Figure 2 illustrates distribution of WHO classifications within the original iFS, and after addition of erythroid evaluation.

Robustness of the MDS-FC results Interpretation of FC data in MDS is considered to require a high level of expertise. To check solidity of our MDS-FC based conclusions, 25% of the MDS cases were analyzed blindly by an independent MDS-FC expert center (VHJvdV and JtM). The scores were calculated in the same data files. Results of the diagnostic score revealed a concordance of 100% and 89% for the 4-color and 8-color analysis, respectively. Analysis of the iFS revealed a concordance of 89% and 86%, for the 4-color analysis and the 8-color analysis, respectively. Addition of erythroid evaluation did not influence the concordance of the MDS-FC models.

Figure 2 WHO-classifications within different MDS-FC groups. Distribution of WHO-classifications within the original iFS categories, and iFS categories after the addition of erythroid evaluation. With the addition of the erythroid compartment, patients shift into a higher MDS-FC category. Category A ‘no MDS-related features’, B ‘limited number of changes associated with MDS’, or C ‘features consistent with MDS’. Absolute patient numbers are provided in the Online Supplementary Files (Table S2)

324

haematologica | 2017; 102(2)


Erythroid lineage analysis by FC in diagnosis MDS

Discussion The evaluation of dyserythropoiesis by a flow cytometric (FC) approach is not included in most of todayâ&#x20AC;&#x2122;s MDSFC models. The International/European LeukemiaNet Working Group for Flow Cytometry in MDS (IMDS-flow) proposed a method for evaluation of the erythroid compartment by FC. In the current study, we validated erythroid evaluation and investigated the value of the introduced erythroid evaluation in two previously validated MDS-FC approaches. We analyzed 167 bone marrow aspirates, 106 patients with MDS, and 61 cytopenic controls for which the IMDS-Flow erythroid score, diagnostic score, and integrated FC score (iFS) were calculated.13,16 Originally, the erythroid score was designed as a weighted score. It can also be applied as a numerical score (one point per parameter) in which â&#x2030;Ľ2 points identifies MDS-associated erythroid aberrancies. The exception made in the tandem-manuscript is to be noted: if the 2 points are based on the combination of decreased MFI of CD71 and abnormal percentage of CD117, an additional aberrancy is warranted. The latter was not seen in this cohort. Results from erythroid evaluation confirmed the results of the IMDSflow report since we showed a strong significant correlation between MDS-associated erythroid aberrancies assessed by FC and MDS. Investigation of the correlations between cytomorphological results and FC results suggested that FC detected less erythroid aberrancies compared with cytomorphology results. Here, the fact that both techniques investigate different aspects needs to be considered. FC mainly evaluates cell surface characteristics, whilst cytomorphology also evaluates features within the cell, such as nuclear bridging. It is unknown whether these dysplastic features result in altered antigen expression. The FC method is however rather specific as, for example, it did not report MDS-associated erythroid aberrancies where cytomorphology described dyserythropoiesis in patients with a vitamin B12 deficiency. This indicates that both techniques provide supplementary information and complement, rather than contradict, one another. The goal of the study was to increase the sensitivity of currently applied MDS-FC models. Indeed, the addition of erythroid lineage analysis to the currently applied diagnostic score demonstrated an increased sensitivity (from 74% to 86%), without a major loss in specificity (87% to 84%). These results support the findings of Mathis et al., who tested the addition of erythroid evaluation by FC in nonlysed samples (RED score) to the diagnostic score.22 The combination was analyzed in a cohort of 101 patients (83 MDS patients and only 18 controls) and resulted in a sensitivity and specificity of 88% and 89%, respectively. The RED score and the erythroid score described by the IMDSflow both comprised evaluation of CD36 CV and CD71 CV. Differences were, however, i) a non-lysed method in the RED score, ii) the addition of hemoglobin level in the RED score, iii) the added value of percentage of CD117, and iii) added value of expression level of CD71. As illustrated by Mathis and colleagues, hemoglobin showed a strong negative correlation with the other markers in the RED score. Note, hemoglobin might be subject to confounders, e.g., transfusion requirements, and as a non-FC parameter less suitable in a MDS-FC model. The second diagnostic MDS-FC model evaluated in the current study was the iFS; a more extensive model, comhaematologica | 2017; 102(2)

prising the diagnostic score and evaluation of frequently described aberrancies on (im)mature myelo-/monocytic cells. Addition of erythroid markers to this score led to an increased sensitivity (from 69 to 80%), without substantially affecting the specificity (from 98 to 95%). The combination of the iFS with the IMDS-flow erythroid score showed the highest specificity; higher than the other described scores. Most described MDS-FC scores were designed and validated in large patient cohorts. However, interpretation of results within individual patients can be challenging. To our knowledge, the iFS is the only MDS-FC algorithm that has proven its power in individual patients, demonstrated by its high specificity in patients with cytopenias of unknown origin followed over time.17 After addition of erythroid lineage evaluation, its specificity remained high and, therefore, it might be expected that the new model is applicable in individual patient analysis. To not overcall patients with cytopenia of unknown origin as MDS, one would prefer to apply the most specific model. However, in an era where cost-effectiveness is becoming increasingly important, a limited panel might be preferred. To improve the four-parameter diagnostic score, Bardet and colleagues advised the addition of CD7 (on myeloid progenitors) and CD56 (on monocytes) to the diagnostic score.28 Specificity of this adjusted score was 87%; however, the sensitivity was low (66%). Here, the addition of selected erythroid markers might improve the sensitivity of FC. The addition of analysis of mutation in genes involving splicing factors, epigenetic regulators, signal transduction or the cohesion complex, to diagnostic evaluation is suggested.29,30 However, none of the mutations is disease spe-

Table 3. Comparison of dyserythropoiesis as assessed by cytomorphology and flow cytometry. Control group Alcohol abuse Aplastic anemia Auto-immune cytopenia Chronic disease Eosinophilia Iron deficiency Iron incorporation disorder Medication caused cytopenia Vitamin B12 deficiency MDS group RCUD RARS RCMD RCMD-RS RAEB-1 Isolated del(5q) MDS-U CMML

By FC (N)

%

By CM (N)*

%

7/61 0/2 0/3 3/10 0/3 0/1 4/26 0/10 0/3 0/3 68/106 3/4 17/20 13/23 19/27 10/14 4/12 0/2 2/4

11 0 0 30 0 0 15 0 0 0 64 75 85 57 70 71 33 0 50

6/60 0/2 0/2** 2/10 0/3 0/1 0/26 1/10 0/3 3/3 81/97*** 3/3 20/20 17/20 26/26 11/12 1/10 0/2 4/4

10 0 0 20 0 0 0 10 0 100 84 100 100 85 100 92 10 0 100

*Less than 10% erythroid dysplasia and therefore not enough for diagnosis MDS or >10% and classified MDS according to WHO criteria; **For one aplastic anemia patient there were not enough erythroid cells for proper evaluation; ***Cytomorphological details absent in 9 patients.

325


E.M.P. Cremers et al.

cific, and some mutations appeared to be present in low frequency in the elderly population.31 Therefore, more research regarding their role in the diagnostic setting in MDS is warranted. Until then, FC has proven to be a valuable diagnostic tool, which can fill in the gaps where cytomorphology and cytogenetic results are less certain of a diagnosis. It has shown to be highly specific in the diagnosis of MDS, so can exclude patients from unnecessary follow-up. MDS-FC is described to be less sensitive in MDS recognition. Our study, however, showed that addition of erythroid evaluation to currently applied MDS-FC models

References 1. Vardiman JW, Thiele J, Arber DA, et al. The 2008 revision of the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia: rationale and important changes. Blood. 2009;114(5):937–951. 2. Malcovati L, Hellström-Lindberg E, Bowen D, et al. Diagnosis and treatment of primary myelodysplastic syndromes in adults: recommendations from the European LeukemiaNet. Blood. 2013; 122 (17):2943– 2964. 3. van de Loosdrecht AA, Alhan C, Béné MC, et al. Standardization of flow cytometry in myelodysplastic syndromes: report from the first European LeukemiaNet working conference on flow cytometry in myelodysplastic syndromes. Haematologica. 2009;94(8):1124–1134. 4. Porwit A, van de Loosdrecht AA, Bettelheim P, et al. Revisiting guidelines for integration of flow cytometry results in the WHO classification of myelodysplastic syndromes-proposal from the International/European LeukemiaNet Working Group for Flow Cytometry in MDS. Leukemia. 2014;28(9):1793–1798. 5. Chopra A, Pati H, Mahapatra M, et al. Flow cytometry in myelodysplastic syndrome: analysis of diagnostic utility using maturation pattern-based and quantitative approaches. Ann Hematol. 2012; 91(9):1351–1362. 6. Matarraz S, López A, Barrena S, et al. Bone marrow cells from myelodysplastic syndromes show altered immunophenotypic profiles that may contribute to the diagnosis and prognostic stratification of the disease: a pilot study on a series of 56 patients. Cytometry B Clin Cytom. 2010; 78(3):154– 168. 7. Wells D, Benesch M, Loken MR, et al. Myeloid and monocytic dyspoiesis as determined by flow cytometric scoring in myelodysplastic syndrome correlates with the IPSS and with outcome after hematopoietic stem cell transplantation. Blood. 2003;102(1):394–403. 8. van de Loosdrecht AA, Ireland R, Kern W, et al. Rationale for the clinical application of flow cytometry in patients with myelodysplastic syndromes: position paper of an International Consortium and the European LeukemiaNet Working Group. Leuk Lymphoma. 2013; 54(3):472–475. 9. Kern W, Haferlach C, Schnittger S, Haferlach T. Clinical utility of multiparameter flow cytometry in the diagnosis of 1013 patients with suspected myelodys-

326

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

increased the sensitivity of FC in the detection of MDS. We postulate, therefore, that MDS-FC is ready for general clinical application. Acknowledgments The authors would like to thank Yvonne van der Veeken, Elsbeth Römers, and Adri Zevenbergen (VU University Medical Center, Cancer Center Amsterdam, Amsterdam) for their technical assistance, and all the participating centers within DutchBelgian Hematology Oncology Cooperative Trial Group (HOVON) for including patients in the HOVON89 trial.

plastic syndrome: correlation to cytomorphology, cytogenetics, and clinical data. Cancer. 2010;116(19):4549–4563. Lorand-Metze I, Ribeiro E, Lima CSP, et al. Detection of hematopoietic maturation abnormalities by flow cytometry in myelodysplastic syndromes and its utility for the differential diagnosis with nonclonal disorders. Leuk Res. 2007; 31(2):147– 155. Malcovati L, Della Porta MG, Lunghi M, et al. Flow cytometry evaluation of erythroid and myeloid dysplasia in patients with myelodysplastic syndrome. Leukemia. 2005;19(5):776–783. Xu F, Li X, Chang C-K, et al. Establishment and Validation of an Updated Diagnostic FCM Scoring System Based on Pooled Immunophenotyping in CD34+ Blasts and Its Clinical Significance for Myelodysplastic Syndromes. PLoS One. 2014;9(2):e88706. Della Porta MG, Picone C, Pascutto C, et al. Multicenter validation of a reproducible flow cytometric score for the diagnosis of low-grade myelodysplastic syndromes: results of a European LeukemiaNET study. Haematologica. 2012;97(8):1209–1217. Ogata K, Della Porta MG, Malcovati L, et al. Diagnostic utility of flow cytometry in low-grade myelodysplastic syndromes: a prospective validation study. Haematologica. 2009;94(8):1066–1074. Westers TM, Ireland R, Kern W, et al. Standardization of flow cytometry in myelodysplastic syndromes: a report from an international consortium and the European LeukemiaNet Working Group. Leukemia. 2012;26(7):1730–1741. Van de Loosdrecht A, Westers T. Cutting edge: flow cytometry in myelodysplastic syndromes. J Natl Compr Canc Netw. 2013;11(7):892–902. Cremers EMP, Westers TM, Alhan C, et al. Multiparameter flow cytometry is instrumental to distinguish myelodysplastic syndromes from non-neoplastic cytopenias. Eur J Cancer. 2016;54:49–56. Germing U, Strupp C, Giagounidis A, et al. Evaluation of dysplasia through detailed cytomorphology in 3156 patients from the Düsseldorf Registry on myelodysplastic syndromes. Leuk Res. 2012;36(6):727–734. Xu F, Wu L, He Q, et al.Immunophenotypic analysis of erythroid dysplasia and its diagnostic application in myelodysplastic syndromes. Intern Med J. 2012;42(4):401–411. Fajtova M, Kovarikova A, Svec P, et al. Immunophenotypic profile of nucleated erythroid progenitors during maturation in regenerating bone marrow. Leuk

Lymphoma. 2013;54(11):2523–2530. 21. Eidenschink Brodersen L, Menssen AJ, Wangen JR, et al. Assessment of erythroid dysplasia by “Difference from normal” in routine clinical flow cytometry workup. Cytometry B Clin Cytom. 2015;88(2):125135. 22. Mathis S, Chapuis N, Debord C, et al. Flow cytometric detection of dyserythropoiesis: a sensitive and powerful diagnostic tool for myelodysplastic syndromes. Leukemia. 2013;27(10):1981–1987. 23. Vardiman JW, Harris NL, Brunning RD. The World Health Organization (WHO) classification of the myeloid neoplasms. Blood. 2002;100(7):2292–2302. 24. Westers TM, van der Velden VHJ, Alhan C, et al. Implementation of flow cytometry in the diagnostic work-up of myelodysplastic syndromes in a multicenter approach: report from the Dutch Working Party on Flow Cytometry in MDS. Leuk Res. 2012;36(4):422–430. 25. Cutler J, Wells D, van de Loosdrecht AA, et al. Phenotypic abnormalities strongly reflect genotype in patients with unexplained cytopenias. Cytometry B Clin Cytom. 2011;80(3):150–157. 26. Maassen A, Strupp C, Giagounidis A, et al. Validation and proposals for a refinement of the WHO 2008 classification of myelodysplastic syndromes without excess of blasts. Leuk Res. 2013; 37(1):64– 70. 27. Della Porta MG, Travaglino E, Boveri E, et al. Minimal morphological criteria for defining bone marrow dysplasia: a basis for clinical implementation of WHO classification of myelodysplastic syndromes. Leukemia. 2015;29(1):66–75. 28. Bardet V, Wagner-Ballon O, Guy J, et al. Multicentric study underlining the interest of adding CD5, CD7 and CD56 expression assessment to the flow cytometric Ogata score in myelodysplastic syndromes and myelodysplastic/myeloproliferative neoplasms. Haematologica. 2015;100(4):472– 478. 29. Haferlach T, Nagata Y, Grossmann V, et al. Landscape of genetic lesions in 944 patients with myelodysplastic syndromes. Leukemia. 2014;28(2):241-247. 30. Papaemmanuil E, Gerstung M, Malcovati L, et al. Clinical and biological implications of driver mutations in myelodysplastic syndromes. Blood. 2013;122(22):3616-3627. 31. Genovese G, Kähler AK, Handsaker RE, et al. Clonal Hematopoiesis and BloodCancer Risk Inferred from Blood DNA Sequence. N Engl J Med. 2014 Dec 25;371(26):2477-2487.

haematologica | 2017; 102(2)


ARTICLE

Myeloproliferative Disorders

Primary analysis of a phase II open-label trial of INCB039110, a selective JAK1 inhibitor, in patients with myelofibrosis

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

John O. Mascarenhas,1 Moshe Talpaz,2 Vikas Gupta,3 Lynda M. Foltz,4 Michael R. Savona,5 Ronald Paquette,6* A. Robert Turner,7 Paul Coughlin,8 Elliott Winton,9 Timothy C. Burn,10 Peter O'Neill,10 Jason Clark,10 Deborah Hunter,10 Albert Assad,10 Ronald Hoffman1 and Srdan Verstovsek11 1 The Icahn School of Medicine at Mount Sinai, New York, NY, USA; 2University of Michigan Cancer Center, Ann Arbor, MI, USA; 3Princess Margaret Cancer Centre, Toronto, ON, Canada; 4St. Paul's Hospital, University of British Columbia, Vancouver, BC, Canada; 5Vanderbilt-Ingram Cancer Center, Vanderbilt University Medical Center, Nashville, TN, USA; 6University of California, Los Angeles, CA, USA; 7Cross Cancer Institute Edmonton, AB, Canada; 8Monash University, VIC, Australia; 9Winship Cancer Institute, Emory University, Atlanta, GA, USA; 10Incyte Corporation, Wilmington, DE, USA and 11The University of Texas MD Anderson Cancer Center, Houston, TX, USA

*Current affiliation: Cedars-Sinai, Los Angeles, CA, USA

Haematologica 2017 Volume 102(2):327-335

ABSTRACT

C

ombined Janus kinase 1 (JAK1) and JAK2 inhibition therapy effectively reduces splenomegaly and symptom burden related to myelofibrosis but is associated with dose-dependent anemia and thrombocytopenia. In this open-label phase II study, we evaluated the efficacy and safety of three dose levels of INCB039110, a potent and selective oral JAK1 inhibitor, in patients with intermediate- or high-risk myelofibrosis and a platelet count ≥50×109/L. Of 10, 45, and 32 patients enrolled in the 100 mg twice-daily, 200 mg twice-daily, and 600 mg once-daily cohorts, respectively, 50.0%, 64.4%, and 68.8% completed week 24. A ≥50% reduction in total symptom score was achieved by 35.7% and 28.6% of patients in the 200 mg twice-daily cohort and 32.3% and 35.5% in the 600 mg once-daily cohort at week 12 (primary end point) and 24, respectively. By contrast, two patients (20%) in the 100 mg twice-daily cohort had ≥50% total symptom score reduction at weeks 12 and 24. For the 200 mg twice-daily and 600 mg once-daily cohorts, the median spleen volume reductions at week 12 were 14.2% and 17.4%, respectively. Furthermore, 21/39 (53.8%) patients who required red blood cell transfusions during the 12 weeks preceding treatment initiation achieved a ≥50% reduction in the number of red blood cell units transfused during study weeks 1-24. Only one patient discontinued for grade 3 thrombocytopenia. Non-hematologic adverse events were largely grade 1 or 2; the most common was fatigue. Treatment with INCB039110 resulted in clinically meaningful symptom relief, modest spleen volume reduction, and limited myelosuppression.

Correspondence: john.mascarenhas@mssm.edu

Received: June 23, 2016. Accepted: October 17, 2016. Pre-published: October 27, 2016. doi:10.3324/haematol.2016.151126

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/327

Introduction Myelofibrosis is a myeloproliferative neoplasm associated with progressive bone marrow fibrosis and ineffective hematopoiesis.1,2 Patients with myelofibrosis have a reduced life expectancy and commonly experience splenomegaly and a variety of constitutional and/or spleen-related symptoms, including early satiety, night sweats, pruritus, and abdominal discomfort.3 Although the symptoms of myelofibrosis vary among patients, for many the overall symptom burden is substantial and often has a debilitating effect on quality of life.4-6 Although most patients with myelofibrosis have primary disease, some develop myelofibrosis secondary to other myeloproliferative neoplasms, including polyhaematologica | 2017; 102(2)

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

327


J.O. Mascarenhas et al.

cythemia vera and essential thrombocythemia. Patients with myelofibrosis may have multiple somatic hematopoietic stem cell mutations, and nearly all carry mutations in one of three genes that have been linked to overactivation of the Janus kinase/signal transducers and activators of transcription (JAK-STAT) signaling pathway. These mutations include JAK2V617F, a gain-of-function mutation causing constitutive JAK2 activation, MPL mutations that facilitate ligand-independent thrombopoietin receptor signaling through JAK2, and mutations in CALR.7,8 In addition to overactive JAK2 signaling in malignant hematopoietic stem cells,8 the dysregulation of cytokine networks that control inflammatory processes and rely on JAK1-mediated signal transduction is an important pathogenic mechanism.9,10 A large number of inflammatory and immune mediators, such as interleukin-6, interleukin-2, and interferon-γ, are known to signal primarily through the association of either JAK1/JAK2 or JAK1/JAK3 pairing.9 The only therapy currently approved for the treatment of patients with intermediate- or high-risk myelofibrosis is ruxolitinib, an equipotent JAK1 and JAK2 inhibitor that has been shown in phase III clinical trials [COntrolled MyeloFibrosis Study With ORal JAK Inhibitor Treatment I (COMFORT-I)11 and COMFORT-II12] to be highly effective in reducing myelofibrosis-related splenomegaly and symptom burden. The relative contribution of JAK1 and JAK2 inhibition to the therapeutic benefit of ruxolitinib is unclear. Patients treated with ruxolitinib experience rapid reductions in the circulating levels of inflammatory cytokines, including tumor necrosis factor α and interleukin-6.11,13 It is likely that inhibition of JAK-STAT signaling beyond solely inhibiting neoplastic myeloproliferation is responsible for the treatment effects of ruxolitinib and, given the importance of JAK1 in the transduction of proinflammatory signals, JAK1 inhibition may be an important basis for therapeutic effects of ruxolitinib.9 A potential benefit of selective JAK1 inhibition without JAK2 inhibition would be the elimination of myelosuppression attributable to the inhibition of JAK2-mediated hematopoiesis. To explore this possibility, we sought to determine the therapeutic benefits of INCB039110, a potent and selective inhibitor of JAK1 with low in vitro affinity for JAK2 (>20-fold selectivity for JAK1 over JAK2) and other members of the JAK family (>100-fold selectivity for JAK1 over JAK3 and TYK2) (data on file). Here we report the results of a phase II clinical study that evaluated the efficacy and safety of INCB039110 in patients with intermediate- or high-risk myelofibrosis.

Methods Patients and study design This open-label phase II study (ClinicalTrials.gov identifier NCT01633372) included adults with intermediate- or high-risk (according to the Dynamic International Prognostic Scoring System) primary myelofibrosis, post–polycythemia vera myelofibrosis, or post–essential thrombocythemia myelofibrosis, regardless of JAK2 mutation status, with platelet count ≥50×109/L, hemoglobin ≥8.0 g/dL (transfusions permitted), absolute neutrophil count ≥1×109/L, palpable spleen or prior splenectomy, and active myelofibrosis-related symptoms. Key exclusion criteria were impaired liver function, end-stage renal disease requiring dialysis, clinically significant concurrent infections requiring therapy, unstable cardiac function, and invasive malignancies in the preceding 2 years. Prior treatment for myelofibrosis, including JAK inhibitor 328

therapy, was not an exclusion criterion; however, eligible patients had to discontinue all prior treatment for myelofibrosis no later than 14 days before receipt of the first dose of study drug. The study used a Simon two-stage design to assess the efficacy and safety of different doses of INCB039110 (Online Supplementary Figure S1). The first 20 patients were randomized to two dose cohorts (100 or 200 mg twice daily). Based on emerging data from these cohorts, patients were assigned to an additional dose cohort (600 mg once daily), with the goal of optimizing starting doses for efficacy and safety. A dose cohort was expanded to enroll an additional seven patients if at least three of the first ten patients met the primary endpoint. Based on efficacy and safety data, a dose cohort could be expanded further, by up to 40 additional patients. The primary endpoint was the proportion of patients in each dose group with a ≥50% reduction from baseline to week 12 in total symptom score (TSS), consisting of the sum of all individual scores for six myelofibrosis-related symptoms: night sweats, pruritus, abdominal discomfort, pain under the left ribs, early satiety, and bone or muscle pain. Secondary endpoints included the proportion of patients with a ≥50% reduction in TSS from baseline to week 24, the proportions of patients with a ≥35% reduction in spleen volume from baseline to weeks 12 and 24, the percentage changes from baseline to weeks 12 and 24 in TSS and spleen volume, the proportion of patients requiring red blood cell (RBC) transfusions before study entry (i.e. ≥2 units in the 12 weeks preceding study day 1) who exhibited a ≥50% decrease in transfusion frequency over any 12-week period during the study, and safety and tolerability. Changes in the percentage of patients with RBC transfusion needs were determined ad hoc. Additional exploratory analyses included the determination of changes from baseline in JAK2V617F allele burden at weeks 12 and 24 and changes from baseline in plasma cytokine levels at week 4. The study protocol was approved by the local institutional review boards and ethics committees, and the study was conducted in accordance with the ethical principles laid out in the Declaration of Helsinki, the guidelines of the International Conference on Harmonisation for Good Clinical Practice, and all applicable federal and local regulations. All patients provided written informed consent before treatment initiation.

Assessments and analyses Patients had scheduled study visits at screening, baseline, day 1, and weeks 4, 8, 12, 16, 20, and 24. Patients assessed the severity of disease-related symptoms daily, using the modified Myelofibrosis Symptom Assessment Form v3.0 electronic diary, in which each symptom is ranked on a scale of 0 (absent) to 10 (worst imaginable). Spleen volume was assessed by magnetic resonance imaging or computed tomography at weeks 12 and 24. Patient Global Impression of Change (PGIC) of myelofibrosis symptoms was assessed at post-baseline visits using a questionnaire with a 7point rating scale (1: very much improved, 2: much improved, 3: minimally improved, 4: no change, 5: minimally worse, 6: much worse, 7: very much worse). Non-hematologic adverse events were reported using the Common Terminology Criteria for Adverse Events v4.03. Reporting of hematologic adverse events was based on laboratory data, except that grade 4 anemia was reported based on Common Terminology Criteria for Adverse Events v4.03. Genomic DNA isolation from peripheral blood and JAK2V617F assays were performed as described previously14 on samples collected at baseline and weeks 12 and 24. The percentage change from baseline in allele burden was calculated for all patients harboring the JAK2V617F mutation at baseline. Cytokines were measured in plasma samples collected at baseline and week 4 using the Myriad RBM HumanMap® Panel haematologica | 2017; 102(2)


JAK1 inhibitor treatment of myelofibrosis

(Myriad RBM, Austin, TX, USA). Plasma cytokine levels were analyzed using OmicSoft’s Array Studio software (OmicSoft Corporation, Cary, NC, USA). Changes at week 4 relative to baseline were determined and converted to Log2 scale and heat maps were generated using hierarchical clustering. For calculations of the proportion of patients achieving ≥50% reduction from baseline in TSS and those achieving ≥10% reduction from baseline in spleen volume, all patients who had baseline data were included in the analyses. Patients who discontinued treatment before the scheduled time of post-baseline response assessments (week 12 or 24) were considered to be non-responders. All other analyses at specific time-points were based on observed cases, and patients who discontinued treatment before the time of the post-baseline assessment were excluded from analysis at those time-points. Adverse events were assessed in the safety evaluable population, which included all patients who took at least one dose of study medication.

(11 patients, mostly due to a perceived lack of response to therapy as judged by the physician or patient). Both deaths were considered by the investigators to be unrelated to treatment. The mean age of the patients was 64 years and 84% had intermediate-1 (37%) or intermediate-2 (47%) risk myelofibrosis according to the Dynamic International Prognostic Scoring System. The mean spleen volume at baseline was 2442.7 cm3 and the mean hemoglobin concentration was 10.2 g/dL. As shown in Table 1, patients’ demographics and disease characteristics at baseline were similar across cohorts, except for myelofibrosis disease type and platelet counts. A post hoc analysis showed that only baseline platelet count varied significantly (P=0.0111 for differences between dose groups, analysis of variance); the differences in myelofibrosis disease type were not statistically significant (P=0.1745, chi‐square test).

Symptom improvement and spleen volume reduction Results Patients Enrollment is complete and 87 patients have been treated with INCB039110, including 10, 45, and 32 patients in the 100 mg twice-daily, 200 mg twice-daily, and 600 mg oncedaily cohorts, respectively. A total of 10, 42, and 31 patients in the 100 mg twice-daily, 200 mg twice-daily, and 600 mg once-daily cohorts, respectively, were evaluable for the primary endpoint. Overall, 5/10 (50.0%) patients in the 100 mg twice-daily cohort, 16/45 (35.6%) patients in the 200 mg twice-daily cohort, and 10/32 (31.3%) patients in the 600 mg once-daily cohort discontinued treatment prior to week 24. Primary reasons for discontinuation of therapy included adverse events (7 patients), disease progression (6 patients), consent withdrawal (5 patients), death (2 patients), and other reasons

Table 1. Patients’ demographic and disease characteristics at baseline by cohort.

Characteristic

100 mg twice daily (n=10)

200 mg twice daily (n=45)

600 mg once daily (n=32)

Age, mean (range), years 65.9 (56-80) MF type, n (%)* Primary MF 7 (70.0) Post–essential thrombocythemia MF 2 (20.0) Post–polycythemia vera MF 1 (10.0) DIPSS risk, n (%) Intermediate-1 4 (40.0) Intermediate-2 4 (40.0) High 2 (20.0) JAK2V617F-positive, n (%) 5† (55.6) Prior JAK inhibitor therapy, n (%) 4 (40.0) 3 Spleen volume, mean ± SD, cm 2429±1074 Hemoglobin, mean ± SD, g/L 96±14 Platelet count, mean ± SD 109.3±61.2 (range), × 109/L‡ (15-202)

63.2 (35-84)

64.0 (35-81)

28 (62.2) 8 (17.8) 9 (20.0)

13 (40.6) 13 (40.6) 6 (18.8)

18 (40.0) 19 (42.2) 8 (17.8) 31 (68.9) 14 (31.1) 2588±1709 102±20 211.0±192.5 (57-1084)

10 (31.3) 18 (56.3) 4 (12.5) 22 (68.8) 2 (6.3) 2247±1456 105±20 339.7±309.0 (58-1470)

DIPSS: Dynamic International Prognostic Scoring System; MF: myelofibrosis; SD: standard deviation. *P=0.1745 for differences between dose groups (chi‐square test). †Nine patients in the 100 mg twice-daily cohort were evaluable for JAK2V617F status. ‡P<0.05 for differences between dose groups (analysis of variance).

haematologica | 2017; 102(2)

The 100 mg twice-daily cohort was not expanded beyond stage 1 because only two of ten (20.0%) patients achieved the primary endpoint (≥50% reduction in TSS from baseline to week 12). In both patients, ≥50% reduction in TSS was maintained at week 24. The 200 mg twice-daily and 600 mg once-daily cohorts met the criteria for expansion; 15/42 (35.7%) and 10/31 (32.3%) patients, respectively, achieved the primary endpoint. The corresponding proportions of patients with a ���50% reduction in TSS from baseline at week 24 were 28.6% and 35.5% (Figure 1A). Based on data evaluable at the respective time-points, patients in the 200 mg twice-daily cohort had a median reduction in TSS of 45.8% (n=34) and 48.6% (n=25) at week 12 and week 24, respectively, and the corresponding values for the 600 mg once-daily cohort were 37.2% (n=28) and 46.7% (n=23). Median reductions in TSS or spleen volume were not calculated for the 100 mg twice-daily cohort because of the high discontinuation rate in this cohort. The majority of patients with evaluable data at weeks 12 and/or 24 had some degree of symptom improvement (Figure 1B). Two patients experienced percentage increases in TSS outside the range shown due to very low TSS values at baseline; one patient had an increase from 0.29 at baseline to 1.86 at week 12 (representing a 550% increase), and the other had an increase from 2.29 at baseline to 7.14 at week 24 (representing a 212.5% increase). Based on the results of the PGIC questionnaire, many patients experienced substantial symptom improvement within the first 4 weeks of therapy, with 50.0% and 46.9% of evaluable patients in the 200 mg twice-daily and 600 mg once-daily cohorts, respectively, reporting that their myelofibrosis-related symptoms were much or very much improved. Among patients who remained on therapy, improvements on the PGIC were generally maintained throughout the 24-week duration of the study (Online Supplementary Figure S2). For the 200 mg twice-daily and 600 mg once-daily cohorts, median spleen volume reductions for patients with evaluable data were 14.2% (n=37) and 14.5% (n=31), respectively, at week 12 and 17.4% (n=31) and 17.1% (n=24) at week 24. The proportions of patients in the 200 mg twice-daily and 600 mg once-daily cohorts who had a ≥10% reduction in spleen volume at week 12 were 52.3% and 59.4%, respectively; the corresponding proportions at week 24 were 52.3% and 46.9% (Figure 2A). Across all three cohorts, a ≥35% spleen volume reduction was 329


J.O. Mascarenhas et al.

achieved by five patients at week 12 and ten patients at week 24 (Figure 2B).

Safety and tolerability The median (range) exposure to INCB039110 was 102 (23-519) days in the 100 mg twice-daily cohort, 268 (22535) days in the 200 mg twice-daily cohort, and 197 (58343) days in the 600 mg once-daily cohort. The most common non-hematologic adverse events regardless of causality are shown in Table 2; most were grade 1 or 2. Grade ≥3 non-hematologic adverse events that occurred in more than one patient were: pneumonia, dyspnea, and hypertension (3 patients each), and congestive heart failure, rectal hemorrhage, asthenia, pyrexia, urinary tract infection, hyperkalemia, increased alkaline phosphatase, and acute renal failure (2 patients each). These 25 events occurred in 18 unique patients. Although infections were common (44.8%), including upper respiratory tract infections in 19.5% of patients (Online Supplementary Table S1), most were mild or moderate, and only four cases (1 each of bronchitis, folliculitis, Herpes simplex, and urinary tract infection) were considered treatment-related by the investigator. All of these four cases were grade 2 and not considered serious, and all resolved without changes to study treatment. Two patients (both in the 600 mg once-daily cohort) died during the study: a 62-year old patient died of pneumonia after approximately 5 months on therapy, and

A

a 61-year old patient died of unspecified causes potentially related to disease progression after slightly less than 4 months on therapy. Both deaths were considered by the investigator to be unrelated to treatment. New or worsening grade 3 or 4 hematologic adverse events by dose group are shown in Table 3. Overall, 32.5% of patients experienced grade 3 anemia, the majority of whom had grade 2 anemia at baseline (Online Supplementary Table S2), and none experienced grade 4 anemia. New or worsening grade 3 or 4 thrombocytopenia occurred in 24.4% and 4.7% of patients, respectively; many of these patients had grade 2 thrombocytopenia at baseline (Online Supplementary Table S3). One patient discontinued therapy because of grade 3 thrombocytopenia. Twenty-six patients had bleeding events, including two (20.0%), 15 (33.3%), and nine (28.1%) patients in the 100 mg twice-daily, 200 mg twice-daily, and 600 once-daily cohorts, respectively. Of those, six patients had grade 3 bleeding events: rectal hemorrhage (n=2) and upper gastrointestinal hemorrhage, epistaxis, hematoma, and hematochezia (n=1 each). Hematochezia occurred when the patient had a platelet count of <50×109/L; the other bleeding events did not occur in the presence of grade ≥3 thrombocytopenia. There were no grade 4 bleeding events. In general, hemoglobin levels were stable over the course of treatment, regardless of transfusion status at baseline and during the study (Figure 3A-C). Of 48

Figure 1. Treatment effects on total symptom score (TSS). Results are shown by cohort: 100 mg twice daily (orange), 200 mg twice daily (green), and 600 mg once daily (purple). (A) Proportion of patients with a ≥50% improvement in TSS from baseline. Patients without baseline data were not included in the responder analysis; patients who discontinued therapy prior to the week 12 or week 24 visit were considered nonresponders at those time-points. (B) Individual patients’ changes from baseline in TSS.

B

330

haematologica | 2017; 102(2)


JAK1 inhibitor treatment of myelofibrosis

patients who did not require RBC transfusions during the 12 weeks preceding treatment initiation, only three patients required ≥2 units of RBC transfusions during both weeks 1-12 and weeks 13-24. A total of 39 patients required RBC transfusions during the 12 weeks preceding treatment initiation, receiving a median of 4 RBC units. Of those, six (15.4%) patients did not require RBC transfusions during the treatment period. Furthermore, 21/39 (53.8%) patients achieved a ≥50% reduction in RBC units transfused during the study, including 1/3 (33.3%) in the 100 mg twice-daily cohort with a reduction of eight RBC units, 15/24 (62.5%) in the 200 mg twice-daily cohort with a median reduction of three units (range, 2 to 10 units) and 5/12 (41.7%) in the 600 mg once-daily cohort with a median reduction of seven units (range, 2 to 28 units). Mean platelet counts decreased slightly over time in the 100 mg twice-daily cohort and remained stable in the 200 mg twice-daily cohort. In the 600 mg once-daily cohort, mean platelet counts gradually decreased during the 24week study period but remained within normal limits (Figure 3D).

JAK2V617F allele burden and cytokine profile Overall, 68% (58/85) of evaluable patients were JAK2V617F-positive at baseline; the mean JAK2V617F allele bur-

den was 65.6% among these patients (Table 1). Mean percentage changes from baseline (standard deviation) in JAK2V617F allele burden were −4% (14%) at week 12 and −4% (18%) at week 24, with no significant difference in allele burden changes between the 200-mg and 600-mg cohorts at either time-point. Analysis of the expression levels of a large number of plasma markers revealed global changes from baseline to week 4 in cytokine expression patterns in all patients. The plasma levels of a number of key inflammatory markers, such as C-reactive protein, interleukin-6, interleukin-10, CD40 ligand, RANTES, and vascular endothelial growth factor, decreased in most patients following 4 weeks of treatment (Figure 4).

Discussion In this study, patients with myelofibrosis treated with the selective JAK1 inhibitor INCB039110 at doses of 200 mg twice daily or 600 mg once daily experienced clinically meaningful improvements in myelofibrosis-related symptoms at weeks 12 and 24. Although the 100 mg twicedaily cohort was not selected for expansion, two patients experienced a ≥50% reduction in myelofibrosis-related symptoms at week 12, which was maintained at week 24.

A Figure 2. Treatment effects on spleen volume. Results are shown by cohort: 100 mg twice daily (orange), 200 mg twice daily (green), and 600 mg once daily (purple). (A) Proportion of patients with ≥10% reduction in spleen volume from baseline. Patients without baseline data were not included in the responder analysis; patients who discontinued therapy prior to the week 12 or 24 visit were considered non-responders at those time-points. (B) Individual patients’ changes from baseline in spleen volume.

B

haematologica | 2017; 102(2)

331


J.O. Mascarenhas et al.

INCB039110 treatment resulted in modest spleen size reduction, an expected result given the lack of JAK2 inhibition. Notably, INCB039110 overall was less effective in providing spleen size reductions than ruxolitinib and some JAK2 inhibitors tested in phase III clinical trials.11,12,15,16 Few patients in our study experienced ≥35% spleen volume reduction; however, among patients who remained on treatment, the proportion achieving ≥35% spleen volume reduction increased from week 12 to 24. Additionally, more than half of the patients treated with 200 mg twice daily and nearly half of those treated with 600 mg once daily experienced a ≥10% spleen volume reduction at week 24, which was associated with clinically meaningful symptom improvement in patients treated with ruxolitinib in COMFORT-I.17 Consistent with its mechanism of action, INCB039110 had only minimal effects on JAK2V617F allele burden during the 24-week study period. Our findings reinforce the theory that JAK1 inhibition is an important mechanism of symptom control in patients with myelofibrosis. In our study, 28.6% and 35.5% of patients in the 200 mg twice-daily and 600 mg once-daily cohorts, respectively, achieved ≥50% reduction in TSS at week 24. Although studies cannot be directly compared given differences in study design, symptom assessment methods, and patient populations, overall, these percentages are similar to corresponding values reported for fedratinib, a selective JAK2 inhibitor, and pacritinib, a JAK2/FLT3 inhibitor, in large randomized phase III studies. In JAKARTA, a placebo-controlled phase III study of fedratinib in patients with intermediate-2 or high-risk myelofibrosis, the proportion of patients treated with fedratinib who achieved ≥50% reduction in TSS (assessed with the modified Myelofibrosis Symptom Assessment Form) at week 24 was 34% or 36% depending on dose (7% for placebo).15 In PERSIST-1, a phase III study of pacritinib versus best available therapy (excluding ruxolitinib) in patients with intermediate- or high-risk myelofibrosis, 24.5% of patients in the pacritinib arm (6.5% for best available therapy) achieved a ≥50% reduction in TSS (assessed with the Myeloproliferative Neoplasm Symptom Assessment Form) at week 24.16 The propor-

Table 2. Non-hematologic adverse events of any grade (regardless of causality).*

Patients, n (%) Fatigue Nausea Upper respiratory tract infection Constipation Cough Diarrhea Dyspnea Peripheral edema Pyrexia Pain in extremity Abdominal pain Night sweats Vomiting Dizziness *In >10% of all patients.

332

100 mg twice 200 mg twice 600 mg once daily (n=10) daily (n=45) daily (n=32) 3 (30.0) 3 (30.0) 3 (30.0) 1 (10.0) 1 (10.0) 2 (20.0) 1 (10.0) 1 (10.0) 1 (10.0) 2 (20.0) 2 (20.0) 1 (10.0) 0 0

18 (40.0) 9 (20.0) 8 (17.8) 10 (22.2) 10 (22.2) 8 (17.8) 6 (13.3) 9 (20.0) 6 (13.3) 7 (15.6) 5 (11.1) 5 (11.1) 8 (17.8) 6 (13.3)

4 (12.5) 6 (18.8) 6 (18.8) 4 (12.5) 4 (12.5) 5 (15.6) 6 (18.8) 3 (9.4) 5 (15.6) 2 (6.3) 2 (6.3) 3 (9.4) 1 (3.1) 3 (9.4)

tions of patients with this degree of symptom response in JAKARTA, PERSIST-1, and the current study were smaller than the corresponding proportion seen with ruxolitinib in the COMFORT-I study (45.9% versus 5.3% with placebo).11 The degree of symptom response was also high in a phase I/II trial of the JAK1/JAK2 inhibitor momelotinib, in which 60% to 100% of patients experienced ≥50% reductions in some individual symptoms at 3 and 6 months; however, TSS was not assessed.18 Overall, these data support the notion that JAK1 and JAK2 inhibition contribute to improvement in myelofibrosis-related symptoms; however, dual inhibition of JAK1 and JAK2 may provide more effective symptom mitigation than is achievable with inhibition of either JAK isoform alone. Consistent with its effect on symptoms, INCB039110 treatment rapidly led to global reductions in plasma inflammatory makers. Notably, expression levels of Creactive protein, a biomarker of acute inflammation, were markedly reduced in most patients. These changes in inflammatory markers are consistent with those seen in ruxolitinib-treated patients with myelofibrosis11,13 and support the notion that JAK1 inhibition can play a vital role in reducing disease-related inflammation in patients with myelofibrosis. INCB039110 was generally well tolerated, and nonhematologic adverse events were generally grade 1 or 2. Only four patients had infections that were considered treatment related, but of those none was serious or affected study treatment. Immunosuppression, primarily affecting T-cell function, has been offered as an explanation for the occurrence of treatment-related infections in the case of ruxolitinib,19 and it is likely that ruxolitinib-mediated immunosuppression is at least in part attributable to JAK1 inhibition. The potential immunosuppressive effects of INCB039110 have not been investigated. Our findings suggest that selective JAK1 inhibition is not associated with high degrees of myelosuppression. Mean platelet counts at baseline varied among dose groups. In the group of patients who received 100 mg twice daily, the mean platelet count decreased slightly from a low baseline value (109.3×109/L) over the course of the study, whereas it remained relatively stable in the cohort that received 200 mg twice daily. The mean platelet count in the 600 mg once-daily cohort decreased over time from a high baseline value (339.7×109/L) and remained within the normal range over the 24-week period of the study. Given that the study included patients with low platelet counts at base-

Table 3. Summary of grade ≥3 hematologic adverse events.*

Patients, n (%) Anemia Grade 3 Grade 4* Thrombocytopenia Grade 3 Grade 4 Neutropenia Grade 3 Grade 4

100 mg twice daily

200 mg twice daily

600 mg once daily

n=9 3 (33.3) 0 n=9 4 (44.4) 0 n=10 1 (10.0) 0

n=42 16 (38.1) 0 n=45 13 (28.9) 3 (6.7) n=43 2 (4.7) 1 (2.4)

n=32 8 (25.0) 0 n=32 4 (12.5) 1 (3.1) n=32 0 0

*Events were graded based on laboratory data, except for the reporting of grade 4 anemia, which was based on Common Terminology Criteria for Adverse Events v4.03 classification.

haematologica | 2017; 102(2)


JAK1 inhibitor treatment of myelofibrosis

line (≥50×109/L to <100×109/L), it is noteworthy that only one patient discontinued treatment because of thrombocytopenia (grade 3). Mean hemoglobin levels remained relatively stable from baseline to week 24 in the group of 48 patients who did not require transfusions prior to study entry, and only 6% of these patients required two or more units of RBC transfusions during both weeks 1-12 and weeks 13-24 of the study. Importantly, slightly more than half of the patients who required RBC transfusions before beginning INCB039110 treatment experienced clinically meaningful reductions in RBC transfusions while on therapy. In COMFORT-I, patients treated with the JAK1/JAK2 inhibitor ruxolitinib had temporary decreases in hemoglobin values during the first 8 to 12 weeks of therapy, and a decrease in platelet count was identified as a doselimiting adverse effect of ruxolitinib therapy.11,20 The proportions of patients who experienced grade ≥3 anemia and grade ≥3 thrombocytopenia with ruxolitinib were 45.2% and 12.9%, respectively (placebo, 19.2% and 1.3%, respectively); however, very few patients discon-

A

C

tinued therapy because of cytopenias.11 Furthermore, results of the similarly designed phase III JAKARTA study showed high rates of grade ≥3 anemia (fedratinib, 43% and 60%, depending on dose, versus 25% for placebo) and grade ≥3 thrombocytopenia (fedratinib, 17% and 27% versus 9% for placebo).15 These findings from randomized placebo-controlled studies are consistent with myelosuppression as an expected consequence of JAK2 inhibition, as JAK2 signaling is required for normal thrombopoiesis and erythropoiesis. However, according to data from PERSIST-1, presented at the 2015 annual meeting of the American Society of Clinical Oncology, 14.5% and 2.3% of patients in the pacritinib arm had grade 3 or 4 anemia, respectively (versus best available therapy, 12.3% and 2.8%, respectively), and 5.5% and 6.4% had grade 3 or 4 thrombocytopenia, respectively (versus best available therapy, 6.6% and 2.8%, respectively). Although the assessment of hematologic adverse events in PERSIST-1 was based on investigator reports and not on laboratory analyses as in the ruxolitinib and fedratinib studies, these data suggest that the hematolog-

B

D

Figure 3. Mean hemoglobin level and platelet count over time by dose cohort. Mean hemoglobin for (A) all patients, (B) patients who did not require transfusion before study entry, and (C) patients who did not require transfusion before study entry or during the study. (D) Mean platelet count over time.

haematologica | 2017; 102(2)

333


J.O. Mascarenhas et al.

Figure 4. Changes from baseline in plasma cytokine levels at week 4. Individual patients are presented in each column, and each row is a plasma marker. Green denotes decreased levels at week 4 versus baseline and red denotes increased levels at week 4 versus baseline, with the color intensity representing the magnitude of the change on a log2 scale. Gray denotes missing data points that were not available for analysis. bid: twice daily; qd, once daily.

ic toxicity of JAK2 inhibitors may vary substantially, likely due to off-target effects that may offset myelosuppression mediated by JAK2 inhibition. In this context it is worth noting that the JAK1 and JAK2 inhibitor momelotinib has been reported to provide anemia responses.18 Based on preliminary evidence, this benefit may be attributable to the off-target inhibition of activin receptor-like kinase-2 and the consequent reduction of hepcidin production leading to enhanced erythropoiesis.21 In the current study, 32.5% of patients experienced grade 3 anemia (defined by laboratory values) while on study, the majority of whom had grade 2 anemia at baseline, and no patients experienced grade 4 anemia. Grade 3 or 4 thrombocytopenia (defined by laboratory values) occurred in 29.1% of patients, most commonly in the 100 mg twice-daily cohort, which had a median platelet count at baseline of 105.5Ă&#x2014;109/L. In light of its limited hematologic toxicity and its efficacy in myelofibrosis symptom relief, INCB039110 may be useful in the future as part of combination therapy with 334

agents that have complementary activity or dose-limiting hematologic toxicities, such as Hedgehog pathway inhibitors (e.g., sonidegib), phosphatidylinositol 3-kinase inhibitors, histone deacetylase inhibitors (e.g., panobinostat), and hypomethylating agents (e.g., 5-azacytidine).22,23 Combinations involving selective JAK1 inhibition instead of JAK1/JAK2 inhibition may offer the potential for similar efficacy with less myelosuppression, although this has not been confirmed in clinical trials. Although a long-term, placebo-controlled trial or comparative studies with other JAK inhibitors would provide a more complete clinical picture and a deeper understanding of the relative contributions of JAK1 and JAK2 inhibition in myelofibrosis, the findings of this phase II study provide preliminary evidence that selective JAK1 inhibition with INCB039110 can provide effective relief of myelofibrosis-related symptoms with limited hematologic toxicity. Furthermore, the results of this study contribute to a better understanding of JAK1 inhibition as a therapeutic target. haematologica | 2017; 102(2)


JAK1 inhibitor treatment of myelofibrosis

Acknowledgments Medical writing assistance was provided by Roland Tacke, PhD, CMPP, of Evidence Scientific Solutions, Philadelphia, PA, and funded by Incyte Corporation.

References 1. Mesa RA, Green A, Barosi G, Verstovsek S, Vardiman J, Gale RP. MPN-associated myelofibrosis (MPN-MF). Leuk Res. 2011;35(1):12-13. 2. Gregory SA, Mesa RA, Hoffman R, Shammo JM. Clinical and laboratory features of myelofibrosis and limitations of current therapies. Clin Adv Hematol Oncol. 2011;9(9 Suppl 22):1-16. 3. Mesa RA, Schwager S, Radia D, et al. The Myelofibrosis Symptom Assessment Form (MFSAF): an evidence-based brief inventory to measure quality of life and symptomatic response to treatment in myelofibrosis. Leuk Res. 2009;33(9):1199-1203. 4. Mitra D, Kaye JA, Piecoro LT, et al. Symptom burden and splenomegaly in patients with myelofibrosis in the United States: a retrospective medical record review. Cancer Med. 2013;2(6):889-898. 5. Mesa RA, Shields A, Hare T, et al. Progressive burden of myelofibrosis in untreated patients: assessment of patientreported outcomes in patients randomized to placebo in the COMFORT-I study. Leuk Res. 2013;37(8):911-916. 6. Mesa RA, Niblack J, Wadleigh M, et al. The burden of fatigue and quality of life in myeloproliferative disorders (MPDs): an international Internet-based survey of 1179 MPD patients. Cancer. 2007;109(1):68-76. 7. Vannucchi AM, Lasho TL, Guglielmelli P, et al. Mutations and prognosis in primary myelofibrosis. Leukemia. 2013;27(9):18611869. 8. Rampal R, Al-Shahrour F, Abdel-Wahab O,

haematologica | 2017; 102(2)

9.

10.

11.

12.

13.

14.

15.

16.

Funding This study (NCT01633372) was sponsored by Incyte Corporation. MD Anderson receives a cancer center support grant from the NCI of the National Institutes of Health (P30 CA016672).

et al. Integrated genomic analysis illustrates the central role of JAK-STAT pathway activation in myeloproliferative neoplasm pathogenesis. Blood. 2014;123(22):e123133. Quintรกs-Cardama A, Kantarjian H, Cortes J, Verstovsek S. Janus kinase inhibitors for the treatment of myeloproliferative neoplasias and beyond. Nat Rev Drug Discov. 2011;10(2):127-140. Mascarenhas J. Selective Janus associated kinase 1 inhibition as a therapeutic target in myelofibrosis. Leuk Lymphoma. 2015;56(9):1-5. Verstovsek S, Mesa RA, Gotlib J, et al. A double-blind, placebo-controlled trial of ruxolitinib for myelofibrosis. N Engl J Med. 2012;366(9):799-807. Harrison C, Kiladjian JJ, Al-Ali HK, et al. JAK inhibition with ruxolitinib versus best available therapy for myelofibrosis. N Engl J Med. 2012;366(9):787-798. Verstovsek S, Kantarjian H, Mesa RA, et al. Safety and efficacy of INCB018424, a JAK1 and JAK2 inhibitor, in myelofibrosis. N Engl J Med. 2010;363(12):1117-1127. Collier P, Patel K, Waeltz P, et al. Validation of standards for quantitative assessment of JAK2 c.1849G>T (p.V617F) allele burden analysis in clinical samples. Genet Test Mol Biomarkers. 2013;17(5):429-437. Pardanani A, Harrison C, Cortes JE, et al. Safety and efficacy of fedratinib in patients with primary or secondary myelofibrosis: a randomized clinical trial. JAMA Oncol. 2015;1(5):643-651. Mesa RA, Egyed M, Szoke A, et al. Results of the PERSIST-1 phase III study of pacritinib (PAC) versus best available therapy

17.

18.

19.

20.

21.

22.

23.

(BAT) in primary myelofibrosis (PMF), post-polycythemia vera myelofibrosis (PPV-MF), or post-essential thrombocythemia-myelofibrosis (PET-MF). J Clin Oncol. 2015;33(15): Abstract LBA7006. Mesa RA, Gotlib J, Gupta V, et al. Effect of ruxolitinib therapy on myelofibrosis-related symptoms and other patient-reported outcomes in COMFORT-I: a randomized, double-blind, placebo-controlled trial. J Clin Oncol. 2013;31(10):1285-1292. Pardanani A, Laborde RR, Lasho TL, et al. Safety and efficacy of CYT387, a JAK1 and JAK2 inhibitor, in myelofibrosis. Leukemia. 2013;27(6):1322-1327. Parampalli Yajnanarayana S, Stubig T, Cornez I, et al. JAK1/2 inhibition impairs T cell function in vitro and in patients with myeloproliferative neoplasms. British journal of haematology. 2015;169(6):824-833. Verstovsek S, Gotlib J, Gupta V, et al. Management of cytopenias in patients with myelofibrosis treated with ruxolitinib and effect of dose modifications on efficacy outcomes. OncoTargets Ther. 2013;7(1321. Asshoff M, Warr M, Haschka D, et al. The Jak1/Jak2 inhibitor momelotinib inhibits Alk2, decreases hepcidin production and ameliorates anemia of chronic disease (ACD) in rodents. Blood. 2015;126(23): 538-538. Stein BL, Cervantes F, Giles F, Harrison CN, Verstovsek S. Novel therapies for myelofibrosis. Leuk Lymphoma. 2015;56(10):27682778. Mascarenhas J. Rationale for combination therapy in myelofibrosis. Best Pract Res Clin Haematol. 2014;27(2):197-208.

335


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Myeloid Leukemia

Ferrata Storti Foundation

Distinct global binding patterns of the Wilms tumor gene 1 (WT1) -KTS and +KTS isoforms in leukemic cells

Tove Ullmark,1 Linnea Järvstråt,1 Carl Sandén,2 Giorgia Montano,1 Helena Jernmark-Nilsson,1 Henrik Lilljebjörn,2 Andreas Lennartsson,3 Thoas Fioretos,2 Kristina Drott,1 Karina Vidovic,1 Björn Nilsson,1 and Urban Gullberg1

Division of Hematology and Transfusion Medicine, Department of Laboratory Medicine, Lund University; 2Division of Clinical Genetics, Department of Laboratory Medicine, Lund University and 3Department of Biosciences and Nutrition, Karolinska Institutet, Huddinge; Sweden

1

Haematologica 2017 Volume 102(2):336-345

ABSTRACT

T

Correspondence: urban.gullberg@med.lu.se

Received: May 20, 2016. Accepted: September 5, 2016. Pre-published: September 9, 2016. doi:10.3324/haematol.2016.149815

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/336

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

336

he zinc finger transcription factor Wilms tumor gene 1 (WT1) acts as an oncogene in acute myeloid leukemia. A naturally occurring alternative splice event between zinc fingers three and four, removing or retaining three amino acids (±KTS), is believed to change the DNA binding affinity of WT1, although there are conflicting data regarding the binding affinity and motifs of the different isoforms. Increased expression of the WT1 -KTS isoform at the expense of the WT1 +KTS isoform is associated with poor prognosis in acute myeloid leukemia. We determined the genome-wide binding pattern of WT1 -KTS and WT1 +KTS in leukemic K562 cells by chromatin immunoprecipitation and deep sequencing. We discovered that the WT1 -KTS isoform predominantly binds close to transcription start sites and to enhancers, in a similar fashion to other transcription factors, whereas WT1 +KTS binding is enriched within gene bodies. We observed a significant overlap between WT1 -KTS and WT1 +KTS target genes, despite the binding sites being distinct. Motif discovery revealed distinct binding motifs for the isoforms, some of which have been previously reported as WT1 binding sites. Additional analyses showed that both WT1 -KTS and WT1 +KTS target genes are more likely to be transcribed than non-targets, and are involved in cell proliferation, cell death, and development. Our study provides evidence that WT1 -KTS and WT1 +KTS share target genes yet still bind distinct locations, indicating isoform-specific regulation in transcription of genes related to cell proliferation and differentiation, consistent with the involvement of WT1 in acute myeloid leukemia.

Introduction The Wilms tumor gene 1 (WT1) was discovered as a tumor suppressor in the pediatric kidney malignancy Wilms tumor,1 and later identified as an oncogene in many solid tumors and in leukemia.2 WT1 mutations occur in 10% of cases of acute myeloid leukemia (AML), and carry a poor prognosis.3-6 Wild-type WT1 is expressed in acute leukemia cells,2 and high expression correlates with a poor outcome in AML.7-9 The ubiquitous overexpression in leukemia has spurred the development of vaccines aimed at raising a T-cell response against WT1 for therapeutic purposes.10 In mice, overexpression of WT1 induces myeloproliferation, while overexpression of WT1 plus the RUNX1/RUNX1T1 (AML1-ETO) fusion gene induces AML.11 Four major WT1 isoforms have been identified. These result from two alternative splicing events, leading to inclusion/exclusion of 17 amino acids (17AA) from exon five, and inclusion/exclusion of the three amino acids KTS between zinc fingers three and four. The KTS insertion is thought to hamper DNA binding, as it might displace zinc finger four and thereby decrease DNA binding affinity.12 A perturbed ratio between isoforms may contribute to pathological function; increased expression of haematologica | 2017; 102(2)


Global binding patterns of the WT1 -KTS and +KTS isoforms

WT1 +17AA/ -KTS relative to WT1 +17AA/+KTS correlates with resistance to therapy in AML.13 Additionally, high expression of WT1 +17AA/-KTS has been linked to a higher risk of relapse.14 Despite these observations, the functional differences between WT1 isoforms remain unknown. Several DNA binding motifs have been identified for WT1 (listed in Online Supplementary Table S1). No motif has been conclusively shown to bind only WT1 -KTS or only WT1 +KTS protein, and there are conflicting results regarding the binding affinity and the ability to influence target gene transcription of the different isoforms. In the present study, we carried out the first analysis of the global DNA binding patterns of WT1 -KTS and WT1 +KTS isoforms in leukemic cells. Our analysis shows that the two isoforms bind to many shared target genes and that binding of either isoform is correlated with active transcription, yet the binding sites of the isoforms are distinct.

Methods Chromatin immunoprecipitation and streptavidin capture For establishment of K562 clones expressing biotinylated WT1 -KTS or WT1 +KTS, we used a previously published protocol15 with some modifications as described in detail in the Online Supplementary Methods. Briefly, cells were transfected with E. coli biotin protein ligase (BirA) enzyme and the respective isoform of WT1. Two clones were then chosen for subsequent chromatin immunoprecipitation (ChIP) analysis, aiming for expression levels of biotinylated WT1 that were comparable with those of wildtype WT1. ChIP and streptavidin capture were performed with nuclear extracts from the K562 clones expressing BirA and tagged WT1 +17AA/-KTS or WT1 +17AA/+KTS, and from the K562 clones expressing BirA only as background control. Further methodological details are given in the Online Supplementary Methods.

Library preparation and sequencing From immunoprecipitated DNA, libraries were prepared using the ThruPlex DNA-seq kit (Rubicon Genomics, Ann Arbor, Michigan, USA). Library purification removed the fragments above 700 bp, and below 200 bp. The sequencing was performed on a NextSeq 500 sequencer (Illumina, San Diego, CA, USA) using Illuminaâ&#x20AC;&#x2122;s NextSeq 500/550 High Output Kit v2 (75 cycles). Some samples were sequenced on an Illumina platform at the Science for Life Laboratory core facility in Uppsala, Sweden. Further methodological details are given in the Online Supplementary Methods. Data for all samples are available in the Gene Expression Omnibus (GEO) under accession number GSE81009.

Data analysis Data were analyzed as described by Sanden et al.16 with some modifications and extensions as detailed in the Online Supplementary Methods.

Results WT1 -KTS peaks are enriched close to transcription start sites, and WT1 +KTS peaks inside gene bodies To characterize the genome-wide binding of WT1 -KTS and WT1 +KTS we performed streptavidin capture haematologica | 2017; 102(2)

of the biotinylated proteins with attached chromatin followed by deep sequencing (ChIP-Seq). Two independent experiments yielded 2,009 overlapping peaks of WT1 binding for WT1 -KTS, and three independent experiments yielded 21,831 overlapping binding peaks for WT1 +KTS. According to ENCODE-recommendations for background control in ChIP-Seq experiments with tagged proteins,17 the clone containing BirA only and no tagged protein, from which both clones expressing tagged WT1 were made, was sequenced after streptavidin capture to the same depth as the other samples. Three independent experiments yielded only five recurrent peaks, none of which matched peaks from our WT1 samples. From this background analysis we conclude that background binding is negligible. Analysis of the distance from the center of each WT1 -KTS peak to the closest transcription start site (TSS) revealed that the WT1 -KTS binding sites are primarily located around the TSS of genes, with as many as 26% of all peaks being located within 500 bp, and 42% within 2,000 bp, up- or down-stream of a TSS (Figure 1A). The preferential localization of WT1 -KTS peaks close to the TSS was mirrored by the relationship of the WT1 -KTS peaks to distinct gene elements. This analysis showed that the peaks are found primarily within or immediately adjacent to genes, with a strong overrepresentation of binding to the 2 kb just upstream of the TSS, to exons (particularly first exons), and to first introns and junctions between exons and introns (Figure 1C). Fewer intergenic peaks were found than could be expected from a comparison with a randomly generated set of 30,000 control genomic locations (Figure 1C). Taken together, our data demonstrate preferential binding of WT1 -KTS close to TSS, consistent with a transcription factor role for WT1 -KTS. All differences, as compared to the randomized genomic positions, with the exception of that for the 2-30 kb upstream of the TSS and the 2-5 kb downstream of the transcription end, were statistically significant (P<0.001). For WT1 +KTS, a minor enrichment around the TSS could also be seen (Figure 1C), but it was not as prominent as for WT1 -KTS. WT1 +KTS peaks were enriched, as compared to the 30,000 random locations, within the exons of target genes, the intron-exon junctions, and to a lesser extent in introns, up to 30 kb upstream of the TSS, and up to 5 kb downstream of the transcription end (Figure 1c). All differences, as compared to the random positions, were statistically significant (P<0.001). To investigate whether or not WT1 binds enhancer regions, we utilized a set of 43,011 enhancer regions defined by the FANTOM consortium after cap analysis of gene expression (CAGE) analyses.18 Upon comparison of our WT1 peaks with the defined enhancer regions, we found that WT1 -KTS peaks clearly co-localize with enhancers (7.3% of the WT1 -KTS peaks coincided with these enhancer regions), but that only 0.7% of the WT1 +KTS peaks did so (Figure 1D). The small WT1 +KTS enrichment was, however, statistically significant. We conclude that WT1 -KTS binds enhancer regions to a much larger extent than does WT1 +KTS.

WT1 -KTS peaks show more similarity than WT1 +KTS peaks to other transcription factor tracks We compared our WT1 peaks to tracks of peaks from other factors in the ENCODE database.19,20 The similarity 337


T. Ullmark et al.

of our peaks and the ENCODE track under comparison is based on nucleotide bases present in both peak tracks and as such describes a strong similarity between the binding patterns.16 Table 1 lists the 20 tracks that were most alike the WT1 -KTS and WT1 +KTS peaks. All tracks used in the comparison were derived from the K562 cell line and are available in the ENCODE database (Online Supplementary Table S5A provides a complete list of all ENCODE tracks showing similarity to those of WT1 â&#x20AC;&#x201C;KTS, while Online Supplementary Table S5B lists those for WT1 +KTS). For WT1 -KTS, the most similar track was the one of EGR-1, which is not surprising considering that the two proteins can bind the same DNA sequences.21 The 20 tracks that most resembled the WT1 -KTS peaks included other transcription factors, consistent with our observation that WT1 -KTS preferentially binds close to TSS and in enhancer regions. Interestingly, no track in the ENCODE database showed a high degree of similarity to the WT1 +KTS

peaks. There were, however, several tracks for which the similarity, though small, was significant. The transcription factor track most resembling that of WT1 +KTS, was Zinc Finger Protein 263 (ZNF263), which, like WT1 +KTS, was found to bind to a large extent within its target genes rather than to promoter regions.22 Methylation and acetylation of lysine residues on histones can contribute to activation or repression of gene transcription. In some cases, the modifications appear to be secondary to transcription rather than contributing to it. Thus, the histone modifications can be used in ChIPSeq experiments as markers not only for chromatin states, but also for transcriptional activity. Trimethylation of histone H3 at lysine 27 (H3K27me3) is a well-established mark of transcriptional repression and heterochromatin that is regulated by the Polycomb complex 2.23 On the top 20 list of the tracks that most resemble WT1 +KTS, five tracks are for this repressive histone mark. There are two more histone marks on the WT1 +KTS list, one track for trimethylation of histone H3 on

A

B

C

D

Figure 1. WT1 -KTS, and to a lesser extent WT1 +KTS, peaks are enriched around the transcription start sites of target genes and in enhancers. After chromatin immunoprecipitation, sequencing and peak calling, the peak centers were annotated in reference to the closest transcription start site in the genome. The graphs depict the distribution of (A) WT1 -KTS peaks and (B) WT1 +KTS peaks, (distance from the closest TSS in base pairs). In (C) all peaks are shown in reference to gene bodies in the genome. (D) Fraction of peaks that correspond to enhancer areas, as defined by CAGE data.18 "Randomized" refers to 30,000 positions randomly distributed across the genome. (TSS: transcription start site; TE: transcription end).

338

haematologica | 2017; 102(2)


Global binding patterns of the WT1 -KTS and +KTS isoforms

lysine 36 (H3K36me3), which is considered a marker for the gene body of actively transcribed genes23 and three tracks for monomethylation of histone H3 on lysine 4 (H3K4me1), a histone mark of active enhancers and gene bodies of actively transcribed genes.23 Furthermore, an

open chromatin track comes first on the list of most similar tracks. Together, these histone marks reveal WT1 +KTS as a factor that co-localizes with both repressive and activating histone marks, perhaps indicating a potential for both repression and activation of target genes.

Table 1. Similarity between WT1 peaks and ENCODE ChIP-Seq tracks.

WT1 -KTS ENCODE tracks wgEncodeHaibTfbsK562Egr1V0416101PkRep1.strack wgEncodeHaibTfbsK562Cbx3sc101004V0422111PkRep1.strack wgEncodeHaibTfbsK562Zbtb7asc34508V0416101PkRep2.strack wgEncodeOpenChromDnaseK562G1phasePk.strack wgEncodeOpenChromDnaseK562G2mphasePk.strack wgEncodeOpenChromDnaseK562PkV2.strack wgEncodeHaibTfbsK562Hey1Pcr1xPkRep1.strack wgEncodeHaibTfbsK562MaxV0416102PkRep2.strack wgEncodeOpenChromDnaseK562SahactrlPk.strack wgEncodeOpenChromDnaseK562Saha1u72hrPk.strack wgEncodeOpenChromDnaseK562NabutPk.strack wgEncodeOpenChromDnaseK562Pk.strack wgEncodeHaibTfbsK562MaxV0416102PkRep1.strack wgEncodeHaibTfbsK562E2f6sc22823V0416102PkRep2.strack wgEncodeHaibTfbsK562E2f6V0416102PkRep2.strack wgEncodeOpenChromChipK562CmycPk.strack wgEncodeHaibTfbsK562Hey1Pcr1xPkRep2.strack wgEncodeOpenChromSynthK562Pk.strack wgEncodeOpenChromFaireK562Pk.strack

Track description

wgEncodeBroadHistoneK562H2azStdPk.strack

Egr1 transcription factor Cbx3 heterochromatin component Zbtb7 transcription factor Open chromatin DNaseI G1 phase cells Open chromatin DNase I G2/M phase cells Open chromatin DNase I Hey1 transcription factor Max transcription factor Open chromatin DNase I SAHA control (only DMSO) treatment Open chromatin DNase I SAHA (Vorinostat) 1 uM 72 hours Open chromatin DNase I sodium butyrate 72 hours Open Chromatin DNase I Max transcription factor E2f6 transcription factor E2f6 transcription factor Cmyc transcription factor Hey1 transcription factor Open chromatin (synthesis of FAIRE, ChIP tracks and DNase I) Open chromatin (formaldehyde assisted isolation of regulatory elements) H2az (histone 2 variant)

WT1 +KTS ENCODE tracks

Track description

wgEncodeOpenChromSynthK562Pk.strack wgEncodeHaibMethyl450K562SitesRep1.strack wgEncodeSydhTfbsK562Znf263UcdPk.strack wgEncodeBroadHistoneK562H3k4me1StdPk.strack wgEncodeBroadHistoneK562Chd1a301218aStdPk.strack wgEncodeBroadHistoneK562CtcfStdPk.strack wgEncodeSydhHistoneK562bH3k4me1UcdPk.strack wgEncodeSydhHistoneK562H3k4me1UcdPk.strack wgEncodeUwHistoneK562H3k27me3StdPkRep1.strack wgEncodeUwHistoneK562H3k27me3StdPkRep2.strack wgEncodeUwHistoneK562H3k27me3StdPkRep2.strack wgEncodeSydhHistoneK562bH3k27me3bUcdPk.strack wgEncodeSydhHistoneK562H3k27me3bUcdPk.strack wgEncodeBroadHistoneK562Cbx3sc101004Pk.strack wgEncodeHaibTfbsK562Mef2aV0416101PkRep1.strack wgEncodeSydhTfbsK562Znf143IggrabPk.strack wgEncodeHaibTfbsK562Pu1Pcr1xPkRep1.strack wgEncodeSydhTfbsK562Atf3StdPk.strack wgEncodeUwHistoneK562H3k36me3StdPkRep1.strack wgEncodeSydhTfbsK562Gata2UcdPk.strack

Open chromatin (synthesis of FAIRE, ChIP tracks and DNase I) Methylated DNA Znf263 transcription factor H3K4me1 histone modification Chd1 chromatin remodeller Ctcf H3K4me1 histone modification H3K4me1 histone modification H3K27me3 histone modification H3K27me3 histone modification H3K27me3 histone modification H3K27me3 histone modification H3K27me3 histone modification Cbx3 heterochromatin component Mef2a Znf143 Pu1 Atf3 H3K36me3 histone modification Gata2

Similarity score

P value

0.108274 0.058714 0.056838 0.04972 0.049081 0.048757 0.048251 0.047713 0.047693 0.047583 0.045986 0.044686 0.042228 0.041029 0.041029 0.040484 0.037512 0.035282 0.034536

9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04

0.028038

9.99E-04

Similarity score

P value

0.003739 0.003115 0.002099 0.001946 0.001739 0.001564 0.001531 0.001531 0.001092 0.001065 0.001065 0.00094 0.00094 0.000936 0.000692 0.000608 0.000549 0.000548 0.000535 0.000502

9.99E-04 9.99E-04 9.99E-04 2.50E-02 7.99E-03 9.99E-03 9.99E-04 9.99E-04 9.99E-04 9.99E-04 9.99E-04 8.99E-03 8.99E-03 9.99E-04 9.99E-04 2.00E-03 2.00E-03 3.80E-02 9.99E-04 2.00E-03

The score for similarity between our peaks and those of ENCODE K562 tracks, and the P values, were calculated as described in the Methods section.

haematologica | 2017; 102(2)

339


T. Ullmark et al.

Another indication that this could be the case is that WT1 +KTS, but not WT1 -KTS, significantly co-localizes with a track for methylated DNA in Table 1. The positive correlations for WT1 -KTS to DNA methylation are far down the list of similar tracks, and the similarity score for the top hit is better for WT1 +KTS. In contrast to the WT1 +KTS isoform, the WT1 -KTS track does not match the repressive H3K27me3 mark (where five H3K27me3 tracks are positive for WT1 +KTS, only one is positive for WT1 -KTS and it scores at position 207 in similarity to WT1 –KTS). WT1 -KTS is also a worse match for H3K36me3 (only 1 track is positive for WT1 –KTS as compared to 2 for WT1 +KTS, and it scores at position 213). Rather, WT1 -KTS shows greater similarity with multiple open chromatin tracks as well as the H3K4me1 mark (3 tracks for H3K4me1 are positive for each isoform, but WT1 –KTS has a higher similarity score). The H3K4me3 mark has five tracks positive for co-localization with WT1 –KTS and two very low-scoring positives for WT1 +KTS, consistent with this mark and WT1 –KTS both appearing mainly around the TSS, which is not the case for WT1 +KTS. Online Supplementary Table S5A gives a complete list of all tracks showing similarity to those of WT1 –KTS and Online Supplementary Table S5B to those of WT1 +KTS. In summary, both transcription factors co-localize with histone marks for active transcription, but only WT1 +KTS peaks co-localize with a transcriptional repression mark. The binding pattern of WT1 -KTS is similar to those of other transcription factors, whereas the binding pattern of WT1 +KTS is not. This observation is consistent with our previous finding, that WT1 +KTS localizes only to a small extent near the TSS and to enhancer regions.

WT1 -KTS and WT1 +KTS bind different locations within the same genes To avoid uncertain gene annotations we limited our investigation of target genes to the genes in which the peak was situated within the promoter area, defined as the 2,000 bp upstream of the TSS, or within the gene body. For WT1 -KTS, 1,141 unique target genes could be identified in this way, and for WT1 +KTS 5,775 genes. Of these, 508 genes were found to be shared by the two isoforms, a highly significant overlap (Figure 2). Not surprisingly, considering the differences in general binding pattern, the peaks of the WT1 -KTS and WT1 +KTS isoforms on these shared target genes rarely coincided. Only for eight (1.6%) of the genes, did the binding sites overlap, and only for a further 23 (4.5%) of the genes, was there a WT1 -KTS and a WT1 +KTS peak within 1 kb of each other. Upon closer inspection of the target genes that we identified to be common to both isoforms, we found that they include several previously known WT1 targets, including BCL2-Like 1 (BCL2L1/BCLXL),24 DNA (Cytosine-5)-Methyltransferase 3 Alpha (DNMT3A),25 Platelet-Derived Growth Factor Alpha Polypeptide (PDGFA),26 SMAD Family Member 3 (SMAD3),27 and Vitamin D Receptor (VDR).28 In conclusion, among our identified target genes that include previously known targets, WT1 -KTS and WT1 +KTS isoforms bind, to a large extent, to the same target genes, but in distinct locations, suggesting at least partially different molecular mechanisms. 340

Known motifs of WT1 binding are found within peaks To investigate the binding pattern of the WT1 isoforms further, we searched the peaks for predicted WT1 binding sites, using the FIMO tool and the two WT1 motif position weight matrices from the TRANSFAC database.37,38 The first motif, called GCGGGGGGGT, with a matrix for a total of 17 bases, was found at least once within 44% of WT1 -KTS peaks, while the second motif, called GGGGshort, consisting of a matrix for 12 bases, was even more frequent and was present in 78% of WT1 -KTS peaks (Figure 3). For WT1 +KTS, the motif frequencies were much lower, 17% and 21%, respectively. When the motif search was performed on the entire genome both motifs proved so scarce that less than 1% of the WT1 peaks would be expected to have motif occurrence, making both WT1 -KTS and WT1 +KTS peaks highly enriched for the TRANSFAC WT1 motifs (Figure 3). To search for enrichment of other motifs as well, we used the DREME software that performs unbiased searches for nucleotide patterns within the peaks. Table 2 lists the top ten identified motifs ranked according to significance of enrichment among WT1 -KTS peaks (all significantly enriched motifs are listed in Online Supplementary Table S2, and a list of previously published WT1 motifs is given in Online Supplementary Table S1). Motifs number 1, 2, 4, 8 and 11 found for WT1 -KTS are matches for four different previously published motifs21,29-31 (see alignment in Online Supplementary Figure S3). Several of the enriched motifs in Table 2 are annotated to EGR-1, which is not surprising since the three zinc fingers of EGR-1 are homologous to number two through four or WT1’s four zinc fingers,21 and most WT1 motifs defined so far are EGR-1-like. The TOMTOM database did not contain a WT1 motif, explaining why no WT1 annotations can be seen in Table 2. Taken together, these motif analyses reveal enrichment in the WT1 -KTS peaks of several different motifs previously found to bind the WT1 protein. For WT1 +KTS, only one among the enriched motifs listed in Table 3, motif number 3, is a match for a core of a previously known motif31 (the complete list of significantly enriched motifs is given in Online Supplementary

Figure 2. WT1 -KTS and WT1 +KTS have highly significant overlap in target genes. Taking into account only genes with at least one peak within the promoter area or the gene body, 44.5% of the WT1 -KTS target genes are also target genes of WT1 +KTS.

haematologica | 2017; 102(2)


Global binding patterns of the WT1 -KTS and +KTS isoforms

Table S2, the core sequence in Online Supplementary Table S1, and alignment in Online Supplementary Figure S3). This core motif was found in 50% of our WT1 +KTS peaks. In contrast to WT1 -KTS, EGR-1 binding sites were not found among the WT1 +KTS peaks, consistent with a recurrent hypothesis of WT1 +KTS not binding the EGR-1 motif. None of the enriched motifs for WT1 +KTS was identified as the binding motif for any other transcription factor, but further investigations are required in order to determine whether the enriched motifs bind directly to WT1 +KTS, or might instead bind a protein partner. We conclude that the two TRANSFAC WT1 motifs are highly enriched in our WT1 peaks. These motif occurrences were more numerous for WT1 -KTS than for WT1 +KTS. Apart from the TRANSFAC motifs, we also found four previously published motifs enriched within our WT1 -KTS peaks, and one previously published motif within our WT1 +KTS peaks.

WT1 binds to genes related to development and oncogenic processes To understand the function of the target genes better, we performed a Gene Ontology (GO) analysis with a GO Slim Mapper tool. Results from the analysis mode ”process” are presented in Online Supplementary Table S3 and those for ”function” in Online Supplementary Table S4. Again, we limited the analysis to genes with peaks within the promoter or gene body areas. The GO groups found for the two isoforms are very similar. Several categories found by the GO analysis describe processes previously known to be influenced by WT1. WT1’s role in embryonic development and in the transitions between mesenchymal and epithelial states necessary for embryonic development32 is consistent with enrichment of genes in groups such as anatomical structure development, but also groups such as cell motility, cell junction organization and cytoskeleton organization. WT1 is indispensable for the embryonic development of

A

reproductive organs,33 and the genes in the GO reproduction group were also enriched in our peaks. Several gene groups are consistent with a role in oncogenesis, including cell division, proliferation, response to stress, cell-cell signaling, and cell death. This enrichment indicates that WT1 has the ability to influence many of the properties of malignancy through transcriptional regulation of target genes, consistent with the well-documented role for WT1 in tumor development.2,32,34 Another interesting group is nucleocytoplasmic transport. Accumulating evidence suggests that transport from the nucleus to cytoplasm is tightly regulated and at least in some instances selective for certain transcripts,35 potentially providing WT1 with further possibilities to influence protein production. Enrichment of groups such as transcription factor activity, transcription factor binding, and histone binding indicates that WT1, through its target genes, might also have the ability to further influence transcription downstream, and possibly also splicing. In summary, the target genes of both WT1 isoforms are found in GO groups for embryonic development and homeostasis, as well as functions central to oncogenesis such as cell death, cell proliferation and differentiation.

WT1 -KTS and WT1 +KTS target genes are more likely to be transcribed than non-target genes Making use of FANTOM consortium CAGE data on gene expression (http://fantom.gsc.riken.jp/5; accessed 27 January 2016),36 we compared the expression pattern of all genes in K562 cells with that of our identified target genes of the WT1 isoforms, as described in the Online Supplementary Methods. Both WT1 -KTS and WT1 +KTS target genes were enriched among the intermediately expressed genes, at the expense of the non-expressed gene group (Figure 4). For WT1 -KTS there was also an enrichment in the highly expressed group. These data are consistent with the notion of WT1 as an activator of transcription.

B

Figure 3. WT1 motif occurrence differs between isoforms. We performed a search within our WT1 -KTS and WT1 +KTS peaks for two known WT1 motif matrices from the TRANSFAC database.37,38 (A) The fraction of peaks with at least one occurrence of the GCGGGGGGGT motif. (B) The fraction of peaks with at least one occurrence of the GGGG short motif.

haematologica | 2017; 102(2)

341


T. Ullmark et al. Table 2. Unbiased motif search within the WT1 -KTS peaks identified several WT1 motifs. 1 2 3 4 5 6 7 8 9 10

Motif

E-value

Peaks with motif

GYGKGGGM CCTCCYMC ARATA ACTCCCAC GGAARY TGATWA ABAAA CCCCGCCC CCGCCKCC GCYGGGA

1.30E-267 9.50E-38 4.30E-31 4.30E-31 3.80E-10 1.70E-09 5.00E-08 2.50E-05 8.40E-04 3.40E-03

89.9% 44.1% 26.5% 4.2% 58.4% 7.7% 47.3% 14.5% 16.0% 23.0%

TOMTOM matches EGR1, EGR2 SP1, ZNF263, EGR1, SP2 EGR1, SP1, KLF5, SP2, KLF4, EGR2, E2F4, E2F6, E2F3, E2F1, KLF1 -

Using DREME software, the WT1 -KTS peaks were analyzed to find short nucleotide sequences enriched in our material as compared to the entire genome. The top ten most enriched motifs of a total of 13 are listed. The enriched sequences were then investigated using TOMTOM software to identify proteins whose DNA binding motif matched the sequences. Sequences number 1 (Nakagama 1995),29 2 (Wang 1990, Bickmore 1992),30,31 4 (Nakagama 1995)29 and 8 (Rauscher 1990)21 match previously published WT1 motifs (for alignment see Online Supplementary Figure S3).

Trimethylation of histone H3 at lysine 4 (H3K4me3) is a good marker for actively transcribed genes, having the ability to recruit parts of the transcription complex directly.23 We, therefore, conducted a ChIP-Seq analysis of the H3K4me3 pattern and compared it to that of our WT1 target genes. In our investigation, 60% of the WT1 -KTS target genes in the K562 cells had a H3K4me3 peak center within 1 kb of the TSS, as compared to 41% of all genes in the tagged WT1 -KTS/BirA K562 cells (Figure 5). For WT1 +KTS, 69% of the target genes were positive for H3K4 trimethylation around the TSS, as compared to 59% of all genes in the tagged WT1 +KTS/BirA K562 cells (Figure 5). Both differences were statistically significant. We conclude that the target genes of both WT1 -KTS and WT1 +KTS are more highly expressed than genes not bound by WT1, as shown by CAGE transcription data from K562 cells and by the H3K4me3 mark of active transcription.

Discussion WT1 is an oncogene recurrently mutated and overexpressed in AML. Here, we report the findings of the first ChIP-Seq of WT1 performed in human cells, the very first in hematopoietic cells, as well as the first that is isoform specific. Our investigation revealed that WT1 -KTS and WT1 +KTS bind specifically to partly overlapping, partly unique, sets of target genes in leukemic cells. Both isoforms, but WT1 -KTS much more than WT1 +KTS, bind the TRANSFAC37,38 motifs; however, the alternative motifs identified are not shared. The ability of WT1 +KTS to bind WT1 -KTS target sequences has been the subject of much research and discussion, and our results indicate that the two proteins bind quite separately. This almost complete lack of shared peaks cannot be fully accounted for by the motif analysis, and might instead be due to smaller differences between motifs still fitting the same motif matrix, to differences in protein partners, and/or to differences in flanking sequences. One ChIPchip (chromatin immunoprecipitation coupled with DNA microarrays) and three ChIP-Seq investigations into the target genes of WT1, although not distinguishing WT1 -KTS and WT1 +KTS from each other, have been 342

Table 3. Unbiased motif search within the WT1 +KTS peaks identified one known WT1 motif.

1 2 3 4 5 6 7 8 9 10

Motif

E-value

Peaks with motif

TOMTOM matches

TAWTTTTW TARTCCCA GAGGCBGA CAGGAGAW GGKTTCAY GTKAGCCR GTAGAGAY CGCCCGSC GCTACTY TGCAGTGR

1.4e-2917 1.5e-2706 1.0e-2550 2.4e-2412 1.4e-2463 6.7e-2119 1.7e-2064 2.2e-1871 1.8e-1663 8.4e-1946

35.5% 40.0% 50.3% 42.5% 42.0% 42.2% 31.9% 35.3% 38.9% 36.2%

-

Using DREME software, the WT1 +KTS peaks were analyzed to find short nucleotide sequences enriched in our material as compared to the genome. The top ten most enriched motifs of a total of 26 are listed. The enriched sequences were then investigated using TOMTOM software to identify proteins whose DNA binding motif matched the sequences. For neither of the sequences was a matching motif found within the TOMTOM database. Sequence number 3 matches a previously published WT1 motif (Bickmore 1992)31 (for alignment see Online Supplementary Figure S3).

published.39-42 All studies were performed on mouse kidney cells. The motifs found to be most enriched for the WT1 -KTS isoform in our investigation are similar to those found in the four previous studies, closely resembling the established EGR-1 and WTE motifs. Our distribution of -KTS peaks, centering around the TSS, resembles the results of Dong et al.41 and Kann et al.,42 with comparable concentration in the promoter area (23.1% in our investigation as compared to 21% found by Kann et al. and 16.9% by Dong et al.) although due to differences in the analysis software exact comparisons are not possible. Motamedi et al.40 found a different distribution of peaks, with only 20% of peaks located within 5 kb of a TSS, whereas we found 46.5% of all -KTS peaks in this area. Moreover, several of the GO groups enriched in our investigation were also found to be enriched in the earlier ChIP-Seqs and ChIP-chip investigations into WT1, such as cytoskeleton and adhesion,39,41,42 cell cycle,39,42 and development.39 We found that the WT1 -KTS isoform binds as a traditional transcription factor around the TSS and to known haematologica | 2017; 102(2)


Global binding patterns of the WT1 -KTS and +KTS isoforms

enhancer elements, whereas the WT1 +KTS binding pattern is enriched predominantly within gene bodies, with no other transcription factors closely resembling its binding pattern. With the exception of EGR-1, which binds some of WT1’s target sequences, the proteins whose binding most closely resemble that of WT1 -KTS are ZBTB7, HEY1 and MYC and its binding partner MAX. However, binding motifs annotated for these transcription factors did not appear in our analysis (Table 2 and Online Supplementary Table S2). The reason for this is unclear. One explanation could be that while peaks of different tracks overlap (and thus score as similar tracks), the overlap may not necessarily contain the binding motifs (and thus not be detected in the motif-analysis). Another possibility is indirect binding to DNA, suggesting that they are possible protein partners of WT1, inviting further investigation. Interestingly, MYC is a known target gene of WT1,43 although we did not find any peaks in our material that can be safely assigned to the MYC gene. HEY1 has also been found to be a WT1 target gene in murine cells.39 The combination of a binding partner and target gene of WT1 has been described for Zinc Finger 224 (ZNF224).44,45 Intriguingly, we found ten times more peaks for the WT1 +KTS isoform, as compared to the number of WT1 -KTS peaks. During peak calling, we noticed that the fewer peaks obtained with WT1 -KTS were higher, as compared to the larger number of peaks obtained with WT1 +KTS, suggesting that the occupancy of WT1 was higher at WT1 -KTS binding sites than at the binding sites of WT1 +KTS. This could indicate that WT1 -KTS isoforms are more concentrated on fewer high affinity sites, while WT1 +KTS isoforms are more dispersed on a larger number of binding sites with lower affinity. This speculative notion would be consistent with some pub-

lished data showing lower affinity to DNA for the WT1 +KTS isoform, but further experiments are needed to draw definitive conclusions. For our identified WT1 –KTS and WT1 +KTS target genes, both CAGE data36 from K562 cells, and the H3K4me3 pattern around TSS, indicate that the target genes are more likely to be transcribed than non-target genes. This pattern suggests a mostly activating function for both isoforms, consistent with data from two of the earlier ChIP-Seq investigations41,42 in which WT1 was found to be predominantly activating in mouse kidney glomeruli cells. When, instead, we analyzed the localization of all the WT1 peaks from our investigation, and compared the localizations with those of peaks from ChIP-Seq investigations present in the ENCODE database, we found that tracks for methylated DNA and the repressive histone mark H3K27me3 overlapped with our WT1 +KTS peaks to the extent of becoming top similarity hits. This was not the case for our WT1 –KTS peaks. Both isoforms are known from previous reports to be capable of both activation and repression.43,46-50 Our finding could indicate that, even though the dominant influence of both isoforms on the target genes seems to be activating, in our cellular context WT1 +KTS co-localizes more often than WT1 –KTS with repressive complexes. Importantly, it should be pointed out that our investigation was restricted to WT1 +17AA isoforms. The 17AA isoform could significantly modify the transactivating properties of WT1, and although an influence of the 17AA isoform domain on DNA binding is not supposed, it cannot be totally excluded. How could WT1 +KTS, binding within gene bodies, affect transcription? Our own ChIP-Seq analysis shows that WT1 +KTS binding is only weakly enriched within enhancers (enhancers as defined by CAGE), compared to

Figure 4. WT1 target genes are enriched among the expressed genes in K562 cells. Comparing CAGE data in tags per million (TPM), from the publicly available FANTOM5 consortium datasets,36 of all genes in K562 cells with the genes bound by WT1 -KTS and WT1 +KTS, the WT1 target genes were significantly enriched among the expressed genes at the expense of the silenced genes. Expression levels of the target genes of the two isoforms also differed significantly from each other. The P-values refer to differences of gene distribution in expression bins for all genes in K562 cells, -KTS target genes, and +KTS target genes, respectively.

Figure 5. Genes with a H3K4me3 peak around the transcription start site (TSS) are enriched among the WT1 target genes for both -KTS and +KTS. Compared to all genes, a larger proportion of WT1 -KTS and +KTS target genes had a H3K4me3 ChIP-Seq peak, a mark of active transcription, within 1 kb of the TSS.

haematologica | 2017; 102(2)

343


T. Ullmark et al.

the strong enrichment of WT1 -KTS peaks (Figure 1D), arguing against WT1 +KTS binding to enhancers (as defined by CAGE). Nevertheless, target genes of WT1 +KTS are highly enriched for H3K4me3 around TSS (Figure 5), although WT1 +KTS peaks and H3K4me3 peaks do not overlap. Moreover, H3K4me1 and H3K36me3 score high in the list of tracks similar to WT1 +KTS (Table 1). These observations possibly indicate that WT1 +KTS binds within the gene body in an enhancer-like fashion, affecting chromatin modification also around the TSS to enhance transcription. Many of the GO groups enriched in our study indicate functions central to oncogenesis, such as cell proliferation, cell death, locomotion, cell adhesion, and cell-cell signaling. The GO analysis also showed enrichment in groups involved in formation of anatomical structures and embryo development. The key to WT1’s role in embryogenesis lies in its effect on transitions between the epithelial and mesenchymal states. These transitions have a well-established role in metastasis of solid tumors and also a role in the generation of cancer stem cells.32 Not much is known about their function in leukemia, but a recent study found that epithelial-mesenchymal transition genes were upregulated in promyelocytic leukemia,

References 1. Haber DA, Buckler AJ, Glaser T, et al. An internal deletion within an 11p13 zinc finger gene contributes to the development of Wilms' tumor. Cell. 1990;61(7):1257-1269. 2. Yang L, Han Y, Suarez Saiz F, Minden MD. A tumor suppressor and oncogene: the WT1 story. Leukemia. 2007;21(5):868-876. 3. Hollink IH, van den Heuvel-Eibrink MM, Zimmermann M, et al. Clinical relevance of Wilms tumor 1 gene mutations in childhood acute myeloid leukemia. Blood. 2009;113(23):5951-5960. 4. Renneville A, Boissel N, Zurawski V, et al. Wilms tumor 1 gene mutations are associated with a higher risk of recurrence in young adults with acute myeloid leukemia: a study from the Acute Leukemia French Association. Cancer. 2009;115(16):37193727. 5. Virappane P, Gale R, Hills R, et al. Mutation of the Wilms' tumor 1 gene is a poor prognostic factor associated with chemotherapy resistance in normal karyotype acute myeloid leukemia: the United Kingdom Medical Research Council Adult Leukaemia Working Party. J Clin Oncol. 2008;26(33):5429-5435. 6. Owen C, Fitzgibbon J, Paschka P. The clinical relevance of Wilms tumour 1 (WT1) gene mutations in acute leukaemia. Hematol Oncol. 2010;28(1):13-19. 7. Kim HJ, Choi EJ, Sohn HJ, Park SH, Min WS, Kim TG. Combinatorial molecular marker assays of WT1, survivin and TERT at initial diagnosis of adult acute myeloid leukemia. Eur J Haematol. 2013;91(5):411-422. 8. Bergmann L, Miething C, Maurer U, et al. High levels of Wilms' tumor gene (wt1) mRNA in acute myeloid leukemias are associated with a worse long-term outcome. Blood. 1997;90(3):1217-1225. 9. Nomdedéu JF, Hoyos M, Carricondo M, et

344

10.

11.

12.

13.

14.

15.

16.

17.

and that one of the genes found to be overexpressed in the network was WT1.51 In conclusion, our ChIP-Seq investigation into the binding pattern of WT1 -KTS and WT1 +KTS in leukemic cells reveals that the binding of both WT1 isoforms is correlated to active transcription, and that WT1 -KTS and WT1 +KTS share target genes, although the binding motifs are mostly different and the peaks of the two isoforms with very few exceptions do not overlap. The target genes for both isoforms reveal functions central to oncogenesis, consistent with an effector role in leukemia. Acknowledgments This work was supported by research grants from the Swedish Cancer Foundation and Children Cancer Foundation, the Gunnar Nilssons Cancer Foundation, the Physiographic Society of Lund, the Swedish Foundation for Strategic Research (ICA08-0057), the Marianne and Marcus Wallenberg Foundation (2010.0112), the Knut and Alice Wallenberg Foundation (2012.0193), ALF grants from Region Skåne, the Medical Faculty at Lund University, and the Swedish Society of Medicine. The funding bodies had no role in designing, carrying out or reporting the study.

al.; CETLAM Group. Bone marrow WT1 levels at diagnosis, post-induction and postintensification in adult de novo AML. Leukemia. 2013;27(11):2157-2164. Rein LA, Chao NJ. WT1 vaccination in acute myeloid leukemia: new methods of implementing adoptive immunotherapy. Expert Opin Investig Drugs. 2014;23(3):417-426. Nishida S, Hosen N, Shirakata T, et al. AML1-ETO rapidly induces acute myeloblastic leukemia in cooperation with the Wilms tumor gene, WT1. Blood. 2006;107(8):3303-3312. Laity JH, Dyson HJ, Wright PE. Molecular basis for modulation of biological function by alternate splicing of the Wilms' tumor suppressor protein. Proc Natl Acad Sci USA. 2000;97(22):11932-11935. Lopotová T, Polák J, Schwarz J, Klamová H, Moravcová J. Expression of four major WT1 splicing variants in acute and chronic myeloid leukemia patients analyzed by newly developed four real-time RT PCRs. Blood Cells Mol Dis. 2012;49(1):41-47. Kramarzova K, Stuchly J, Willasch A, et al. Real-time PCR quantification of major Wilms' tumor gene 1 (WT1) isoforms in acute myeloid leukemia, their characteristic expression patterns and possible functional consequences. Leukemia. 2012;26(9):20862095. Kim J, Cantor AB, Orkin SH, Wang J. Use of in vivo biotinylation to study protein-protein and protein-DNA interactions in mouse embryonic stem cells. Nat Protoc. 2009;4(4):506-517. Sandén C, Järvstråt L, Lennartsson A, Brattås PL, Nilsson B, Gullberg U. The DEK oncoprotein binds to highly and ubiquitously expressed genes with a dual role in their transcriptional regulation. Mol Cancer. 2014;13:215. Landt SG, Marinov GK, Kundaje A, et al. ChIP-seq guidelines and practices of the

18.

19. 20.

21.

22.

23. 24.

25.

26.

ENCODE and modENCODE consortia. Genome Res. 2012;22(9):1813-1831. Andersson R, Gebhard C, Miguel-Escalada I, et al.; FANTOM Consortium. An atlas of active enhancers across human cell types and tissues. Nature. 2014;507(7493):455461. ENCODE at UCSC. ftp://hgdownload.cse. ucsc.edu/goldenPath/hg19/encodeDCC/. Accessed 14 Feb 2013. ENCODE Project Consortium. An integrated encyclopedia of DNA elements in the human genome. Nature. 2012;489(7414): 57-74. Rauscher FJ 3rd, Morris JF, Tournay OE, Cook DM, Curran T. Binding of the Wilms' tumor locus zinc finger protein to the EGR1 consensus sequence. Science. 1990;250 (4985):1259-1262. Frietze S, Lan X, Jin VX, Farnham PJ. Genomic targets of the KRAB and SCAN domain-containing zinc finger protein 263. J Biol Chem. 2010;285(2):1393-1403. Kimura H. Histone modifications for human epigenome analysis. J Hum Genet. 2013;58(7):439-445. Bansal H, Seifert T, Bachier C, et al. The transcription factor Wilms tumor 1 confers resistance in myeloid leukemia cells against the proapoptotic therapeutic agent TRAIL (tumor necrosis factor -related apoptosisinducing ligand) by regulating the antiapoptotic protein Bcl-xL. J Biol Chem. 2012;287(39):32875-32880. Szemes M, Dallosso AR, Melegh Z, et al. Control of epigenetic states by WT1 via regulation of de novo DNA methyltransferase 3A. Hum Mol Genet. 2013;22(1):7483. Gashler AL, Bonthron DT, Madden SL, Rauscher FJ 3rd, Collins T, Sukhatme VP. Human platelet-derived growth factor A chain is transcriptionally repressed by the Wilms tumor suppressor WT1. Proc Natl Acad Sci USA. 1992;89(22):10984-10988.

haematologica | 2017; 102(2)


Global binding patterns of the WT1 -KTS and +KTS isoforms

27. Sitaram RT, Degerman S, Ljungberg B, et al. Wilms' tumour 1 can suppress hTERT gene expression and telomerase activity in clear cell renal cell carcinoma via multiple pathways. Br J Cancer. 2010;103(8):1255-1262. 28. Lee TH, Pelletier J. Functional characterization of WT1 binding sites within the human vitamin D receptor gene promoter. Physiol Genomics. 2001;7(2):187-200. 29. Nakagama H, Heinrich G, Pelletier J, Housman DE. Sequence and structural requirements for high-affinity DNA binding by the WT1 gene product. Mol Cell Biol. 1995;15(3):1489-1498. 30. Wang ZY, Qiu QQ, Enger KT, Deuel TF. A second transcriptionally active DNA-binding site for the Wilms tumor gene product, WT1. Proc Natl Acad Sci USA. 1993;90(19):8896-8900. 31. Bickmore WA, Oghene K, Little MH, Seawright A, van Heyningen V, Hastie ND. Modulation of DNA binding specificity by alternative splicing of the Wilms tumor wt1 gene transcript. Science. 1992;257(5067): 235-237. 32. Chau YY, Hastie ND. The role of Wt1 in regulating mesenchyme in cancer, development, and tissue homeostasis. Trends Genet. 2012;28(10):515-524. 33. Hutson JM, Grover SR, O'Connell M, Pennell SD. Malformation syndromes associated with disorders of sex development. Nat Rev Endocrinol. 2014;10(8):476-487. 34. Qi XW, Zhang F, Wu H, Liu JL, Zong BG, Xu C, Jiang J. Wilms' tumor 1 (WT1) expression and prognosis in solid cancer patients: a systematic review and metaanalysis. Sci Rep. 2015;5:8924. 35. Wickramasinghe VO, Laskey RA. Control of mammalian gene expression by selective

haematologica | 2017; 102(2)

36.

37. 38.

39.

40.

41.

42.

43.

44.

mRNA export. Nat Rev Mol Cell Biol. 2015;16(7):431-442. FANTOM Consortium and the RIKEN PMI and CLST (DGT), Forrest AR, Kawaji H, Rehli M, et al. A promoter-level mammalian expression atlas. Nature. 2014;507(7493):462-470. TRANSFAC in gene-regulation.com. h t t p : / / w w w. g e n e - r e g u l a t i o n . c o m / pub/databases.html. Accessed 19 Jan 2015. Wingender E, Chen X, Hehl R, et al. TRANSFAC: an integrated system for gene expression regulation. Nucleic Acids Res. 2000;28(1):316-319. Hartwig S, Ho J, Pandey P, et al. Genomic characterization of Wilms' tumor suppressor 1 targets in nephron progenitor cells during kidney development. Development. 2010;137(7):1189-1203. Motamedi FJ, Badro DA, Clarkson M, et al. WT1 controls antagonistic FGF and BMPpSMAD pathways in early renal progenitors. Nat Commun. 2014;5:4444. Dong L, Pietsch S, Tan Z, et al. Integration of cistromic and transcriptomic analyses identifies Nphs2, Mafb, and Magi2 as Wilms' tumor 1 target genes in podocyte differentiation and maintenance. J Am Soc Nephrol. 2015;26(9):2118-2128. Kann M, Ettou S, Jung YL, et al. Genomewide analysis of Wilms' tumor 1-controlled gene expression in podocytes reveals key regulatory mechanisms. J Am Soc Nephrol. 2015;26(9):2097-2104. Han Y, San-Marina S, Liu J, Minden MD. Transcriptional activation of c-myc protooncogene by WT1 protein. Oncogene. 2004;23(41):6933-6941. Montano G, Cesaro E, Fattore L, et al. Role of WT1-ZNF224 interaction in the expres-

45.

46.

47.

48.

49.

50.

51.

sion of apoptosis-regulating genes. Hum Mol Genet. 2013;22(9):1771-1782. Montano G, Vidovic K, Palladino C, et al. WT1-mediated repression of the proapoptotic transcription factor ZNF224 is triggered by the BCR-ABL oncogene. Oncotarget. 2015;6(29):28223-28237. Rupprecht HD, Drummond IA, Madden SL, Rauscher FJ 3rd, Sukhatme VP. The Wilms' tumor suppressor gene WT1 is negatively autoregulated. J Biol Chem. 1994;269(8):6198-6206. Miyamoto Y, Taniguchi H, Hamel F, Silversides DW, Viger RS. A GATA4/WT1 cooperation regulates transcription of genes required for mammalian sex determination and differentiation. BMC Mol Biol. 2008;9:44. Vidovic K, Svensson E, Nilsson B, et al. Wilms' tumor gene 1 protein represses the expression of the tumor suppressor interferon regulatory factor 8 in human hematopoietic progenitors and in leukemic cells. Leukemia. 2010;24(5):992-1000. Duarte A, Caricasole A, Graham CF, Ward A. Wilms' tumour-suppressor protein isoforms have opposite effects on Igf2 expression in primary embryonic cells, independently of p53 genotype. Br J Cancer. 1998;77(2):253-259. Ryan G, Steele-Perkins V, Morris JF, Rauscher FJ 3rd, Dressler GR. Repression of Pax-2 by WT1 during normal kidney development. Development. 1995;121(3):867875. Percio S, Coltella N, Grisanti S, Bernardi R, Pattini L. A HIF-1 network reveals characteristics of epithelial-mesenchymal transition in acute promyelocytic leukemia. Genome Med. 2014;6(12):84

345


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Lymphoblastic Leukemia

Ferrata Storti Foundation

The role of ZAP70 kinase in acute lymphoblastic leukemia infiltration into the central nervous system

Ameera Alsadeq,1 Henning Fedders,1 Christian Vokuhl,2 Nele M. Belau,1 Martin Zimmermann,3 Tim Wirbelauer,1 Steffi Spielberg,1 Michaela Vossen-Gajcy,1 Gunnar Cario,1 Martin Schrappe1 and Denis M. Schewe1

Haematologica 2017 Volume 102(2):346-355

Department of General Pediatrics, ALL-BFM Study Group, Christian-Albrechts University Kiel and University Medical Center Schleswig-Holstein, Kiel; 2Kiel Pediatric Tumor Registry, Department of Pediatric Pathology, University Medical Center SchleswigHolstein and 3Pediatric Hematology and Oncology, Hannover Medical School, Germany

1

ABSTRACT

C

Correspondence: denis.schewe@uksh.de

Received: April 12, 2016. Accepted: September 22, 2016. Pre-published: September 29, 2016. doi:10.3324/haematol.2016.147744

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/346

Š2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

346

entral nervous system infiltration and relapse are poorly understood in childhood acute lymphoblastic leukemia. We examined the role of zeta-chain-associated protein kinase 70 in preclinical models of central nervous system leukemia and performed correlative studies in patients. Zeta-chain-associated protein kinase 70 expression in acute lymphoblastic leukemia cells was modulated using short hairpin ribonucleic acid-mediated knockdown or ectopic expression. We show that zeta-chain-associated protein kinase 70 regulates CCR7/CXCR4 via activation of extracellular signal-regulated kinases. High expression of zeta-chain-associated protein kinase 70 in acute lymphoblastic leukemia cells resulted in a higher proportion of central nervous system leukemia in xenografts as compared to zeta-chain-associated protein kinase 70 low expressing counterparts. High zeta-chain-associated protein kinase 70 also enhanced the migration potential towards CCL19/CXCL12 gradients in vitro. CCR7 blockade almost abrogated homing of acute lymphoblastic leukemia cells to the central nervous system in xenografts. In 130 B-cell precursor acute lymphoblastic leukemia and 117 T-cell acute lymphoblastic leukemia patients, zeta-chain-associated protein kinase 70 and CCR7/CXCR4 expression levels were significantly correlated. Zetachain-associated protein kinase 70 expression correlated with central nervous system disease in B-cell precursor acute lymphoblastic leukemia, and CCR7/CXCR4 correlated with central nervous system involvement in T-cell acute lymphoblastic leukemia patients. In multivariate analysis, zeta-chain-associated protein kinase 70 expression levels in the upper third and fourth quartiles were associated with central nervous system involvement in B-cell precursor acute lymphoblastic leukemia (odds ratio=7.48, 95% confidence interval, 2.06-27.17; odds ratio=6.86, 95% confidence interval, 1.86-25.26, respectively). CCR7 expression in the upper fourth quartile correlated with central nervous system positivity in T-cell acute lymphoblastic leukemia (odds ratio=11.00, 95% confidence interval, 2.00-60.62). We propose zeta-chain-associated protein kinase 70, CCR7 and CXCR4 as markers of central nervous system infiltration in acute lymphoblastic leukemia warranting prospective investigation. Introduction Diagnosis and treatment of central nervous system (CNS) infiltration and/or CNS relapse in childhood acute lymphoblastic leukemia (ALL) remain challenging.1,2 CNS leukemia is mainly a leptomeningeal disease, but leukemic cells may be detectable in the cerebrospinal fluid.3 Survival and homing mechanisms of leukemic cells into haematologica | 2017; 102(2)


ZAP70 facilitates CNS infiltration of ALL

the CNS remain largely unclear. Factors associated with CNS involvement are peripheral hyperleukocytosis, a proB- or T-cell immunophenotype,4,5 and certain cytogenetics (e. g., BCR-ABL in B-cell precursor ALL (BCP-ALL)).4,6 Additionally, the E2A-PBX1 fusion (t(1;19)(q23;p13)) has been associated with an increased risk of CNS relapses in BCP-ALL,7 and we recently identified the MER tyrosine kinase as a marker for CNS involvement in that entity.8 Most chemotherapeutic drugs applied for ALL treatment poorly penetrate the blood-brain barrier. Accordingly, current pediatric treatment protocols advocate extensive intrathecal and systemic chemotherapy,9-11 which has been associated with long-term neurologic sequelae.11,12 Thus, more precise diagnostic and prognostic markers for CNS involvement in ALL are needed, not only to control CNS infiltration but also to avoid systemic overtreatment. Zeta-chain-associated protein kinase 70 (ZAP70), a tyrosine kinase mainly expressed in T cells and natural killer (NK) cells, is also found to be expressed at low levels in B cells.13 The role of ZAP70 in BCP-ALL is poorly described. Few reports have shown that ZAP70 is expressed in some ALL cell lines and in a report of 5 patients with the E2A-PBX1 fusion.14-16 ZAP70 has been shown to be overexpressed in B-cell chronic lymphocytic leukemia (B-CLL), indicating an aggressive course of the disease.17 ZAP70 expression and phosphorylation have been associated with B-cell receptor (BCR) signaling in B-CLL.18,19 Further, ZAP70 may function as an adaptor protein and enhance BCR signaling independently of its kinase activity.20 ZAP70 has been shown to enhance the migration of malignant B cells to the bone marrow (BM) by upregulating adhesion molecules and chemokine receptors (CCRs),21,22 and is also required for C-X-C motif chemokine ligand 12 (CXCL12)-mediated T-cell transendothelial migration.23,24 In T-cell acute lymphoblastic leukemia (T-ALL), the chemokine receptors CCR7 and CXCR4 have been associated with an increased capability of T-ALL cells to enter the CNS, mainly in cell line models.25,26 We hypothesized that ZAP70 mediates the infiltration and the survival of ALL cells in the CNS. Downregulating ZAP70 in ALL cell lines resulted in a reduced CCR7/CXCR4 expression and an impaired CNS infiltration in NSG mice via the regulation of ERK. In contrast, up-regulating ZAP70 caused an enhanced CCR7/CXCR4 expression and an improved migratory capacity towards chemokine ligand 19 (CCL19) and CXCL12 gradients. Blocking CCR7 with a monoclonal antibody resulted in a decreased homing capacity of ALL cells in vivo, including the CNS. High ZAP70 expression in patient samples correlated with CNS infiltration in xenografts. Furthermore, patient CNS-positivity correlated with a high ZAP70 expression in BCP-ALL, and with a high CCR7/CXCR4 expression in T-ALL patients. Multivariate analysis confirmed that a high expression of ZAP70 and CCR7 conferred an increased risk for CNS involvement in these patients. Our data suggest an important role of these mechanisms for homing and survival of ALL cells in the CNS niche.

Kiel, Germany). Cell viability was measured using trypan blue. CD19, CD3ε, hCD45, mCD45, CCR7, CXCR4 and ZAP70 antibodies were purchased from eBioscience or Santa Cruz Biotechnology, p-ERK/ERK and b-tubulin antibodies from Cell Signaling Technology, anti-Immunoglobulin M (IgM) F(ab)2 from SouthernBiotech, and U0126 from LC Laboratories.

Patients 130 BCP-ALL and 117 T-ALL patients were treated according to the Berlin-Frankfurt-Münster (BFM) ALL 2000 or 2009 protocols. Informed consent was obtained according to institutional regulations, in accordance with the Declaration of Helsinki.

Xenografts NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) and NOD/SCID mice were purchased from Charles River Laboratories and bred. The mice were maintained as approved by the governmental animal care and use committees. Eight to twelve week old female mice were injected intrafemorally with 1 × 106 ALL cells from patient BM (>90% blasts), or intravenously with 1 x 104 - 1 x 105 ALL cells from cell lines.8,27 Animals were sacrificed upon detection of >75% leukemic blasts or clinical leukemia (weight or activity loss, organomegaly, hind limb paralysis). CNS-leukemic cells were recovered from meninges and separated on 30% percoll.28

Expression assays Western blotting was performed as described.27 Mouse heads were decalcified, paraffin-embedded and cut. CNS infiltration was scored negative (−), intermediate (+) and high (++) as described.8 Quantitative real-time polymerase chain reaction (qRT-PCR) analyses were performed on an ABI7900HT using QuantiTect primer assays (Qiagen) and real-time quantitative PCR value (RQ) (2^-DCCT) was calculated.

Knockdown and upregulation of ZAP70 A short hairpin ribonucleic acid (shRNA) against ZAP70 (TRCN0000000438) was cloned into pSicoR-Ef1a-mCh (Addgene vector 31847). pSicoR-Ef1a-mCh expressing an shRNA against green fluorescent protein (GFP) was used as control (Addgene vector 31849). 697, REH and JURKAT cells were transduced with viruses and sorted for mCherry-positivity.8,27 pMIG or pMIGZAP70 vectors, provided by Prof. Michael Reth (University of Freiburg, Germany), were used to transduce leukemic cells.29

Chemotaxis Chemotaxis was measured as described.21 5 × 105 cells were added to the top chamber of a transwell culture insert (Corning) and allowed to migrate toward media containing 2 μg/ml CCL19 or 0.1 μg/ml CXCL12 (ImmunoTools) for 16 h. The cell number in the lower chamber was determined.

Antibody treatment in vivo NSG mice were injected with 10,000 697 or JURKAT cells/animal (day 0). 1 mg/kg of anti-CCR7 antibody (clone 150503, R&D Systems) was applied on day +1, +3 and +7. All mice were sacrificed when control animals showed signs of leukemia. Leukemic infiltration was assayed by FACS or histology.

Statistical analysis Methods Cell lines, antibodies REH and 697 (EU-3) cell lines were purchased from DSMZ. JURKAT cells were provided by Dr. Renate Burger (University of haematologica | 2017; 102(2)

Statistical tests are indicated in the Figure legends. A P-value of <0.05 was considered significant. For the in vivo histology (Figures 1 and 2) one-sided testing was chosen, since death rates at a time point in the control groups were known and we aimed to detect a difference in one direction. In vitro panels are representative of at least 2 independent experiments. Densitometry was performed 347


A. Alsadeq et al. using ImageJ.30 The association between gene expression and CNS status was examined by unconditional logistic regression to calculate odds ratios (ORs) and 95% confidence intervals (CIs).

Results ZAP70 expression influences CNS infiltration in vivo First, we studied the effects of ZAP70 downregulation in ALL cell lines known to infiltrate the CNS of NSG and NOD/SCID mice. 697 and REH BCP-ALL and JURKAT TALL cells were transduced with a non-targeting shRNA (shGFP) or with an shRNA targeting ZAP70 (shZAP70). ZAP70 messenger RNA (mRNA) expression was reduced by >70% in all shZAP700 cells (Figure 1A). 697- and JURKAT-shGFP and -shZAP70 cells were then injected into NSG and REH-shGFP and -shZAP70 cells into NOD/SCID mice, a model for an almost isolated CNS leukemia with this cell line. Mice were sacrificed when hind limb paralysis as a symptom of CNS leukemia was

visible. Median BM infiltration in shZAP70 xenografts was not different from shGFP controls in any cell line (Online Supplementary Figure S1A). For 697 and JURKAT, both groups had a similar median survival (Online Supplementary Figures S1B and S1C). For 697, the survival difference between 697-shGFP and 697-shZAP70 was statistically significant but small (median 28 vs. 26 days, Online Supplementary Figure S1B). Standardized histological semi-quantitative scoring of CNS infiltration8 (Figure 1B) revealed that 9/10 animals (90%) in the 697-shGFP group were CNS++ or CNS+, and only 2/7 animals (29%) in the 697-shZAP70 group showed a CNS-positive status in the xenografts (Figure 1C). In the REH-shGFP controls, 8/8 animals (100%) were CNS-positive as compared to 4/8 (50%) in the REH-shZAP70 test group (Figure 1C). For JURKAT, scoring was not significantly different in JURKAT-shGFP vs. JURKAT-shZAP70 cells (Figure 1C), however, the total number of blasts recovered from the CNS of xenografts in controls was significantly higher than those from JURKAT-shZAP70 (Figure 1D). For REH

A

B

C

E

348

D

F

Figure 1. ZAP70 expression is associated with CNS infiltration in vivo. A: Efficiency of ZAP70 knockdown by shRNA, as determined by qRT-PCR. RQ=2^-DDCT. B: Semiquantitative measurement of CNS infiltration as described in Methods. Hematoxylin and Eosin staining of a negative (-, left), an intermediate (+, middle), and a high (++, right) sample as examples for the scoring method, 200x magnification. * signifies leukemic infiltration. C: Mice were xenografted with 697 (n=10 per group), REH (n=8 per group) and JURKAT (n=10 per group) cells bearing an shRNA against ZAP70 (shZAP70) or a control (shGFP). Mice were sacrificed upon appearance of hind limb paralysis, and semi-quantitative scoring of the xenograft CNS is shown in the table (Fisher´s exact test, one-sided P-values. *P is significant). In the 697shZAP70 group, 1 cage containing 3 mice was accidentally eliminated from the experiment on day +2. These animals were not included in the statistical analysis. D: Number of leukemic blasts recovered from the CNS of mice xenografted with either JURKAT-shGFP (n=4) or JURKAT-shZAP70 (n=3) (unpaired t-test, one-sided *P-value 0.041). E: 1x105 REH-shGFP or REHshZAP70 cells/mouse were xenografted into 10 NOD/SCID mice/group by tail vein injection, xenograft survival (Kaplan-Meier log-rank test, **P=0.0021). F: 10 primary samples of pediatric BCP-ALL patients chosen according to ZAP70 expression (5 ZAP70lo and 5 ZAP70hi) were xenografted into duplicate NSG-mice. Mice were sacrificed when leukemic symptoms were visible and semi-quantitative scoring of the CNS was performed. Patient and xenograft characteristics are depicted in the Online Supplementary Table S1. *P<0.05; **P<0.01. CSF: cerebrospinal fluid; DM: dura mater; CNS: central nervous system; BCP-ALL: B-cell precursor acute lymphoblastic leukemia; Xeno: xenograft; shGFP: shRNA against GFP(green fluorescent protein); shZAP70: shRNA targeting ZAP70 (zeta-chain-associated protein kinase 70).

haematologica | 2017; 102(2)


ZAP70 facilitates CNS infiltration of ALL

cells, ZAP70 knockdown significantly prolonged the median xenograft survival by 25 days in the REHshZAP70 group as compared to REH-shGFP (104 vs. 79 days, Figure 1E). These data suggest that despite a similar leukemic burden in the periphery, knockdown of ZAP70 hampers the infiltration of 697 and REH BCP-ALL cells into the CNS. In the REH NOD/SCID model, CNS

leukemia outweighs peripheral leukemia and mice show low levels of blood, BM and splenic engraftment, but massive CNS infiltration. Our data show that knockdown of ZAP70 in REH cells delays CNS infiltration as a cause of xenograft death in vivo. The data also suggest that knockdown of ZAP70 is not sufficient to prevent CNS infiltration in xenografts bearing JURKAT cells.

A

B

C

D

E

F

G

Figure 2. ZAP70 regulates CCR7 and CXCR4 expression, and CCR7 inhibition is relevant for homing processes. A: CCR7 and CXCR4 expression in 697-shZAP70 in comparison to 697-shGFP cells as measured by qRT-PCR (left) and by extracellular FACS staining (right), black: isotype, red: 697-shGFP and blue: 697-shZAP70. Unpaired t-test, two-sided P-value. B: ZAP70, CCR7 and CXCR4 expression in 697 cells transduced with either pMIG or pMIG-ZAP70, unpaired t-test, two-sided P-value. C: 697-shGFP or -shZAP70 cells were subjected to a migration assay toward CCL19 or CXCL12. Cells in the lower chamber were counted by flow cytometry or trypan blue. The migration index is defined as the number of transmigrating cells in the presence of the chemokine divided by the number of transmigrating cells to control medium. Results are shown as the mean ¹ SEM of 4 independent experiments (unpaired t-test, two-sided P-value). D: Migration assay of 697-pMIG and 697-pMIG-ZAP70 toward CCL19 or CXCL12 (unpaired t-test, two-sided P-value). E: 697 or JURKAT cells were injected into NSG mice, 5 mice were treated with 1 mg/kg anti-CCR7 antibody on day +1, +3 and +7, 5 mice were treated with vehicle alone. Spleen volume as assessed by the formula longest length x highest height x broadest width (unpaired t-test, two-sided P-value). F: Spleen and BM infiltration by human leukemic blasts in control and treated animals (unpaired t-test, two-sided Pvalue). G: Semi-quantitative CNS scoring of control and treated animals (Fisher´s exact test, one-sided P-value. *P is significant). *P<0.05; **P<0.01, ***P<0.001, n.s. = not significant. BM: bone marrow; CNS: central nervous system; Sp: spleen; shGFP: shRNA against GFP(green fluorescent protein); shZAP70: shRNA targeting ZAP70 (zeta-chain-associated protein kinase 70).

haematologica | 2017; 102(2)

349


A. Alsadeq et al.

remaining patients as ZAP70lo. 1 x 106 primary cells (>90% blasts) of 5 ZAP70hi and 5 ZAP70lo patients were injected into replicate NSG mice. The clinical characteristics of these patients are shown in the Online Supplementary Table S1. Interestingly, 7/10 (70%) mice injected with ZAP70hi

We next investigated if ZAP70 influences the ability of primary patient cells to infiltrate the CNS of NSG mice. ZAP70 mRNA levels were measured in pediatric BCP-ALL patients. Patients with ZAP70 expression levels higher than the median were considered as ZAP70hi, and the

A

B

C

D

E

F

G

Figure 3. ZAP70 regulates CCR7 and CXCR4 via ERK. A: Bone marrow cells from initial diagnosis from 9 BCP-ALL patients were xenografted into NSG mice, and leukemic blasts from xenograft spleens were analyzed by western blotting for ZAP70, p-ERK, ERK and b-Tubulin (left). The ZAP70/p-ERK ratios were calculated using Image J software for each individual lane (Spearman´s rank correlation). B: 697 (left) and JURKAT cells (right) were serum starved for 30 min, then left without treatment or stimulated with one or more of the following: CCL19 1μg/ml, anti-IgM F(ab)2 10μg/ml, and anti-CD3 10μg/ml for 5 or 30 min. Samples were analyzed by western blotting for p-ERK, ERK and b-Tubulin. Densitometry was performed using ImageJ software. C: Primary leukemic blasts from xenograft spleens were treated and analyzed as the cell lines in Panel B. Densitometry was performed using ImageJ software. D: 697-shGFP and 697-shZAP70 cells (left) and JURKA-shGFP and JURKAT-ZAP70 cells (right) were lysed and studied by western blotting for the indicated proteins. E: 697-shGFP and 697-shZAP70 cells were stimulated with CCL19 or CXCL12 as indicated. Cells were then lysed and studied by western blotting for the indicated proteins. F: CCR7 and CXCR4 expression was quantified via qRT-PCR of 697 and JURKAT cells in the presence or absence of U1026 (20μM) for 48 h. G: Cells were treated with U1026 (20μM) for 1 h, then subjected to a migration assay for 16 h toward CCL19 or CXCL12 and cells in the lower chamber were counted by FACS or trypan blue. *P<0.05. Xen: xenograft; RQ: real-time quantitative PCR value; AU: arbitrary unit; IgM: immunoglobulin M; PV: pervanadate, positive control; nt: shGFP (shRNA against GFP(green fluorescent protein)); kd: shZAP70 (shRNA targeting ZAP70 (zeta-chain-associated protein kinase 70)).

350

haematologica | 2017; 102(2)


ZAP70 facilitates CNS infiltration of ALL

primary cells showed CNS++ or CNS+ phenotypes by histology, whereas only 1/10 (10%) mice bearing ZAP70lo patient cells were CNS positive (Figure 1F, Online Supplementary Table S1). These data suggest that ZAP70 expression levels play a role for the homing and/or the survival of primary patient BCP-ALL cells in the CNS, and corroborate our cell line data.

ZAP70 regulates CCR7 and CXCR4 ZAP70 was shown to upregulate the expression of adhesion molecules and chemokine receptors in B-CLL.21,24 Chemokine receptors CCR7 and CXCR4 in turn have been associated with an increased capability of T-ALL cells to enter the CNS.25,26 We hypothesized that ZAP70 mediated upregulation of CCR7 and CXCR4 may enhance the homing and the survival of CNS-prone BCP-ALL cells in the CNS niche. Downregulation of ZAP70 in 697 cells resulted in a reduced CCR7 and CXCR4 mRNA expression and a reduction of the proteins on the cell surface (Figure 2A). On the other hand, ectopic overexpression of ZAP70 in 697 cells resulted in a 3-fold upregulation of CCR7 and CXCR4 mRNA levels (Figure 2B). In order to test if ZAP70 mediates the migration of ALL cells via CCR7, we performed transwell assays with or without CCL19 or CXCL12, which are known ligands of CCR7 and CXCR4. The migratory capacity of 697-shZAP70 cells was reduced by ≈50% as compared to controls with CCL19, and by ≈80% with CXCL12 (Figure 2C). Similarly, 697 cells expressing pMIG-ZAP70 showed a ≈2-fold higher migration index toward these cytokines (Figure 2D). These data show that CCR7 and CXCR4 expression, in

this cell line, are ZAP70 dependent, which correlates with migratory properties. In order to exemplify that CCR7 is required for CNS homing, we injected 697 and JURKAT cells into NSG mice (day 0). Treatment with 1 mg/kg of anti-CCR7 antibody on day +1, +3 and +7 was then initiated in the test group. On day +23 3/5 (60%) of the 697 mice had developed signs of CNS-leukemia, and on day +35 2/5 (40%) of the JURKAT control mice had developed signs of CNS-leukemia; all mice in both groups were sacrificed. Interestingly, spleens from control mice were enlarged, whereas spleen volumes in the treatment group were normal (Figure 2E). Mice from both groups showed leukemic engraftment, however median blast percentages in the spleen and BM were lower in the anti-CCR7 groups (8.6% vs. 12.7% and 13.6% vs. 32.2% for 697 and 39.8% vs. 65.8% and 81.8% vs. 92.9% for JURKAT, respectively, Figure 2F). The difference in BM engraftment for JURKAT was not statistically significant, suggesting that leukemic burden at the end of the experiment was equal in both groups. Most importantly, CNS histology showed that 5/5 (100%) control mice (both 697 and JURKAT) were CNS++/+ as compared to 1/5 (20%) in both treatment groups (Figure 2G). These data suggest that CCR7 inhibition in 697 and JURKAT cells reduces the homing to hematopoietic organs, including CNS infiltration.

ZAP70 regulates CCR7 and CXCR4 via ERK CNS infiltration in BCP-ALL has been associated with Ras-mutations activating mitogen-activated protein kinase (MAPK) pathways.31 ZAP70 was shown to enhance the migration of B-CLL cells via ERK1/2 activation.21

Table 1. Univariate and multivariate associations for ZAP70 expression quartiles and CNS status in 130 childhood BCP-ALL patients, and for CCR7 or CXCR4 expression quartiles and CNS status in 117 childhood T-ALL patients, as indicated.

BCP-ALL ZAP70 Quartile* I II III IV

T-ALL CCR7 Quartile* I II III IV

T-ALL CXCR4 Quartile* I II III IV

CNS neg. (n=82) No.

CNS pos. (n=48) %

Univariate No.

%

OR†

27 22 17 16

33.0 26.8 20.7 19.5

5 11 16 16

10.4 23.0 33.3 33.3

1.000§ 2.700 5.082 5.400

CNS neg. (n=90) No.

CNS pos. (n=27) %

No.

%

OR†

26 23 24 17

28.9 25.6 26.7 18.8

2 7 6 12

7.4 25.9 22.2 44.5

1.000§ 3.957 3.250 9.176

CNS neg. (n=90) No.

CNS pos. (n=27) %

No.

%

OR†

25 26 22 17

27.8 28.9 24.4 18.9

4 4 7 12

14.8 14.8 25.9 44.4

1.000§ 0.962 1.989 4.412

Multivariate 95% Cl 0.815-8.943 1.572-16.429 1.660-17.561

P

OR†

95% Cl

P

0.104 0.007 0.005

1.000§ 3.155 7.479 6.860

0.858-11.602 2.059-27.172 1.863-25.264

0.084 0.002 0.004

Univariate

Multivariate 95% Cl 0.746-20.989 0.597-17.679 1.822-46.229

P

OR†

95% Cl

P

0.106 0.173 0.007

1.000§ 3.836 3.791 11.001

0.666-22.081 0.649-22.138 1.996-60.622

0.132 0.139 0.006

Univariate

Multivariate 95% Cl 0.217 - 4.269 0.513 - 7.713 1.216-16.002

P

OR†

95% Cl

P

0.959 0.320 0.024

1.000§ 0.820 2.094 3.760

0.165 – 4.076 0.468 - 9.368 0.896-15.788

0.809 0.334 0.070

CNS status is described in the Online Supplementary Table S2. §Reference category. *Based on expression as measured by RT-PCR of ZAP70 in all 130 BCP-ALL and of CCR7 or CXCR4 in all 117 T-ALL patients. †Multivariate OR controlled for age and WBC count at diagnosis, and TEL-AML and BCR-ABL positivity in BCP-ALL. OR: odds ratio; Cl: Confidence interval; BCP-ALL: B-cell precursor acute lymphoblastic leukemia; T-ALL: T-cell acute lymphoblastic leukemia; CNS: central nervous system; neg: negative; pos: positive; ZAP70: zeta-chain-associated protein kinase 70.

haematologica | 2017; 102(2)

351


A. Alsadeq et al.

Furthermore, ERK1/2 is required for the migration of naĂŻve T lymphocytes toward CCL21, another ligand of CCR7.32 We hypothesized that ZAP70 activates MAPK pathways to regulate CCR7 and CXCR4 and to mediate CCL19- and CXCL12-induced migration. In order to examine ZAP70 expression and ERK activation, we analyzed 9 patient

xenografts by western blotting (Figure 3A, left). ZAP70 expression correlated with ERK1/2 phosphorylation (Figure 3A, right). To examine the role of ERK1/2 in response to CCL19, ALL cells were stimulated with CCL19 (Figure 3B). In order to mimic pre-BCR and pre-Tcell receptor (TCR) activation as controls, 697 cells were

A

B

C

D

E

Figure 4. Correlation of ZAP70, CCR7 and CXCR4 with CNS status in ALL patients. Correlation of ZAP70, CCR7 and CXCR4 with CNS status in ALL patients. ZAP70, CCR7 and CXCR4 expression levels were determined with qRT-PCR in 130 pediatric BCP-ALL and 117 pediatric T-ALL patients. Correlation analysis between ZAP70 and CCR7 (A) and ZAP70 and CXCR4 (B) for BCP-ALL (left) and T-ALL (right) patients. Correlation between ZAP70 in BCP-ALL (C) and CCR7 (D)/CXCR4 (E) in T-ALL patients and CNS group (CNS negative/no relapse, CNS positive/no relapse, CNS negative/relapse, CNS positive/relapse). Further definitions are provided in the Online Supplementary Table S2. BCP-ALL: B-cell precursor acute lymphoblastic leukemia; T-ALL: T-cell acute lymphoblastic leukemia; CNS: central nervous system; n.s: not significant; neg: negative; pos: positive; rel: relapse; RQ: real-time quantitative PCR value; ZAP70: zeta-chain-associated protein kinase 70.

352

haematologica | 2017; 102(2)


ZAP70 facilitates CNS infiltration of ALL

additionally stimulated with anti-IgM F(ab)2 and JURKAT cells with anti-CD3 (Figure 3B). Upon stimulation with CCL19, 697 cells displayed an increase in p-ERK1/2 as compared to unstimulated cells, but to a lesser extent than in response to anti-IgM F(ab)2 (Figure 3B, left). Similarly, JURKAT cells showed an increase in p-ERK after 5 min of CCL19 stimulation (Figure 3B, right), which declined after 30 min (data not shown). In BCP-ALL primary xenograft cells, p-ERK1/2 was also enhanced by 30 min of CCL19 stimulation in most samples (Figure 3C, Online Supplementary Figure S2A). p-ERK was also induced by CXCL12 in 697 and JURKAT cells (Online Supplementary Figure S2B) as well as in BCP-ALL primary xenograft cells (Online Supplementary Figure S2C). These data suggest that CCL19/CCR7 and CXCL12/CXCR4 activate ERK in 697, JURKAT and BCP-ALL primary cells. In order to examine p-ERK in the context of ZAP70 downregulation, we analyzed p-ERK in 697-shZAP70 and JURKAT-shZAP70 cells. Both cell lines bearing shZAP70 showed a reduction in basal p-ERK (Figure 3D). Furthermore, the induction of p-ERK by CCL19 and CXCL12 was entirely abrogated (Figure 3E). The treatment of 697 and JURKAT cells with the MAPK kinase (MEK) inhibitor U0126, thereby abrogating ERK activation, resulted in a â&#x2030;&#x2C6;2-fold downregulation of CCR7 mRNA in 697 and JURKAT cells, and a >2fold downregulation of CXCR4 mRNA (Figure 3F). Additionally, 697 and JURKAT cells treated with U0126 showed significantly reduced migration indices towards both cytokines, as compared to cells treated with vehicle alone (Figure 3G). Our data show that high expression of ZAP70 correlates with ERK activation. Furthermore, CCL19 and CXCL12 induce p-ERK. Finally, the inhibition of ZAP70 reduces p-ERK, and MEK inhibition results in a downregulation of CCR7 and CXCR4 and impaired migration towards CCL19 and CXCL12. The data suggest that ERK may be a missing link between ZAP70 and CCR7/CXCR4.

Association of ZAP70 and CCR7 with CNS involvement in ALL patients To test whether ZAP70/CCR7/CXCR4 is associated with CNS involvement in patients, we analyzed the mRNA levels of these markers in diagnostic BM samples of patients with BCP-ALL and T-ALL. ZAP70 mRNA expression correlated with protein levels in 9 patient xenografts tested as examples (Spearman r=0.5667, Online Supplementary Figure S3), however the P-value did not reach statistical significance (P=0.0603). The cohort of 130 BCP-ALL patients (Online Supplementary Table S2, top) was selected to contain 46 patients that were initially CNSpositive and developed no CNS relapse (CNS positive/no relapse), 18 patients that were CNS-negative at diagnosis but developed CNS relapse (CNS negative/relapse) as well as 2 patients that were CNS-positive at diagnosis and developed CNS relapse (CNS positive/relapse), matched to 64 CNS-negative patients that developed no CNS relapse at all (CNS negative/no relapse). The cohort of 117 T-ALL patients (Online Supplementary Table S2, bottom) included 24 CNS positive/no relapse, 6 CNS negative/relapse, 3 CNS positive/relapse and 84 CNS negative/no relapse patients. There were no statistical differences in sex, age, prednisone response or cytogenetics between the CNS negative/no relapse, CNS positive/no relapse and CNS negative/relapse groups. CNS positive/relapse patients were not included in the analysis haematologica | 2017; 102(2)

due to low numbers. However, in the T-ALL CNS positive/no relapse group, patients had a significantly higher initial white blood cell (WBC) count as compared to CNS negative/no relapse control patients. In BCP-ALL patients, there were a significantly higher number of patients in the CNS negative/relapse group stratified into the minimal residual disease (MRD) intermediate-risk (IR) group than in CNS negative/no relapse control patients. These data confirm the importance of hyperleukocytosis as a risk factor for CNS involvement in T-ALL, and show that outcomes in the MRD-IR group are difficult to predict. ZAP70 expression levels in patient samples were correlated with CCR7 expression levels in the BCP-ALL cohort as well as in the T-ALL cohort, but to a lesser extent in the latter cohort (Figure 4A). ZAP70 also correlated with CXCR4 in both cohorts, however, more markedly in TALL (Figure 4B). Importantly, ZAP70 expression was found to be significantly elevated in CNS positive/no relapse as compared to CNS negative/no relapse BCP-ALL, but not T-ALL patients (Figure 4C, Online Supplementary Figure S4A). CNS positive/no relapse T-ALL patients showed a significantly higher CCR7 expression as compared to CNS negative/no relapse patients, which was not detectable in BCP-ALL patients (Figure 4D, Online Supplementary Figure S4B). Similarly, CXCR4 expression was higher in CNS positive/no relapse T-ALL in comparison to CNS negative/no relapse patients (Figure 4E). Again, this difference was not detectable in BCP-ALL (Online Supplementary Figure S4C). ZAP70 may thus be important for CNS infiltration in BCP-ALL, and the upregulation of chemokine receptors may be more relevant in T-ALL patients, which matches the discrepancies seen in our in vivo experiments between BCP-ALL and T-ALL cells. CNS negative/relapse patients showed no elevations in ZAP70 or CCR7 mRNA levels as compared to CNS negative/no relapse patients, neither in BCP-ALL nor in T-ALL (Figure 4C,D, Online Supplementary Figure S4A,S4B). In fact, CNS negative/relapse BCP-ALL patients had significantly lower ZAP70 mRNA expression levels than CNS negative/no relapse patients (Figure 4C). This suggests that ZAP70/CCR7 expression in ALL blasts may be relevant as a homing/survival mechanism initially, and not as a predictor of CNS relapse in the course of the disease. For CXCR4, CNS negative/relapse T-ALL patients showed significantly higher levels than CNS negative/no relapse patients (Figure 4E). Interestingly, the 2 CNS positive/relapse BCP-ALL patients showed the highest expression levels of both ZAP70 and CCR7 (Figure 4C, Online Supplementary Figure S4B), however, evaluation of more patients is needed before conclusions can be drawn. In order to evaluate the independent impact of high ZAP70 and CCR7/CXCR4 mRNA levels on CNS status, we performed a binary logistic regression analysis controlling for age, WBC and the presence of certain cytogenetics (BCR-ABL or TEL-AML1) in BCP-ALL with CNS infiltration at the initial diagnosis (yes/no) as the dependent variable. In BCP-ALL patients, ZAP70 expression levels in the upper two quartiles conferred a â&#x2030;&#x2C6;7-fold increased risk of CNS-positivity compared with the lowest expression quartile (OR=7.48, 95% CI 2.06-27.17, P=0.002 and OR=6.86, 95% CI 1.86-25.26, P=0.004, respectively) (Table 1). Similarly, in T-ALL patients, CCR7 expression levels in the highest quartile conferred an 11-fold increased risk of CNS3 status compared with the lowest quartile (OR=11.00, 95% CI 2.00-60.62, P=0.006) (Table 353


A. Alsadeq et al.

1). In T-ALL patients, CXCR4 expression levels in the highest quartile conferred a â&#x2030;&#x2C6;4-fold increased risk of CNS3 status compared with the lowest quartile in univariate analysis (OR=4.41, 95% CI 1.22-16.00, P=0.024), however, this effect did not hold up in multivariate testing (P=0.070, Table 1). Taken together, these data suggest that the expression of ZAP70 is higher in leukemic cells from diagnostic BM samples of BCP-ALL with overt CNS involvement, and that the same is true for CCR7 and CXCR4 in T-ALL. Also, ZAP70 has an independent predictive impact on the CNS status in BCP-ALL and CCR7 in T-ALL.

Discussion CNS involvement in ALL is linked to leukemic relapses in the CNS. The ability of some ALL blasts to penetrate the blood brain barrier exposes these cells to suboptimal levels of chemotherapeutic drugs, contributing to their survival in a protected niche.33,34 We hypothesized that ZAP70 regulates CNS-leukemia. ZAP70 and SYK, the two members of the Syk kinase family, exert overlapping and distinct functions in lymphocytes. The ZAP70 tyrosine kinase, constitutively expressed in T cells and NK cells, mediates signals downstream of the pre-TCR and mature TCR.35 ZAP70 is involved in early B-lymphocyte development and activation,13 and is also important for B-cell signaling. A recent report identified a subset of ALLs that depend on tonic pre-BCR signaling and is selectively sensitive to the inhibition of SYK and SRC downstream of the pre-BCR.36 In the same study, ZAP70 was shown to be a target of the E2A-PBX1 fusion.36 An upregulation of ZAP70 in E2A-PBX1 positive BCP-ALL has been observed.16,37 Downregulating the expression of ZAP70 in the E2APBX1-positive 697 cell line reduced CNS-positivity in xenografts (Figure 1C). This also applied to REH BCP-ALL cells bearing the TEL-AML1 translocation. Similarly, t(1;19) xenografts bearing ZAP70hi patients exhibited a pronounced CNS-positive phenotype as compared to ZAP70lo patients (Figure 1F). The reduction in ZAP70 expression did not alter the migration of ALL cells to the BM, but rather to the CNS. This observation differs from the situation in B-CLL, where cells with lower ZAP70 expression showed a reduced BM homing.21,22 In JURKAT T-ALL cells, downregulation of ZAP70 did not prevent CNS infiltration; however, it was associated with a reduction in CNS blasts recovered from xenografts (Figure 1D). Downregulating ZAP70 in JURKAT cells did not prolong xenograft survival. On the other hand, animals bearing 697-shZAP70 lived slightly longer and REH-shZAP70 xenografts lived considerably longer than controls (Figure 1E, Online Supplementary Figure S1B,S1C). This may be explained by the REH-model, which causes an almost isolated CNS leukemia in NOD/SCID mice. To investigate how ZAP70 facilitates the migration of ALL cells to the CNS, we hypothesized that ZAP70 regulates chemokine receptors. Chemokines control leukocyte homing into organs, including the CNS.38,39 ALL cells upregulate chemokine receptors, such as CXCR4, CCR3, CCR4 and CCR7,25,40,41 and ZAP70 was shown to enhance the migration of lymphocytes towards chemokines, e.g., CXCL12, CCL21 or CCL19.21,23,24 Herein we show that CCR7/CXCR4 expression was reduced in 697-shZAP70 cells, which impaired their migration toward 354

CCL19/CXCL12. Similarly, enforced ZAP70 expression in 697 cells increased the expression of CCR7/CXCR4 and subsequent migration toward their ligand (Figure 2). Importantly, CCR7 blockade resulted in a reduction in homing processes in vivo analogous to B-CLL42 and, importantly, a reduction in CNS infiltration (Figure 2E-G), which is an extremely rare event in B-CLL.43 Stimulating cells with CCL19/CXCL12 increased p-ERK in 697, JURKAT and BCP-ALL primary cells. Also, basal and CXCL19/CXCL12-inducible p-ERK was reduced in cells bearing a ZAP70 knockdown. CCR7/CXCR4 expression and in vitro migration toward CCL19/CXCL12 were reduced when cells were pre-treated with a MEK inhibitor (Figure 3F,G). The RAS/RAF/MEK/ERK cascade is essential for many cancer cells,44 and hyper-activation of this pathway is common in malignancies due to genetic alterations in the RAS pathway45 and oncogenic tyrosine kinases.46 The use of MEK inhibitors reduced RAS-mutated CNS leukemia in xenografts.31 On the other hand, ERK negative feedback is also a mechanism of leukemia progression in ALL.47 We propose that ZAP70 functions as a regulator of CNS leukemia in ALL via CCR7/CXCR4 and ERK in preclinical in vitro and in vivo models. In addition to providing a chemotactic stimulus, this may confer a niche-specific survival advantage. Most notably, we demonstrated an importance of these markers in large exploratory patient cohorts (Online Supplementary Table S2). We found a correlation between ZAP70 and CCR7/CXCR4 expression (Figure 4A,B) and showed that CNS positive/no relapse BCP-ALL patients expressed higher levels of ZAP70 (Figure 4C). Similarly, CNS positive/no relapse T-ALL patients expressed higher levels of CCR7/CXCR4 (Figure 4D,E). However, there were no significant correlations between CCR7/CXCR4 expression and CNS-positivity in BCP-ALL, and ZAP70 expression and CNS-positivity in T-ALL patients (Online Supplementary Figure S4). Even though we demonstrated a ZAP70-mediated regulation of CCR7/CXCR4 in pre-clinical models, the situation in patients differs. In T-ALL, CCR7/CXCR4 may be regulated by additional mechanisms. In BCP-ALL, the activation of pre-BCR signaling by ZAP70 may be more important than a regulation of chemokine receptors. It is critical to emphasize that a correlation between ZAP70 and CCR7/CXCR4 was observed for all patients, regardless of the immunophenotype. Furthermore, ZAP70 in BCP-ALL and CCR7 in T-ALL hold up in multivariate analyses, suggesting that high marker expression predicts initial CNS infiltration (Table 1). In contrast to recent experimental findings,48 our data suggest that chemokine receptors and their regulation are relevant in CNS leukemia. As our analysis is limited by patient selection in order to analyze sufficient numbers of CNSpositive cases, prospective validation will be important. We conclude that ZAP70 plays a role for the homing to and/or the survival of ALL cells in the CNS and that ZAP70 may represent a therapeutic target. Furthermore, targeting CCR7/CXCR4 may be particularly promising in treating T-ALL.49,50 Acknowledgments D. M. S. is supported by the Max-Eder group leader program by the Deutsche Krebshilfe. We thank Michael Reth and David Medgyesi for providing the pMIG plasmids. We thank Katrin Timm-Richert, Katrin Neumann, Juliane Schmäh and Birthe Fedders for the excellent technical assistance. haematologica | 2017; 102(2)


ZAP70 facilitates CNS infiltration of ALL

References 1. Forestier E, Heyman M, Andersen MK, et al. Outcome of ETV6/RUNX1-positive childhood acute lymphoblastic leukaemia in the NOPHO-ALL-1992 protocol: frequent late relapses but good overall survival. Br J Haematol. 2008;140(6):665-672. 2. Gandemer V, Chevret S, Petit A, et al. Excellent prognosis of late relapses of ETV6/RUNX1-positive childhood acute lymphoblastic leukemia: lessons from the FRALLE 93 protocol. Haematologica. 2012;97(11):1743-1750. 3. Price RA, Johnson WW. The central nervous system in childhood leukemia. I. The arachnoid. Cancer. 1973;31(3):520-533. 4. Burger B, Zimmermann M, Mann G, et al. Diagnostic cerebrospinal fluid examination in children with acute lymphoblastic leukemia: significance of low leukocyte counts with blasts or traumatic lumbar puncture. J Clin Oncol. 2003;21(2):184188. 5. Howard SC, Gajjar AJ, Cheng C, et al. Risk factors for traumatic and bloody lumbar puncture in children with acute lymphoblastic leukemia. Jama. 2002;288(16): 2001-2007. 6. Gajjar A, Harrison PL, Sandlund JT, et al. Traumatic lumbar puncture at diagnosis adversely affects outcome in childhood acute lymphoblastic leukemia. Blood. 2000;96(10):3381-3384. 7. Jeha S, Pei D, Raimondi SC, et al. Increased risk for CNS relapse in pre-B cell leukemia with the t(1;19)/TCF3-PBX1. Leukemia. 2009;23(8):1406-1409. 8. Krause S, Pfeiffer C, Strube S, et al. Mer tyrosine kinase promotes the survival of t(1;19)positive acute lymphoblastic leukemia (ALL) in the central nervous system (CNS). Blood. 2015;125(5):820-830. 9. Pinkel D. Five-year follow-up of "total therapy" of childhood lymphocytic leukemia. Jama. 1971;216(4):648-652. 10. Silverman LB, Declerck L, Gelber RD, et al. Results of Dana-Farber Cancer Institute Consortium protocols for children with newly diagnosed acute lymphoblastic leukemia (1981-1995). Leukemia. 2000;14(12):2247-2256. 11. Clarke M, Gaynon P, Hann I, et al. CNSdirected therapy for childhood acute lymphoblastic leukemia: Childhood ALL Collaborative Group overview of 43 randomized trials. J Clin Oncol. 2003;21(9):1798-1809. 12. Pui CH, Evans WE. Treatment of acute lymphoblastic leukemia. N Engl J Med. 2006;354(2):166-178. 13. Au-Yeung BB, Deindl S, Hsu LY, et al. The structure, regulation, and function of ZAP70. Immunol Rev. 2009;228(1):41-57. 14. Chakupurakal G, Bell A, Griffiths M, et al. Analysis of ZAP70 expression in adult acute lymphoblastic leukaemia by real time quantitative PCR. Mol Cytogenet. 2012;5(1):22. 15. Wandroo F, Bell A, Darbyshire P, et al. ZAP70 is highly expressed in most cases of childhood pre-B cell acute lymphoblastic leukemia. Int J Lab Hematol. 2008;30(2):149157. 16. Chiaretti S, Guarini A, De Propris MS, et al. ZAP-70 expression in acute lymphoblastic leukemia: association with the E2A/PBX1 rearrangement and the pre-B stage of differentiation and prognostic implications. Blood. 2006;107(1):197-204. 17. Rassenti LZ, Huynh L, Toy TL, et al. ZAP-70

haematologica | 2017; 102(2)

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30. 31.

32.

33. 34.

compared with immunoglobulin heavychain gene mutation status as a predictor of disease progression in chronic lymphocytic leukemia. N Engl J Med. 2004;351(9):893901. Chen L, Apgar J, Huynh L, et al. ZAP-70 directly enhances IgM signaling in chronic lymphocytic leukemia. Blood. 2005;105(5): 2036-2041. Chen L, Widhopf G, Huynh L, et al. Expression of ZAP-70 is associated with increased B-cell receptor signaling in chronic lymphocytic leukemia. Blood. 2002;100(13): 4609-4614. Gobessi S, Laurenti L, Longo PG, et al. ZAP70 enhances B-cell-receptor signaling despite absent or inefficient tyrosine kinase activation in chronic lymphocytic leukemia and lymphoma B cells. Blood. 2007; 109(5):20322039. Calpe E, Codony C, Baptista MJ, et al. ZAP70 enhances migration of malignant B lymphocytes toward CCL21 by inducing CCR7 expression via IgM-ERK1/2 activation. Blood. 2011;118(16):4401-4410. Calpe E, Purroy N, Carpio C, et al. ZAP-70 promotes the infiltration of malignant Blymphocytes into the bone marrow by enhancing signaling and migration after CXCR4 stimulation. PLoS One. 2013;8 (12):e81221. Ticchioni M, Charvet C, Noraz N, et al. Signaling through ZAP-70 is required for CXCL12-mediated T-cell transendothelial migration. Blood. 2002;99(9):3111-3118. Ottoson NC, Pribila JT, Chan AS, et al. Cutting edge: T cell migration regulated by CXCR4 chemokine receptor signaling to ZAP-70 tyrosine kinase. J Immunol. 2001; 167(4):1857-1861. Buonamici S, Trimarchi T, Ruocco MG, et al. CCR7 signalling as an essential regulator of CNS infiltration in T-cell leukaemia. Nature. 2009;459(7249):1000-1004. Jost TR, Borga C, Radaelli E, et al. Role of CXCR4-mediated bone marrow colonization in CNS infiltration by T-cell acute lymphoblastic leukemia. J Leukoc Biol. 2016;99(6):1077-1087. Alsadeq A, Strube S, Krause S, et al. Effects of p38alpha/beta inhibition on acute lymphoblastic leukemia proliferation and survival in vivo. Leukemia. 2015;29(12):23072316. Mohammad MG, Tsai VWW, Ruitenberg MJ, et al. Immune cell trafficking from the brain maintains CNS immune tolerance. J Clin Invest. 2014;124(3):1228-1241. Kulathu Y, Hobeika E, Turchinovich G, et al. The kinase Syk as an adaptor controlling sustained calcium signalling and B-cell development. Embo J. 2008;27(9):1333-1344. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 2012;9(7):671-675. Irving J, Matheson E, Minto L, et al. Ras pathway mutations are prevalent in relapsed childhood acute lymphoblastic leukemia and confer sensitivity to MEK inhibition. Blood. 2014;124(23): 3420-3430. Shannon LA, Calloway PA, Welch TP, et al. CCR7/CCL21 migration on fibronectin is mediated by phospholipase Cgamma1 and ERK1/2 in primary T lymphocytes. J Biol Chem. 2010;285(50):38781-28787. Pinkel D. CNS relapse in childhood leukaemia. Lancet. 1981;2(8255):1115. Pui CH, Howard SC. Current management and challenges of malignant disease in the CNS in paediatric leukaemia. Lancet Oncol.

2008;9(3):257-268. 35. Wang H, Kadlecek TA, Au-Yeung BB, et al. ZAP-70: an essential kinase in T-cell signaling. Cold Spring Harb Perspect Biol. 2010;2(5):a002279. 36. Geng H, Hurtz C, Lenz KB, et al. Self-enforcing feedback activation between BCL6 and pre-B cell receptor signaling defines a distinct subtype of acute lymphoblastic leukemia. Cancer Cell. 2015;27(3):409-425. 37. Duque-Afonso J, Wei MC, Lin C-H, et al. Oncogenic role for the Lck/ZAP70/PLCG2 signaling pathway in pre-B-ALL pathogenesis. Blood. 2015;126(23):810. 38. Colmone A, Amorim M, Pontier AL, et al. Leukemic cells create bone marrow niches that disrupt the behavior of normal hematopoietic progenitor cells. Science. 2008;322(5909):1861-1865. 39. Gomez AM, Martinez C, Gonzalez M, et al. Chemokines and relapses in childhood acute lymphoblastic leukemia: A role in migration and in resistance to antileukemic drugs. Blood Cells Mol Dis. 2015;55(3):220-227. 40. Wong S, Fulcher D. Chemokine receptor expression in B-cell lymphoproliferative disorders. Leukemia Lymphoma. 2004; 45(12):2491-2496. 41. Ishida T, Utsunomiya A, Iida S, et al. Clinical significance of CCR4 expression in adult Tcell leukemia/lymphoma: its close association with skin involvement and unfavorable outcome. Clin Cancer Res. 2003;9(10 Pt 1):3625-3634. 42. Cuesta-Mateos C, Loscertales J, Kreutzman A, et al. Preclinical activity of anti-CCR7 immunotherapy in patients with high-risk chronic lymphocytic leukemia. Cancer Immunol Immunother. 2015;64(6):665-676. 43. Akintola-Ogunremi O, Whitney C, Mathur SC, et al. Chronic lymphocytic leukemia presenting with symptomatic central nervous system involvement. Ann Hematol. 2002;81(7):402-404. 44. Steelman LS, Abrams SL, Shelton JG, et al. Dominant roles of the Raf/MEK/ERK pathway in cell cycle progression, prevention of apoptosis and sensitivity to chemotherapeutic drugs. Cell Cycle. 2010;9(8):1629-1638. 45. Zhang J, Mullighan CG, Harvey RC, et al. Key pathways are frequently mutated in high-risk childhood acute lymphoblastic leukemia: a report from the Children's Oncology Group. Blood. 2011; 118(11): 3080-3087. 46. Fielding AK. How I treat Philadelphia chromosome-positive acute lymphoblastic leukemia. Blood. 2010;116(18):3409-3417. 47. Shojaee S, Caeser R, Buchner M, et al. Erk negative feedback control enables pre-B cell transformation and represents a therapeutic target in acute lymphoblastic leukemia. Cancer Cell. 2015;28(1):114-128. 48. Williams MT, Yousafzai YM, Elder A, et al. The ability to cross the blood-cerebrospinal fluid barrier is a generic property of acute lymphoblastic leukaemia blasts. Blood. 2016;127(16):1998-2006. 49. Alfonso-PĂŠrez M, LĂłpez-Giral S, Quintana NE, et al. Anti-CCR7 monoclonal antibodies as a novel tool for the treatment of chronic lymphocyte leukemia. J Leukoc Biol. 2006;79(6):1157-1165. 50. Kashyap MK, Kumar D, Jones H, et al. Ulocuplumab (BMS-936564 / MDX1338): a fully human anti-CXCR4 antibody induces cell death in chronic lymphocytic leukemia mediated through a reactive oxygen speciesdependent pathway. Oncotarget. 2016;7(3): 2809-2822.

355


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Non-Hodgkin Lymphoma

Ferrata Storti Foundation

Haematologica 2017 Volume 102(2):356-363

Early treatment intensification with R-ICE and 90Y-ibritumomab tiuxetan (Zevalin)-BEAM stem cell transplantation in patients with high-risk diffuse large B-cell lymphoma patients and positive interim PET after 4 cycles of R-CHOP-14

Mark Hertzberg,1 Maher K. Gandhi,2,3 Judith Trotman,4 Belinda Butcher,5 John Taper,6 Amanda Johnston,7 Devinder Gill,3 Shir-Jing Ho,8 Gavin Cull,9 Keith Fay,10 Geoff Chong,11 Andrew Grigg,12 Ian D. Lewis,13 Sam Milliken,14 William Renwick,15 Uwe Hahn,16 Robin Filshie,17 George Kannourakis,18 Anne-Marie Watson,19 Pauline Warburton,20 Andrew Wirth,21 John F. Seymour,22 Michael S. Hofman23 and Rodney J. Hicks;23 on behalf of the Australasian Leukaemia Lymphoma Group (ALLG)

Department of Haematology, Prince of Wales Hospital and University of NSW, Randwick, NSW; 2The University of Queensland Diamantina Institute Woolloongabba, Brisbane, QLD and 3Department of Haematology, Princess Alexandra Hospital Brisbane, QLD; 4Department of Haematology, Repatriation General Hospital Concord and University of Sydney, NSW; 5WriteSource Medical Pty Ltd., Lane Cove, NSW; 6Nepean Cancer Care Centre, Nepean Hospital Nepean, NSW; 7Department of Haematology, Westmead Hospital, NSW; 8Department of Haematology, St George Hospital Kogarah, NSW; 9Department of Haematology, Sir Charles Gairdner Hospital Perth, WA; 10 Department of Haematology, Royal North Shore Hospital, St Leonard's, NSW; 11Olivia Newton John Cancer & Wellness Centre, Austin Hospital, Heidelberg, VIC; 12Department of Haematology, Austin Hospital, Heidelberg, VIC; 13Department of Haematology, Royal Adelaide Hospital Adelaide, SA; 14Department of Haematology, St Vincent's Hospital Darlinghurst, NSW; 15Department of Haematology, Royal Melbourne Hospital Parkville, VIC; 16Department of Haematology, The Queen Elizabeth Hospital, SA; 17Department of Haematology, St Vincent’s Hospital Melbourne, VIC; 18Ballarat Oncology and Haematology Service and Fiona Elsey Cancer Research Institute, Ballarat, VIC; 19 Department of Haematology, Liverpool Hospital, Liverpool, NSW; 20Department of Haematology, Wollongong Hospital, Wollongong, NSW; 21Department of Radiation Oncology, Peter MacCallum Cancer Centre East Melbourne, VIC; 22Department of Haematology, Peter MacCallum Cancer Centre East Melbourne and University of Melbourne, Parkville, VIC and 23Department of Cancer Imaging, Peter MacCallum Cancer Centre East Melbourne, VIC, Australia 1

Correspondence: mhertzberg10@gmail.com

Received: August 11, 2016. Accepted: November 3, 2016. Pre-published: November 10, 2016. doi:10.3324/haematol.2016.154039

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/2/356

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

356

ABSTRACT

I

n the treatment of diffuse large B-cell lymphoma, a persistently positive [18F]fluorodeoxyglucose positron emission tomography (PET) scan typically carries a poor prognosis. In this prospective multi-center phase II study, we sought to establish whether treatment intensification with R-ICE (rituximab, ifosfamide, carboplatin, and etoposide) chemotherapy followed by 90Y-ibritumomab tiuxetan–BEAM (BCNU, etoposide, cytarabine, and melphalan) for high-risk diffuse large B-cell lymphoma patients who are positive on interim PET scan after 4 cycles of R-CHOP-14 (rituximab, cyclophosphamide, doxorubicin, and prednisone) can improve 2-year progression-free survival from a historically unfavorable rate of 40% to a rate of 65%. Patients received 4 cycles of R-CHOP-14, followed by a centrally-reviewed PET performed at day 1720 of cycle 4 and assessed according to International Harmonisation Project criteria. Median age of the 151 evaluable patients was 57 years, with 79% stages 3-4, 54% bulk, and 54% International Prognostic Index 3-5. Among the 143 patients undergoing interim PET, 101 (71%) were PET-negative (96 of whom completed R-CHOP), 42 (29%) were PETpositive (32 of whom completed R-ICE and 90Y-ibritumomab tiuxetanBEAM). At a median follow up of 35 months, the 2-year progressionfree survival for PET-positive patients was 67%, a rate similar to that for haematologica | 2017; 102(2)


Treatment intensification in interim PET-positive DLBCL

PET-negative patients treated with R-CHOP-14 (74%, P=0.11); overall survival was 78% and 88% (P=0.11), respectively. In an exploratory analysis, progression-free and overall survival were markedly superior for PET-positive Deauville score 4 versus score 5 (P=0.0002 and P=0.001, respectively). Therefore, diffuse large B-cell lymphoma patients

who are PET-positive after 4 cycles of R-CHOP-14 and who switched to R-ICE and 90Y-ibritumomab tiuxetan-BEAM achieved favorable survival outcomes similar to those for PET-negative R-CHOP14-treated patients. Further studies are warranted to confirm these promising results. (Registered at: ACTRN12609001077257).

Introduction

We chose to evaluate a change to HDT since it is the most widely accepted curative strategy for patients with DLBCL failing R-CHOP. We hypothesized that improved clinical outcomes for high-risk DLBCL patients with poor prognosis as identified by iPET after 4 chemotherapy cycles would be achieved with early HDT and ASCT delivered when there is a lower burden of chemo-resistant disease than if instituted at the time of radiological progression. Given the favorable reports of the use of radioimmunotherapy combined with ASCT for patients with relapsed/refractory DLBCL, the study combined 90Y-ibr