Haematologica, Volume 102, issue 10

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 Settembre 2017.

haematologica calendar of events

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

9th International Nurses Study Day EBMT - Nurses Group October 5, 2017 Manchester, UK ESH 4th International Conference on Acute Myeloid Leukemia "Molecular and Translational": Advances in Biology and Treatment European School of Haematology (ESH) Chairs: B Löwenberg, H Döhner, M Tallman October 5-7, 2017 Estoril, Portugal JACIE Inspector Training Course The European Group for Blood and Marrow Transplantation (EBMT) Chair: E McGrath October 5-6, 2017 Barcelona, Spain 1st Nurses Research Day The European Group for Blood and Marrow Transplantation (EBMT) October 6, 2017 Manchester, UK 20th IDWP Educational Course STATE-OF-THE-ART-2017 The European Group for Blood and Marrow Transplantation (EBMT) October 12-13, 2017 Poznan, Poland EHA Scientific Meeting on Challenges in the Diagnosis and Management of Myeloproliferative Neoplasms Chairs: J Kiladjian and C Harrison October 12-14, 2017 Budapest, Hungary 46° Congresso Nazionale SIE Società Italiana di Ematologia October 15-18, 2017 Chair: F Pane Rome, Italy Crash course on diagnosis and treatment of non-infectious complications after HCT EMBT – Complications and Quality of Life Working Party October 19-20, 2017 Granada, Spain EHA Tutorial on Biology and Management of Myeloid Malignancies October 20-21, 2017 Yerevan, Armenia Russian Onco-Hematology Society's Conference on Malignant Lymphoma - Joint Symposium October 25-26, 2017 Moscow, Russian Federation The 4th International Congress on Controversies in Stem Cell Transplantation and Cellular Therapies COSTEM Chairs: N Kröger, A Nagler October 27-29, 2017 Berlin, Germany Turkish Society of Hematology - EHA Joint Symposium November 1 - 4, 2017 Antalya, Turkey

28th Congress of the Hellenic Society of Haematology Hellenic Society of Haematology Chairs: P Panayotidis, E Terpos November 2-4, 2017 Athina, Greece International 6th ESLHO Symposium: New developments in MRD diagnostics European Scientific foundation for Laboratory Hemato Oncology (ESLHO) Chairs: M Brüggemann, J Trka, O Ottmann, K Döhner, B Durie, A Orfao, C Pott, M Ladetto, T Szczepanski November 9-10, 2017 Leiden, The Netherlands Transfusion-Transmitted Infectious Diseases and Blood Safety European School of Transfusion Medicine (ESTM) Chairs: S Sauleda, M. Schmidt.

November 10-12, 2017 Barcelona, Spain 14th International Conference on Thalassaemia and other Haemoglobinopathies & 16th TIF Conference for Patients and Parents Thalassaemia International Federation (TIF) Chairs: D Loukopoulos, A Taher, J Porter, MD Cappellini November 17-19, 2017 Menemeni, Greece Argentinian Society of Hematology - EHA Joint Education Day November 17-18, 2017 Mar del Plata, Argentina EHA-SWG Scientific Meeting on Shaping the Future of Mesenchymal Stromal Cells Therapy Chairs: W Fibbe, F Dazzi November 23-25, 2017 Amsterdam, The Netherlands EHA-SWG Scientific Meeting on Integrated Diagnosis Strategies in Oncohematology for the management of cytopenias and leukocytosis Chair: MC Béné February 8-10, 2018 Barcelona, Spain EuroClonality Workshop: Clonality assessment in Pathology European Scientific foundation for Laboratory Hemato Oncology (ESLHO) Chairs: PJTA Groenen, F Fend, AW Langerak February 19-21, 2018 Nijmegen, The Netherlands EHA-SWG Scientific Meeting on New Molecular Insights and Innovative Management Approaches for Acute Lymphoblastic Leukemia Chair: N Gökbuget April 12-14, 2018 Location TBC

Calendar of Events updated on August 31, 2017

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

Table of Contents Volume 102, Issue 10: October 2017 Cover Figure Image generated by www.somersault1824.com.

Editorial 1627

Rapamycin targets several pathophysiological features of immune-mediated bone marrow failure in murine models Wendy W. Weston et al.

Review Article 1629

Incidence and management of toxicity associated with ibrutinib and idelalisib: a practical approach Iris de Weerdt et al.

Articles Iron Metabolism & its Disorders

1640

Residual erythropoiesis protects against myocardial hemosiderosis in transfusion-dependent thalassemia by lowering labile plasma iron via transient generation of apotransferrin Maciej W. Garbowski et al.

Blood Transfusion

1650

Amotosalen/ultraviolet A pathogen inactivation technology reduces platelet activatability, induces apoptosis and accelerates clearance Simona Stivala et al.

Platelet Biology & its Disorders

1661

Bone marrow pathologic abnormalities in familial platelet disorder with propensity for myeloid malignancy and germline RUNX1 mutation Rashmi Kanagal-Shamanna et al.

Bone Marrow Failure

1671

Endotoxemia shifts neutrophils with TIMP-free gelatinase B/MMP-9 from bone marrow to the periphery and induces systematic upregulation of TIMP-1 Jennifer Vandooren et al.

1683

Incidence and outcome of acquired aplastic anemia: real-world data from patients diagnosed in Sweden from 2000–2011 Krista Vaht et al.

1691

Rapamycin is highly effective in murine models of immune-mediated bone marrow failure Xingmin Feng et al.

Chronic Myeloid Leukemia

1704

Prognostic discrimination based on the EUTOS long-term survival score within the International Registry for Chronic Myeloid Leukemia in children and adolescents Frédéric Millot et al.

Acute Myeloid Leukemia

1709

Vosaroxin in combination with decitabine in newly diagnosed older patients with acute myeloid leukemia or high-risk myelodysplastic syndrome Naval Daver et al.

Haematologica 2017; vol. 102 no. 10 - October 2017 http://www.haematologica.org/

haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation 1718

Long non-coding RNA expression profile in cytogenetically normal acute myeloid leukemia identifies a distinct signature and a new biomarker in NPM1-mutated patients Etienne De Clara et al.

Acute Lymphoblastic Leukemia

1727

Prolonged versus standard native E. coli asparaginase therapy in childhood acute lymphoblastic leukemia and non-Hodgkin lymphoma: final results of the EORTC-CLG randomized phase III trial 58951 Veerle Mondelaers et al.

1739

Loss-of-function but not dominant-negative intragenic IKZF1 deletions are associated with an adverse prognosis in adult BCR-ABL-negative acute lymphoblastic leukemia Benjamin Kobitzsch et al.

Hodgkin Lymphoma

1748

Secondary malignant neoplasms, progression-free survival and overall survival in patients treated for Hodgkin lymphoma: a systematic review and meta-analysis of randomized clinical trials Dennis A. Eichenauer et al.

Non-Hodgkin Lymphoma

1758

Exome sequencing identifies recurrent BCOR alterations and the absence of KLF2, TNFAIP3 and MYD88 mutations in splenic diffuse red pulp small B-cell lymphoma Laurent Jallades et al.

Plasma Cell Disorders

1767

Impact of prior therapy on the efficacy and safety of oral ixazomib-lenalidomide-dexamethasone vs. placebo-lenalidomide-dexamethasone in patients with relapsed/refractory multiple myeloma in TOURMALINE-MM1 MarĂ­a-Victoria Mateos et al.

1776

The BET bromodomain inhibitor CPI203 improves lenalidomide and dexamethasone activity in in vitro and in vivo models of multiple myeloma by blockade of Ikaros and MYC signaling Tania DĂ­az et al.

Cell Therapy & Immunotherapy

1785

Notch2 blockade enhances hematopoietic stem cell mobilization and homing Weihuan Wang et al.

Complications in Hematology

1796

Characterization of atrial fibrillation adverse events reported in ibrutinib randomized controlled registration trials Jennifer R. Brown et al.

Letters to the Editor Letters are available online only at www.haematologica.org/content/102/10.toc

e379

Temporal contribution of the platelet body and balloon to thrombin generation Ejaife O. Agbani et al. http://www.haematologica.org/content/102/10/e379

e382

Long-term risk of cancer development in adult patients with idiopathic aplastic anemia after treatment with anti-thymocyte globulin Joost G.K. van der Hem et al. http://www.haematologica.org/content/102/10/e382

Haematologica 2017; vol. 102 no. 10 - October 2017 http://www.haematologica.org/

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

e384

Feasibility study of online yoga for symptom management in patients with myeloproliferative neoplasms Jennifer Huberty et al. http://www.haematologica.org/content/102/10/e384

e389

Acute lymphoblastic leukemia cells create a leukemic niche without affecting the CXCR4/CXCL12 axis Bob de Rooij et al. http://www.haematologica.org/content/102/10/e389

e394

Calreticulin as a novel B-cell receptor antigen in chronic lymphocytic leukemia Elisa ten Hacken et al. http://www.haematologica.org/content/102/10/e394

e397

Ibrutinib may impair serological responses to influenza vaccination Abby P. Douglas et al. http://www.haematologica.org/content/102/10/e397

e400

The specific Bruton tyrosine kinase inhibitor acalabrutinib (ACP-196) shows favorable in vitro activity against chronic lymphocytic leukemia B cells with CD20 antibodies JosĂŠe Golay et al. http://www.haematologica.org/content/102/10/e400

e404

Reliable subtype classification of diffuse large B-cell lymphoma samples from GELA LNH2003 trials using the Lymph2Cx gene expression assay Jean-Philippe Jais et al. http://www.haematologica.org/content/102/10/e404

e407

Changes in the incidence of candidemia and related mortality in patients with hematologic malignancies in the last ten years. A SEIFEM 2015-B report Livio Pagano et al. http://www.haematologica.org/content/102/10/e407

e411

Genome-wide association study of clinical parameters in immunoglobulin light chain amyloidosis in three patient cohorts Iman Meziane et al. http://www.haematologica.org/content/102/10/e411

Case Reports Case Reports are available online only at www.haematologica.org/content/102/10.toc

e415

Red blood cell Gardos channel (KCNN4): the essential determinant of erythrocyte dehydration in hereditary xerocytosis RaphaĂŤl Rapetti-Mauss et al. http://www.haematologica.org/content/102/10/e415

e419

An immunocompetent patient with a recurrence-free Epstein-Barr virus positive plasmacytoma possesses robust Epstein-Barr virus specific T-cell responses Bithi Chatterjee et al. http://www.haematologica.org/content/102/10/e419

e423

Failure of long-term lamivudine prophylaxis in patients with resolved hepatitis B infection undergoing chemotherapy and allogenic hematopoietic stem cell transplantation for hematological malignancies: two case reports Glenda Grossi et al. http://www.haematologica.org/content/102/10/e423

Haematologica 2017; vol. 102 no. 10 - October 2017 http://www.haematologica.org/

EDITORIALS Rapamycin targets several pathophysiological features of immune-mediated bone marrow failure in murine models Wendy W. Weston,1,2 Vesna Jurecic1 and Roland Jurecic1 1

Department of Microbiology and Immunology, Miller School of Medicine, University of Miami, FL and 2Cell Therapy Institute, College of Medicine, Nova Southeastern University, Fort Lauderdale, FL, USA E-mail: rjurecic@med.miami.edu doi:10.3324/haematol.2017.175497

I

n this issue of Haematologica, Feng et al.1 compare the efficacy of treatment with cyclosporine A (CsA) and rapamycin to ameliorate pancytopenia, improve bone marrow (BM) cellularity, and extend survival in murine models of immune-mediated aplastic anemia (AA). Interestingly, while the efficacy of CsA and rapamycin to attenuate immune-mediated bone marrow failure (BMF) in murine AA models is similar, CsA and rapamycin achieve their effects through different mechanisms.1 Immune-mediated aplastic anemia (AA) is an acquired form of BMF and is characterized by an abnormally low number of BM cells (hypoplasia) and severe reduction in blood cells (pancytopenia), which in the severe form of AA (SAA) can be life-threatening. The immune and hematologic pathophysiology of AA is quite complex and includes: a) development and oligoclonal expansion of autoreactive T cells, including CD8+ cytotoxic T cells, CD4+ Th1 cells, and Th17 cells; b) effector T-cell-mediated apoptosis and depletion of hematopoietic stem and progenitor cells (HSPCs) and mature blood cells, leading to BM hypoplasia and pancytopenia; c) production of proinflammatory cytokines (e.g.

TNFÎą and IFNÎł); d) severe reduction and functional impairment of immunosuppressive regulatory T cells (Tregs); and e) karyotype abnormalities, genomic instability, and somatic mutations in different myeloid cancer-associated genes that positively and negatively correlate with response to immunosuppresive therapy (IST) and risk of development of myelodysplasia and acute myeloid leukemia (AML).2-6 Current standard treatments for AA include IST with horse anti-thymocyte globulin (ATG) and cyclosporine A (CsA), or allogeneic HLA-matched sibling or well-matched unrelated donor BM transplant. While IST is effective in 6070% of AA patients, a significant proportion of patients who responded to IST undergo relapse after CsA withdrawal or are refractory to IST. Moreover, IST is not effective in treating refractory and relapsed AA.7-12 Recently, combined application of eltrombopag, a thrombopoietin mimetic, and standard IST has proven to be very effective in treating patients with refractory and severe AA. However, relapse and clonal evolution remain important post-therapy concerns.13,14 Different murine models were developed to study the etiology and pathophysiology of AA, and the MHC partially mismatched lymphocyte infusion models, based on alloantigen recognition, are the best characterized and most relevant pre-clinical AA models. The induced AA in these models exhibits many of the clinical and pathological features of acquired AA in patients, and can be modulated using IST and Treg cell therapies. These models provided important insights into the cellular and molecular immune effectors implicated in AA, and are a powerful and relevant in vivo syshaematologica | 2017; 102(10)

tem for testing new drugs and therapeutic approaches for treating SAA.15,16 The AA in lymphocyte infusion models is induced by infusing parental lymph node cells (LNCs) from H2b C57BL/6 mice into MHC partially mismatched non-irradiated or sublethally irradiated F1 hybrid H2b/d B6D2F1 (C57BL6 x DBA/2J) or CByB6F1 (C57BL6 x BALB/c) recipients.

Among mismatched minor-H antigens, H60 contributes the most to AA development in the C.B10 mouse AA model, which is generated by infusion of LNCs from BL6 mice into C.B10 mice which are pre-conditioned with 5 Gy of sublethal total body irradiation (TBI).15,16 Feng et al. have shown that treatment of AA mice with rapamycin for 12 days and treatment with CsA for nine days resulted in similar and statistically significant improvements in BM cellularity, number of white blood cells (WBCs) and platelets (PLTs), and 100-day survival in comparison to untreated AA mice and control mice that received 5 Gy TBI. Temporal studies of recovery of complete blood counts (CBCs) and BM cellularity in rapamycin-treated mice and TBI control mice revealed a similar degree of recovery at days (d)28, 42 and 100, except for a delayed WBC recovery in rapamycin-treated mice.1 Importantly, delayed treatment with rapamycin was also effective in decreasing pancytopenia and BMF in mice with ongoing AA, with a better response from treatment initiated at d5 versus d7 after LN cell infusion. Moreover, the therapeutic effects of a 5-day delay of rapamycin treatment lasted for ten weeks in this experiment, albeit with a significantly slower recovery of WBCs.1 Subsequent experiments have demonstrated that both CsA and rapamycin rescued mice from BM failure by suppressing CD8+ and CD4+ T-cell infiltration in the BM. However, treatment of AA mice with rapamycin led to an increase in functional regulatory T cells in the BM, lymph node and spleen in comparison to mice treated with CsA and untreated AA mice. Furthermore, rapamycin more efficiently eradicates CD8+ effector T cells in a CByB6F1 AA model and antigen-specific CD8+ effector T cells in a minor histocompatibility antigen H60 mismatched AA model.

Additional in vitro and in vivo experiments have revealed that rapamycin treatment is more efficient in reducing memorylike and effector T cells than CsA treatment.1 Transcriptome analyses of BM CD4+ and CD8+ T cells from BMF mice with or without rapamycin or CsA treatment have discovered important differences in the transcription profile of effector molecules important for immune activity of cells, indicating that rapamycin and CsA exert their immuno-modulating effects through different molecular pathways. Furthermore, the analysis of cytokine secretory profiles of T cells from rapamycin and CsA-treated mice 1627

Editorials

revealed significant differences in the effects of rapamycin and CsA on cytokines related to Th1 and Th2 immune responses.1 These important mechanistic findings warrant further molecular and functional studies to uncover the full spectra of molecular and physiological mechanisms of immunosuppression through which rapamycin and CsA ameliorate BMF. Through significant depletion of effector T cells and increase in Treg cells, treatment with rapamycin resulted in improved BM cellularity, significantly lower pancytopenia, and significantly increased numbers of HSPCs in the BM and long-term survival of mice with AA. Thus, similar to other experimental anti-inflammatory and immunosuppressive approaches,16,17 treatment with rapamycin simultaneously and efficiently targets several pathophysiological features of AA in murine models.1 The analysis of HSPC populations in the BM of control, TBI, untreated, CsA-treated and rapamycin-treated mice on d13, has revealed that rapamycin treatment resulted in a 2-3-fold increase in the frequency and numbers of cKit+Sca1+Lin- (KSL), c-Kit-Sca1+Lin- and KSLCD150+ BM cells, which greatly surpass the numbers observed in control mice. Interestingly, in contrast to control, TBI, untreated and CsA-treated mice, c-Kit+Sca1+Lin- (KSL) and c-Kit-Sca1+Lin- cells from rapamycin-treated mice exhibited significantly increased numbers of cells with high expression of Sca-1 marker. It is well established that inflammatory conditions (radiation, chemotherapy, infections) and inflammatory cytokines such as IFNs and TNFα increase Sca-1 expression on HSPCs.18,19 Thus, it is unclear at this point what the cause of significant Sca-1 upregulation is, since the cytokine profiling of plasma from rapamycin-treated mice on d13 has shown that rapamycin significantly down-regulated both IFNγ and TNFα. It would be very interesting to analyze cKit+Sca1+Lin- (KSL), c-Kit-Sca1+Lin- and KSLCD150+ BM cells in rapamycin-treated mice at later time points during their long-term survival.

Analysis of effects of CsA and rapamycin on mTOR and NFAT signaling pathways suggests that CsA suppresses immune activity by interfering with the NFAT1 signaling pathway, whereas rapamycin promotes differentiation of Th2 effector lineages and suppresses proinflammatory Th1 and Th17 T cell lineages by modulating mTOR activity.

In conclusion, the study by Feng et al. demonstrates that, similar to treatment with standard dose of CsA, rapamycin effectively and reproducibly attenuated immune-mediated BM failure in mouse models of AA.1 Although treatment of AA patients with standard IST and rapamycin in a recent clinical trial was not more effi-

1628

cient than standard IST,20 due to its immunosuppressive activity and tolerogenic role in organ transplantation, rapamycin has clinically relevant potential and will be tested in an upcoming phase II clinical trial as a prophylactic treatment of AA patients at high risk of relapse after withdrawal of CsA treatment.

References 1. Feng X, Lin Z, Sun W, Hollinger MK, et al. Rapamycin is highly effective in murine models of immune-mediated bone marrow failure.

Haematologica. 2017;102(10):1691-1703.

2. Young NS. Current concepts in the pathophysiology and treatment of aplastic anemia. Hematology Am Soc Hematol Educ Program.

2013;2013:76-81.

3. Zeng Y, Katsanis E. The complex pathophysiology of acquired aplastic anaemia. Clin Exp Immunol. 2015;180(3):361-370.

4. Boddu PC, Kadia TM. Updates on the pathophysiology and treatment of aplastic anemia: a comprehensive review. Expert Rev Hematol. 2017;10(5):433-448.

5. Ogawa S. Clonal hematopoiesis in acquired aplastic anemia. Blood.

2016;128(3):337-347.

6. Stanley N, Olson TS, Babushok DV. Recent advances in understanding clonal haematopoiesis in aplastic anaemia. Br J Haematol.

2017;177(4):509-525.

7. Georges GE, Storb R. Hematopoietic stem cell transplantation for acquired aplastic anemia. Curr Opin Hematol. 2016;23(6):495-500.

8. Risitano AM. Immunosuppressive therapies in the management of acquired immune-mediated marrow failures. Curr Opin Hematol.

2012;19(1):3-13.

9. Marsh JC, Kulasekararaj AG. Management of the refractory aplastic anemia patient: what are the options? Hematology Am Soc Hematol Educ Program. 2013;2013:87-94.

10. Schrezenmeier H, Körper S, Höchsmann B. Immunosuppressive therapy for transplant-ineligible aplastic anemia patients. Expert Rev Hematol. 2015;8(1):89-99.

11. Dietz AC, Lucchini G, Samarasinghe S, et al. Evolving hematopoietic stem cell transplantation strategies in severe aplastic anemia. Curr Opin Pediatr. 2016;28(1):3-11.

12. Bacigalupo A, Giammarco S, Sica S. Bone marrow transplantation versus immunosuppressive therapy in patients with acquired severe aplastic anemia. Int J Hematol. 2016;104(2):168-174.

13. Townsley DM, Scheinberg P, Winkler T, et al. Eltrombopag Added to Standard Immunosuppression for Aplastic Anemia. N Engl J Med. 2017;376(16):1540-1550.

14. Lum SH, Grainger JD. Eltrombopag for the treatment of aplastic anemia: current perspectives. Drug Des Devel Ther. 2016;10:2833-2843.

15. Chen J. Animal models for acquired bone marrow failure syndromes. Clin Med Res. 2005;3(2):102-108.

16. Scheinberg P, Chen J. Aplastic anemia: what have we learned from animal models and from the clinic. Semin Hematol. 2013;50(2):156164.

17. Weston W, Gupta V, Adkins R, et al. New therapeutic approaches for protecting hematopoietic stem cells in aplastic anemia. Immunol Res. 2013;57(1-3):34-43.

18. King KY, Goodell MA. Inflammatory modulation of HSCs: viewing the HSC as a foundation for the immune response. Nat Rev Immunol. 2011;11(10):685-692.

19. Schuettpelz LG, Link DC. Regulation of hematopoietic stem cell activity by inflammation. Front Immunol. 2013;4:204.

20. Scheinberg P, Wu CO, Nunez O, et al. Treatment of severe aplastic anemia with a combination of horse antithymocyte globulin and cyclosporine, with or without sirolimus: a prospective randomized study. Haematologica. 2009;94(3):348-354.

haematologica | 2017; 102(10)

REVIEW ARTICLE

Incidence and management of toxicity associated with ibrutinib and idelalisib: a practical approach

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Iris de Weerdt,1,2 Suzanne M. Koopmans,3 Arnon P. Kater1,4 and Michel van Gelder3

1 Department of Hematology, Academic Medical Center, Amsterdam; 2Department of Experimental Immunology, Academic Medical Center, Amsterdam; 3Division of Hematology, Department of Internal Medicine, Maastricht University Medical Center, Maastricht and 4Lymphoma and Myeloma Center Amsterdam, LYMMCARE, the Netherlands

Haematologica 2017 Volume 102(10):1629-1639

ABSTRACT

T

he use of novel B-cell receptor signaling inhibitors results in high response rates and long progression-free survival in patients with indolent B-cell malignancies, such as chronic lymphocytic leukemia, follicular lymphoma, mantle cell lymphoma and Waldenström macroglobulinemia. Ibrutinib, the first-in-class inhibitor of Bruton tyrosine kinase, and idelalisib, the first-in-class inhibitor of phosphatidylinositol 3-kinase δ, have recently been approved for the treatment of several indolent B-cell malignancies. These drugs are especially being used for previously unmet needs, i.e., for patients with relapsed or refractory disease, high-risk cytogenetic or molecular abnormalities, or with comorbidities. Treatment with ibrutinib and idelalisib is generally well tolerated, even by elderly patients. However, the use of these drugs may come with toxicities that are distinct from the side effects of immunochemotherapy. In this review we discuss the most commonly reported and/or most clinically relevant adverse events associated with these B-cell receptor inhibitors, with special emphasis on recommendations for their management.

Introduction Recently, a new class of drugs has been introduced for the treatment of various B-cell malignancies, including chronic lymphocytic leukemia (CLL), small lymphocytic lymphoma, follicular lymphoma, mantle cell lymphoma, marginal zone lymphoma and Waldenström macroglobulinemia. These drugs inhibit Bruton tyrosine kinase (BTK) or phosphatidylinositol 3-kinase (PI3K), key components of the B-cell receptor signaling pathway that is crucial for proliferation, survival and homing of malignant B cells.1-6 They are highly effective with respect to induction of remission and prolongation of progression-free survival compared to standard therapies in patients with relapsed or refractory disease, high-risk disease (e.g. CLL with deletion of 17p) or elderly or comorbid patients unfit for immunochemotherapy.

Ibrutinib is currently approved for the treatment of mantle cell lymphoma in patients who have received at least one prior therapy, CLL, Waldenström macroglobulinemia [United States Federal Drug Agency (FDA), European Medicine Agency (EMA)] and marginal zone lymphoma (FDA), and idelalisib is approved for previously treated CLL in combination with rituximab and for follicular lymphoma and small lymphocytic lymphoma in patients who have received at least two prior therapies (FDA, EMA).7-13 Ibrutinib covalently inhibits BTK, which is essential for B-cell homeostasis.

Genetic loss of BTK, as occurs in X-linked agammaglobulinemia, results in the absence of B cells and hypogammaglobulinemia.14 Inhibition of BTK in malignant B cells induces diminished proliferation, decreased survival and impaired adhesion and migration of the malignant B cells to their growth-promoting microenvironment.1-4 Idelalisib is a reversible inhibitor of PI3Kδ. PI3K is a cytoplasmic tyrosine haematologica | 2017; 102(10)

Correspondence: a.p.kater@amc.nl

Received: January 11, 2017.

Accepted: July 6, 2017.

Pre-published: August 3, 2017. doi:10.3324/haematol.2017.164103 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1629 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1629

I. de Weerdt et al.

kinase involved in various signaling pathways, most importantly activating the AKT/mTOR pathway. The δ isoform is ubiquitously expressed in leukocytes.

Inhibition of PI3Kδ induces disruption of interactions between malignant B cells and their microenvironment.

The use of these drugs comes with side effects that are uncommon for immunochemotherapy-based regimens, and in this review an overview is given of their nature and management. Richter transformation is not discussed extensively as it is not an adverse event, although it is important to be aware that Richter transformation is occasionally observed during treatment with B-cell receptor inhibitors.15,16 We performed extensive searches in PubMed and screened published abstracts of the American Society of Hematology, the European Hematology Association and American Society of Clinical Oncology from 2014 up to January 2017 using the search term ‘ibrutinib’ or ‘idelalisib’. We incorporated reports of clinical trials, real-world analyses, meta-analyses, original articles about mechanisms of action or resistance, and articles on specific side effects of interest. Information from clinical trials was used either from the most recent publication, or, when appropriate, from earlier reports in the case that the required details were only given there.

Ibrutinib The currently approved daily dose is 560 mg for patients with mantle cell lymphoma and 420 mg for those with CLL/ small lymphocytic lymphoma and Waldenström macroglobulinemia.9,11,17-19 Ibrutinib has also been combined with the anti-CD20 monoclonal antibodies rituximab or ofatumumab10,20 and with bendamustine plus rituximab in clinical trials.21,23 Ibrutinib is often associated with asymptomatic lymphocytosis upon initiation of treatment. Lymphocytosis has been recognized to be inherent to its mechanism of action, as ibrutinib disrupts integrin-mediated adhesion and homing of malignant B cells to the lymphoid microenvironment, and does not require any specific management even when persistent for months.24

Drug interactions, dose and discontinuation Ibrutinib is metabolized by CYP3A4, and concomitant use of a CYP3A4 inhibitor (e.g. antifungal azoles, macrolides and diltiazem) or CYP3A4 inducer (e.g. rifampicin or carbamazepine) has been demonstrated to have profound effects on serum ibrutinib levels in healthy volunteers.25 Ibrutinib can also increase the levels of P-glycoprotein substrates (e.g. digoxin, dabigatran).26 It has not been definitively established that dose (or serum level) affects tolerability, but two observations suggest that it does. The first observation is the higher discontinuation rate due to adverse events in CLL patients on a higher dose (840 mg/day) than those on the current standard dose of 420 mg/day, although the cohorts were rather small [4/34 (12%) versus 2/51 (4%)].27 The second observation is that patients who experienced inacceptable toxicity were able to continue ibrutinib treatment after dose reduction without progression-free survival being affected.28,29 It seems safe to discontinue ibrutinib for at least 8-14 days without this affecting progression-free survival, e.g.

1630

in the case of invasive procedures (see section on bleeding).29,30 Dose reduction because of adverse events allows the continuation of ibrutinib without affecting progression-free survival.28,29 The discontinuation rate because of adverse events in prospective studies with ibrutinib monotherapy increased over time to 20% after a median time on study of 46 months.31 The incidence of dose modification in two realworld analyses was 19% and 26% (median follow-up of 17 and 16 months, respectively).29,32 The reported incidence of permanent discontinuation varied greatly in realworld experience: two studies reported 11% and 18% discontinuation rates due to adverse events (median followup 10 and 16 months)29,33 and one reported a 51% discontinuation rate due to adverse events (median follow-up 17 months).32 The most frequent reasons for discontinuation or dose reduction varied between the studies and included, in alphabetical order: arthralgia, atrial fibrillation (AF), bleeding, second malignancy, general debility, infection and pneumonitis.29,32,33 Fatal adverse events have been reported in 1-9% of patients on single-agent ibrutinib.8,9,19,33-35

Bruising and clinically relevant bleeding Incidence and severity Safety concerns on the combination of ibrutinib and anticoagulant/antiplatelet (AC/AP) therapy were raised by the company during the first trials. The concerns were based on the observation of incidental severe bleeding, including subdural hematomas and post-invasive procedural bleeding, although precise information on the number of patients and concomitant AC or AP therapy was not released.9,11,27 The observed bleeding events subsequently led to the exclusion of patients on vitamin K antagonist therapy in trials and the strong recommendation to avoid combining ibrutinib with vitamin K antagonists outside clinical trials.8,10,21,36 Additionally, it was advised to withhold ibrutinib 3-7 days before and after invasive procedures depending on the bleeding risk.11,26 In vitro studies demonstrated a collagen-dependent platelet activation defect and absent adherence to von Willebrand factor in 7/14 patients after starting ibrutinib, of whom five had bruising.37 Intriguingly, however, another study found that the platelet function assay already showed impaired aggregation at baseline in 22/85 tested patients, i.e. before starting ibrutinib, with the proportion increasing to 41/85 after starting ibrutinib.38 After initiation of ibrutinib treatment, this study also found inhibition of collagen-induced platelet aggregation, whereas the ADPinduced platelet aggregation improved on ibrutinib therapy. None of the 99 patients in these two studies had major bleeding. These preclinical and clinical findings all raised interest in reporting the incidence and severity of bleeding in patients on ibrutinib.

With the abovementioned restrictions and precautions, lower grade (<3) bleeding (mainly ecchymosis and petechiae presenting during the first 6 months) occurred in 28% of 50 patients unequivocally reported not to be on simultaneous AP or AC therapy.9 A systematic review of four randomized controlled trials confirmed an increased incidence of any grade bleeding, with a 2.93-fold increase (P=0.03) in the ibrutinib compared to the control arms.

The relative risk of major bleeding was 1.72 in the ibrutinib compared to the control arms (P=0.07).39 The addition of ibrutinib to bendamustine-rituximab did not result in a haematologica | 2017; 102(10)

Ibrutinib and idelalisib toxicity

higher incidence of any grade or major bleeding.40 Grade ≥3 bleeding occurred in 2-4% in the studies unambiguously reporting on patients not on concomitant AP or AC therapy (11/392)9,41,42 which seems to be of a similar magnitude as observed in treated patients before the targeted therapy era (6%/year).43 The rate of major hemorrhagic events (grade ≥3 and intracranial bleeding) was similar (3.8%) among the 287 patients treated with ibrutinib and bendamustine-rituximab, without a difference between patients on or not on concomitant AP or AC treatment.21 The incidence of major bleeding in patients simultaneously treated with AP, but not AC, treatment was 2.5% (8/318).33,41,42,44 This is comparable to the 2.2-2.7% risk of major bleeding per year in patients treated with long-term, low-dose aspirin (up to 325 mg) and is of the same magnitude as that in patients on aspirin and clopidogrel (3.7%).45,46 The incidence of major bleeding was 3.2% (2/62) in patients being concomitantly treated with AC [mainly low-molecular-weight heparin or directly acting oral anticoagulants (DOAC)], but not AP.33,41,42,44 In the report with the highest number of patients on concomitant DOAC therapy and detailed information on bleeding incidence, none of the 15 patients developed major bleeding.47 The 3.2% risk of major bleeding among patients on concomitant AC therapy is within the range reported for long-term vitamin K antagonists (3.1%-3.4% per year) or DOAC

(2.1%-3.6% per year) treatment for AF.48-50 As only a few patients had received vitamin K antagonists concomitantly with ibrutinib in the referenced studies, it is uncertain whether co-treatment with a vitamin K antagonist may result in a higher risk of major bleeding.

Experience with ibrutinib in combination with both AP and AC treatment is limited. Major bleeding was reported in 10/48 patients (21%).41,42,44 This incidence seems higher than that reported for dual/triple AP and AC therapy in patients not on ibrutinib treatment (2.6%-14%),46,51,52 despite the possibility that major bleeding in patients on ibrutinib is overestimated due to the low number of patients.

Management The clinically most relevant issues are summarized in Figure 1.

Although grade 1 bruising is very frequent, it does not need to be considered a precursor of major bleeding, nor should bruising lead to ibrutinib discontinuation as in the vast majority of patients it will not advance beyond grade 1 severity and will disappear spontaneously.

The concomitant use of either AP or AC with ibrutinib does not increase the risk of major bleeding, based on the limited follow-up currently available, and does not, therefore, require any specific precautions. Nonetheless, the need for AP or AC therapy should be reconsidered in every case, particularly since it is not unusual that the indi-

Table 1. Adverse events reported during ibrutinib use.

Total (number) Diarrhea, any grade Grade ≥3 Fatigue, any grade Grade ≥3 Arthralgia, any grade Grade ≥3 Bleeding, any grade Grade ≥3 * AF, any grade Grade ≥3 Neutropenia, any grade Grade ≥3 Anemia, any grade Grade ≥3 Thrombocytopenia, any grade Grade ≥3 Infection, any grade Grade ≥3 Febrile neutropenia, any grade Pneumonia, any grade URTI, any grade Cataract, any grade

Previously untreated (19, 62)

Previously treated(8, 9, 11, 20, 34, 65, 92)

165 42-68 4-13 30-32 1-3 16-23 0 NR 4 6 1 16 10-17 16-19 0-6 13 2-3 NR 10 2 NR 17-26 NR

730 29-82 0-7 21-98 2-4 17 0-1 10-50 6-8 4-14 2-12 16-48 0-11 16-48 0-16 17-52 4-13 70-78 24-28 3 10-20 16-28 3

Values represent percentages of patients affected. AF: atrial fibrillation, URTI: upper respiratory tract infection, NR: not reported.

haematologica | 2017; 102(10)

Figure 1. Summary of relevant issues relating to bleeding and anticoagulation during ibrutinib treatment.

1631

I. de Weerdt et al.

cation has expired. As experience with concomitant vitamin K antagonists is almost absent, and because of the warnings in the early days, patients should switch to either low-molecular-weight heparin or a DOAC. If dual or triple therapy with AP and AC is required, alternative antineoplastic therapy should be considered, when available, because of the high risk of major bleeding.

In the event of severe bleeding ibrutinib should be interrupted, although there is no evidence about the efficacy of ibrutinib interruption. However, since temporary discontinuation does not compromise progression-free survival, it seems rational in these cases.

In line with safety measures in clinical trials, perioperative withdrawal of ibrutinib for 3-7 days should be considered for invasive procedures, although interruption may not always be necessary if mechanical hemostasis is feasible. In vitro studies showed that platelet aggregation is fully restored within 5-7 days after ibrutinib cessation, which coincides with the time of physiological platelet restoral.37,53 In the case of serious bleeding, platelet transfusion should be considered even in the absence of thrombocytopenia. As platelet transfusion is expected to be most effective after the ibrutinib half-life interval, repeated platelet transfusions â&#x2030;Ľ3 hours after the last ibrutinib dosage may be considered, although no evidence is available to support this strategy.

(HAS-BLED score),60,61 a DOAC is preferred over a vitamin K antagonist because of the above-mentioned considerations (see section on bleeding) and because of the favorable risk-benefit profile of DOACs in AF patients (Figure 2). Dual or triple AC and AP therapy with concomitant ibrutinib should be avoided, and in these cases alternative anti-lymphoproliferative disease treatment is encouraged.

Hypertension Incidence and severity

The incidence of grade â&#x2030;Ľ3 hypertension requiring medical treatment among patients on ibrutinib therapy increased over time to 26% after 46 months.7,31 After initiation of antihypertensive medication, dose reduction or discontinuation of ibrutinib due to hypertension was not reported to be necessary.19,62

Management Blood pressure should be monitored regularly, especially since hypertension may be co-causal for the development of AF. Hypertension should be managed as usual. The dose of ibrutinib does not need to be reduced nor does the ibrutinib need to be discontinued.

Atrial fibrillation Incidence and severity The incidence of AF was 9% with a median time on ibrutinib of 46 months.31 A meta-analysis of four trials with a median follow-up of 26 months found an incidence of AF of 3.3/100 person-years in patients receiving ibrutinib, and 0.8 in the non-ibrutinib-treated patients.54 The latter incidence is in the same range as that found in 2444 non-ibrutinib-treated patients (1/100 person-years).55 In both ibrutinib- and non-ibrutinib-treated patients, older age, male sex, a history of AF, hypertension and pre-existing cardiac disease increased the likelihood of developing AF.40,55 In a retrospective analysis of 56 AF events during ibrutinib treatment 42% of the patients had grade 3-4 AF (i.e. symptomatic or requiring urgent treatment) and AF was paroxysmal in 64%.47 Ibrutinib treatment also results in an increased incidence of ventricular arrhythmia, which was estimated to be 2/100 person-years versus 0 in non-ibrutinib-treated CLL patients in the randomized clinical trials.56

Management Based on currently available information, it cannot be recommended to withhold ibrutinib when AF develops because this does not seem to result in a higher resolution rate of the AF,47 but does compromise progression-free survival and overall survival (see above). Likewise, dose reduction does not alter the resolution rate of AF.47 Given the observation that once AF has developed ibrutinib withdrawal does not change its course, appropriate treatment of AF should be started as would be done in nonibrutinib-treated patients. The pharmacological interactions with P-glycoprotein substrates (e.g. digoxin, dabigatran), CYP3A4-inhibiting anti-arrhythmic drugs (e.g. verapamil, amiodarone) and certain DOACs (e.g. apixaban, rivaroxaban) should be taken into account (see section on pharmacokinetics).25,57-59 If AC therapy is indicated based on the risk of stroke (CHA2DS2-VASc score) and bleeding 1632

Figure 2. Flowchart for management of atrial fibrillation during ibrutinib use. a e.g. digoxin be.g. verapamil, diltiazem. DOAC: directly acting oral anticoagulants; VKA: vitamin K antagonists.

haematologica | 2017; 102(10)

Ibrutinib and idelalisib toxicity

Severe infections Incidence and severity

Grade ≥3 infections occurred in 10-13% of 60 treatmentnaïve patients7,62 and 24-52% of 407 relapsed/refractory patients on ibrutinib monotherapy.7-9 Addition of ibrutinib to bendamustine-rituximab did not lead to an increased incidence of severe infections, as the exposure-adjusted incidence of severe infections was 2.3 per 100 patientmonths in both groups.21,40 Improved IgA levels (>50% over baseline) are associated with a decreased risk of infection.63 Infection prophylaxis with intravenous immunoglobulin administration that had been started before ibrutinib therapy was stopped in 55% of the patients with relapsed/refractory CLL.7 Of note, although IgG levels remained stable during initial therapy, IgG levels declined after 12 months of ibrutinib.27,63 Five cases with PCR-evidence of Pneumocystis jiroveci pneumonia (PJP; all grade ≤2) were found in one cohort of 96 patients,64 although no other studies reported PJP in ≥1% of their patients.9,63,65

necessary.7-9 Beneficial effects of reducing the dose of ibrutinib in combination with antimotility drugs have occasionally been reported.7,8 However, prolonged discontinuation of ibrutinib (>8-14 days) is not recommended.29,30

Rash Incidence and severity Rash occurs frequently and is generally classified as grade 1 or 2.8,10,21,62,69

Management Rash often recovers spontaneously without any specific treatment. Pruritic rash may require topical corticosteroid therapy and oral histamines. Treatment interruption was

Management For patients on ibrutinib presenting with fever or other signs of infection a thorough work-up should be started to identify the focus and etiological microorganism, including opportunistic pathogens. Treatment of bacterial infections should be based on local resistance patterns. The estimated low incidence of PJP during ibrutinib treatment does not justify PJP prophylaxis.66

Hematologic complications Incidence and severity

Grade ≥3 neutropenia occurred in 10-17% of the patients on ibrutinib monotherapy, usually in the initial months of therapy.8,9,11,19,31,34 Grade ≥3 anemia and thrombocytopenia each occurred in approximately 5% of the patients.7,8,19,34 Ibrutinib did not increase the incidence of cytopenias when added to bendamustine-rituximab.21

Management Dose reduction because of cytopenia has been reported in some patients (with unknown benefit), as has the use of growth factors. Discontinuation of ibrutinib because of cytopenia has seldom been judged necessary.

Autoimmune cytopenia Autoimmune cytopenias that needed treatment before starting ibrutinib may resolve completely, while in some patients a temporary flare or first episode has been observed.67,68 Autoimmune cytopenias could typically be managed with continuation of ibrutinib and temporary addition of standard immunosuppressive treatment (e.g. glucocorticoids, rituximab).68

Diarrhea Incidence and severity Diarrhea has been frequently reported in patients on ibrutinib, but its severity rarely exceeds grade 1.7,8,10,19,62,69 It occurs most often during the first 6 months of treatment,70 and its median duration is 20 days.7

Management Diarrhea is usually self-limiting, and antimotility drugs are only occasionally required.7,9 Dose reduction or discontinuation of ibrutinib because of diarrhea has rarely been haematologica | 2017; 102(10)

Figure 3. Flowchart for management idelalisib-induced diarrhea. GI: gastrointestinal tract. IV: intravenous.

1633

I. de Weerdt et al.

judged necessary in some patients only, but ibrutinib dose-reduction or discontinuation has not been reported for this rash.

Hair and nail alterations Hair alterations were described in 26% of 66 patients during ibrutinib therapy.71 The hair changes were characterized by softening and straightening. Brittle fingernails or splitting of the nails developed in 67%, usually at 6 months after starting ibrutinib, which is consistent with the growth time of nails.

There is only anecdotal evidence that biotin supplementation resulted in some benefit.71

Cataract Although animal studies initially raised concern over an increased incidence of cataract formation during ibrutinib treatment, the observed cataract rate in serial ophthalmological examinations in clinical trials in 506 patients was similar to that observed in the age-matched population.27,72-74

Idelalisib The approved dosage of idelalisib is 150 mg twice daily.

In Europe, idelalisib is currently approved in combination with rituximab for patients with CLL and as monotherapy for patients with relapsed/refractory follicular lymphoma.

Asymptomatic lymphocytosis is frequently seen at the beginning of idelalisib treatment in CLL and small lymphocytic lymphoma, with no need for specific management.

Drug interactions, dose and discontinuation Dose reductions of concomitant CYP3A4 substrates may be needed, since the main metabolite of idelalisib is a potent CYP3A4 inhibitor.75 Strong CYP3A4 inducers (e.g. rifampicin) can decrease idelalisib levels.75 In phase I-II trials, 9-20% of the patients discontinued treatment because of adverse events.13,76-78 A phase III CLL trial (n=110) reported treatment discontinuation because of adverse events in 8% of the patients.12 Serious adverse events, most commonly pneumonia, fever and febrile neu-

Figure 4. Flowchart for management of respiratory complaints during idelalisib use. BAL: broncho-alveolar lavage; HRCT: highresolution computed tomography.

1634

haematologica | 2017; 102(10)

Ibrutinib and idelalisib toxicity

tropenia, occurred in 40% of the patients. Of note, in a trial of idelalisib in combination with rituximab in treatment-naive patients the discontinuation rate due to adverse events was considerably higher, at 45%, mainly due to a higher incidence of diarrhea and/or colitis and pneumonitis.79 Potential explanations include the addition of rituximab, although the rate of adverse events appears higher than in other trials of idelalisib and rituximab, and the fact that patients were previously untreated,2 and thus had a more intact immune system, together with histopathological findings supporting the hypothesis that these adverse events resulted from an autoimmune mechanism.35,80-82 Fatal non-progression adverse events have been reported in 3-8% across trials.13,76-79 It is of note that published data on long-term safety of idelalisib treatment are lacking, as the median follow-up time in the published trials ranges from 3.5 to 9.7 months only.12,13,76,78 Results of a recent planned interim analysis of three ongoing randomized idelalisib trials pointed towards decreased overall survival in the idelalisib arms. The fatal adverse events observed were mostly of an infectious nature, including opportunistic infections, specifically PJP and cytomegalovirus infections, and were more commonly seen in the idelalisib study arms, which led to a major drug warning by Gilead.83 Following completion of an EMA review, the benefit-risk balance of idelalisib in combination with rituximab for the treatment of relapsed CLL, including patients with 17p deletion or TP53 mutation, and idelalisib monotherapy for the treatment of refractory follicular lymphoma remained positive albeit with the strong recommendation to implement safety measures, specifically PJP prophylaxis and regular cytomegalovirus monitoring.

Table 2. Adverse events reported during idelalisib use.

Previously untreated (79) Total (number) Diarrhea and/or colitis, any grade Grade ≥3 Fatigue, any grade Grade ≥3 Cough, any grade Grade ≥3 URTI, any grade Grade ≥3 Pneumonia, any grade Grade ≥3 Pneumonitis, any grade Grade ≥3 AST and/or ALT increased, any grade Grade ≥3 Neutropenia, any grade Grade ≥3 Anemia, any grade Grade ≥3 Thrombocytopenia, any grade Grade ≥3 Febrile neutropenia, any grade

64 64 42 31 0 33 2 NR NR 28 19 3 3 67 23 53 28 23 3 14 2 5

Previously treated (12, 13, 76-78) 393 14-43 4-18 24-36 2-3 13-29 0-4 14-20 0 11-22 6-20 2 2 24-60 2-20 30-57 10-43 23-37 2-11 17-30 5-17 3-11

Values represent percentage of patients affected. URTI: upper respiratory tract infection; AST: aspartate transaminase; ALT: alanine transaminase; NR: not reported.

Diarrhea Incidence and severity Diarrhea can occur at any time after initiation of idelalisib and its incidence is higher in treatment-naïve patients (42%)79 than in patients with relapsed/refractory disease (4-18%).12,13,76,77 Diarrhea that occurs within the first 8 weeks of idelalisib use is usually grade 1-2 (i.e. an increase in stools of up to six stools per day over baseline).

Late-onset diarrhea is generally grade ≥3, with a median time to onset of 7.1 months and there are no accompanying symptoms such as cramps, blood or mucus.82,84 Colonoscopy shows macroscopic, in some cases ulcerative, colitis, and histology shows lymphocytic colitis in combination with characteristic epithelial cell apoptosis and neutrophilic cryptitis.82 Idelalisib-induced intestinal perforation is rare (<0.5%).84 Although a definitive underlying mechanism for idelalisib-associated diarrhea is unknown, PI3Kδ inhibition has been associated with immune dysregulation resulting in inhibition of regulatory T cells and increased damage via CD8+ cytotoxic T cells.35,79-81

blunting and increased intra-epithelial lymphocytes were observed; thus a lactose-free diet may be worth consideration.85 In patients with grade ≥3 diarrhea, or grade 2 diarrhea that is unresolved after 24-48 h, it is advisable to interrupt idelalisib treatment and to start oral or intravenous corticosteroids. The median time to resolution of diarrhea after idelalisib interruption ranged from 1 week to 1 month in various trials. Interruption of idelalisib and concurrent initiation of oral budesonide in 23 patients with grade 3 diarrhea led to resolution in all cases after a mean of 12 days.84 Rechallenge was attempted in 71 patients with grade 3 idelalisib-related diarrhea (out of 106); and 58% were reported to be able to continue idelalisib, although no information on the duration of continuation was provided.84

Pneumonia and pneumonitis Incidence and severity

Management Management of grade 1-2 diarrhea with antidiarrheal agents is usually successful (see Figure 3).82,84 Corticosteroids can be prescribed for ongoing grade 1-2 diarrhea with negative cultures. In patients without upper gastrointestinal tract involvement (e.g. nausea, vomiting), distally released oral corticosteroids may be considered (i.e. budesonide). In a small series of duodenal biopsies in patients with idelalisib-induced diarrhea (n=8), villous haematologica | 2017; 102(10)

Infectious pneumonia is common during idelalisib use with a reported incidence of approximately 20% (n=292); the majority of cases are grade ≥3.12,76,77,79 PJP has been reported in a small number of patients on idelalisib treatment, including a few fatal cases.12 Non-infectious pneumonitis was seen in 3% (n=760) mainly during the first 6 months of idelalisib therapy, and was usually severe, with some fatal cases.84 Clinical symptoms include coughing, dyspnea and fever progressing over weeks. Various abnor1635

I. de Weerdt et al.

malities are observed with computed tomography, including ground-glass opacities, consolidation and pleural effusion.86

Management If patients present with respiratory complaints clinically and radiologically compatible with lobar bacterial pneumonia, empiric antibiotic treatment should be started promptly. Interruption of idelalisib is not routinely advised, since idelalisib is not presumed to cause bacterial pneumonia and no beneficial effects of idelalisib interruption or dose reduction have been reported.

In patients with grade â&#x2030;Ľ2 respiratory complaints and no clear bacterial pneumonia or lack of clinical response to empiric antibiotic treatment, high-resolution computed tomography should be performed. In the presence of imaging abnormalities incompatible with lobar pneumonia, broncho-alveolar lavage should be performed to exclude infectious causes, which would require markedly different treatment, and idelalisib should be interrupted while awaiting the results of culture of the lavage fluid, as treatment continuation may be fatal in idelalisib-induced pneumonitis (see Figure 4). In the absence of high-resolution computed tomography abnormalities, pulmonary function testing, including oxygen diffusion capacity, may be considered and inhaled steroids could be prescribed.

When pneumonia is excluded and pneumonitis is highly suspected, individual reports have described beneficial effects of corticosteroids in addition to cessation of idelalisib.84,86 Among 13 patients with pneumonitis who were rechallenged with idelalisib (out of 24), two-thirds were able to continue idelalisib.87 Idelalisib should not be reintroduced if the idelalisib-induced pneumonitis was lifethreatening Almost all cases of PJP occurred in patients not receiving PJP prophylaxis, which prompted the EMA to recommend PJP prophylaxis for up to 2 to 6 months after treatment discontinuation, depending on concurrent immunosuppressive drug use and neutropenia.83,88

Figure 5. Flowchart for management of transaminitis during idelalisib treatment. AST: aspartate transaminase; ALT: alanine transaminase; ULN: upper limit of normal; BID: bis in die.

Hepatotoxicity Incidence and severity Hepatotoxicity is most often seen during the first 3 months of idelalisib treatment and is characterized by an elevation of alanine transaminase (ALT) and aspartate transaminase (AST) blood levels. The incidence of ALT and AST elevations of any grade is 50%, with grade â&#x2030;Ľ3 increases occurring in 16%.84,87 Among 1192 patients treated in idelalisib clinical trials, one fatal case (<0.1%) of hepatoxicity occurred in a patient treated with idelalisib and ofatumumab.84 Hepatotoxicity was more prevalent in younger, previously untreated patients.35,79 The median time to onset of grade â&#x2030;Ľ3 ALT/AST elevations was 1.4 months.

Management ALT and AST should be monitored frequently, especially in the first months of treatment. If hepatotoxicity occurs, the liver enzymes should be monitored every week until it is resolved (Figure 5). Idelalisib treatment can be continued if ALT/AST elevations three to five times the upper limit of normal (ULN) occur, with close monitoring of the liver enzymes. Idelalisib should be discontinued if ALT and AST elevations reach 5-20 times the ULN. 1636

Idelalisib can be reinitiated at a lower dose of 100 mg twice daily when ALT and AST levels have returned to normal. If ALT and AST elevations do not recur at the idelalisib dose of 100 mg twice daily, re-escalating the idelalisib dose to 150 mg twice daily can be considered at the discretion of the treating physician. After dose interruption, the elevations in liver enzymes are reversible in the majority of patients and do not recur after reinitiating idelalisib at a lower dose.89 Idelalisib should be permanently discontinued if ALT/AST levels reach more than 20 times the ULN.84,90 Idelalisib is well-tolerated in patients with pre-existing moderate or severe hepatic impairment.91 Therefore, dose adjustment beforehand is not necessary in patients with prior hepatic impairment, and it is advised to monitor patients as described above.

Hematologic complications Incidence and severity Neutropenia is common during the first months of idelalisib treatment. Any grade neutropenia occurs in 44-57% haematologica | 2017; 102(10)

Ibrutinib and idelalisib toxicity

of the patients, with the neutropenia being grade 3-4 in 23-43%.12,76,77 Across trials, GM-CSF was administered to 16-25% of the patients, whereas dose reduction was rarely judged necessary (1%) and the drug was withheld in <0.05% because of neutropenia.84 Anemia occurred in 23-37% (grade 3: 3-11%) of the patients during idelalisib reatment. Similarly, thrombocytopenia occurred in 1730% (grade 3: 5-17%).12,13,76,77

Management Blood counts should be monitored frequently during the first months of idelalisib treatment. In the case of persistent neutropenia, temporary growth factor support can be considered.

Rash Incidence and severity Any grade rash was reported in 10-22% of patients with relapsed or relapsed/refractory disease, with grade â&#x2030;Ľ3 rash occurring in 0-2%.12,77 The reported frequency of rash was considerably higher in treatment-naĂŻve patients at 58% (grade 3: 13%).79

Management If serious cutaneous reactions occur during idelalisib treatment, the drug should be discontinued. The efficacy of antihistamines or steroids has not been described.

Conclusion Novel B-cell receptor inhibitors have been shown to be effective in the treatment of indolent B-cell malignancies.

Ibrutinib and idelalisib, the first two approved B-cell receptor pathway inhibitors, are administered orally and continuously. Their use results in high response rates and long progression-free survival even in patients with highrisk, relapsed or refractory disease.

Clinical trials have shown acceptable safety profiles of these drugs. Nonetheless, both agents have toxicity profiles that are different from those of immunochemotherapy (Figure 6). Moreover, the safety profile of ibrutinib is clearly distinct from that of idelalisib and this should be taken into consideration when making individual treatment decisions.

During ibrutinib treatment, bleeding and AF can pose especially complex treatment dilemmas, whereas diarrhea, pneumonitis and ALT/AST elevations are challenging during idelalisib treatment. Appropriate management and awareness of these adverse events is especially important in the light of continuous use of B-cell receptor inhibitors.

References 1. Ponader S, Chen SS, Buggy JJ, et al. The Bruton tyrosine kinase inhibitor PCI-32765 thwarts chronic lymphocytic leukemia cell survival and tissue homing in vitro and in vivo. Blood. 2012;119(5):1182-1189.

2. de Rooij MF, Kuil A, Geest CR, et al. The clinically active BTK inhibitor PCI-32765 targets B-cell receptor- and chemokine-con-

haematologica | 2017; 102(10)

Figure 6. Recommendations for the clinic summarizing important toxicity-related issues during therapy with ibrutinib or idelalisib. DOAC: directly acting oral anticoagulants; LMWH: low-molecular-weight heparin; PJP: Pneumocystis jiroveci pneumonia; CMV: cytomegalovirus.

Continued monitoring of toxicity associated with B-cell receptor inhibitors is essential and should be preferably reported using common terminology (i.e. Common Terminology Criteria for Adverse Events). This is particularly important for long-term safety data, as they are currently largely absent. Mechanistic insights and increased experience will likely lead to improved management strategies for the prevention of serious complications.

Acknowledgments The authors would like to thank F. Moenen, M.D., for searching literature on baseline incidence of major bleeding in patients on antiplatelet and anticoagulant therapy.

trolled adhesion and migration in chronic lymphocytic leukemia. Blood. 2012;119(11): 2590-2594.

3. Herman SE, Mustafa RZ, Gyamfi JA, et al.

Ibrutinib inhibits BCR and NF-kappaB signaling and reduces tumor proliferation in tissue-resident cells of patients with CLL. Blood. 2014;123(21):3286-3295.

4. Cheng S, Ma J, Guo A, et al. BTK inhibition targets in vivo CLL proliferation through its effects on B-cell receptor signaling activity.

Leukemia. 2014;28(3):649-657.

5. Hoellenriegel J, Meadows SA, Sivina M, et al. The phosphoinositide 3'-kinase delta inhibitor, CAL-101, inhibits B-cell receptor signaling and chemokine networks in chronic lymphocytic leukemia.

Blood.

2011;118(13):3603-3612.

6. Lannutti BJ, Meadows SA, Herman SE, et al.

CAL-101, a p110delta selective phosphatidylinositol-3-kinase inhibitor for the treatment of B-cell malignancies, inhibits

1637

I. de Weerdt et al.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

1638

PI3K signaling and cellular viability. Blood.

2011;117(2):591-594.

Byrd JC, Furman RR, Coutre SE, et al. Threeyear follow-up of treatment-naive and previously treated patients with CLL and SLL receiving single-agent ibrutinib. Blood.

2015;125(16):2497-2506.

Byrd JC, Brown JR, O'Brien S, et al. Ibrutinib versus ofatumumab in previously treated chronic lymphoid leukemia. N Engl J Med. 2014;371(3):213-223.

Wang ML, Blum KA, Martin P, et al. Longterm follow-up of MCL patients treated with single-agent ibrutinib: updated safety and efficacy results. Blood. 2015;126(6):739745.

Burger JA, Keating MJ, Wierda WG, et al.

Safety and activity of ibrutinib plus rituximab for patients with high-risk chronic lymphocytic leukaemia: a single-arm, phase 2 study. Lancet Oncol. 2014;15(10):10901099.

Treon SP, Tripsas CK, Meid K, et al. Ibrutinib in previously treated Waldenstrom's macroglobulinemia. N Engl J Med. 2015;372(15):1430-1440.

Furman RR, Sharman JP, Coutre SE, et al.

Idelalisib and rituximab in relapsed chronic lymphocytic leukemia. N Engl J Med. 2014;370(11):997-1007.

Kahl BS, Spurgeon SE, Furman RR, et al. A phase 1 study of the PI3Kdelta inhibitor idelalisib in patients with relapsed/refractory mantle cell lymphoma (MCL). Blood.

2014;123(22):3398-3405.

Mohamed AJ, Yu L, Backesjo CM, et al.

Bruton's tyrosine kinase (Btk): function, regulation, and transformation with special emphasis on the PH domain. Immunol Rev. 2009;228(1):58-73.

Wiestner A. The role of B-cell receptor inhibitors in the treatment of patients with chronic lymphocytic leukemia.

Haematologica. 2015;100(12):1495-1507.

Woyach JA. Patterns of resistance to B cellreceptor pathway antagonists in chronic lymphocytic leukemia and strategies for management. Hematology Am Soc Hematol Educ Program. 2015;2015:355-360.

Wang ML, Rule S, Martin P, et al. Targeting BTK with ibrutinib in relapsed or refractory mantle-cell lymphoma. N Engl J Med. 2013;369(6):507-516.

Byrd JC, Furman R, Coutre S, et al. The Bruton's tyrosine kinase (BTK) inhibitor PCI-32765 (P) in treatment-naive (TN) chronic lymphocytic leukemia (CLL) patients (pts): interim results of a phase Ib/II study [abstract]. J Clin Oncol. 2012;30.

Burger JA, Tedeschi A, Barr PM, et al.

Ibrutinib as initial therapy for patients with chronic lymphocytic leukemia. N Engl J Med. 2015;373(25):2425-2437.

Jaglowski SM, Jones JA, Nagar V, et al.

Safety and activity of BTK inhibitor ibrutinib combined with ofatumumab in chronic lymphocytic leukemia: a phase 1b/2 study.

Blood. 2015;126(7):842-850.

Chanan-Khan A, Cramer P, Demirkan F, et al. Ibrutinib combined with bendamustine and rituximab compared with placebo, bendamustine, and rituximab for previously treated chronic lymphocytic leukaemia or small lymphocytic lymphoma (HELIOS): a randomised, double-blind, phase 3 study.

Lancet Oncol. 2016;17(2):200-211.

Maddocks K, Christian B, Jaglowski S, et al.

A phase 1/1b study of rituximab, bendamustine, and ibrutinib in patients with untreated and relapsed/refractory non-Hodgkin lymphoma. Blood. 2015;125(2):242-248.

23. Brown JR, Barrientos JC, Barr PM, et al. The Bruton tyrosine kinase inhibitor ibrutinib with chemoimmunotherapy in patients with chronic lymphocytic leukemia. Blood.

2015;125(19):2915-2922.

24. Woyach JA, Smucker K, Smith LL, et al.

Prolonged lymphocytosis during ibrutinib therapy is associated with distinct molecular characteristics and does not indicate a suboptimal response to therapy. Blood.

2014;123(12):1810-1817.

25. de Zwart L, Snoeys J, De Jong J, et al.

Ibrutinib dosing strategies based on interaction potential of CYP3A4 perpetrators using physiologically based pharmacokinetic modeling. Clin Pharmacol Ther. 2016;100 (5):548-557.

26. Prescribing information: IMBRUVICA (ibrutinib) capsules, for oral use. 2016 [cited 2016 September 3]; Available from: https://www.janssenmd.com/pdf/imbruvica/PI-Imbruvica.pdf 27. Byrd JC, Furman RR, Coutre SE, et al.

Targeting BTK with ibrutinib in relapsed chronic lymphocytic leukemia. N Engl J Med. 2013;369(1):32-42.

28. Mato AR, Timlin C, Ujjani C, et al.

Comparable outcomes in chronic lymphocytic leukaemia (CLL) patients treated with reduced-dose ibrutinib: results from a multicentre study. Br J Haematol. 2017 Feb 21.

[Epub ahead of print] 29. Forum UC. Ibrutinib for relapsed/refractory chronic lymphocytic leukemia: a UK and Ireland analysis of outcomes in 315 patients.

Haematologica. 2016;101(12):1563-1572.

30. Barr PM, Brown JR, Hillmen P, et al. Impact of ibrutinib dose adherence on therapeutic efficacy in patients with previously treated CLL/SLL. Blood. 2017;129(19):2612-2615.

31. O'Brien SM, Furman RR, Coutre SE, et al.

Five-year experience with single-agent ibrutinib in patients with previously untreated and relapsed/refractory chronic lymphocytic leukemia/small lymphocytic leukemia.

Blood. 2016;128:233.

32. Mato AR, Hill BT, Lamanna N, et al.

Optimal sequencing of ibrutinib, idelalisib, and venetoclax in chronic lymphocytic leukemia: results from a multi-center study of 683 patients. Ann Oncol. 2017;28(5): 1050-1056.

33. Winqvist M, Asklid A, Andersson PO, et al.

Real--world results of ibrutinib in patients with relapsed or refractory chronic lymphocytic leukemia: data from 95 consecutive patients treated in a compassionate use program. Haematologica. 2016;101(12):15731580.

34. Dreyling M, Jurczak W, Jerkeman M, et al.

Ibrutinib versus temsirolimus in patients with relapsed or refractory mantle-cell lymphoma: an international, randomised, openlabel, phase 3 study. Lancet. 2016;387 (10020):770-778.

35. Lampson BL, Kasar SN, Matos TR, et al.

Idelalisib given front-line for treatment of chronic lymphocytic leukemia causes frequent immune-mediated hepatotoxicity.

Blood. 2016;128(2):195-203.

36. Imbruvica: summary of product characteristics. 2016 [cited 2016 November 13]; Available from: http://www.ema.europa.eu/ docs/en_GB/document_library/EPAR__Product_Information/human/003791/WC5 00177775.pdf 37. Levade M, David E, Garcia C, et al. Ibrutinib treatment affects collagen and von Willebrand factor-dependent platelet functions. Blood. 2014;124(26):3991-3995.

38. Lipsky AH, Farooqui MZ, Tian X, et al.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

Incidence and risk factors of bleeding-related adverse events in patients with chronic lymphocytic leukemia treated with ibrutinib.

Haematologica. 2015;100(12):1571-1578.

Yun S, Vincelette ND, Acharya U, Abraham I. Risk of atrial fibrillation and bleeding diathesis associated with ibrutinib treatment: a systematic review and pooled analysis of four randomized controlled trials. Clin Lymphoma Myeloma Leuk. 2017;17(1):3137 e13.

Chanan-Khan A, Cramer P, Demirkan F, et al. Insights into the management of adverse events in patients with previously treated chronic lymphocytic leukemia/small lymphocytic lymphoma: experience from the phase 3 HELIOS study of ibrutinib combined with bendamustine/rituximab. Blood.

2015;126(23):2936-2936.

Pavlik A, Barr H, Dotson E, et al. Major bleeding complications among patients treated with ibrutinib and concomitant antiplatelet, anticoagulant, or supplemental therapy. Blood. 2016; 128:4387.

Jones JA, Hillmen P, Coutre S, et al. Use of anticoagulants and antiplatelet in patients with chronic lymphocytic leukaemia treated with single-agent ibrutinib. Br J Haematol.

2017 Apr 10. [Epub ahead of print] Gifkins DM, Matcho A, Yang H, et al.

Incidence of major hemorrhage among CLL and MCL patients compared to the general elderly population: an analysis of the US SEER-Medicare linked database. Blood.

2015;126:3268.

Kunk PR, Mock J, Devitt ME, et al. Major bleeding with ibrutinib: more than expected. Blood. 2016;128:3229.

Yusuf S, Zhao F, Mehta SR, et al. Effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without STsegment elevation. N Engl J Med. 2001;345(7):494-502.

Warfarin Antiplatelet Vascular Evaluation Trial Investigators, Anand S, Yusuf S, et al.

Oral anticoagulant and antiplatelet therapy and peripheral arterial disease. N Engl J Med. 2007;357(3):217-227.

Thompson PA, Levy V, Tam CS, et al. Atrial fibrillation in CLL patients treated with ibrutinib. An international retrospective study.

Br J Haematol. 2016;175(3):462-466.

Connolly SJ, Ezekowitz MD, Yusuf S, et al.

Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361(12):1139-1151.

Patel MR, Mahaffey KW, Garg J, et al.

Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011;365 (10):883-891.

Granger CB, Alexander JH, McMurray JJ, et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2011; 365(11):981-992.

Rossini R, Musumeci G, Lettieri C, et al.

Long-term outcomes in patients undergoing coronary stenting on dual oral antiplatelet treatment requiring oral anticoagulant therapy. Am J Cardiol. 2008;102 (12):1618-1623.

Hermosillo AJ, Spinler SA. Aspirin, clopidogrel, and warfarin: is the combination appropriate and effective or inappropriate and too dangerous? Ann Pharmacother. 2008;42(6): 790-805.

Kamel S, Horton L, Ysebaert L, et al.

Ibrutinib inhibits collagen-mediated but not ADP-mediated platelet aggregation.

Leukemia. 2015;29(4):783-787.

Leong DP, Caron F, Hillis C, et al. The risk of atrial fibrillation with ibrutinib use: a sys-

haematologica | 2017; 102(10)

Ibrutinib and idelalisib toxicity

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

tematic review and meta-analysis. Blood.

2016;128(1):138-140.

Shanafelt TD, Parikh SA, Noseworthy PA, et al. Atrial fibrillation in patients with chronic lymphocytic leukemia (CLL). Leuk Lymphoma. 2016:1-10.

Lampson BL, Yu L, Glynn RJ, et al.

Ventricular arrhythmias and sudden death in patients taking ibrutinib.

Blood.

2017;129(18):2581-2584. de Jong J, Skee D, Murphy J, et al. Effect of CYP3A perpetrators on ibrutinib exposure in healthy participants. Pharmacol Res Perspect. 2015;3(4):e00156.

Scheers E, Leclercq L, de Jong J, et al.

Absorption, metabolism, and excretion of oral (14)C radiolabeled ibrutinib: an open-label, phase I, single-dose study in healthy men. Drug Metab Dispos. 2015; 43(2):289-297.

Lambert Kuhn E, Leveque D, Lioure B, Gourieux B, Bilbault P. Adverse event potentially due to an interaction between ibrutinib and verapamil: a case report. J Clin Pharm Ther. 2016;41(1):104-105.

Vrontikis A, Carey J, Gilreath JA, et al.

Proposed algorithm for managing ibrutinibrelated atrial fibrillation. Oncology (Williston Park). 2016;30(11):970-974.

Kirchhof P, Benussi S, Kotecha D, et al. 2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Eur Heart J. 2016;37:2893-2962.

O'Brien S, Furman RR, Coutre SE, et al.

Ibrutinib as initial therapy for elderly patients with chronic lymphocytic leukaemia or small lymphocytic lymphoma: an open-label, multicentre, phase 1b/2 trial.

Lancet Oncol. 2014;15(1):48-58.

Sun C, Tian X, Lee YS, et al. Partial reconstitution of humoral immunity and fewer infections in patients with chronic lymphocytic leukemia treated with ibrutinib. Blood.

2015;126(19):2213-2219.

Ahn IE, Jerussi T, Farooqui M, et al. Atypical Pneumocystis jirovecii pneumonia in previously untreated patients with CLL on singleagent ibrutinib. Blood. 2016;128(15):19401943.

O'Brien S, Jones JA, Coutre SE, et al.

Ibrutinib for patients with relapsed or refractory chronic lymphocytic leukaemia with 17p deletion (RESONATE-17): a phase 2, open-label, multicentre study. Lancet Oncol.

2016;17(10):1409-1418.

Green H, Paul M, Vidal L, Leibovici L. Prophylaxis of Pneumocystis pneumonia in immunocompromised non-HIV-infected patients: systematic review and meta-analysis of randomized controlled trials. Mayo Clin Proc. 2007;82(9):1052-1059.

Rogers KA, Ruppert AS, Bingman A, et al.

haematologica | 2017; 102(10)

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

Incidence and description of autoimmune cytopenias during treatment with ibrutinib for chronic lymphocytic leukemia.

Leukemia. 2016;30(2):346-350.

Vitale C, Ahn IE, Sivina M, et al.

Autoimmune cytopenias in patients with chronic lymphocytic leukemia treated with ibrutinib. Haematologica. 2016;101(6):e254258.

Farooqui MZ, Valdez J, Martyr S, et al.

Ibrutinib for previously untreated and relapsed or refractory chronic lymphocytic leukaemia with TP53 aberrations: a phase 2, single-arm trial. Lancet Oncol. 2015;16(2): 169-176.

Barrientos JC, O'Brien S, Brown JR, et al.

Hematologic and immunologic function and patient well-being for the phase III RESONATETM study of ibrutinib vs ofatumumab in relapsed/refractory chronic lymphocytic leukemia/small lymphocytic lymphoma. Blood. 2014;124(21):4696-4696.

Bitar C, Farooqui MZ, Valdez J, et al. Hair and nail changes during long-term therapy with ibrutinib for chronic lymphocytic leukemia. JAMA Dermatol. 2016;152(6): 698-701.

Klein BE, Klein R, Lee KE. Incidence of agerelated cataract over a 10-year interval: the Beaver Dam Eye Study. Ophthalmology.

2002;109(11):2052-2057.

Neffendorf JE, Gout I, Hildebrand GD. Ibrutinib in relapsed chronic lymphocytic leukemia. N Engl J Med. 2013;369(13):1277.

Byrd JC, O'Brien S, James DF. Ibrutinib in relapsed chronic lymphocytic leukemia. N Engl J Med. 2013;369(13):1278-1279.

Jin F, Robeson M, Zhou H, et al. Clinical drug interaction profile of idelalisib in healthy subjects. J Clin Pharmacol.

2015;55(8):909-919.

Flinn IW, Kahl BS, Leonard JP, et al. Idelalisib, a selective inhibitor of phosphatidylinositol 3-kinase-delta, as therapy for previously treated indolent non-Hodgkin lymphoma.

Blood. 2014;123(22):3406-3413.

Brown JR, Byrd JC, Coutre SE, et al.

Idelalisib, an inhibitor of phosphatidylinositol 3-kinase p110delta, for relapsed/refractory chronic lymphocytic leukemia. Blood.

2014;123(22):3390-3397.

Gopal AK, Kahl BS, de Vos S, et al. PI3Kdelta inhibition by idelalisib in patients with relapsed indolent lymphoma. N Engl J Med. 2014;370(11):1008-1018.

O'Brien SM, Lamanna N, Kipps TJ, et al. A phase 2 study of idelalisib plus rituximab in treatment-naive older patients with chronic lymphocytic leukemia. Blood. 2015;126(25): 2686-2694.

Patton DT, Garden OA, Pearce WP, et al.

Cutting edge: the phosphoinositide 3-kinase

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

p110 delta is critical for the function of CD4+CD25+Foxp3+ regulatory T cells. J Immunol. 2006;177(10):6598-6602.

Balagoni H, Chaudhari D, Reddy C, Young M. Idelalisib: a rare cause of enterocolitis.

Ann Gastroenterol. 2016;29(2):233-235.

Weidner AS, Panarelli NC, Geyer JT, et al.

Idelalisib-associated colitis: histologic findings in 14 patients. Am J Surg Pathol.

2015;39(12):1661-1667.

Direct Healthcare Professional Communication: Restrictions on the use of Zydelig for the treatment of CLL and relapsed FL following new clinical trial results. 2016 [cited 2016 September 3]; Available from: h t t p s : / / w w w. h p r a . i e / d o c s / d e f a u l t source/default-document-library/importantsafety-information---zydelig(idelalisib).pdf?sfvrsn=0 Coutre SE, Barrientos JC, Brown JR, et al.

Management of adverse events associated with idelalisib treatment: expert panel opinion. Leuk Lymphoma. 2015;56(10):27792786.

Louie CY, DiMaio MA, Matsukuma KE, et al. Idelalisib-associated enterocolitis: clinicopathologic features and distinction from other enterocolitides. Am J Surg Pathol.

2015;39(12):1653-1660.

Haustraete E, Obert J, Diab S, et al.

Idelalisib-related pneumonitis. Eur Respir J. 2016;47(4):1280-1283.

Coutre SE, Barrientes JC, Brown JR et al.

Safety of idelalisib in B-cell malignancies: integrated analysis of eight clinical trials. J Clin Oncol. 2015;33(supp; abstr e18030).

CHMP assessment report: Zydelig. 2016 [cited 2016 September 18]; Available from: http://www.ema.europa.eu/docs/ e n _ G B / d o c u m e n t _ l i b r a r y / E PA R _ _Assessment_Report_-_Variation/human/ 003843/WC500216147.pdf Wierda WG Coutre S, Flinn I, et al.

Management of transaminase elevations in patients receiving idelalisib. J Clin Oncol.

2016;34(suppl; abstr 7532).

Prescribing information: ZYDELIG (idelalisib) tablets, for oral use. 2016 [cited 2016 September 3]; Available from: http://www.gilead.com/~/media/Files/pdfs/ medicines/oncology/zydelig/zydelig_pi.pdf Jin F, Robeson M, Zhou H, Hisoire G, Ramanathan S. The pharmacokinetics and safety of idelalisib in subjects with moderate or severe hepatic impairment. J Clin Pharmacol. 2015;55(8):944-952.

Wang ML, Lee H, Chuang H, et al. Ibrutinib in combination with rituximab in relapsed or refractory mantle cell lymphoma: a single-centre, open-label, phase 2 trial. Lancet Oncol. 2016;17(1):48-56.

1639

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Iron Metabolism & Its Disorders

Ferrata Storti Foundation

Haematologica 2017 Volume 102(10):1640-1649

Residual erythropoiesis protects against myocardial hemosiderosis in transfusiondependent thalassemia by lowering labile plasma iron via transient generation of apotransferrin

Maciej W. Garbowski,1,2 Patricia Evans,1 Evangelia Vlachodimitropoulou,1 Robert Hider3 and John B. Porter1,2

Research Haematology Department, Cancer Institute, University College London; University College London Hospitals and 3Institute of Pharmaceutical Sciences, Kingâ&#x20AC;&#x2122;s College London, UK 1 2

ABSTRACT

C

Correspondence: maciej.garbowski@ucl.ac.uk

Received: April 14, 2017.

Accepted: June 20, 2017.

Pre-published: June 22, 2017.

doi:10.3324/haematol.2017.170605 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1640 Š2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1640

ardiosiderosis is a leading cause of mortality in transfusiondependent thalassemias. Plasma non-transferrin-bound iron and its redox-active component, labile plasma iron, are key sources of iron loading in cardiosiderosis. Risk factors were identified in 73 patients with or without cardiosiderosis. Soluble transferrin receptor-1 levels were significantly lower in patients with cardiosiderosis (odds ratio 21).

This risk increased when transfusion-iron loading rates exceeded the erythroid transferrin uptake rate (derived from soluble transferrin receptor-1) by >0.21 mg/kg/day (odds ratio 48). Labile plasma iron was >3fold higher when this uptake rate threshold was exceeded, but nontransferrin-bound iron and transferrin saturation were comparable. The risk of cardiosiderosis was decreased in patients with low liver iron, ferritin and labile plasma iron, or high bilirubin, reticulocyte counts or hepcidin. We hypothesized that high erythroid transferrin uptake rate decreases cardiosiderosis through increased erythroid re-generation of apotransferrin. To test this, iron uptake and intracellular reactive oxygen species were examined in HL-1 cardiomyocytes under conditions modeling transferrin effects on non-transferrin-bound iron speciation with ferric citrate. Intracellular iron and reactive oxygen species increased with ferric citrate concentrations especially when iron-to-citrate ratios exceeded 1:100, i.e. conditions favoring kinetically labile monoferric rather than oligomer species. Excess iron-binding equivalents of apotransferrin inhibited iron uptake and decreased both intracellular reactive oxygen species and labile plasma iron under conditions favoring monoferric species. In conclusion, high transferrin iron utilization, relative to the transfusion-iron load rate, decreases the risk of cardiosiderosis. A putative mechanism is the transient re-generation of apotransferrin by an active erythron, rapidly binding labile plasma iron-detectable ferric monocitrate species.

Introduction Cardiosiderosis or myocardial hemosiderosis (MH) is a leading cause of mortality in transfusion-dependent thalassemias.1 Suggested risk factors for MH and/or consequent cardiomyopathy have included: sustained high serum ferritin levels,2 high liver iron concentrations (LIC),3 poor adherence to chelation therapy,4 as well as genetic susceptibility factors.5 However, while improved monitoring by magnetic resonance imaging and better use of iron chelators6 have led to falling frequencies of MH, as judged by cardiac magnetic resonance studies, there remains a variable prevalence haematologica | 2017; 102(10)

Thalassemic erythropoiesis, LPI, and cardiac iron

between populations7 and between individuals that is not fully understood.

The conduit through which MH develops is plasma nontransferrin-bound iron (NTBI), as transferrin-mediated iron uptake by cardiomyocytes is relatively 'insignificant'.8 NTBI is detectable as chelatable iron9,10 or redox-active iron (labile plasma iron, LPI)11 at transferrin saturations >75% for NTBI12 or 100% in the case of LPI.13 Animal data suggest that NTBI is taken into the myocardium through L-type calcium channels,14 and are supported by recent clinical data indicating that MH is inhibited in transfusion-dependent thalassemia patients by the calcium channel blocker amlodipine.15 Plasma NTBI is not a single entity, however, being a heterogeneous, multispeciated pool of ferric iron (not bound to high-affinity transferrin iron-binding sites) containing monomeric, oligomeric, and polymeric iron citrate species16 with weak albumin binding,17 or stronger binding, where post-translational modifications to albumin occur.18 There are a number of potential non-protein ligands for NTBI, including phosphate, acetate, amino acids, pyrophosphate, and citrate. However, phosphate, acetate, and amino acids cannot compete with the hydroxyl ion for iron(III) at pH 7.4. Pyrophosphate is potentially a potent iron(III) ligand but has a vanishingly small concentration in plasma after accounting for the effect of magnesium and calcium binding, thus citrate is the dominant ligand. The iron-to-citrate ratio determines the mix of species present.

As the plasma citrate concentration is roughly constant at 100 mM, the iron-to-citrate ratio is determined by the plasma NTBI concentration. At 1 µM NTBI the citrate excess is 100-fold and chelatable iron citrate species predominate, whereas, at ≥10 mM NTBI, the proportion of chelatable iron drops substantially.17,19 LPI refers to the redox-active fraction of NTBI, hence its term ‘labile’, which is chelatable,20 and has been implicated in organ hemosiderosis.21 Its chemical nature is not characterized, although it is predicted to comprise both monomeric and oligomeric ferric citrate, albumin iron complexes, and possibly partially coordinated iron chelates, e.g. of deferiprone,22 which are able to form hydroxyl radicals in the presence of ascorbate and hydrogen peroxide.

The balance between the rate of transferrin iron utilization by the erythron and the transfusion iron loading rate is key to determining levels of plasma NTBI.23 Blood transfusion delivers a mean of 0.4 mg/kg/day but with a wide range (0.2-0.6 mg/kg/day),24 exceeding by 10-fold the gut iron loading rate seen in non-transfusion-dependent thalassemia.25 Transfused red cells are ultimately catabolized within macrophages of the spleen, liver, and bone marrow, with iron released onto transferrin and - when the latter approaches saturation - forming plasma NTBI. Transferrin iron uptake by the erythron via transferrin receptor-1 (TfR1) liberates iron from transferrin during receptor-mediated endocytosis, whereupon iron-free apotransferrin is recycled back to plasma, as described in detail by Gkouvatsos et al.26 The extracellular domain of TfR1 is shed by red cell progenitors and circulates bound to holotransferrin (sTfR1).27 TfR1 expression in the erythron is transcriptionally regulated such that levels increase in iron deficiency.28 However, in thalassaemias, sTfR1 levels primarily reflect the degree of expansion of the erythron.29–31 Blood transfusion suppresses expansion and activity of the erythron, thereby decreasing sTfR1 proportionately to the pre-transfusion hemoglobin.32 In this study, we examined clinical factors associated haematologica | 2017; 102(10)

with MH as well as the mechanisms underlying these associations. In particular, we focused on how transferrin-iron utilization by the erythron marked by sTfR1, relative to the transfusion-iron loading rate, affects the risk of MH. We considered how apotransferrin formed after endocytosis in the bone marrow may bind NTBI species, decreasing their availability for myocardial uptake.

Methods Patients A cohort of 73 transfusion-dependent thalassemia patients on deferasirox, with known transfusion-iron load rate,24 was divided into those with MH (n=24, cardiac T2*<20 ms, R2*>50 s-1) and those without MH (n=49, T2*>20 ms, R2* <50 s-1).33 Details are provided in the Online Supplement. Patients gave written informed consent to participate in the study and approval was obtained from The Joint University College London/University College London Hospitals Committees on Ethics of Human Research.

Biomarkers Pre-transfusion samples were tested, after a 48-h washout from deferasirox chelation, for iron metabolism and routine biochemical variables, and compared in both groups (±MH). The tests included assays for NTBI,9 LPI,11 hepcidin, urea-gel transferrin saturation, sTfR1, and growth differentiation factor 15 (GDF-15), as published previously,23 as well as routine clinical variables (hemoglobin, absolute reticulocyte count, nucleated red cells, serum ferritin, bilirubin, and serum iron), as described in detail in the Online Supplement.

Cell-line experiments Murine HL-1 cardiomyocytes (ATCC number CRL-12197) were grown in Claycomb medium (Sigma); the culture protocol is described in the Online Supplement. Buffered ferric citrate17,16 or ferric ammonium citrate (FAC) was used to model NTBI. Ferric nitrilotriacetate was used to saturate transferrin. Total cellular iron was assayed by ferrozine. The level of cytosolic reactive oxygen species, tested using 2,7-dichlorofluorescein diacetate, was calculated from slopes of fluorescence-versus-time data. Detailed methods are provided in the Online Supplement.

Statistics Continuous data are presented as mean ± SD or median ± interquartile range and compared using the t-test or nonparametric tests, depending on assumed distribution, unless otherwise specified. Categorical variables are compared using the χ2-test. A P value <0.05 is considered statistically significant.

Results Factors associated with myocardial hemosiderosis To determine factors associated with MH, biomarkers were compared in transfusion-dependent thalassemia patients with (n=24) and without (n=49) MH. The sTfR1 emerged as the factor most significantly associated with MH (P<0.001, Figure 1A-C) being >3-fold higher in controls than in patients with MH (medians 4.06 versus 1.26 mg/mL, P=0.0005) with an area-under-curve for the receiver operating characteristic curve (AUCROC) of 0.8 (P=0.0004). A threshold of 1.77 mg/mL was predictive of MH with 89.7% sensitivity and 64.7% specificity (normal range for sTfR1: 0.3-1.65 mg/mL). Other biomarkers signif1641

M.W. Garbowski et al.

icantly associated with MH were bilirubin, reticulocyte count, hepcidin, LIC, and serum ferritin (Table 1).

Although the median LPI differed insignificantly between patients with and without MH, a cut-off threshold of >0.31 mM was significantly associated with MH (χ2 test, P=0.04).

Differences in NTBI, LPI, transferrin saturation (TfSat), and transfusion iron load rate (ILR) were insignificant (Figure 1D-G), and no such thresholds were found for NTBI or TfSat. High LIC and, to a lesser extent, serum ferritin were also significant risk factors for MH in our study, which is of interest, as this was not seen in earlier studies.33 However, determination of liver iron in earlier studies was not optimal34 and many patients had recently undergone intesive iron chelation which removes iron faster from the liver than from the heart,35 thereby obscuring such a relationship.36 A further significant, though weak, risk factor for MH was high plasma hepcidin. Hepcidin did not correlate with markers of iron overload such as LIC (as hepatic R2*, P=0.4), or ferritin (P=0.52), but was inversely and significantly correlated with sTfR1 (Spearman r=-0.54, P<0.0001). sTfR1 did not correlate with LIC or ferritin

A

D

G

(P=0.4 and P=0.52, respectively; Online Supplementary Figure S1). Associations between MH an nucleated red cells, total serum iron, age, weight, ILR, TfSat, and GDF-15 were insignificant.

Relationship of transferrin-iron utilization to myocardial hemosiderosis in transfusion-dependent thalassemia To gain insight into the relationship between MH and sTfR1, we utilized understandings from ferrokinetic studies29,30 which established the erythron transferrin uptake (ETU) as proportional to plasma sTfR1. The regression equation (r2=0.84) for ETU was calculated from published data:29 ETU[mmol Fe/L plasma/day]=0.013*sTfR1[mg/L]+2.25 Assuming that the relationship remains temporally stable, sTfR1 represents erythropoiesis quantitatively, given that plasma levels are proportionate to tissue transferrin receptors of the whole organism,31 of which erythroid cells are the main dynamic component. ETU can be expressed as

B

C

E

F

H

I

Figure 1. Biomarkers in patients with and without myocardial siderosis. Comparison of means and medians in patients with and without cardiac iron, MH(+) and MH(-), respectively, shown as box and whisker (min-max) plots. (A) Soluble transferrin receptor-1, sTfR1: 4.06 vs. 1.26 mg/mL, MW P=0.0005. (B) Receiver operating characteristic (ROC) curve for sTfR1, AUCROC=0.8±0.07 (P=0.0004). (C) Plot of cardiac R2* vs. sTfR1, Spearman correlation coefficient r = –0.61, P<0.0001. (D) Nontransferrin-bound iron, NTBI: 2.86 vs. 2.7 mM, MW P=0.59. (E) Transferrin saturation, TfSat: 100 vs. 100%, MW P=0.94. (F) Labile plasma iron, LPI: 0.1 vs. 0.27 mM, MW P=0.13. (G) Transfusion-iron load rate, ILR: 0.32 vs. 0.35 mg Fe/kg/day, t-test P=0.19. (H) Growth-differentiation factor-15, GDF-15: 6702 vs. 4430 pg/mL, MW P=0.4 (I) Plasma hepcidin 6.8 vs. 14.3 nM, MW P=0.04. MW, Mann-Whitney test; AUC, area under curve, MH, myocardial hemosiderosis.

1642

haematologica | 2017; 102(10)

Thalassemic erythropoiesis, LPI, and cardiac iron

a rate corrected for estimated blood volume (BV) [70 mL/kg (males), 65 mL/kg (females)] and the patient’s weight (erythroid transferrin uptake rate, ETUR).

ETUR[mg/kg/day]=ETU[µmol/L/day]*55.845[g/mol]/10 00*BV[mL]/1000/weight[kg].

Although transfusion does not directly relate to cardiac iron, when plasma sTfR1 levels were expressed as ETUR (i.e. transferrin-iron utilization rate) and compared to the transfusion ILR (Figure 2A-C), MH was present only when the ‘net ILR’ (ILR-ETUR) exceeded 0.21 mg/kg/day (P<0.001). Above and below this threshold, LIC,34 TfSat, and NTBI were similar (9.7 versus 10.1 mg/g dry weight, 96 versus 98%, 3 versus 2.6 mM, P=ns), while LPI was >3-fold higher (0.35 versus 0.1 mM, P=0.01). Thus, LPI was high only in patients in whom the net ILR exceeded 0.21 mg/kg/day, whereas no differences were seen in LIC, TfSat, and NTBI between patients with and without MH, even when adjusted for the net ILR (Figure 2C).

This suggests a link between iron clearance from transferrin - a process implying the formation of transient levels of apotransferrin - and the low propensity to MH in transfusion-dependent thalassemia. We hypothesized that increased transient local concentrations of apotransferrin in the marrow and circulating reticulocytes37 could decrease the uptake of NTBI into the myocardium. With similar TfSat and NTBI in MH-positive and MH-negative patients, this effect of recycled apotransferrin was expected to influence the speciation of NTBI.

Effects of iron-to-citrate ratio on iron detection in the labile plasma iron assay As LPI was found to differ between MH-positive and MH-negative patients, we wished to characterize which fraction of NTBI was detected by the LPI assay. In particular, we wished to determine whether LPI was most detectable under conditions in which either monomer or oligomer ferric citrate species predominated. As shown in Figure 3A, at constant citrate concentrations (100 mM), the proportion of iron detectable in the LPI assay decreases as

iron concentration increases. Conversely, LPI is proportionately most detectable at 1000-fold citrate excesses. Under these clinically relevant conditions, the citrate excess favors the formation of monomer rather than oligomer species. As LPI was associated with an increased risk of MH in our clinical observations, these findings suggest that the monomer LPI might be the species most readily taken into myocardial cells.

Effect of iron-to-citrate ratio on iron uptake into HL-1 cardiomyocytes To confirm this interpretation, we examined iron uptake in HL-1 cardiomyocytes at varying iron-to-citrate ratios.

Figure 3B shows the effect of increasing ferric iron at constant citrate concentration (100 mM) on iron uptake into HL-1 cells at 24 h. A clear plateau effect is seen whereby iron uptake does not increase further when citrate exceeds iron by less than 100-fold. Under these conditions oligomer rather than monomer ferric citrate species are favored.17 This suggests that oligomer iron citrate species are less available than the monomer species for uptake by cardiomyocytes.

Effect of iron-to-citrate ratio on iron binding to transferrin We also wished to determine how the iron-to-citrate ratio affected the availability of ferric iron to bind transferrin. Apotransferrin (at 35.5 mM or 71 mM iron binding equivalents, IBE) was incubated with buffered ferric citrate at 0 to 4000-fold citrate excess (Figure 3C). Iron binding to transferrin, determined by spectroscopy, increased to a maximum of 40% as the citrate excess increased. Thus binding of iron to transferrin is more rapid under conditions favoring monomer ferric citrate species than under conditions in which oligomers predominate.

Low (nanomolar) concentrations of apotransferrin substantially decrease labile plasma iron The preferential binding of the monomer ferric citrate species by transferrin was predicted to decrease those same species responsible for activity in the LPI assay. We

Table 1. Significant clinical and laboratory variables associated with myocardial hemosiderosis in transfusion-dependent thalassemias.

Variables

Threshold

Chelation start age [years] Age [years] Unchelated transfusion years [%] Transfusion dependency start age [years] ETUR-ILR [mg/kg/day] Total bilirubin [mM] Hepatic R2* [s-1] sTfR1 [mg/mL] ETU [mgFe/L whole blood/day] Absolute reticulocyte count [×109/L] LPI [µM] Hepcidin [nM] Ferritin [mg/L]

≥7.5 ≥35.92 ≥5 ≥1.5 <0.21 ≥16.5 <1421 ≥1.34 ≥1.45 ≥7.4 <0.31 <6.78 <1700

OR±SE

P value

MH(-)

MH(+)

P-value

8.42±0.81 5.9±0.61 4.68±0.6 3.64±0.58 48.4±1.15 21±1.1 21±1.06 20.81±0.87 16.04±0.73 13.75±1.14 10.5±1.02 9±0.81 7.63±0.62

0.009 0.005 0.018 0.046 0.00008 0.0014 0.0007 0.00016 0.0001 0.02 0.04 0.006 0.001

7.5 (3.5-12) 35.67 ± 10.57 8 (3-18) 1.5 (0.67-5.5) 0.09 (-0.08-0.17) 35 (25-52) 139 (79-365) 4.06 (2.55-5.95) 14.06 (6.53-18.26) 44.4 (19.95-128.9) 0.1 (0.02-0.17) 6.8 (3.18-22.13) 1309 (844-3057)

2.5 (2-6) 32.11 ± 6.92 4.5 (3-9) 0.9 (0.5-1.88) 0.27 (0.24-0.34) 22 (14-29.75) 326 (235-625) 1.26 (0.68-2.71) 3.89 (1.82-4.86) 20.6 (7.1-30.7) 0.27 (0.05-0.45) 14.3 (7.58-23.65) 2926 (1958-4284)

0.0039 0.12 0.15 0.02 0.0002 0.0004 0.003 0.0003 0.0001 0.029 0.13 0.04 0.0013

Medians with 25th-75th percentile range or mean ± SD are provided for groups of patients without myocardial hemosiderosis MH(-) or with MH(+), compared with a MannWhitney or t-test, as well as the threshold protecting from MH and its odds ratio (OR). 1LIC=4.41 mg/g dry weight34. ETUR: erythroid transferrin uptake rate; ILR: transfusion iron load rate; sTfRI: soluble transferrin receptor-1; ETU: erythroid transferrin uptake; LPI: labile plasma iron.

haematologica | 2017; 102(10)

1643

M.W. Garbowski et al.

tested this; the results are shown in Figure 3D. The control without transferrin showed increasing LPI values with increasing ferric iron concentration (at a constant 100 mM concentration of citrate). When apotransferrin was added at remarkably low concentrations (30 nM or 60 nM IBE), LPI detectability decreased substantially across all ratios.

This effect exceeded results expected from simple stoichiometric binding. A higher concentration of apotransferrin (10 mM) completely abrogated LPI detectability up to 10 mM iron. These findings show that the redox-active iron species responsible for the majority of LPI detectability is present at a very low concentration (nanomolar). This species is most likely to be ferric monocitrate because apotransferrin is known to bind ferric monocitrate most avidly.

are those that most inhibit iron uptake in HL-1 cardiomyocyte cells.

We first examined the effect of transferrin on uptake from FAC, which is a stable form of monomeric ferric iron coordinated by two citrate molecules.38 Figure 3E shows that iron uptake from 1 mM FAC was almost completely inhibited by apotransferrin at 1 mM (2 mM IBE). Figure 3F illustrates the effect of a constant transferrin concentration with varying percentage saturations on iron uptake from 5 mM FAC at 24 h. Uptake was completely inhibited by physiologically relevant transferrin concentrations of 37 mM (74 mM IBE) at saturations below 99%.

We then wished to examine the inhibitory effect of apotransferrin on iron uptake at varying iron-to-citrate ratios.

In order to study short-time intervals we developed an intracellular reactive oxygen species assay and validated this against total cellular iron at 24 h (Figure 4A). The control iron uptake marked by intracellular reactive oxygen species at 60 min in Figure 4B (closed circles) showed a plateau, as previously noted for total iron uptake at 24 h (Figure 3B). Thus, under conditions in which oligomer

Effect of apotransferrin on HL-1 cardiomyocyte iron uptake from ferric citrate species Having established the conditions favoring LPI detectability and apotransferrin binding of ferric citrate species, we wanted to test whether these same conditions

A

B

C

D

Figure 2. Balance between transfusion-iron load rate and erythroid transferrin uptake rate derived from soluble transferrin receptor-1 relates to cardiac iron. (A) Cardiac R2* plotted against transfusion-iron load rate (ILR), gray circles mark patients with cardiac iron (cR2*>50 s-1), open circles – patients without cardiac iron; no relationship overall. (B) Cardiac R2* (cR2*) plotted against ILR, open circles mark patients with cardiac iron (cR2*>50 s-1), gray circles – patients without cardiac iron; no relationship overall, relationship present in patients with cardiac iron only, r2=0.61; points differ in size according to sTfR1 level [mg/mL], see inset. (C) same as in panel B but the x-axis shows ILR corrected for utilization rate derived from sTfR1 according to Beguin et al. 199329 (ETU[mmol/L/day]=0.013×sTfR1 [mg/L]+2.25; ETUR[mg/kg/day]=ETU×55.845[g/mol]/1000×blood volume[mL]/1000/body weight[kg]). Highly discriminant threshold 0.21mg/kg/day P<0.0001, 100% sensitive, 83% specific for myocardial hemosiderosis (positive predictive value 71%, negative predictive value 100%) above and below which the liver iron concentration (LIC), transferrin saturation (TfSat) and non-transferrin-bound iron (NTBI) do not differ (P=ns) but labile plasma iron is 0.35 mM vs. 0.1 mM P=0.01, respectively. (D) same as panel C but points differ in size according to total body iron derived from LIC values.

1644

haematologica | 2017; 102(10)

Thalassemic erythropoiesis, LPI, and cardiac iron

species increasingly predominate, uptake into HL-1 cells forms a plateau. In the same experiment (Figure 4B, triangles), the addition of 5 mM apotransferrin inhibited uptake up to the point where ferric citrate approached 10 mM, at which concentration iron uptake exceeded that from control. Importantly, this contrasts with the effects of apotransferrin on iron uptake from FAC (Figure 3F) where uptake was inhibited at every apotransferrin concentration. The key difference between Figure 3B,E and F is the form of iron citrate presented to the cells. Freshly prepared FAC is a fully coordinated monomeric iron dicitrate,38 whereas at the 1:10 iron-to-citrate ratio used in the experiments shown in Figure 3B, mixtures of oligomer and monomer iron species were present.17,16 The increased uptake for apotransferrin shown in Figure 4B might be interpreted superficially as increased uptake from holotransferrin, formed from the addition of apotransferrin to iron citrate. However control experiments,

including those illustrated in Figure 4C and Online Supplementary Figure S2, showed that incubation of HL-1 cardiomyocytes with holotransferrin does not increase iron uptake: increasing concentrations of 95%-saturated transferrin actually inhibited iron accumulation.

An alternative explanation for the findings presented in Figure 4B is that the interaction of apotransferrin with iron citrate alters the proportions of citrate species. To test this hypothesis further, we examined the iron uptake at increasing concentrations of apotransferrin but constant iron citrate concentrations that favor oligomer species (10 mM ferric iron and 100 mM citrate) (Figure 4D).

Apotransferrin concentrations between 1 and 8 mM increased iron uptake but at higher apotransferrin concentrations (8 mM or 16 mM) uptake was inhibited. This is more consistent with catalytic depolymerization of oligomers to monomers as elaborated in the discussion.

We therefore suggest that the presence of apotransferrin

A

B

C

D

E

F

Figure 3. Effect of ferric citrate speciation on detectability of labile plasma iron, total cellular iron and transferrin binding. (A) Labile plasma iron (LPI) as a percentage of ferric citrate with constant citrate (100 mM) and variable iron (0-10 mM, shown next to data points in µM), as a function of excess citrate (fold), representative of three experiments. This illustrates what proportion of a given ratio species is LPI-detectable. (B) Total cellular iron dose response in confluent HL-1 cells to 24 h incubation in Claycomb medium (CM) with ferric citrate in MOPS pH=7.4 at variable Fe:citrate ratios and constant citrate at 100 mM, r2=0.95, EC50=0.31 mM, EC90=1.09 mM; shown as mean±SD, n=6. (C) Percentage of ferric citrate iron that binds to 35.5 mM apotransferrin (ApoTf) to form ferrotransferrin (over 2 h incubation at 37°C) as a function of citrate:Fe ratio. 0-30 mM iron and 100 mM citrate were increased 25.5-fold to keep the same ratio (see legend: 0-765 mM and 2550 mM f.c. respectively) in order to better resolve absorbance peaks of holotransferrin formation (at 465 nm, inset), and compared to 35.5 mM 100% saturated ferrotransferrin absorbance of 1.69. Percentage of iron bound to transferrin was calculated from transferrin iron content/nominal concentrations prepared (see inset, mean±SD, n=2). (D) Apotransferrin-dependent inhibition of LPI in ferric citrate. Apotransferrin (ApoTf) with 25 mM bicarbonate was added at 0-10 mM (f.c.) together with ferric citrate, subsequent DHR oxidation was followed up for 1 h. LPI values are interpolated from the standard curve, mean±SD, n=3. (E) HL-1 cells grown to confluence and incubated for 24 h in CM with 1 mM ferric ammonium citrate (FAC) ±0-2.5 mM ApoTf; ANOVA P<0.001, mean±SD, n=2, Bonferroni post-test significant for ApoTf effect in the FAC group. Global fit r2=0.91, ApoTf IC50=0.81 mM, curves significantly different (P<0.0001). (F) Transferrin saturation-dependent model of NTBI uptake (as FAC) into HL-1 cells grown to confluence then incubated in CM with apotransferrin and 100% saturated holotransferrin at relevant ratios to obtain a 37.5 mM TfSat model ±5 mM FAC; difference vs. control (no NTBI) only detectable for TfSat=100%, i.e. when no apotransferrin is present (multiple t-test with Holm-Sidak correction for multiple comparisons, P=0.004, mean±SD, n=6).

haematologica | 2017; 102(10)

1645

M.W. Garbowski et al.

favors the formation of citrate species that are more rapidly taken into cells, namely monomeric iron species. Only when the iron binding capacity of apotransferrin exceeds the iron content of the ferric citrate (here above about 10 mM), is iron uptake inhibited. We conclude that speciation of iron citrate is critical to cardiomyocyte iron uptake and that apotransferrin alters this speciation.

Discussion In this stduy we sought to identfy risk factors for MH in transfused thalassemia patients. A key novel finding is that low levels of sTfR1 appear to be a powerful predictor for MH with an odds ratio of 21. It is unlikely that sTfR1 has a direct mechanistic role in iron distribution, as mice transfected with sTfR1 showed similar iron absorption and hepcidin to controls.39 Furthermore, the circulating concentration of sTfR1 (nanomolar) is three logs less than that of diferric transferrin and thus unlikely to be a significant receptor trap for transferrin iron utilization. We also found a clear relationship between the transfusion-iron loading

A

B

rate and myocardial iron, as determined by magnetic resonance imaging (mR2*), in MH-positive patients (Figure 2B).

This suggests that the balance between the transfusion iron-loading rate and erythron iron utilization might be key to MH. We therefore developed a model building on data linking sTfR1 to quantitate iron utilization by the erythron (the ETUR),29,30 and found that the difference between the ILR and the ETUR predicted MH with an odds ratio of 48. We identified a threshold for this difference of 0.21 mg/kg/day, above which MH was more likely and below which MH was not seen (Figure 2C), even in the presence of high total body iron (Figure 2D). Hence a high transfusion iron intake relative to endogenous erythropoiesis puts patients at increased risk of MH. The high transfusion iron intake relative to endogenous erythropoiesis as a risk for MH is supported by other clinical observations. Additional factors that we found correlated with lower MH risk, such as high bilirubin concentration and high reticulocyte count, are also consistent with pronounced activity of the erythron. Others have shown that a reticulocyte count below 5% in sickle cell disease actually predicted premature development of cardiac iron

C

D

Figure 4. Effect of apotransferrin and citrate speciation on the iron uptake marked by cytosolic reactive oxygen species. (A) The relationship of intracellular reactive oxygene species (ROS) levels by the DCF method and total cellular iron by the ferrozine method, r=0.91, P<0.0001; Deming regression slope 1.14×10-5±1.68×10-6, intercept 2.69×10-5±5.08×10-6, mean±SD, n=6. (B) Cytosolic ROS levels in HL-1 cardiomyocytes as a function of control ferric citrate (iron 0-30 mM, citrate 100 mM) in MOPS pH=7.4 with (triangles) and without (circles) 5 mM apotransferrin (ApoTf) in 25 mM bicarbonate; mean±SD, n=5. (C) Dose response of cytosolic ROS levels in HL-1 cardiomyocytes to 95% saturated transferrin (0.37-15 mM) ±100 mM iron-binding equivalents of CP40, an extracellular chelator, mean±SD, n=4. (D) Cytosolic ROS formation in HL-1 cardiomyocytes showing recent uptake of iron from predominantly oligomer species (10 mM iron, 100 mM citrate) or control (citrate only) under dose-response effect of 0-16 mM apotransferrin in 30 mM bicarbonate/phosphate-buffered saline; mean±SD, n=4.

1646

haematologica | 2017; 102(10)

Thalassemic erythropoiesis, LPI, and cardiac iron

deposition.40 Erythropoietin, although not measured in this cohort, would be predicted to negatively correlate with MH because it drives sTfR1 levels29 and correlates with them closely in transfusion-dependent thalassemia. The relationship of age at onset of transfusion dependence to MH suggests that an initial expansion of the bone marrow, by under-transfusion or late introduction of transfusion, is hard to suppress even after many subsequent years of hypertransfusion (Online Supplementary Figure S3). Others have indirectly linked the absence of MH with extramedullary hematopoiesis in transfused thalassemia patients, using extramedullary masses by magnetic resonance imaging as surrogate markers of erythroid mass,41 which also proves the point that even life-long hypertransfusion does not suppress erythropoiesis completely. A high risk of MH has also been identified in another transfusiondependent condition with low iron utilization and extremely low sTfR1, namely Diamond-Blackfan anemia.23,42 Conversely, non-transfusion-dependent tha-

lassemia patients have very high sTfR1 levels13 with high ineffective erythropoiesis but a low risk of MH,43,44 despite substantial iron overload. Factors in addition to iron uptake may influence net MH. For example, our findings of increased plasma hepcidin in MH-positive patients are consistent with recent work in animal models of iron overload implicating iron exit through cardiac ferroportin channels as important to myocardial iron retention.45,46 There was a significant relationship between hepcidin and sTfR1, but not with GDF-15, nonetheless it would be of value to look at this relationship with erythroferrone when the assay becomes available.

What is the role of the LPI fraction of NTBI in MH risk?

A notable feature of non-transfusion-dependent thalassemias, found elsewhere in a large cohort of 155 patients, is that, even in the presence of high levels of body iron and raised NTBI, LPI is typically within the normal range, while the few cases with increased LPI have no apotransferrin.13 This is consistent with this studyâ&#x20AC;&#x2122;s findings in

A

B

C

Figure 5. Model of apotransferrin-dependent re-speciation of polymeric ferric citrate. The paradoxical effect of apotransferrin seen in Figure 4D, in which uptake from 10 mM ferric citrate is increased before it is abolished by apotransferrin, is consistent with the sub-equivalent concentration of apotransferrin disrupting ferric citrate oligomers and releasing from them ferric monocitrate species. As apotransferrin binds a portion of iron in high ratio ferric citrate, this decreases the amount of iron per citrate, so effectively changes the iron-to-citrate ratio, i.e. re-speciates it. In the presence of apotransferrin and bicarbonate, which forms a ternary complex with citrate, oligomer complexes of ferric citrate become a source of ferric monocitrate species. These are subject to competition with uptake mechanisms for cellular entry (dotted line), with apotransferrin for the formation of ferrotransferrin and with citrate for the formation of ferric dicitrate, citrate also competing with apotransferrin for ferric monocitrate. Kinetic differences between ferrotransferrin formation and cellular uptake from ferric citrate may explain the additional iron uptake from newly released mononuclear species before apotransferrin can chelate them altogether. The coordination sites on the iron (shown in bold) are typically occupied by water, but they are labile sites and can also bind oxygen and H2O2, rendering the species susceptible to redox chemistry (marked in bold cursive on the right). They are also the sites of condensation with other iron complexes.

haematologica | 2017; 102(10)

1647

M.W. Garbowski et al.

trasfusion-dependent patients in whom a >3-fold difference in LPI was seen across the ILR-ETUR threshold, while plasma NTBI levels and transferrin saturations were indistinguishable (Figure 2C). This suggests that LPI-detectable species are those taken up by cardiomyocytes, consistent with previous clinical observations linking LPI-detectable iron to MH.21 As the LPI assay detects only the redox-active sub-fraction of NTBI, we wished to determine whether LPIdetectable NTBI species were those particularly prone to cardiomyocyte uptake. We used buffered ferric citrate as a model for NTBI at a constant citrate concentration of 100 mM to represent plasma concentration. Detectability of iron by the LPI assay was less using iron citrate as the source of iron than ferric nitrilotriacetate (Online Supplementary Figure S4), indicating that only a proportion of ferric citrate species is detectable by the LPI assay.

Furthermore, LPI was proportionately most detectable when citrate exceeded ferric iron concentrations by >100fold; ratios known to be associated with the appearance of monoferric citrate species.16,17 We, therefore, hypothesized that monoferric species are taken into cardiomyocytes most readily since, in our study, MH associated with LPI, rather than with total NTBI. Monoferric species include ferric monocitrate and ferric dicitrate16,17 (Figure 5, compounds A and C). Redox-cycling of iron species is dependent on the ability of reductants and oxidants to gain access to the iron center by displacing the monodentate water molecules which occupy the ‘free’ coordination sites of iron complex that is only partially coordinated by ligands such as one tridentate citrate molecule or one bidentate deferiprone molecule22 (Figure 5, coordination sites in bold). Ferric dicitrate, with its iron center fully hexa-coordinated by two tridentate citrate molecules, has a high stability constant for iron(III),16 is less redox-active,22 and thus less of a candidate for uptake (a process requiring reduction14,47,48). We therefore predicted that it is ferric monocitrate which is most available for cardiomyocyte uptake as well as transferrin binding, as discussed below.

We found that cardiomyocyte iron uptake from citrate and its inhibition by apotransferrin are speciation-dependent. Cardiomyocyte uptake of ferric citrate did indeed occur most rapidly when citrate excess was high (Figure 3B), under conditions favoring monomer rather than oligomer species. This contrasts with previous findings for hepatocytes and cardiomyocytes8,47,49 using ferric nitrilotriacetate, a non-physiological but typically monomeric iron source, where the speciation dependence of uptake was not studied.

Very low concentrations of apotransferrin, and transferrin saturations ≤99% inhibited cardiomyocyte iron uptake from FAC (Figure 3E,F). This suggests that the NTBI species partaking in cardiac uptake constitute only a small fraction of the total NTBI. The small magnitude of this fraction is consistent with monocitrate being the predominant species for cardiomyocyte iron uptake. Furthermore, because apotransferrin decreased LPI in our assays, and iron monocitrate was the species most rapidly chelated by apotransferrin, as described above, ferric monocitrate is most likely the predominant LPI species. Very low transferrin concentrations (nanomolar) were all that were necessary to inhibit LPI-detectable iron citrate species (Figure 3D), which we identified above as being monoferric citrate. This suggests that this LPI-detectable species is present at a very low concentration but is kinetically labile.

1648

Surprisingly, we found that at very high ferric citrate concentrations (10 mM, Figure 4D), under conditions that favor oligomer species, rarely observed clinically, sub-stoichiometric concentrations of apotransferrin, insufficient to bind all the ferric citrate, actually increased cellular iron uptake. This was not due to increased uptake by holotransferrin, as negligible iron was delivered to cardiomyocytes by holotransferrin (Figure 4C), consistent with previous observations in cardiomyocytes.8 We deduced therefore that increased uptake from ferric citrate in the presence of apotransferrin under these conditions may occur by accelerating the dissociation of oligomer to monomer citrate species (Figure 5). This deduction is consistent with the observation that the transition from absence to excess of transferrin binding capacity is associated with the generation of kinetically labile iron citrate species that are rapidly taken into HL-1 cardiomyocytes (Figure 4D). This effect is notable only when apotransferrin is present at intermediate concentrations allowing depolymerization of oligomers to their constituent monomers (i.e. ferric monocitrate, see Figure 5). Previous work, consistent with this model, showed that the rate-limiting step for exchange of polymeric ferric citrate iron with apotransferrin was the depolymerization and release of monomeric (mononuclear) ferric citrate.50 Ferric dicitrate was unreactive towards apotransferrin unless converted to an active intermediate, which the authors had supposed to be the monocitrate.50 Importantly, such conditions of very high NTBI are unlikely to occur clinically so that inhibitory effects of apotransferrin on NTBI uptake will predominate.

In conclusion, we propose a mechanism of MH inhibition by the generation of apotransferrin during erythropoiesis. Taken together our clinical and in vitro data point to increased generation of apotransferrin by an active bone marrow (marked by high sTfR1) as a key mechanism for decreasing the risk of MH in transfusion-dependent thalassemias. We suggest that a local excess of apotransferrin in the bone marrow, around sinusoids and reticulocytes, chelates monomeric ferric citrate species, the same species most rapidly taken into the myocardium. The clinical implications of this are that a critical balance appears to exist between the transfusion rate, endogenous erythropoiesis, MH risk and NTBI speciation. We suggest that prospective longitudinal data collection, including sequential sTfR1 measurements, would be valuable in order that clear recommendations could be made about whether reducing transfusion exposure decreases the risk of MH. Furthermore, due to the marked geographic variability in MH risk,7 which cannot be related solely to chelation practices, cross-sectional studies on the impact of regional transfusion policies on ETUR and MH risk could be indicated.

Acknowledgments The authors would like to thank Dr. Sukhvinder Bansal from the Department of Pharmacy at King’s College London for performing the hepcidin assay. MG would like to thank Dr. Farrukh Shah for Ph.D. co-supervision; the British Society for Haematology, Sickle Cell Society and UK Thalassaemia Society for the Haemoglobinopathy Fellowship Grant, as well as the Leukaemia and Blood Diseases Appeal for grant support. JP would like to thank UCL Biomedical Research Centre for Cardiometabolic Programme support. All authors would also like to thank the Wellcome Trust for grant support (WT093209MA). haematologica | 2017; 102(10)

Thalassemic erythropoiesis, LPI, and cardiac iron

References 1. Borgna-Pignatti C, Cappellini MD, De Stefano P, et al. Survival and complications in thalassemia. Ann NY Acad Sci. 2005;1054: 40–47.

2. Olivieri NF, Nathan DG, MacMillan JH, et al.

Survival in medically treated patients with homozygous beta-thalassemia. N Engl J Med. 1994;331(9):574–578.

3. Telfer PT, Prestcott E, Holden S, Walker M, Hoffbrand AV, Wonke B. Hepatic iron concentration combined with long-term monitoring of serum ferritin to predict complications of iron overload in thalassaemia major.

Br J Haematol. 2000;110(4):971–977.

4. Gabutti V, Piga A. Results of long-term ironchelating therapy.

Acta Haematol.

1996;95(1):26–36.

5. El Beshlawy A, El Tagui M, Hamdy M, et al.

Low prevalence of cardiac siderosis in heavily iron loaded Egyptian thalassemia major patients. Ann Hematol. 2014;93(3):375–379.

6. Pennell DJ, Berdoukas V, Karagiorga M, et al.

Randomized controlled trial of deferiprone or deferoxamine in beta-thalassemia major patients with asymptomatic myocardial siderosis. Blood. 2006;107(9):3738–3744.

7. Aydinok Y, Porter JB, Piga A, et al. Prevalence and distribution of iron overload in patients with transfusion-dependent anemias differs across geographic regions: Results from the CORDELIA study. Eur J Haematol. 2015;95 (3):244–253.

8. Liu Y, Parkes JG, Templeton DM. Differential accumulation of non-transferrin-bound iron by cardiac myocytes and fibroblasts. J Mol Cell Cardiol. 2003;35(5):505–514.

9. Singh S, Hider RC, Porter JB. A direct method for quantification of non-transferrinbound iron. Anal Biochem. 1990;186(2):320– 323.

10. Garbowski MW, Ma Y, Fucharoen S, Srichairatanakool S, Hider R, Porter JB. Clinical and methodological factors affecting non-transferrin-bound iron values using a novel fluorescent bead assay. Transl Res. 2016;177:19–30.e5.

11. Esposito BP, Breuer W, Sirankapracha P, Pootrakul P, Hershko C, Cabantchik ZI. Labile plasma iron in iron overload: redox activity and susceptibility to chelation.

Blood. 2003;102(7):2670–2677.

12. Gosriwatana I, Loreal O, Lu S, Brissot P, Porter J, Hider RC. Quantification of nontransferrin-bound iron in the presence of unsaturated transferrin. Anal Biochem.

1999;273(2):212–220.

13. Porter J, Cappellini M, Kattamis A, et al. Iron overload across the spectrum of non-transfusion-dependent thalassaemias: role of erythropoiesis, splenectomy and transfusion. Br J Haematol. 2017;176: 288–299.

14. Oudit GY, Sun H, Trivieri MG, et al. L-type Ca2+ channels provide a major pathway for iron entry into cardiomyocytes in iron-overload cardiomyopathy.

Nat Med. 2003;9(9):1187–1194.

15. Fernandes JL, Loggetto SR, Veríssimo MPA, et al. A randomized trial of amlodipine in addition to standard chelation therapy in patients with thalassemia major. Blood.

2016;128(12):1555–1561.

16. Silva AM, Kong X, Parkin MC, Cammack R, Hider RC. Iron(III) citrate speciation in aqueous solution. Dalt Trans. 2009;(40):8616– 8625.

17. Evans RW, Rafique R, Zarea A, et al. Nature of non-transferrin-bound iron: studies on iron citrate complexes and thalassemic sera. J

haematologica | 2017; 102(10)

Biol Inorg Chem. 2008;13(1):57–74.

18. Silva AM, Hider RC. Influence of non-enzymatic post-translation modifications on the ability of human serum albumin to bind iron.

Implications for non-transferrin-bound iron speciation. Biochim Biophys Acta.

2009;1794(10):1449–1458.

19. Evans P, Kayyali R, Hider RC, Eccleston J, Porter JB. Mechanisms for the shuttling of plasma non-transferrin-bound iron (NTBI) onto deferoxamine by deferiprone. Transl Res. 2010;156(2):55–67.

20. Zanninelli G, Breuer W, Cabantchik ZI. Daily labile plasma iron as an indicator of chelator activity in thalassaemia major patients. Br J Haematol. 2009;147(5):744– 751.

21. Wood JC, Glynos T, Thompson A, et al.

Relationship between labile plasma iron, liver iron concentration and cardiac response in a deferasirox monotherapy trial.

Haematologica. 2011;96(7):1055–1058.

22. Devanur LD, Neubert H, Hider RC. The Fenton activity of iron(III) in the presence of deferiprone. J Pharm Sci. 2008;97(4):1454– 1467.

23. Porter JB, Walter PB, Neumayr LD, et al.

Mechanisms of plasma non-transferrin bound iron generation: insights from comparing transfused Diamond Blackfan anaemia with sickle cell and thalassaemia patients. Br J Haematol. 2014;167(5):692– 696.

24. Cohen AR, Glimm E, Porter JB. Effect of transfusional iron intake on response to chelation therapy in β-thalassemia major.

Blood. 2008;111(2):583–587.

25. Taher AT, Porter J, Viprakasit V, et al.

Deferasirox reduces iron overload significantly in nontransfusion-dependent thalassemia: 1-year results from a prospective, randomized, double-blind, placebo-controlled study. Blood. 2012;120(5):970–977.

26. Gkouvatsos K, Papanikolaou G, Pantopoulos K. Regulation of iron transport and the role of transferrin. Biochim Biophys Acta.

2012;1820(3):188–202.

27. Hikawa A, Nomata Y, Suzuki T, Ozasa H, Yamada O. Soluble transferrin receptortransferrin complex in serum: measurement by latex agglutination nephelometric immunoassay. Clin Chim Acta. 1996;254(2): 159–172.

28. Ponka P, Lok CN. The transferrin receptor: role in health and disease. Int J Biochem Cell Biol. 1999;31(10):1111–1137.

29. Beguin Y, Clemons GK, Pootrakul P, Fillet G. Quantitative assessment of erythropoiesis and functional classification of anemia based on measurements of serum transferrin receptor and erythropoietin. Blood. 1993;81(4): 1067–1076.

30. Huebers HA, Beguin Y, Pootrakul P, Einspahr D, Finch CA. Intact transferrin receptors in human plasma and their relation to erythropoiesis. Blood. 1990;75(1):102–107.

31. R’zik S, Beguin Y. Serum soluble transferrin receptor concentration is an accurate estimate of the mass of tissue receptors. Exp Hematol. 2001;29(6):677–685.

32. Cazzola M, De Stefano P, Ponchio L, et al.

Relationship between transfusion regimen and suppression of erythropoiesis in betathalassaemia major. Br J Haematol.

1995;89(3):473–478.

33. Anderson LJ, Holden S, Davis B, et al.

Cardiovascular T2-star (T2*) magnetic resonance for the early diagnosis of myocardial iron overload. Eur Heart J. 2001;22(23):2171– 2179.

34. Garbowski MW, Carpenter JP, Smith G, et al.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

Biopsy-based calibration of T2* magnetic resonance for estimation of liver iron concentration and comparison with R2 Ferriscan. J Cardiovasc Magn Reson. 2014;16(1):40.

Anderson LJ, Westwood MA, Holden S, et al.

Myocardial iron clearance during reversal of siderotic cardiomyopathy with intravenous desferrioxamine: a prospective study using T2* cardiovascular magnetic resonance. Br J Haematol. 2004;127(3):348–355.

Noetzli LJ, Carson SM, Nord AS, Coates TD, Wood JC. Longitudinal analysis of heart and liver iron in thalassemia major. Blood.

2008;112(7):2973–2978.

Finch C, Huebers H, Eng M, Miller L. Effect of transfused reticulocytes on iron exchange.

Blood. 1982;59(2):364–369.

Matzapetakis M, Raptopoulou CP, Tsohos A, Papaefthymiou V, Moon N, Salifoglou A. Synthesis, spectroscopic and structural characterization of the first mononuclear, water soluble iron−citrate complex, (NH4)5Fe(C6H4O7)2·2H2O. J Am Chem Soc. 1998;120(10):13266–13267.

Flanagan JM, Peng H, Wang L, et al. Soluble transferrin receptor-1 levels in mice do not affect iron absorption. Acta Haematol.

2006;116(4):249–254.

Meloni A, Puliyel M, Pepe A, Berdoukas V, Coates TD, Wood JC. Cardiac iron overload in sickle-cell disease. Am J Hematol.

2014;89(7):678–683.

Ricchi P, Meloni A, Spasiano A, et al.

Extramedullary hematopoiesis is associated with lower cardiac iron loading in chronically transfused thalassemia patients. Am J Hematol. 2015;90(11):1008–1012.

Wood JC. Cardiac iron across different transfusion-dependent diseases. Blood Rev. 2008;22(Suppl. 2):14–21.

Taher AT, Musallam KM, Wood JC, Cappellini MD. Magnetic resonance evaluation of hepatic and myocardial iron deposition in transfusion-independent thalassemia intermedia compared to regularly transfused thalassemia major patients. Am J Hematol.

2010;85(4):288–290.

Roghi A, Cappellini MD, Wood JC, et al.

Absence of cardiac siderosis despite hepatic iron overload in Italian patients with thalassemia intermedia: an MRI T2* study. Ann Hematol. 2010;89(6):585–589.

Lakhal-Littleton S, Wolna M, Carr CA, et al.

Cardiac ferroportin regulates cellular iron homeostasis and is important for cardiac function. Proc Natl Acad Sci USA. 2015;112(10):3164–3169.

Altamura S, Kessler R, Groene HJ, et al.

Resistance of ferroportin to hepcidin binding causes exocrine pancreatic failure and fatal iron overload. Cell Metab. 2014;20(2):359– 367.

Parkes JG, Olivieri NF, Templeton DM. Characterization of Fe2+ and Fe3+ transport by iron-loaded cardiac myocytes.

Toxicology. 1997;117(2–3):141–151.

Liuzzi JP, Aydemir F, Nam H, Knutson MD, Cousins RJ. Zip14 (Slc39a14) mediates nontransferrin-bound iron uptake into cells. Proc Natl Acad Sci USA. 2006;103(37):13612– 13617.

Parkes JG, Randell EW, Olivieri NF, Templeton DM. Modulation by iron loading and chelation of the uptake of non-transferrin-bound iron by human liver cells. Biochim Biophys Acta. 1995;1243(3):373–380.

Bates GW, Billups C, Saltman P. The kinetics and mechanism of iron(III) exchange between chelates and rransferrin. I. The complexes of citrate and nitrilotriacetic acid.

J Biol Chem. 1967;242 (12):2810–2815.

1649

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Blood Transfusion

Ferrata Storti Foundation

Amotosalen/ultraviolet A pathogen inactivation technology reduces platelet activatability, induces apoptosis and accelerates clearance Simona Stivala,1,2 Sara Gobbato,1,2 Laura Infanti,3 Martin F. Reiner,1,2 Nicole Bonetti,1,2 Sara C. Meyer,4 Giovanni G. Camici,5 Thomas F. Lüscher,6 Andreas Buser3 and Jürg H. Beer1,2

Haematologica 2017 Volume 102(10):1650-1660

Laboratory for Platelet Research, Center for Molecular Cardiology, University of Zurich; Department of Internal Medicine, Cantonal Hospital Baden; 3Regional Blood Transfusion Service of the Swiss Red Cross, Basel; 4Division of Hematology and Department of Biomedicine, University Hospital Basel; 5Center of Molecular Cardiology, University of Zurich and 6Department of Cardiology, University Heart Center, University Hospital Zurich, Switzerland 1 2

ABSTRACT

A

Correspondence: hansjuerg.beer@ksb.ch

Received: January 9, 2017.

Accepted: July 13, 2017.

Pre-published: July 20, 2017. doi:10.3324/haematol.2017.164137 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1650 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1650

motosalen and ultraviolet A (UVA) photochemical-based pathogen reduction using the Intercept™ Blood System (IBS) is an effective and established technology for platelet and plasma components, which is adopted in more than 40 countries worldwide. Several reports point towards a reduced platelet function after Amotosalen/UVA exposure. The study herein was undertaken to identify the mechanisms responsible for the early impairment of platelet function by the IBS. Twenty-five platelet apheresis units were collected from healthy volunteers following standard procedures and split into 2 components, 1 untreated and the other treated with Amotosalen/UVA. Platelet impedance aggregation in response to collagen and thrombin was reduced by 80% and 60%, respectively, in IBS-treated units at day 1 of storage.

Glycoprotein Ib (GpIb) levels were significantly lower in IBS samples and soluble glycocalicin correspondingly augmented; furthermore, GpIbα was significantly more desialylated as shown by Erythrina Cristagalli Lectin (ECL) binding. The pro-apoptotic Bak protein was significantly increased, as well as the MAPK p38 phosphorylation and caspase-3 cleavage. Stored IBS-treated platelets injected into immune-deficient nonobese diabetic/severe combined immunodeficiency (NOD/SCID) mice showed a faster clearance. We conclude that the IBS induces platelet p38 activation, GpIb shedding and platelet apoptosis through a caspase-dependent mechanism, thus reducing platelet function and survival. These mechanisms are of relevance in transfusion medicine, where the IBS increases patient safety at the expense of platelet function and survival.

Introduction Platelet transfusion is a cornerstone in today’s medicine in general, and more particularly in hemato-oncology, as illustrated by the 1.3 million platelet units transfused annually in the USA and more than 2.9 million in Europe.1-3 One of the major challenges in transfusion medicine is the reduction of pathogen transmission by blood products, in particular for platelet components, since they need storage at room temperature.4 To circumvent the problem of pathogen contamination of blood products, pathogen inactivation (PI) technologies have been developed and routinely implemented in blood transfusion centers worldwide, including the USA, France and Switzerland.5–7 One such technology, the IBS (Cerus Corporation, Concord, CA, USA), employs a synthetic psoralen (amotosalen, S-59) and UVA light to induce cross-linking of DNA and ribonucleic acid (RNA) molecules, thus blocking replication and pathogen haematologica | 2017; 102(10)

Amotosalen /UVA and Platelet Clearance proliferation8 and rendering γ-irradiation for graft-versushost disease (GvHD) prophylaxis unnecessary. Several studies on the efficacy of non-pathogen-reduced versus IBS-treated platelets reported no cases of transfusion transmitted infections or transfusion associated GvHD, together with a reduction of other transfusion reactions.

On the other hand, some reduction in platelet function, platelet count increments (CI) and corrected count increment (CCI) have been described.9,10 Although 1 trial showed an increase in clinically irrelevant World Health Organization (WHO) grade 2 bleeding,10 other studies did not find an increase in bleeding, thus confirming the safety of the IBS technology.9,11–15 However, evaluating platelet function and survival in vivo is a challenging task due to the multiple and heterogeneous clinical and pharmacological factors affecting platelet function in patients.

Some reports suggest that all pathogen inactivation systems, including the IBS, aggravate the platelet storage lesion (PSL) and reduce the platelet function in vitro; the molecular mechanism behind these observations, however, is unclear16–18 Abonnenc et al. reported a reduced aggregation response to low-dose TRAP and collagen in IBStreated platelets, a finding confirmed in the study by Picker et al.19,20 The latter also described an increased glycolytic flux after pathogen reduction technology (PRT), with lactate accumulation and increased acidity. Schubert and Chen reported an increased phosphorylation of several intracellular kinases and higher caspase activity after riboflavin and ultraviolet B (UVB) treatment (Mirasol), which could be reverted by pre-treatment with specific p38 inhibitors.21,22 The study herein was undertaken in order to test the hypothesis that the IBS leads to reduced platelet activatability in response to certain agonists (i.e., collagen, thrombin and von Willebrand Factor [vWF]), increased platelet apoptosis and, consequently, enhanced clearance from the circulation. We therefore compared platelet function and parameters of apoptosis and clearance of untreated and IBS-treated human platelets in a large number of in vitro and in vivo tests in an immune-deficient mouse model (NOD/SCID). In addition, we analyzed the physiopathologic pathway(s) involved.

Methods

and were stored at standard blood banking conditions (22 ±2°C under gentle agitation).

Adhesion to collagen and vWF under flow Adhesion in the microfluidic chamber (Fluxion Biosciences, San Francisco, CA, USA) was performed on citrate, calcein-stained platelets (4 mM calcein AM, Enzo Life Sciences) at low and high shear rates (10 dyn/cm2 and 100 dyn/cm2, respectively).23 For detailed protocol see the Online Supplementary Methods.

In vivo platelet survival in NOD/SCID mouse Platelets from untreated and IBS-treated AU were incubated with calcein as described above, then pelleted (340 relative centrifugal force (RCF), 10 min) and resuspended in 0.9% sodium chloride (NaCl) at 4x109/ml. Eight-week old NOD/SCID male mice (Charles River, France) were injected intravenously with 100 ml of the platelet suspension.24 Thirty minutes, 2 hours and 5 hours after injection, a 10 ml blood sample was taken from the tail tip and mixed with Aster Jandl anticoagulant, centrifuged (125 RCF, 8 min), and 100 m of the supernatant was immediately analyzed on a Fortessa LSR II (BD Biosciences).25 The 30 minutes sample was set as 100%, and the percentages of human platelets in circulation at 2 and 5 hours were calculated accordingly. Following the final blood sampling, the animals were euthanized and the spleens excised and frozen in optimal cutting temperature (O.C.T) medium (Tissue-Tek O.C.T, Sakura Finetek Europe, AJ Alphen aan den Rijn, The Netherlands). All animal experiments were approved by and in strict compliance with the local Veterinary Office (animal licenses 174/2011 and 035/15).

Impedance aggregometry, flow cytometry, ELISA and Western blotting Detailed protocol for additional methods (impedance platelet aggregometry, flow cytometry staining, Western blotting (WB), glycocalicin enzyme-linked immunosorbent assay (ELISA), immunofluorescence staining) can be found in the Online Supplementary Methods.

Statistical analyses Results are mean ± SEM. Data were analyzed by paired, twotailed Student’s t-test, one- or two-way analysis of variance (ANOVA), followed by Bonferroni post hoc test as appropriate. A P-value of less than 0.05 was considered significant. All calculations were performed with GraphPad Prism 5.04 (GraphPad Software Inc., San Diego, CA, USA).

Platelet collection and processing Platelet apheresis units (AU) were collected from 25 volunteers at the Regional Blood Transfusion Service of the Swiss Red Cross of Basel, Switzerland. A table with the AU characteristics is provided in the Online Supplementary Material. The study was approved by the Institutional Review Board and each donor provided written informed consent. Of the 2 AU obtained from each donor, 1 was kept untreated (non-IBS) and the other treated with IBS on day 1 after collection according to the standard procedure (IBS).8 In some cases (n=10), 3 bags were obtained by splitting the apheresis product from 1 donor: 1 kept untreated (non-IBS), and 2 IBS-treated, of which 1 was injected with a sterile solution of the p38 inhibitor SB203580 (Sigma-Aldrich; final concentration 20 mM, n=5), or the sialidase inhibitor 2,3-dehydro-2-deoxy-Nacetylneuraminic acid (DANA, Calbiochem, final concentration 150 mM, n=5) and the other with an equal volume of vehicle (ethanol). Both were left to incubate overnight before undergoing the IBS procedure. All AU contained about 1/3 plasma and 2/3 the platelet additive solution InterSol (Fenwal, Lake Zurich, IL, USA) haematologica | 2017; 102(10)

Results Amotosalen and UVA photochemical treatment reduces platelet aggregation Untreated and IBS-treated samples were analyzed for aggregation in response to different doses of collagen and thrombin (Figure 1). At day (d)1 of storage, IBS-treated platelets showed a maximum collagen-induced aggregation of 20.5% (5 mg/ml collagen) and 45.2% (10 mg/ml collagen) compared to the non-IBS samples set as 100% (n=20, P<0.0001; Figure 1A,B). In response to thrombin, the IBS platelets showed 40.2% of aggregation compared to the non-IBS samples with a lower dose (0.25 U/ml, n=20, P<0.0001; Figure 1C), but did not show a significant difference with a high dose (0.5 U/ml; Figure 1D).

To further evaluate the effects on shear-induced aggregation, platelets were analyzed under low (10 dyn/cm2) and high (100 dyn/cm2) shear on collagen and human 1651

S. Stivala et al.

vWF-coated channels, respectively. The platelet-covered area from IBS samples showed a 47% reduction in collagen compared to non-IBS (from 38’561 mm2 to 20’718 mm2, n=17; Figure 1E), and a 65% reduction on vWF (from 2490 mm2 to 866 mm2, n=17; Figure 1F), which reached a borderline significance for the area under the curve (AUC) for collagen (P=0.05; Figure 1G), and was significant for vWF (P=0.01; Figure 1H).

of the N-terminal fragment of GpIbα (glycocalicin) in the supernatant was significantly increased by 20% in IBS samples (Figure 2B). This result was confirmed upon adjustment of the platelet count per unit (Glycocalicin Index; Figure 2C). Since the MAPK p38 is directly involved in TNF-α converting enzyme (TACE) activation and GpIb shedding, we compared p38 phosphorylation in IBS and non-IBS platelet lysates and found it to be significantly increased in the IBS samples (Figure 2D).

Amotosalen and UVA photochemical treatment induce desialylation and cleavage of GpIbα, the release of glycocalicin and p38 phosphorylation

Amotosalen and UV light increases Bak protein level and induces apoptosis

While flow cytometric analyses for Annexin V and PAC1 binding and P-selectin exposure showed no significant difference in IBS samples compared to the untreated controls (Online Supplementary Figure S1A-C), IBS platelets had a significantly lower expression of the vWF receptor GpIbα (mean fluorescence intensity (MFI) d1: 2258.9 nonIBS, 1937.4 IBS, n=20, P=0.01; Figure 2A), which could contribute to the reduced platelet aggregation on immobilized vWF observed under flow. Consistently, the amount

On account of the key role played by the proteins of the Bcl-2 family in determining platelet lifespan in vivo,27 we analyzed the expression of Bak and Bcl-XL in non-IBS and IBS platelets. The level of the anti-apoptotic protein BclXL was unchanged (data not shown) but that of the proapoptotic Bak was significantly increased in the IBS platelets (Figure 2E). In order to confirm that the increased level of Bak was inducing platelet apoptosis, we determined the amount of cleaved caspase-3 in platelet

7 day

5

day

day

1

7 day

5 day

1

7 day

1 day

7 day

1

F

day

7 day

day

day

G

5

E

1

D

C

day

7 day

day

day

5

B

1

A

H

Figure 1. IBS treatment reduces platelet function. Platelet aggregation in response to collagen (A and B) and thrombin (C and D) at days 1, 5 and 7 of storage from untreated (non-IBS, ◊) versus IBS-treated (■) AU. Lower panel: platelet aggregation under flow on collagen at 10 dyn/cm2 (E) and vWF at 100 dyn/cm2 (F) was also reduced in IBS samples compared to untreated platelets. Area under the curve (AUC) for the aggregation on collagen reached a borderline significance (G; P=0.05), while it was statistically significant for vWF (H). n=20, *P<0.05, ****P<0.0001. Graphs show mean±SEM. vWF: von Willebrand Factor; IBS: Intercept Blood System; AU: apheresis unit.

1652

haematologica | 2017; 102(10)

Amotosalen /UVA and Platelet Clearance

lysates; as shown in the blot and the relative quantification, it was significantly increased in the samples treated with the IBS as compared to the non-IBS controls (Figure 2F). Immunofluorescence staining of fixed, permeabilized platelets confirmed an increased Bak intensity for the IBS platelets (Figure 2G).

IBS treatment reduces platelet survival in vivo in NOD/SCID mice Next, we tested the physiological relevance of our findings in vitro on platelet survival in vivo. The AUC for platelet survival over 5 hours was significantly lower for the IBS platelets (Figure 3A,B), demonstrating reduced platelet half-life due to accelerated clearance in vivo upon transfusion. Taking the first time point (30 min post-injec-

D

7

C

day

day

day

5

B

1

A

tion) as 100%, 34.4% non-IBS platelets were still circulating compared to 28.2% IBS at 2 hours (n=15, P>0.05); at 5 hours, they were 26.1% versus 11.5%, respectively (P=0.05; Figure 3A).

The area of fluorescent platelets in spleens from injected mice was significantly increased in mice receiving IBS platelets compared to those injected with untreated platelets (Figure 3C and corresponding micrographs in the bottom panel). Under these conditions, we could not detect platelet clearance in the liver of these mice (data not shown).

Since it has been reported that GpIbÎą levels correlate with platelet survival,28 we analyzed its correlation with the in vivo survival of non-IBS and IBS platelets and found a significant positive correlation between GpIb levels and

E

F

G

Figure 2. IBS induces GpIb shedding from platelets, p38 activation and apoptosis through Bak. MFI of surface GpIbÎą measured by flow cytometry is significantly reduced in IBS platelets at day 1 of storage (A) and correspondingly increased in the supernatant when measured by ELISA (B) n=15, P<0.01. (C) Glycocalicin (GC) concentration from AU supernatant, normalized to the platelet count to give the GC Index, was significantly increased in IBS-treated AU; n=15, P=0.03. Western blot quantification of phosphorylated p38 (D), (E) Bak expression and (F) caspase-3 cleavage in platelet lysates from untreated or IBS-treated AUs. (G) Immunofluorescence staining of fixed/permeabilized platelets for Bak (green) and Bcl-XL(red) and relative quantification for Bak. Scale bar = 5mm. *P<0.05, **P<0.01, ****P<0.0001. IBS: Intercept Blood System.

haematologica | 2017; 102(10)

1653

S. Stivala et al.

survival at 2 hours post-injection (r2=0.1993, P=0.028; Figure 3D). It has been recognized that sialic acid on the heavily glycosylated GpIbα plays a relevant role in platelet clearance;26 therefore, we analyzed our samples for desialylation of platelet surface proteins by the fluorescein isothiocyanate (FITC)-conjugated Erythrina Cristagalli Lectin (ECL) binding in flow cytometry experiments (Figure 3E). Erythrina Cristagalli Agglutinin Lectin (ECA) binds to unsialylated galactose (β1-4) on N-acetyl-glucosamine (GlcNAc) and the ECA-binding level is inversely proportional to the level of sialylation. Concordant to our hypothesis, ECA binding was significantly higher in IBStreated platelets compared to non-IBS samples, even after normalization to GpIb levels to account for the increased receptor shedding in IBS samples (n=6, P=0.006; Figure 3E). In order to confirm that GpIbα desialylation was due to an increased neuraminidase exposure/release following Amotosalen/UVA, a specific neuraminidase activity assay

A

B

C

D

7 day

day

day

1654

5

F

1

E

was performed on the supernatants from untreated or IBStreated samples. Cleavage of the specific neuraminidase substrate was significantly higher in supernatants from samples that underwent the Amotosalen/UVA procedure, demonstrating a release of neuraminidase from platelets following the IBS (Figure 3F; P<0.001). Staining of fixed, non-permeabilized platelets for Neu1 revealed a higher fluorescence intensity for IBS-treated samples compared to control ones, while the fluorescence intensity was equivalent following platelet permabilization to reveal total (surface and internal) Neu1 (Online Supplementary Figure S1).

UV light without Amotosalen is sufficient to induce an increase in apoptotic Bak protein through mRNA translation Platelets don’t have a nucleus but they contain messenger (m)RNA and are capable of translation and protein

Figure 3. Platelet clearance in vivo in NOD/SCID mice is increased in IBS samples. (A) NOD/SCID mice were injected intravenously with fluorescently labeled platelets (untreated ◊, or IBS-treated, ■) and the % of circulating human platelets calculated after 2 and 5 hours; n=10. (B) Area under the curve (AUC) for the overall platelet survival of non-IBS versus IBS platelets; *P=0.02. (C) Spleens from mice injected with untreated or IBS-treated platelets were analyzed for the area of fluorescently-labeled platelets; n=10, *P=0.018. (D) Correlation analysis of GpIbα levels and platelet survival 2 hours post-injection in mice. (E) Desialylation of platelets by FITC-conjugated ECA lectin staining and flow cytometry analysis, normalized to GpIb levels; n=6, *P<0.05.

(F) Neuraminidase activity was tested in supernatants from control and IBS samples and was found to be significantly increased after Amotosalen/UVA treatment; n=18, ****P<0.001.

Bottom panel: representative microphotograph (4x magnification) of spleen from mice injected with fluorescent non-IBS or IBS platelets (platelet: green, nuclei: blue). Scale bar = 100 mm. Plotted are mean ±SEM. IBS: Intercept Blood System; plt: platelet; ECA: Erythrina Cristagalli Agglutinin; MFI: mean fluorescence intensity.

haematologica | 2017; 102(10)

Amotosalen /UVA and Platelet Clearance

synthesis.29 We hypothesized, therefore, that the increase in Bak protein after IBS was due to the translation of BAK specific mRNA. Immunoprecipitation of eukaryotic initiation factor 4E (eIF4E) followed by quantitative PCR showed that the relative expression of BAK was increased significantly 24 hours after irradiation, demonstrating an increased association of the specific BAK mRNA with eIF4E (n=4, P=0.01; Figure 4A). This result was confirmed at the protein level, because WB analysis of the platelet lysates showed an increased Bak level 24 hours after UV irradiation compared to non-UV platelets (n=9, P=0.009; Figure 4B). Blockade of mRNA translation with the protein synthesis inhibitor cycloheximide (10 mg/ml final concentration) was able to block the increase in Bak protein after UV irradiation (n=3, P=0.01; Figure 4C).

Inhibition of p38 restores GpIb levels but does not rescue platelet survival Due to the increased p38 phosphorylation observed in the IBS platelets, we reasoned that inhibition of p38 would block the adverse effects caused by the Amotosalen/UVA treatment. However, when injected into NOD/SCID mice, platelets pre-treated with the p38 inhibitor did not survive better as compared to untreated platelets, as shown in Figure 5A. At 2 hours post-injection, 28.5% of SB203580 platelets were circulating compared to 26.4% IBS vehicletreated and 41.8% untreated platelets, respectively; at 5 hours, there were 5.08% of the SB203580 samples versus 5.7% of the IBS platelets and 32.2% for the untreated ones, respectively (n=5, P>0.05; Figure 5A). Analysis of GpIbα expression on untreated, IBS vehicle and IBS-SB203580 platelets revealed that the receptor cleavage was indeed blocked in the SB203580 samples, with receptor levels similar to the untreated samples (Figure 5B). Loss of sialic acid from platelet receptors was also prevented by the p38 inhibitor (Figure 5C). Aggregation to collagen and thrombin, however, was not different from the vehicle-treated IBS samples (data not shown).

A

B

Interestingly, we found that levels of the pro-apoptotic Bak were increased in the SB203580 samples, as shown by WB analysis of platelet lysate (Online Supplementary Figure S2A) and by immunofluorescence staining (Online Supplementary Figure S2B and corresponding microphotograph). Additionally, we found increased levels of cleaved caspase-3, suggesting that the increment in Bak leads to platelet apoptosis (Figure 5D).

Exploratory analysis of potential mechanisms with the neuraminidase inhibitor DANA It has been proposed that desialylation of platelet receptors may in part regulate platelet survival, with the heavily glycosylated GpIbα being a major contributer.30,31 Since we observed increased desialylation of platelets after IBS treatment (Figure 3E), we hypothesized that pre-treatment of platelets with the neuraminidase inhibitor DANA could block sialic acid loss and increase platelet survival. As shown in Figure 6A, 2 hours after injection 31.8% non-IBS platelets were circulating compared to 20.3% IBS-treated platelets. Pre-treatment with DANA was able to restore the circulating platelet value back to 31.4% during the first 2 hours (n=5, P>0.05; Figure 6A). However, this effect was partly lost at the later time point (5 hours), when circulating platelets from the DANA sample were 14.2% compared to 22.9% non-IBS and 10.4% IBS-treated, respectively (Figure 6A), possibly as a result of platelet washing causing removal of the inhibitor before the injection.

GpIbα levels in the DANA samples were similar to those in IBS samples at all days of storage tested (Figure 6B). We also analyzed the level of desialylation and found that preincubation with DANA protected platelets from sialic acid loss, with levels of ECL binding similar to that of the untreated control at all days tested (Figure 6C).

Analysis of p38 phosphorylation, Bak expression and caspase-3 cleavage by WB did not show any significant difference between the samples (Figure 6D-F) in this series of experiments, and the limited number of donors (as per

C

Figure 4. UV irradiation of platelets induces Bak protein expression through mRNA translation. Platelets were isolated and resuspended in 40% plasma/60% Intersol and kept either untreated (no UV) or UVA-irradiated (UV). (A) eIF4E was immunoprecipitated 2 or 24 hours after UV irradiation with a specific or a control antibody, and Bak quantitative PCR performed on the isolated RNA from the IP. 24 hours after UV irradiation, specific Bak RNA in complex with eIF4E is significantly increased (n=4, *P=0.01). (B) Bak protein levels normalized to GAPDH are increased 24 hours after UV irradiation, reflecting an increased protein synthesis (n=9, **P=0.009). (C) Platelets non irradiated or irradiated after pre-treatment with cycloheximide (10 mg/ml) were analyzed for Bak protein expression. Cycloheximide blocked Bak increase induced by UV (n=3, *P=0.01). qPCR: quantitative polymerase chain reaction.

haematologica | 2017; 102(10)

1655

S. Stivala et al.

C

D

day 7 day

5 day

1 day

7

B

1

A

day

The study herein analyses in depth the structural and functional consequences induced by Amotosalen/UVA treatment using the IBS and the underlying mechanisms.

We provide evidence of diminished platelet function, i.e., reduced aggregation and adhesion under flow, and reduced platelet survival in vivo by increased apoptosis

5

Discussion

through Bak upregulation and a caspase-dependent pathway. We propose mechanisms based on our data (Figure 7) and potential interventions to reverse them. Besides the significantly reduced platelet response to physiological agonists in aggregometry, we found a reduced adhesion to vWF and collagen under flow after IBS treatment from the first day of storage (Figure 1), implicating a direct and rapid effect of the IBS on platelet function. This pattern is explained at the molecular level with a significant loss (about 20%) of surface GpIbÎą in IBS-treated platelets, and, accordingly, the corresponding accumulation of the cleaved glycocalicin in the supernatant plasma/Intersol (Figure 2A,B). The reduced aggregation over collagen could be explained by a reduced ability of platelets to

day

ethical approval) did not allow us to increase the number of samples. In aggregation experiments, the response of DANA samples to collagen and thrombin was not different from that of the IBS samples (data not shown).

Figure 5. Pre-incubation of platelets with SB203580 does not improve platelet survival and exacerbates apoptosis. (A) Platelets from AU untreated, IBS-treated and pre-treated with the p38 inhibitor SB203580 (final concentration 20 mM) were injected i.v. into NOD/SCID mice, and their survival in the circulation analyzed by flowcytometry over 5 hours. The p38 inhibitor did not improve platelet survival compared to untreated platelets. (B) GpIbÎą expression levels in platelets pre-treated with SB203580 were comparable to untreated platelets at all days of storage; *P=0.04 IBS vs. non-IBS; **P=0.01 IBS vs. SB203580. (C) When levels of desialylation were measured by FITC-conjugated ECA lectin binding, SB203580 was able to inhibit desialylation compared to IBS platelets at all days of storage; **P<0.01 nonIBS vs. IBS. (D) Pre-treatment of platelets with SB203580 induced caspase-3 cleavage; n=5, *P<0.05, **P<0.01, ***P<0.001. IBS: Intercept Blood System; plt: platelet; hu: human; ECA: Erythrina Cristagalli Agglutinin; MFI: mean fluorescence intensity; ECL: Erythrina Cristagalli Lectin.

1656

haematologica | 2017; 102(10)

Amotosalen /UVA and Platelet Clearance

respond to external stimuli due to increased Bak-dependent apoptosis, and, additionally, to indirect mechanisms caused by vWF “bridging” collagen to GpIb. In addition, the study from Hechler et al.32 found a significant loss of glycoprotein V (GPV) after Amotosalen/UVA treatment, and GPV was found to participate in platelet response to collagen.33 Therefore, we may speculate that IBS-induced GPV shedding could also be responsible for the reduced adhesion and aggregation in response to collagen.

Increased loss of GpIb is associated with the typical PSL in untreated platelets28,34 due to the activation of TACE; our results support the hypothesis of an accelerated lesion induced by the IBS which seems to be independent form the PSL, since it is observed from day 1 of storage. It has

7 day

day

day

7 day

day

day

E

5

D

1

C

5

B

1

A

also been shown that cold storage of platelets induces GpIb desialylation, which primes the receptor for TACEdependent shedding.35 Our results extend this observation to the IBS treatment, since we detected increased platelet desialylation in the treated samples as compared to the untreated platelets (Figure 3E) as well as a significantly increased neuraminidase activity in the supernatant from IBS samples, suggesting release of the enzyme from platelets after the PI; these observations suggest that this is an effect of the IBS (and perhaps of all PI technologies in general) as well as of storage over time. Interestingly, this is in line with our earlier structural and functional observations that deglycosylation of GpIb results in a collapse of GpIb on the membrane and a loss of platelet-vWF inter-

F

Figure 6. Neuraminidase inhibitor DANA partially rescues platelet survival in vivo. (A) control (◊), IBS (■) and DANA (●) platelets were injected in vivo into NOD/SCID mice and their survival analyzed over 5 hours. DANA-treated platelets were protected from clearance at the early time point (2 hours), at which time the clearance matched the non-IBS controls, but not at the later stage (5 hours). (B) GpIbα cleavage in DANA-treated samples was comparable to IBS samples at all days of storage; *P=0.03 IBS/DANA vs. non-IBS. (C) Desialylation was abrogated in the presence of DANA up to 7 days of storage; *P=0.03 IBS vs. non-IBS. (D) Immunofluorescence staining and (E) WB analysis of Bak and cleaved caspase-3 (F) revealed no difference between non-IBS, IBS and DANA samples; n=5, P>0.05. IBS: Intercept Blood System; plt: platelet; hu: human; MFI: mean fluorescence intensity; ECL: Erythrina Cristagalli Lectin; DANA: 2,3-dehydro-2-deoxy-N-acetylneuraminic acid.

haematologica | 2017; 102(10)

1657

S. Stivala et al. action.36 Thus, a dual effect of the IBS on GpIb (cleavage and desialylation) may lead to platelet clearance. Indeed, in vivo, we found a significant correlation of the increased clearance of IBS platelets in the spleen of NOD/SCID mice; platelet clearance correlated with GpIb levels, confirming the important role of this receptor for platelet survival, as also demonstrated by other groups.35,37,38 The sialidase inhibitor DANA seems to reduce, in part, the loss and clearance of IBS-treated platelets in vivo in our platelet survival model in NOD-SCID mice (Figure 5A-C), albeit only at the early time point. DANA pre-treated samples did not show an increase in Bak or in cleaved caspase-3 observed after IBS treatment. Interestingly, and perhaps surprisingly, we could not detect platelet clearance in the liver of these mice, in spite of the important role played by the Ashwell-Morell receptor on hepatocytes in the removal of desialylated platelets26,39 (data not shown); however, the extensive shedding of GpIbα after Amotosalen/UVA could be responsible for blocking the recognition and hepatic clearance of desialylated platelets.

At the intracellular level, we found that the IBS was linked to an increased phosphorylation of the signalling molecule p38 (Figure 2D), in agreement with previous reports of storage of untreated platelets.21,34 Interestingly, p38 is a known TACE activator,40 thus its increased phosphorylation is directly linked to the increment in GpIb cleavage observed in IBS platelets in this study (Figure 2A,B). Pre-incubation with a specific p38 inhibitor (SB203580) reduced GpIb shedding and desialylation upon

IBS treatment but did not improve platelet survival in mice (Figure 5A,B), suggesting that restoring GpIb levels is not sufficient to reduce platelet removal and that other mechanisms play a role in the accelerated clearance of IBS platelets, possibly through the induction of apoptosis, which was worsened by the p38 inhibitor as shown by cleaved caspase-3 levels (Figure 5D).

Other than cleavage, the induction of apoptosis could represent an important mechanism of platelet clearance, as has been shown for the riboflavin/UV light-based (Mirasol) PI.17,22 Expression of the pro-apoptotic protein Bak and cleavage of caspase-3 were significantly increased in IBS samples compared to non-IBS, confirming induction of platelet apoptosis as a mechanism of the reduced platelet function, and accelerated clearance after PI (Figure 2E-G). In contrast to previous studies,19,32,41 we did not detect an increased activation of the fibrinogen receptor GpIIbIIIa (Online Supplementary Figure S1C); this could be partly explained by the different protocol used for platelet collection, which was shown to affect platelet activation.42,43 Schripchenko et al. recently reported that p38 or sialidase inhibition could not block PSL caused by 4°C storage, in agreement with our results.44 However, this is in contrast to the results of other groups, which show amelioration of platelet function after p38 inhibition or GpIb shedding blockade during storage.21,34,45 The reason for these contrasting results remains unclear at this point. An intriguing hypothesis is that p38 activation in response to the stress associated with PI may have a protective role,

Figure 7. Schematic image representing the mechanisms found in our study. (1) IBS leads to phosphorylation of p38 (2), which in turn causes TNF-α converting enzyme (TACE) activation and GpIb shedding (3). The latter is also up-regulated by the release of neuraminidase from intracellular stores, which cleaves sialic acid residues on GpIbα, priming the receptor for cleavage. IBS also induces (likely by UV) upregulation of the pro-apoptotic Bak (4), which results in the induction of apoptosis through a caspase-dependent pathway.

Both processes lead to a reduced response of platelets to agonists and an accelerated clearance in vivo. GC: glycocalicin.

1658

haematologica | 2017; 102(10)

Amotosalen /UVA and Platelet Clearance

which leads to an increased apoptosis when inhibited, as reported by Rukoyatkina et al.46 An interesting observation of our study is that the IBS induces expression of the proapoptotic protein Bak, and this is replicated when freshly isolated platelets are irradiated with UVA without the addition of Amotosalen (Figure 3E,F and Figure 4B). We were also able to show that this occurs through an increased mRNA translation following its association with the protein eIF4E, considering that the protein synthesis inhibitor cycloheximide was able to block the increase in Bak after UV (Figure 4A,C). Since platelets contain mRNA and all the necessary machinery to enable them to translate into proteins,29,47 our results suggest that PI might trigger translation of specific mRNA inducing apoptosis, similar to the way in which it alters mRNA and microRNA expression.18,48 The development of PI represents a major cornerstone in transfusion medicine by reducing the risk of transfusion transmitted diseases in patients receiving blood products. The downside of this technology is the observation that PI exacerbates the PSL and has an impact on platelet function, although one study reported no change in platelet aggregation when washed platelets were used, the significance of which is not clear since the number of platelet concentrates analyzed was low.7,17,22,32,49,50 The study herein clearly demonstrates that platelet treatment with Amotosalen/UVA causes an alteration of platelet function. However, we also observed a detrimental effect with UVA treatment alone, and a nega-

References 1. Hersh EM, Bodey GP, Nies BA, Freireich EJ. Causes of death in acute leukemia: a tenyear study of 414 patients from 1954-1963.

JAMA. 1965;193:105-109.

2. Mayr WR. Blood transfusion in Europe-The White Book 2005: The patchwork of transfusion medicine in Europe. Transfus Clin Biol. 2005;12:357-358.

3. Whitaker B, Rajbhandary S, Kleinman S, Harris A, Kamani N. Trends in United States blood collection and transfusion: results from the 2013 AABB Blood Collection, Utilization, and Patient Blood Management Survey. Transfusion. 2016; 56(9):2173-2183.

4. Heal JM, Singal S, Sardisco E, Mayer T. Bacterial proliferation in platelet concentrates. Transfusion. 1986;26(4):388-390.

5. Butler C, Doree C, Estcourt LJ, et al.

Pathogen-reduced platelets for the prevention of bleeding. Cochrane database Syst Rev. 2013;3(3):CD009072.

6. Solheim BG. Pathogen reduction of blood components. Transfus Apher. Sci. 2008;39(1473):75–82.

7. Picker SM. Current methods for the reduction of blood-borne pathogens: A comprehensive literature review. Blood Transfus.

2013;11:343-348.

8. Irsch J, Lin L. Pathogen inactivation of platelet and plasma blood components for transfusion using the INTERCEPT Blood SystemTM. Transfus Med Hemother. 2011; 38(1):19-31.

9. Lozano M, Knutson F, Tardivel R, et al. A multi-centre study of therapeutic efficacy

haematologica | 2017; 102(10)

10.

11.

12.

13.

14.

15.

16.

tive impact on platelet function has been reported for γirradiation.51 Therefore, whether the IBS has a different or greater effect on platelets remains unclear. Importantly, we provided a mechanistic insight into the pathways involved in the negative effects of the Amotosalen/UVA treatment on platelets.

Although a large number of clinical studies did not demonstrate an inferior clinical efficacy of IBS-treated platelets, further research on the clinical outcomes of IBStreated platelet transfusion, focusing on bleeding, are necessary. The implementation of the IBS in more than 40 countries worldwide shows the necessity of technologies capable of reducing the risk associated with blood products transfusion, in spite of the alterations in platelet function caused by the procedure. Nevertheless, our observations indicate the importance of developing strategies that can be implemented to PI methods (such as new platelet additive solutions) in order to preserve platelet function and thus provide patients with safer, qualitatively optimal transfusion products.52-54 Acknowledgments We are grateful to Alexandra Plattner for her excellent technical support.

Funding This work was supported by the Swiss National Science Foundation grant #310030_144152 to J.H.B.

and safety of platelet components treated with amotosalen and ultraviolet A pathogen inactivation stored for 6 or 7 d prior to transfusion. Br J Haematol.

2011;153(3):393–401.

Kerkhoffs JLH, Van Putten WLJ, Novotny VMJ, et al. Clinical effectiveness of leucoreduced, pooled donor platelet concentrates, stored in plasma or additive solution with and without pathogen reduction. Br J Haematol. 2010;150(2):209–217. van Rhenen D, Gulliksson H, Cazenave J-P, et al. Transfusion of pooled buffy coat platelet components prepared with photochemical pathogen inactivation treatment: the euroSPRITE trial. Blood. 2003; 101(6):2426-2433.

Infanti L, Stebler C, Job S, et al. Pathogeninactivation of platelet components with the INTERCEPT Blood SystemTM: a cohort study. Transfus Apher Sci. 2011; 45(2):175181.

Vamvakas EC. Meta-analysis of the studies of bleeding complications of platelets pathogen-reduced with the Intercept system. Vox Sang. 2012;102(4):302-316.

Sigle J-P, Infanti L, Studt J-D, et al.

Comparison of transfusion efficacy of amotosalen-based pathogen-reduced platelet components and gamma-irradiated platelet components.

Transfusion.

2013; 53(8):1788-1797.

McCullough J, Vesole DH, Benjamin RJ, et al. Therapeutic efficacy and safety of platelets treated with a photochemical process for pathogen inactivation: the SPRINT Trial Blood. 2004; 104(5):1534– 1541.

Picker SM, Oustianskaia L, Schneider V,

17.

18.

19.

20.

21.

22.

Gathof BS. Annexin V release and transmembrane mitochondrial potential during storage of apheresis-derived platelets treated for pathogen reduction. Transfus Med Hemother. 2010;37(1):7-12.

Reid S, Johnson L, Woodland N, Marks DC. Pathogen reduction treatment of buffy coat platelet concentrates in additive solution induces proapoptotic signaling.

Transfusion. 2012;52(10):2094-2103.

Prudent M, D’Alessandro A, Cazenave JP, et al. Proteome changes in platelets after pathogen inactivation-an interlaboratory consensus. Transfus Med Rev. 2014; 28(2):72-83.

Abonnenc M, Sonego G, Kaiser-Guignard J, et al. In vitro evaluation of pathogen-inactivated buffy coat-derived platelet concentrates during storage: psoralen-based photochemical treatment step-by-step. Blood Transfus. 2015;13(2):255-264.

Picker SM, Oustianskaia L, Schneider V, Gathof BS. Functional characteristics of apheresis-derived platelets treated with ultraviolet light combined with either amotosalen-HCl (S-59) or riboflavin (vitamin B2) for pathogen-reduction. Vox Sang.

2009;97(1):26-33.

Schubert P, Coupland D, Culibrk B, Goodrich RP, Devine D V. Riboflavin and ultraviolet light treatment of platelets triggers p38MAPK signaling: inhibition significantly improves in vitro platelet quality after pathogen reduction treatment.

Transfusion. 2013;53(12):3164-3173.

Chen Z, Schubert P, Culibrk B, Devine D V. p38MAPK is involved in apoptosis development in apheresis platelet concentrates after riboflavin and ultraviolet light treat-

1659

S. Stivala et al. ment. Transfusion. 2014;1-10.

23. Stivala S, Reiner MF, Lohmann C, et al.

Dietary -linolenic acid increases the platelet count in ApoE-/- mice by reducing clearance. Blood. 2013;122(6):1026-1033.

24. Boylan B, Berndt MC, Kahn ML, Newman PJ. Activation-independent, antibodymediated removal of GPVI from circulating human platelets: development of a novel NOD/SCID mouse model to evaluate the in vivo effectiveness of anti-human platelet agents. Blood. 2006;108(3):908-914.

25. Dowling MR, Josefsson EC, Henley KJ, Hodgkin PD, Kile BT. Platelet senescence is regulated by an internal timer, not damage inflicted by hits. Blood. 2010; 116(10):17761778.

26. Sørensen AL, Rumjantseva V, NayebHashemi S, et al. Role of sialic acid for platelet life span: exposure of beta-galactose results in the rapid clearance of platelets from the circulation by asialoglycoprotein receptor-expressing liver macrophages and hepatocytes. Blood.

2009;114(8):1645-1654.

27. Kile BT. The role of the intrinsic apoptosis pathway in platelet life and death. J. Thromb. Haemost. 2009;7 Suppl 1:214217.

28. Bergmeier W, Burger PC, Piffath CL, et al.

Metalloproteinase inhibitors improve the recovery and hemostatic function of in vitro-aged or -injured mouse platelets.

Blood. 2003;102(12):4229-4235.

29. Weyrich a S, Dixon D a, Pabla R, et al.

Signal-dependent translation of a regulatory protein, Bcl-3, in activated human platelets. Proc Natl Acad Sci USA. 1998;95(10):5556-5561.

30. Grozovsky R, Hoffmeister KM, Falet H. Novel clearance mechanisms of platelets.

Curr Opin Hematol. 2010;17(6):585-589.

31. Li J, van der Wal DE, Zhu G, et al.

Desialylation is a mechanism of Fc-independent platelet clearance and a therapeutic target in immune thrombocytopenia.

Nat Commun. 2015;6:7737.

32. Hechler B, Ohlmann P, Chafey P, et al.

Preserved functional and biochemical characteristics of platelet components prepared with amotosalen and ultraviolet A for pathogen inactivation. Transfusion. 2013; 53(6):1187-1200.

33. Moog S, Mangin P, Lenain N, et al. Platelet

1660

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

glycoprotein V binds to collagen and participates in platelet adhesion and aggregation.

Blood. 2001;98(4):1038-1046.

Canault M, Duerschmied D, Brill A, et al. p38 mitogen-activated protein kinase activation during platelet storage: consequences for platelet recovery and hemostatic function in vivo. Blood. 2010; 115(9):1835-1842.

Jansen a JG, Josefsson EC, Rumjantseva V, et al. Desialylation accelerates platelet clearance after refrigeration and initiates GPIb metalloproteinase-mediated cleavage in mice. Blood. 2012;119(5):1263-1273.

Moshfegh K, Lengweiler S, Häner M, et al.

Fine structural and functional consequences of deglycosylation of the platelet adhesion receptor GPIb-IX (CD 42b). Biochem.

Biophys Res Commun. 1998;249(3):903909.

Leytin V, Allen DJ, Gwozdz A, Garvey B, Freedman J. Role of platelet surface glycoprotein Ibalpha and P-selectin in the clearance of transfused platelet concentrates.

Transfusion. 2004;44(10):1487-1495.

Hoffmeister KM, Josefsson EC, Isaac N a, et al. Glycosylation restores survival of chilled blood platelets.

Science.

2003; 301(5639):1531-1534.

Rumjantseva V, Grewal PK, Wandall HH, et al. Dual roles for hepatic lectin receptors in the clearance of chilled platelets. Nat Med. 2009;15(11):1273-1280.

Brill A, Chauhan AK, Canault M, et al.

Oxidative stress activates ADAM17/TACE and induces its target receptor shedding in platelets in a p38-dependent fashion.

Cardiovasc Res. 2009;84(1):137-144.

Sandgren P, Diedrich B. Pathogen inactivation of double-dose buffy-coat platelet concentrates photochemically treated with amotosalen and UVA light: Preservation of in vitro function. Vox Sang. 2015; 108(4):340-349.

Ali SF. Platelet activation of platelet concentrates derived from buffy coat and apheresis methods. Transfus Apher Sci. 2011;44(1):11-13.

Böck M, Rahrig S, Kunz D, Lutze G, Heim MU. Platelet concentrates derived from buffy coat and apheresis: Biochemical and functional differences. Transfus Med. 2002; 12(5):317-324.

Skripchenko A, Thompson-Montgomery

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

D, Awatefe H, Turgeon a., Wagner SJ. Addition of sialidase or p38 MAPK inhibitors does not ameliorate decrements in platelet in vitro storage properties caused by 4 °C storage. Vox Sang. 2014;360-367.

Chen W, Liang X, Syed AK, et al. Inhibiting GPIbα shedding preserves post-transfusion recovery and hemostatic function of platelets after prolonged storage.

Arterioscler Thromb Vasc Biol. 2016; ATVBAHA.116.307639.

Rukoyatkina N, Mindukshev I, Walter U, Gambaryan S. Dual role of the p38 MAPK/cPLA2 pathway in the regulation of platelet apoptosis induced by ABT-737 and strong platelet agonists. Cell Death Dis. 2013;4(11):e931.

Lindemann S, Tolley ND, Dixon D a, et al.

Activated platelets mediate inflammatory signaling by regulated interleukin 1beta synthesis. J Cell Biol. 2001;154(3):485-490.

Osman A, Hitzler WE, Meyer CU, et al.

Effects of pathogen reduction systems on platelet microRNAs, mRNAs, activation, and function. Platelets. 2014;1-14.

Van Aelst B, Feys HB, Devloo R, et al.

Riboflavin and amotosalen photochemical treatments of platelet concentrates reduce thrombus formation kinetics in vitro. Vox Sang. 2015;108(4):328-339.

Prudent M, Crettaz D, Delobel J, Tissot J-D, Lion N. Proteomic analysis of Intercepttreated platelets. J Proteomics. 2012;76 Spec No(Srts Vd):316-328.

Julmy F, Ammann RA, Fontana S, et al.

Transfusion efficacy of apheresis platelet concentrates irradiated at the day of transfusion is significantly superior compared to platelets irradiated in advance. Transfus Med Hemotherapy. 2014;41(3):176-181.

Hess JR, Pagano MB, Barbeau JD, Johannson PI. Will pathogen reduction of blood components harm more people than it helps in developed countries?

Transfusion.

2016;56(5):1236-1241.

Devine DV, Schubert P. Pathogen Inactivation Technologies. Hematol Oncol Clin North Am. 2016;30(3):609-617.

Corash L, Benjamin RJ. The role of hemovigilance and postmarketing studies when introducing innovation into transfusion medicine practice: the amotosalen-ultraviolet A pathogen reduction treatment model.

Transfusion. 2016; 56(March):S29-S38.

haematologica | 2017; 102(10)

ARTICLE

Platelet Biology & its Disorders

Bone marrow pathologic abnormalities in familial platelet disorder with propensity for myeloid malignancy and germline RUNX1 mutation Rashmi Kanagal-Shamanna,†1 Sanam Loghavi,1 Courtney D. DiNardo,2 L. Jeffrey Medeiros,1 Guillermo Garcia-Manero,2 Elias Jabbour,2 Mark J. Routbort,1 Rajyalakshmi Luthra,1 Carlos E. Bueso-Ramos1 and Joseph D. Khoury†1

1 Department of Hematopathology and 2Department of Leukemia, The University of Texas M.D. Anderson Cancer Center, Houston, TX, USA

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2017 Volume 102(10):1661-1670

ABSTRACT

A

subset of patients with familial platelet disorder with propensity to myeloid malignancy and germline RUNX1 mutation develops hematological malignancies, often myelodysplastic syndrome/acute myeloid leukemia, currently recognized in the 2016 WHO classification. Patients who develop hematologic malignancies are typically young, respond poorly to conventional therapy, and need allogeneic stem cell transplant from non-familial donors. Understanding the spectrum of bone marrow morphologic and genetic findings in these patients is critical to ensure diagnostic accuracy and develop criteria to recognize the onset of hematologic malignancies, particularly myelodysplastic syndrome. However, bone marrow features remain poorly characterized. To address this knowledge gap, we analyzed the clinicopathologic and genetic findings of 11 patients from 7 pedigrees. Of these, 6 patients did not develop hematologic malignancies over a 22-month follow-up period; 5 patients developed hematologic malignancies (3 acute myeloid leukemia; 2 myelodysplastic syndrome). All patients had thrombocytopenia at initial presentation. All 6 patients who did not develop hematologic malignancies showed baseline bone marrow abnormalities: low-for-age cellularity (n=4), dysmegakaryopoiesis (n=5), megakaryocytic hypoplasia/hyperplasia (n=5), and eosinophilia (n=4).

Two patients had multiple immunophenotypic alterations in CD34-positive myeloblasts; 1 patient had clonal hematopoiesis. In contrast, patients who developed hematologic malignancies had additional cytopenia(s) (n=4), abnormal platelet granulation (n=5), bone marrow hypercellularity (n=4), dysplasia in ≥2 lineages including megakaryocytes (n=3) and acquired clonal genetic aberrations (n=5). In conclusion, our study demonstrated that specific bone marrow abnormalities and acquired genetic alterations may be harbingers of progression to hematological malignancies in patients with familial platelet disorder with germline RUNX1 mutation.

Introduction The widespread use of next-generation sequencing (NGS)-based assays has facilitated an increased recognition of familial clustering of myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML).1 Familial syndromes in which MDS/AML is a primary feature include familial platelet disorder with predisposition to myeloid malignancy (FPDMM) associated with germline RUNX1 mutations, GATA2-associated syndromes, familial AML with CEBPA mutation, and syndromes associated with germline mutations in SRP72, ANKRD26, DDX41, or ETV6.2,3 Accordingly, the 2016 haematologica | 2017; 102(10)

Correspondence: rkanagal@mdanderson.org or jkhoury@mdanderson.org Received: March 17, 2017.

Accepted: June 20, 2017.

Pre-published: June 28, 2017. doi:10.3324/haematol.2017.167726 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1661 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1661

R. Kanagal-Shamanna et al.

revision to the WHO classification system for myeloid neoplasms has incorporated a section on “myeloid neoplasms with germline predisposition”.4 FPDMM (OMIM #601399) is an autosomal-dominant

disorder with variable penetrance genetically defined by the presence of germline RUNX1 mutation. RUNX1 encodes one of the α subunits of a core-binding transcription factor and plays a critical role in hematopoiesis,

Table 1A. Clinical, laboratory and peripheral blood findings on FPDMM patients.

Pedigree

Age/ sex

Family history

Diagnosis

Follow up

WBC

Hgb

MCV

PC

MPV

A II-3

46/ M

+

A&W, 29 months

6.3

12.9

99

119

12.3

A III-1

7/ F

+

9.7

103

63

10.1

53/ F

+

4.9

9.3

83

20

9.1

C II-1

48/F

+

3.2

11

100

69

9.3

C II-3

42/M

+

4

13

103

33

NA

A III-2 C I-1 D I-1 E II-1 F I-1 G I-1

4/ F 70/F 57/ M 39/ M 27/ M 14/ F

+ + + + + -

A&W, 27 months. 23 months post stem cell transplant Died; 56.1 months from HM diagnosis Died, 54 months from HM diagnosis Died, 8 months from HM diagnosis A&W, 25 months A&W, 8 months A&W, 39 months Lost for follow up A&W, 15 months A&W, 22 months

3.9

B II-1

FPDMMHM+ MDS-MLD FPDMMHM+ MDS-EB-1 FPDMMHM+ AML MRC FPDMMHM+ AML MRC, SM FPDMMHM+ AML MRC FPDMMHMFPDMMHMFPDMMHMFPDMMHMFPDMMHMFPDMMHM-

6.8 5.5 6.5 3.2 4.5 5.9

12.5 13.1 14 14.1 14.4 13.1

86 95 89 93 93 88

122 134 92 77 99 88

8.7 9.9 9.5 9.6 7.2 NA

WBC: white blood cell count; Hgb: hemoglobin; MCV: mean corpuscular volume; PC: platelet count; MPV: mean platelet volume. Cytopenia(s) defined by laboratory reference range (matched for age). A&W,: alive and well; AML MRC: acute myeloid leukemia with myelodysplasia-related changes; FPDMMHM+: FPDMM with hematological malignancy; FPDMMHM-: FPDMM without hematological malignancy; MDS-MLD: myelodysplastic syndrome with multilineage dysplasia; MDS-EB-1: myelodysplastic syndrome with excess blasts-1; HM: hematological malignancies; ND: not done; NA: not available; SM: systemic mastocytosis. .

Table 1B. Bone marrow morphologic and flow cytometry immunophenotypic findings on the FPDMM patients.

-

2 2 1

1.00 2.76 83.00

+ + +

-

-

-

↓ + ↑

+ ↑ ↑

0

93.00

+

+

+

+ -

+ + + +

-

+ + -

<1 0 0 0 0 ND

91.00 + 0.80 + 0.08 + 0.60 0.50 + ND ND

+ + + + + ND

-

-

-

-

0

0.70

+

+

Hematogones

CD34

-

+ ↑ ↑

↑ + ↑

+

+

+

+ + + + + ND

+ + ↓ ↓ ↓ ND

+ + + ↑ + ND

+ ↑ + ↑ ↑ ND

+ + ND

+

+

+

+

+

CD38

CD123

CD33

+ + -

+ + too few -

CD117

CD13

-

CD34+ blasts (%)

↓

Ring sideroblasts (%)

1

↓ ↓ ↑ ↑ ↓ too few* ↑

Lymphoid aggregates

G I-1

↑ ↓ ↑ Adequate ↓ ↓

Fibrosis

5 4 2 5 1 1

↓

+ too few too few + + -

Flow cytometry

Eosinophilia

C II-3 A III-2 C I-1 D I-1 E II-1 F I-1

↑

+ + too few too few + + + + + too few* +

Erythroid dysplasia

13

↑ ↓ ↓

Granulocytic dysplasia

C II-1

↑ ↓ ↑

Megakaryocyte dysplasia

7 8 11

Megakaryocyte number

# of BM specimens

A II-3 A III-1 B II-1

Age-matched cellularity

Pedigree

Bone marrow

↓ ↓ +

+ -

*Suboptimal quality bone marrow (insufficient number of megakaryocytes available for evaluation). Dysplasia defined as dysmorphic forms >10% of the megakaryocytes. ND: not done; BM: bone marrow.

1662

haematologica | 2017; 102(10)

Bone marrow pathologic changes in FPDMM

myeloid differentiation and platelet function.5 FPDMM is characterized by abnormalities in platelet number and/or function, namely defective release of δ granules, and a propensity to develop early-onset MDS/AML or, rarely, Tlymphoblastic leukemia/ lymphoma.6 Until now, about 50 pedigrees with germline RUNX1 mutations have been reported.6-15 A subset (median, 35%; range: 22-60%) of FPDMM patients undergoes transformation to hematological malignancies (HM), usually MDS or AML (FPDMMHM+), associat-

ed with the acquisition of additional somatic genetic lesions.3,14,16 FPDMMHM+ respond poorly to conventional therapy and require unique management strategies such as allogeneic stem cell transplant (in all pediatric patients and in eligible adult patients during remission), genetic counseling and work-up and identification of family members with germline RUNX1 mutation.1,3,13,17-19 In this setting, allogeneic stem cell transplant is generally from unrelated donors who need to be carefully screened for germline mutations. Close surveillance and prompt recognition of FPDMMHM+ facili-

Table 2. FPDMM pedigrees with cytogenetic results and somatic mutation analysis using a combination of next-generation based sequencing and conventional techniques.

Pedigree

A II-3

Diagnosis

FPDMMHM+ MDS-MLD

Genetic testing for germline

Somatic mutations*

RUNX1 Mutation

Exon

Somatic mutations

by NGS

c.582A>C p.K194N

6

None

VAF (%)

FPDMMHM+ MDS-EB-1

c.582A>C p.K194N

6

None

B II-1

FPDMMHM+ AML MRC

c.719delC p.Pro240Hisfs c.167T>T p.Leu56Ser (probably benign)

7

NM_001754.4(RUNX1): c.334_339del p.L112_P113del NM_004985.3(KRAS): c.101C>G p.P34R NM_005896.2(IDH1): c.394C>T p.R132C

29.9 20

31.9 30.6 22.4

FPDMMHM+ AML MRC, SM

Partial gene deletion (at least exons 1-6)

1 through 6

C II-3

FPDMMHM+ AML MRC

1 through 6

A III-2

FPDMMHM-

Partial gene deletion (at least exons 1-6) c.582A>C p.K194N

NM_000222.2(KIT): c.2447A>T p.D816V NM_001754.4(RUNX1): c.485G>A p.R162K NM_024426.4(WT1): c.1142dupC p.A382fs None

6

None

C I-1

FPDMMHM-

1 through 6

D I-1 E II-1 F I-1 G I-1

FPDMMHMFPDMMHMFPDMMHMFPDMMHM-

Partial gene deletion (at least exons 1-6) c.836G>A p.W279* c.496C>T p.R166* c.308dup p.T104fs c.1098_1103dupCGGCAT p.I366_G367dup

NM_022552.4(DNMT3A): c.1015-2A>G (splice site) None ND ND None

8 5 4 9

FISH

46,XY,del(11) Deletion of 1 (q13q23)[8]/ 46,XY[14] copy of 11q23 (MLL)

A III-1

C II-1

Karyotype

Cytogenetic studies

46,XX,del(5)(q31q34) [18]/ 46,XX[2]

Deletion (5q)

46,XX,t(2;22)(p23;q13.1), del(7)(q22q32)[20]

t(2;22)(p23;q13.1) (wcp22+;wpc22+) Deletion 7q

46,XX[20]

ND

46,XY,del(7)(q22)[20]

ND

46,XX[20]

46,XX[20]

No trisomy 8, deletions of 5/5q or 7/7q and 11q23 (MLL) ND

46,XY[20] 46,XY,inv(9)(p12q13)[20] 46,XY[20] 46,XX,inv(9)(p12q13)[20]

ND ND ND ND

2.5

14.1

*NGS-based somatic gene mutation analysis using 28-gene myeloid panel, FLT3 ITD and CEBPA. NGS: next-generation sequencing; FPDMM: familial platelet disorder with predisposition to myeloid malignancy; FISH: fluorescence in situ hybridization; AML MRC: acute myeloid leukemia with myelodysplasia-related changes; FPDMMHM+: FPDMM with hematological malignancy; FPDMMHM-: FPDMM without hematological malignancy; MDS-MLD: myelodysplastic syndrome with multilineage dysplasia; MDS-EB-1: myelodysplastic syndrome with excess blasts-1; ND: not done; SM: systemic mastocytosis; VAF: variant allele frequency.

haematologica | 2017; 102(10)

1663

R. Kanagal-Shamanna et al.

tates planning and timely therapeutic interventions before or at the time of leukemic transformation.

The diagnosis of MDS in FPDMM is particularly challenging. Few reports have described dysplastic changes in megakaryocytes due to the underlying germline RUNX1 mutation in asymptomatic FPDMM patients.6,20,21 Additionally, the frequency of clonal hematopoiesis in asymptomatic FPDMM patients below 50 years of age is significantly higher (~67%) compared to that of the healthy general population.2,22 Currently, there are no criteria or guidelines available in the literature for diagnosis, evaluation and monitoring for HM in these patients.4 On the other hand, due to the aggressive therapeutic interventions implicated by the diagnosis of a HM, diagnostic accuracy and avoidance of overcalling MDS is of critical importance.

Thus, there is a need to determine the pathologic features associated with FPDMM and progression to MDS. To begin addressing these gaps in knowledge, a thorough understanding of the bone marrow (BM) features in FPDMM patients and the characteristics associated with progression to HM is required. It is our understanding that no other study has addressed this issue in a systematic manner and this much needed knowledge base is currently lacking for pathologists and the rest of the clinical diagnostic team who are required to diagnose and evaluate patients with FDPMM associated with RUNX1 mutation.

In the study herein, we performed a systematic evaluation of BM morphologic, cytogenetic and molecular findings in 11 patients from 7 distinct FPDMM pedigrees at various stages of disease evolution. We show that baseline BM morphologic and immunophenotypic abnormalities are present in asymptomatic FPDMM patients without MDS/AML. Awareness of these changes is important in order to exert caution in establishing a diagnosis of MDS, an actionable event in this context. We also compared the clinical, morphologic, cytogenetic, immunophenotypic and genetic findings between patients with FPDMMHM- and FPDMMHM+ who had been followed with serial BM examinations over a median interval of 27 months. We identified specific pathologic features and we propose criteria that can facilitate the recognition of MDS in this setting for timely

therapeutic interventions. These findings also highlight the need for baseline and serial BM examination with multimodal ancillary testing to monitor for development of MDS/AML.

Methods Study Group We selected pedigrees of FPDMM with germline RUNX1 mutations that were evaluated at our institution. In some cases, the proband (defined here as the first diagnosed family member) was evaluated at an outside hospital or clinic, whereas other members were referred to our institution following the probandâ&#x20AC;&#x2122;s diagnosis.

This study was approved by the Institutional Review Board and informed consent was obtained from all patients in accordance with the Declaration of Helsinki.

Histopathologic Evaluation Hematoxylin-eosin stained BM core biopsy and/or clot specimens and Wright-Giemsa-stained peripheral blood (PB) and BM aspirate smears and/or touch imprints at baseline and various time points were assessed using standard criteria.23,24 Cytopenia(s) were defined based on institutional laboratory reference ranges. For enumeration of megakaryocytes, we considered 2-6 megakaryocytes per high-power field as a criterion for normal range. Prussian blue staining was used for quantifying ring sideroblasts. In selected cases, immunohistochemistry studies for CD34 and CD61 were performed using standard techniques on automated stainers (Leica Biosystems, Buffalo Grove, IL, USA) using antibodies against CD34 (MY10, 1:40; BD Biosciences, Franklin Lakes, NJ, USA) and CD61 (2F2, 1:100; Cell Marque, Rocklin, CA, USA). The morphologic findings were independently reviewed by 2 independent hematopathologists (RK-S and JDK).

Multiparameter Flow Cytometry Analysis Flow cytometry (FC) immunophenotypic analysis was performed on BM aspirates as described previously.25,26 Aberrancies in expression levels of CD13, CD33, CD34, CD38, CD117, CD123 and additional markers were assessed on CD34+/CD10-/CD19myeloid precursors and hematogones were quantified.

Figure 1. Representative image showing the location of the various types of exonic RUNX1 mutations in this study group.

1664

haematologica | 2017; 102(10)

Bone marrow pathologic changes in FPDMM

Abnormalities in myelomonocytic maturation were assessed as previously described.26

Karyotyping and Fluorescence in situ Hybridization Conventional G-band karyotype analysis and fluorescence in situ hybridization (FISH) on selected cases were performed using standard methods described previously.27,28 All results were reported according to the 2013 International System for Human Cytogenetic Nomenclature.29

Gene Mutation Analysis For germline RUNX1 variant detection, we used genomic DNA extracted from disease-free whole blood/BM samples (in FPDMMHM-) or cultured skin fibroblasts (in FPDMMHM+). For all pedigrees except C, RUNX1 mutation was assessed by polymerase chain reaction (PCR) amplification followed by direct sequencing.

NM_001754.3 was used as the reference sequence for the RUNX1 gene for alignment. For pedigree C, amplicon-based exome sequencing that targeted the ANKRD26, CEBPA. DDX41, ETV6, FLI1, GATA2, RUNX1, SRP72, and TP53 genes using the Illumina system was performed. The deletion was confirmed by exon-level oligo comparative genomic hybridization. Clinical interpretation was performed per published guidelines.30 For assessment of somatic mutations, amplicon-based NGSbased analysis using a clinically-validated 28-gene myeloid panel

was performed on genomic DNA extracted from BM on a MiSeq sequencer (Illumina, San Diego, CA, USA) as described previously.31,32 FLT3 internal tandem duplications and CEBPA mutations were assessed by well-established alternative methods.31

Results Clinical characteristics Our study cohort included 11 patients with FPDMM with germline RUNX1 mutations from 7 unique pedigrees, labeled A through G. There were 6 females and 5 males with a median age of 42 years (range: 4-70) who were tested in various clinical settings and at different stages of clinical progression. The median age at time of diagnosis of MDS/AML in FPDMMHM+ patients was 45 years (range: 753). The patients were either asymptomatic or had a longstanding propensity for bleeding that was often misdiagnosed as immune thrombocytopenic purpura. Upon evaluation, all patients had mild to moderate thrombocytopenia.

The median platelet count was 88 x 109/L (range: 20-134).

The clinical characteristics are presented in Table 1. Six of 7 pedigrees had a family history of thrombocytopenia or leukemia; 1 patient had â&#x20AC;&#x153;sporadicâ&#x20AC;? thrombocytopenia, however, details regarding the family history on the pater-

A

B

C

D

Figure 2. Representative images from the BM biopsy/ aspirate smears of the asymptomatic FPDMM patients from various pedigrees. (A) Pedigree E (II-1): hypocellular for age BM with decreased megakaryocytes that included >10% forms that were small in size with nuclear hypolobation and single lymphoid aggregate; inset, PB smear showing thrombocytopenia with normal sized platelets. (B) Pedigree G (I-1): slightly hypocellular for age BM with increased megakaryocytes, including small hypolobated forms; (C) 4-year old sister (III-2) showing a hypocellular for age marrow with frequent dysmorphic megakaryocytes; inset, aspirate smear arrow showing a small abnormal megakaryocyte. (D) Aspirate smear shows eosinophilia.

haematologica | 2017; 102(10)

1665

R. Kanagal-Shamanna et al.

Table 3. Proposed criteria for diagnosing myelodysplastic syndrome in individuals with familial platelet disorder with propensity for myeloid malignancy and germline RUNX1 mutations.

Major criteria 1. Identification of germline RUNX1 mutation 2. Cytopenia in ≥1 hematopoietic lineage, other than thrombocytopenia 3. Exclusion of non-neoplastic causes of cytopenias 4. Bone marrow and peripheral blood blasts <20%

Minor criteria

A. Morphologic features of myelodysplasia in ≥2 hematopoietic lineages B. Acquired clonal cytogenetic or molecular genetic abnormality All major criteria and one of the minor criteria are required to make a diagnosis of myelodysplastic syndrome. .

nal side were not available for this latter patient. The reasons for germline RUNX1 mutation testing included: evaluation for early-onset MDS, an extensive family history of MDS/AML, FPDMM diagnosis in a relative, and confirmation of a suspected germline mutation identified by NGSbased multi-gene somatic mutation profiling for thrombocytopenia in a patient with no known family history of bleeding or leukemia. The pedigree, diagnosis and genetic alterations, to various extents, in 4 of the 7 pedigrees have been reported previously.13 The detailed pedigrees for each of the families are provided in the Online Supplementary Figures S1 and S2.

FPDMMHM+ patients was lower at 63 (range: 20-119).

Contrary to FPDMMHM-, all patients with FPDMMHM+ had 1 or more cytopenia(s) in addition to thrombocytopenia (anemia in 4 and leukopenia in 2 patients). In 1 patient, the platelet count decreased further at the time of development of MDS. Macrocytosis was present in 4 out of 5 patients.

Absolute eosinophilia was noted in 1 out of 5 patients. PB smears showed platelets with anisocytosis and abnormalities in granulation (hypogranulation and agranulation). Two out of 4 patients showed dysplastic neutrophils that included cytoplasmic hypogranulation and abnormal nuclear segmentation.

Characteristics of germline RUNX1 mutations

BM histologic findings

The types of germline RUNX1 alterations observed in this study cohort included substitutions (1 missense, 2 nonsense), duplications (n=2) and deletions (n=2). The deletion in pedigree C was large and spanned exons 1 through 6.

Four of the 7 RUNX1 germline alterations involved the Runt1 homology domain (RHD); 3 involved the transactivation domain (TAD). Two cases that transformed to AML had additional somatic RUNX1 mutations, both of which involved the Runt domain. The location and type of germline alterations identified in each of the pedigrees are depicted in Figure 1. Five of the detected RUNX1 mutations have not been reported previously.

BM from all 6 FPDMMHM- patients showed baseline morphologic abnormalities. Age-matched BM cellularity was decreased in 4 patients, increased in 1 patient and adequate in 1 patient. Morphologic abnormalities were most apparent in the megakaryocytic lineage. The number of megakaryocytes was increased in 3 patients, decreased in 2 patients and could not be assessed in 1 patient.

Megakaryocytes were dysmorphic, often small with scant cytoplasm and nuclear hypolobation, with asynchronous nuclear cytoplasmic maturation; the dysmorphic forms accounted for more than 10% of the megakaryocytes in 5 patients, barring 1 case in which megakaryocytic dysplasia could not be evaluated due to the poor quality of the specimen. In 1 case megakaryocytic dysplasia was associated with granulocytic dysplasia; however, a diagnosis of MDS was not established because the patient did not have unexplained cytopenia(s) other than mild and stable thrombocytopenia. None of the other patients showed dysplasia in the granulocytic or erythroid lineages. BM eosinophilia was present in 4 out of 6 patients. None of the patients had fibrosis. BM findings in representative FPDMM cases within pedigrees are presented in Figures 2 and 3.

BM samples from 4 out of 5 FPDMMHM+ patients showed increased BM cellularity for age compared to FPDMMHMpatients (80% versus 17%, P=0.08, Fisher’s exact test). The megakaryocytes were adequate in number in 1 patient, and decreased in 4 patients. Three patients had sufficient precursor cells for adequate morphologic evaluation. All patients had dysmegakaryopoiesis with dysplasia in an additional lineage (dyserythropoiesis and/or dysgranulopoiesis). Two FPDMM/AML patients had too few cells to assess for dysplasia due to the presence of many blasts. BM eosinophilia was present in 2 out of 5 patients. None of the 5 patients had BM fibrosis. The diagnoses on FPDMMHM+

PB, BM histologic, immunophenotypic, cytogenetic and molecular findings Within our study group, 6 of 11 patients with FPDMM had no evidence of MDS/AML (FPDMMHM-), and 5 patients developed AML or MDS (FPDMMHM+) over the follow-up time period. The 5 FPDMMHM+ patients included FPDMM with AML (n=3), and FPDMM with MDS (n=2). All 11 patients underwent BM examination; 7 patients (including 3 of 6 FPDMMHM- patients) had BM evaluations performed at multiple time points.

PB findings All 6 FPDMMHM- patients had stable thrombocytopenia (median, 96 x 103 /mL, range: 77-134); 1 patient also had mild leukopenia (no decreased absolute neutrophil count) and no patients had anemia. Mean corpuscular volume (MCV) was within normal range in all patients. Absolute eosinophilia was noted in 3 out of 6 patients. PB smears showed normal sized platelets in all but 1 patient who had normal sized platelets with few large forms that were adequately granulated. In contrast, the median platelet count of 1666

haematologica | 2017; 102(10)

Bone marrow pathologic changes in FPDMM

patients included: MDS with excess blasts (EB)-1 (n=1), MDS with multilineage dysplasia (n=1), and AML with myelodysplasia-related changes (MRC; n=3). AML-MRC was attributable to a history of MDS in 1 patient, morphologic dysplasia in >50% of precursors in at least 2 lineages in a second patient and del(7q) abnormality in a third patient; this patient also developed systemic mastocytosis during remission.

Immunophenotypic findings The immunophenotype of CD34-positive myeloid blasts was analyzed by FC immunophenotyping in 5 FPDMMHMpatients. CD34-positive myeloblasts showed immunophenotypic abnormalities similar to those observed in MDS or a stem cell neoplasm. These included CD13 increased (1/5), CD38 decreased (3/5), CD117 increased (1/5), and CD123 increased (3/5). Hematogones were absent in 2 cases. In 2 out of 5 FPDMMHM- patients, multiple FC aberrations which are typical of MDS or a stem cell neoplasm were noted.

The immunophenotypic findings in the FPDMMHM+ patients showed aberrancies consistent with the diagnosis.

The PB and BM findings are summarized in Table 1.

Somatic clonal cytogenetic and molecular aberrations Karyotype data were available for all patients and NGS mutation data were available for 9 out of 11 patients. None of the 6 FPDMMHM- patients had karyotypic abnormalities.

Two out of 6 patients showed inv(9) chromosomal polymorphism. One out of 4 FPDMMHM- patients who underwent NGS testing showed a somatic DNMT3A splice site mutation (allele frequency 14.1%). In the absence of cytopenia or hematologic malignancy, this finding was consistent with clonal hematopoiesis of indeterminate potential [pedigree C 1-1].

In contrast, all 5 FPDMMHM+ patients had acquired clonal cytogenetic abnormalities and/or somatic gene mutation(s) in addition to germline RUNX1 mutation (4 with karyotype abnormality; 2 with somatic mutations and both of these cases also had a second RUNX1 mutation). These results are summarized in Table 2.

Clinical course and outcome Follow-up data were available for 10 out of 11 patients (Table 1). The median follow-up duration was 27.4 months (range: 8-56.1). The median follow up for FPDMMHMpatients was 22.3 months. Five FPDMMHM- patients with available follow-up data are alive without development of hematological malignancy. All 3 FPDMMHM+ patients with AML (B II-1, C II-1 and C II-3) died. The FPDMMHM+ patient with MDS-EB-1 (A III-1) underwent allogeneic stem cell transplant and is alive and well. The FPDMMHM+ patient (A II-3) who developed MDS with multilineage dysplasia is awaiting therapy with hypomethylating agents.

Using serial BM examinations and comparing certain specific parameters to the baseline BM, 1 patient progressed to overt MDS over a 29-month follow-up period. This asymptomatic patient (A II-3) was evaluated solely due to the diagnosis of FPDMMHM+ in the offspring. The platelet count was minimally decreased and perhaps present life-long, and attributable to germline RUNX1 mutation. Mild anemia (Hgb, 12.9 g/dL) was noted at presentation, but the significance and duration were not clear. The initial baseline BM showed ~10% dysmorphic megakaryocytes without granulocytic or erythroid dysplasia. FC immunophenotypic findings showed aberrant CD34 positive myeloblasts with haematologica | 2017; 102(10)

decreased CD13 and CD38 expression and increased CD123 expression (Table 1). Conventional cytogenetic studies showed a low-level del(11)(q13q23) in 2 out of 30 metaphases, confirmed to involve KMT2A/MLL deletion

by FISH in 9.5% of interphase nuclei. At this time, although del(11q) is an MDS-defining abnormality, due to the lowlevel of the del(11q) clone, in the absence of dysplasia in other lineage(s) other than megakaryocytes, and unclear etiology of anemia, a diagnosis of MDS was not made but a concern was raised. Therefore, the patient was monitored closely with BM exams every 6 months. Over a 23-month follow-up interval, the patient remained anemic and thrombocytopenic, and del(11q) persisted at a low level (1-2 of 20 metaphases). At the 29-month follow up, BM showed additional dysplasia involving the erythroid lineage that coincided with expansion of the del(11q) clone to 8 out of 20 metaphases. At this time, the patient was diagnosed with MDS. At last follow up, he was scheduled to start treatment with hypomethylating agents.

Discussion We describe the spectrum of BM pathologic findings in 11 patients belonging to 7 unique FPDMM pedigrees identified in various clinical settings and at different stages of clinical progression. The findings in this study highlight the importance of initial and serial BM evaluation in FPDMM patients and underscore the need to establish specific criteria for the diagnosis of MDS in patients with a germline predisposition to MDS/AML. Although few studies have described megakaryocytic abnormalities in BM aspirate smears,6,20,21 systematic analysis of BM morphologic, immunophenotypic, and genetic findings in FDPMM are rare. The only other study by Tsang et al. reported serial BM specimens of patients with various etiologies of congenital thrombocytopenia, including a case of FPDMM. The authors describe 3 morphologic patterns at initial presentation: (1) cases with myelodysplastic/ myeloproliferative features such as hypercellularity, myeloid predominance and numerous micromegakaryocytes, (2) cases with hypocellular marrow and small megakaryocytes, as noted in the case with FPDMM, paralleling the observations in our series, and (3) cases with normal morphology.

Similar to the findings in our study group, disease progression in their series was also associated with the development of additional dysplasia in the erythroid and myeloid lineages as well as cytogenetic abnormalities.33 In the study herein we show that FDPMM is characterized by baseline morphologic abnormalities that include low-for-age BM cellularity and dysplastic megakaryopoiesis in the absence of MDS/AML. A subset of these FPDMMHMpatients also showed immunophenotypic aberrancies in CD34-positive myeloid blasts diagnostic of a stem cell neoplasm. However, these patients lacked unexplained cytopenia(s) other than thrombocytopenia, and have been followed up over a median of 22.3 months without development of MDS/AML. Awareness of these baseline BM abnormalities in FPDMM patients is important for diagnosticians not to overcall MDS in FPDMM patients. This is not simply an academic issue; the diagnosis of MDS in FDPMM patients has considerable treatment implications. MDS and AML arising in the context of FDPMM respond poorly to conventional therapy; therefore, aggressive interventions including allogeneic stem cell transplant from non-familial 1667

R. Kanagal-Shamanna et al.

A

B

C

D

Figure 3. Representative images of the BM biopsy/ aspirate smears of FPDMM patients with MDS. (A) 7-year old girl diagnosed with MDS (proband, III-1); BM biopsy is hypocellular for age; (B) BM aspirate smears show increased blasts (dashed arrows) and severe neutrophilic hypogranulation; (C) fatherâ&#x20AC;&#x2122;s (II-3) BM with frequent megakaryocytes that are small and hypolobated; (D) CD61 immunohistochemistry highlights dysmorphic megakaryocytes.

donors is required.

FPDMM patients showed an unusually high proportion of dysmorphic megakaryocytes which were small in size with scant basophilic cytoplasm and hypolobated nuclei.

These abnormalities corroborate the results of published in vitro studies showing the critical role of RUNX1 in terminal differentiation of megakaryocytes17 and the impact of RUNX1 mutations on megakaryopoieisis.34-36 Bluteau et al. demonstrated, in vitro, morphologic abnormalities in megakaryocytes derived from CD34+ hematopoietic progenitors of FPDMM patients. These abnormalities were related to a block in maturation, manifested as small megakaryocytes with a high nucleus/cytoplasm ratio, and decreased ploidy or nuclear lobation.34 Dysmegakaryopoiesis in the BM of FPDMM patients similar to those observed in our study have been described in a few reports.6,20,21,33 Bluteau et al. also demonstrated a 60-80% decrease in the total number of megakaryocytes derived from CD34+ hematopoietic progenitors of FPDMM patients. In contrast, a substantial percentage of patients showed increased numbers of megakaryocytes in the present case series. The variant findings may represent the phenotypic diversity of the various RUNX1 mutations.34 Another interesting finding was PB and BM eosinophilia that were noted in 4 out of 6 FPDMMHM- patients in the study herein. Eosinophilia is characteristic of AML with RUNX1 translocations, such as AML with RUNX1RUNX1T1 or AML with RUNX1â&#x20AC;&#x201C;CBFA2T3.4,37 However, it 1668

is unclear if eosinophilia in FPDMM is directly related to the RUNX1 mutation or secondary to another etiology. One FPDMM patient (C II-1) who progressed to AML subsequently developed systemic mastocytosis with KIT D816V mutation during remission, suggesting an alternative explanation for eosinophilia in this patient. There was no increase in mast cells in the other family members in this pedigree who underwent evaluation (C II-3 and C I-1).

The findings in the study herein illustrate the unique challenges associated with the diagnosis of MDS in patients with FPDMM and the need for developing specific criteria for establishing the diagnosis of MDS in these patients.

However, at this time, there are no guidelines available for this purpose. By comparing the clinicopathologic features of FPDMMHM+ and FPDMMHM- patients in this study, we found certain consistent findings in all FPDMM patients with MDS/AML. These findings included: (1) presence of anemia and/or leukopenia in addition to thrombocytopenia, (2) multilineage dysplasia, and (3) presence of an additional clonal (somatic) event, either a cytogenetic or a molecular aberration. Based on these findings, we propose the following criteria as a helpful guide for the diagnosis of MDS in FPDMM patients (Table 3). However, the series is small, and the proposed criteria need validation in additional studies. The criteria should be used with appropriate clinico-pathologic correlation. The decision to treat MDS is very challenging as it must be tailored to the individual and as such requires input from both oncologists and pathologists. haematologica | 2017; 102(10)

Bone marrow pathologic changes in FPDMM

Other studies have shown that leukemic transformation in FPDMM is always associated with an additional clonal (somatic) event.1,16,38-40 Acquired cytogenetic aberrations include del(5q), +8, del(7q), del(11q23), trisomy 12, and t(2;11)(q31;p15), and reported somatic gene mutations include a second RUNX1 mutation,3,5,41 and mutations in ASXL1, IDH1, TET2, CEBPA and CDC25C.3,39,41-43 CDC25C mutations have not been confirmed by other studies.40 However, acquisition of clonal somatic aberrations is not pathognomonic of leukemic transformation due to the high frequency of clonal hematopoiesis reported in FPDMM. Nevertheless, detection of a new clonal aberration warrants close follow up with comprehensive assessment of PB, BM morphology and cytogenetic and molecular markers and correlation of clinical findings. In the study herein, patient C I-1 had asymptomatic mild thrombocytopenia due to a germline RUNX1 mutation. BM examination showed bilineage dysplasia and a diploid karyotype, and a DNMT3A splice site mutation was detected at 14.1% allelic frequency. In the absence of cytopenia and hematological malignancy, DNMT3A mutation may represent a coincidental clonal hematopoiesis of indeterminate potential. Repeat BM evaluation after 6 months showed persistent dysplasia with stable thrombocytopenia and no evidence of MDS; however, the follow up on this patient is short (~8 months).

The findings of recurrent BM morphologic, immunophenotypic and genetic abnormalities underscore the importance of initial baseline and serial BM evaluation in FPDMM patients. Identification of specific BM abnormalities in FPDMMHM- patients can provide measurable parameters to assess progressive changes during serial follow up for monitoring for potential development of MDS/AML, illustrated by patient II-3 in pedigree A (see Results section).

Identification of the described BM findings can also facilitate the initial recognition of FPDMM. Although most FPDMM patients are identified after the diagnosis of MDS/AML in 1 or more family members, Latger-Cannard et al. have described dysmegakaryopoiesis in a subset of FPDMMHM- patients and suggested that detection of small dysmorphic megakaryocytes in the right clinical context is a clue for early diagnosis of FPDMM.6,20 We believe that dysmegakaryopoiesis in a hypocellular BM with or without an abnormal immunophenotype of CD34-positive blasts should trigger evaluation for germline predisposition syndromes, especially in patients with long-standing thrombocytopenia and normal-sized platelets. This is particularly important in patients without a known family history.

Since the disease has a variable age of presentation and broad spectrum of clinical manifestations, a high level of suspicion provides opportunities for early detection and appropriate genetic counseling for other family members.

We also recommend NGS mutation analysis for the workup of thrombocytopenia, illustrated by patient G I-1 who had no family history. In this patient, NGS-based somatic mutation analysis revealed a RUNX1 variant suggestive of germline origin. Identification of a RUNX1 variant with a near-heterozygous or homozygous allelic frequency, more than 1 RUNX1 variant or biallelic variants, detection of a deleterious RUNX1 variant, or a variant that has been previously reported in FPDMM should prompt evaluation for germline RUNX1 mutation in the appropriate clinicopathologic setting.19,41 RUNX1 and other genes such as ETV6, ANKRD26, DDX41, CEBPA, and GATA2 should be incorpo-

haematologica | 2017; 102(10)

rated in routine NGS panels to facilitate incidental detection of mutations in these genes.39,44 Specifically, germline mutations in ETV6 and ANKRD26 genes can also cause thrombocytopenia. Moreover, multi-gene mutation profiling in FPDMM patients can identify secondary somatic events.

However, not all laboratories may have access to NGS. In this case, Sanger sequencing should be used for mutation analysis of these genes. RUNX1 mutations encompass the whole coding region; hence conventional PCR techniques are not appropriate. However, gene mutation analysis on its own is not sufficient to exclude germline predisposition.

Large deletions spanning numerous exons are frequent in FPDMM, and are often missed by clinical NGS-based somatic mutation analysis alone. This was apparent in pedigree C (Online Supplementary Figure S3). Exon-level oligo-array comparative gene hybridization/ single nucleotide polymorphism arrays or assessment of coverage using exome sequencing is essential.39 In certain cases, identification of a novel variant of uncertain significance may require in vitro functional studies for implicating a diagnosis of FPDMM due to linkage disequilibrium. Pre- and post-test genetic counseling of individuals and family members should be available.

At this time, it is not possible to predict an individual FPDMM patientâ&#x20AC;&#x2122;s risk of developing MDS/AML.3,14,16 FPDMM patients with BM morphologic and FC immunophenotypic abnormalities may have a higher risk of progression and need closer follow up. Large prospective studies are warranted to explore this issue further. Close follow up with a complete blood count (CBC) every 6 months and/or NGS-based mutation studies, if possible, is helpful for monitoring these patients, as there are no alternative criteria or biomarkers to predict the disease course at this time. We recommend an initial BM examination with comprehensive ancillary studies in all FPDMM patients to assess baseline pathologic changes and exclude occult malignancy. Ancillary testing should encompass FC, cytogenetic, and molecular analysis capable of detecting deletions, duplications and rearrangements. Following initial BM examination, patients must be closely monitored for progression to HM by regular BM examination if CBC or NGS studies show abnormalities. We also recommend NGS-based mutation profiling (which includes the RUNX1 gene) for evaluation of patients with long-standing thrombocytopenia without a clear underlying etiology.

In summary, in the study herein we systematically evaluated the BM morphologic, immunophenotypic and genetic findings in a large single institution series of FPDMM patients. Comparison of clinicopathologic and genetic features between FPDMMHM+ and FPDMMHM- patients with a median follow-up duration of over 2 years provided a set of criteria useful for establishing a diagnosis of MDS in these patients; the impact of making a diagnosis of MDS in FPDMM patients is underscored by the significant therapeutic implications including allogeneic stem cell transplantation. The role of precise diagnostic and monitoring criteria using a multimodal approach in the evaluation of patients with FPDMM cannot be overemphasized.

Funding This work was partly supported by institutional startup funds awarded to R K-S, MD Anderson Cancer Center Leukemia SPORE CA 100632 and Charif Souki Cancer Research Fund.

1669

R. Kanagal-Shamanna et al.

References 1. Nickels EM, Soodalter J, Churpek JE, Godley LA. Recognizing familial myeloid leukemia in adults. Ther Adv Hematol.

2013;4(4):254-269.

2. Churpek JE, Pyrtel K, Kanchi KL, et al.

Genomic analysis of germ line and somatic variants in familial myelodysplasia/acute myeloid leukemia. Blood. 2015;126(22): 2484-2490.

3. Babushok DV, Bessler M. Genetic predisposition syndromes: when should they be considered in the work-up of MDS? Best Pract Res Clin Haematol. 2015;28(1):55-68.

4. Arber DA, Orazi A, Hasserjian R, et al. The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia. Blood.

2016;127(20):2391-2405.

5. Schmit JM, Turner DJ, Hromas RA, et al.

Two novel RUNX1 mutations in a patient with congenital thrombocytopenia that evolved into a high grade myelodysplastic syndrome. Leuk Res Rep. 2015;4(1):24-27.

6. Latger-Cannard V, Philippe C, Bouquet A, et al. Haematological spectrum and genotype-phenotype correlations in nine unrelated families with RUNX1 mutations from the French network on inherited platelet disorders. Orphanet J Rare Dis. 2016;11:49.

7. Toya T, Yoshimi A, Morioka T, et al.

Development of hairy cell leukemia in familial platelet disorder with predisposition to acute myeloid leukemia. Platelets.

2014;25(4):300-302.

8. Sorrell A, Espenschied C, Wang W, et al.

Hereditary leukemia due to rare RUNX1c splice variant (L472X) presents with eczematous phenotype. Int J Clin Med. 2012;3(7).

9. Okada Y, Watanabe M, Nakai T, et al.

RUNX1, but not its familial platelet disorder mutants, synergistically activates PF4 gene expression in combination with ETS family proteins. J Thromb Haemost.

2013;11(9):1742-1750.

10. Liew E, Owen C. Familial myelodysplastic syndromes: a review of the literature.

Haematologica. 2011;96(10):1536-1542.

11. Kiraly PA, Kallay K, Marosvari D, et al.

[Clinical and genetic background of familial myelodysplasia and acute myeloid leukemia]. Orv Hetil. 2016;157(8):283-289.

12. Iizuka H, Kagoya Y, Kataoka K, et al.

Targeted gene correction of RUNX1 in induced pluripotent stem cells derived from familial platelet disorder with propensity to myeloid malignancy restores normal megakaryopoiesis.

Exp Hematol.

2015;43(10):849-857.

13. DiNardo CD, Bannon SA, Routbort M, et al. Evaluation of patients and families with concern for predispositions to hematologic malignancies within the Hereditary Hematologic Malignancy Clinic (HHMC).

Clin Lymphoma Myeloma Leuk. 2016; 16(7):417-428 e412.

14. Connelly JP, Kwon EM, Gao Y, et al.

Targeted correction of RUNX1 mutation in FPD patient-specific induced pluripotent stem cells rescues megakaryopoietic defects. Blood. 2014;124(12):1926-1930.

15. Luddy RE, Champion LA, Schwartz AD. A fatal myeloproliferative syndrome in a family with thrombocytopenia and platelet dysfunction. Cancer. 1978;41(5):1959-1963.

16. Owen C, Barnett M, Fitzgibbon J. Familial

1670

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

myelodysplasia and acute myeloid leukaemia: a review. Br J Haematol. 2008; 140(2):123-132.

Owen CJ, Toze CL, Koochin A, et al. Five new pedigrees with inherited RUNX1 mutations causing familial platelet disorder with propensity to myeloid malignancy.

Blood. 2008;112(12):4639-4645.

Buijs A, Poddighe P, van Wijk R, et al. A novel CBFA2 single-nucleotide mutation in familial platelet disorder with propensity to develop myeloid malignancies. Blood.

2001;98(9):2856-2858.

Churpek JE, Godley LA. How I diagnose and manage individuals at risk for inherited myeloid malignancies. Blood. 2016 Jul 28.

[Epub ahead of print] Latger-Cannard V, Philippe C, Jonveaux P, Lecompte T, Favier R. Dysmegakaryopoiesis, a clue for an early diagnosis of familial platelet disorder with propensity to acute myeloid leukemia in case of unexplained inherited thrombocytopenia associated with normal-sized platelets. J Pediatr Hematol Oncol.

2011;33(7):e264-266.

Walker LC, Stevens J, Campbell H, et al. A novel inherited mutation of the transcription factor RUNX1 causes thrombocytopenia and may predispose to acute myeloid leukaemia. Br J Haematol. 2002;117(4):878881.

Steensma DP, Bejar R, Jaiswal S, et al.

Clonal hematopoiesis of indeterminate potential and its distinction from myelodysplastic syndromes.

Blood.

2015;126(1):9-16.

Thiele J, Kvasnicka HM, Facchetti F, et al.

European consensus on grading bone marrow fibrosis and assessment of cellularity.

Haematologica. 2005;90(8):1128-1132.

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.

Ok CY, Leventaki V, Wang SA, et al.

Detection of an abnormal myeloid clone by flow cytometry in familial platelet disorder with propensity to myeloid malignancy.

Am J Clin Pathol. 2016;145(2):271-276.

Tang G, Jorgensen LJ, Zhou Y, et al. Multicolor CD34(+) progenitor-focused flow cytometric assay in evaluation of myelodysplastic syndromes in patients with post cancer therapy cytopenia. Leuk Res. 2012;36(8):974-981.

Khoury JD, Sen F, Abruzzo LV, et al.

Cytogenetic findings in blastoid mantle cell lymphoma. Hum Pathol. 2003;34(10):10221029.

Konoplev S, Medeiros LJ, Lennon PA, et al.

Therapy may unmask hypoplastic myelodysplastic syndrome that mimics aplastic anemia. Cancer. 2007;110(7):15201526.

Simons A, Shaffer LG, Hastings RJ. Cytogenetic nomenclature: changes in the ISCN 2013 compared to the 2009 edition.

Cytogenet Genome Res. 2013;141(1):1-6.

Richards CS, Bale S, Bellissimo DB, et al.

ACMG recommendations for standards for interpretation and reporting of sequence variations: Revisions 2007. Genet Med. 2008;10(4):294-300.

Kanagal-Shamanna R, Luthra R, Yin CC, et al. Myeloid neoplasms with isolated

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

isochromosome 17q demonstrate a high frequency of mutations in SETBP1, SRSF2, ASXL1 and NRAS. Oncotarget. 2016;7 (12):14251-14258.

Luthra R, Patel KP, Reddy NG, et al. Nextgeneration sequencing-based multigene mutational screening for acute myeloid leukemia using MiSeq: applicability for diagnostics and disease monitoring.

Haematologica. 2014;99(3):465-473.

Tsang HC, Bussel JB, Mathew S, et al. Bone marrow morphology and disease progression in congenital thrombocytopenia: a detailed clinicopathologic and genetic study of eight cases. Mod Pathol. 2017; 30(4):486-498.

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):27082718.

Song WJ, Sullivan MG, Legare RD, et al.

Haploinsufficiency of CBFA2 causes familial thrombocytopenia with propensity to develop acute myelogenous leukaemia. Nat Genet. 1999;23(2):166-175.

Ichikawa M, Asai T, Saito T, et al. AML-1 is required for megakaryocytic maturation and lymphocytic differentiation, but not for maintenance of hematopoietic stem cells in adult hematopoiesis. Nat Med. 2004;10(3):299-304.

Park IJ, Park JE, Kim HJ, et al. Acute myeloid leukemia with t (16; 21)(q24; q22) and eosinophilia: case report and review of the literature. Cancer genetics and cytogenetics. 2010;196(1):105-108.

Ripperger T, Steinemann D, Gohring G, et al. A novel pedigree with heterozygous germline RUNX1 mutation causing familial MDS-related AML: can these families serve as a multistep model for leukemic transformation? Leukemia. 2009;23(7):1364-1366.

Haslam K, Langabeer SE, Hayat A, Conneally E, Vandenberghe E. Targeted next-generation sequencing of familial platelet disorder with predisposition to acute myeloid leukaemia. Br J Haematol.

2016;175(1):161-163.

Yoshimi A, Toya T, Nannya Y, et al.

Spectrum of clinical and genetic features of patients with inherited platelet disorder with suspected predisposition to hematological malignancies: a nationwide survey in Japan. Ann Oncol. 2016;27(5):887-895.

Antony-Debre I, Duployez N, Bucci M, et al. Somatic mutations associated with leukemic progression of familial platelet disorder with predisposition to acute myeloid leukemia. Leukemia. 2016; 30(4):999-1002.

Preudhomme C, Renneville A, Bourdon V, et al. High frequency of RUNX1 biallelic alteration in acute myeloid leukemia secondary to familial platelet disorder. Blood.

2009;113(22):5583-5587.

Sakurai M, Kasahara H, Yoshida K, et al.

Genetic basis of myeloid transformation in familial platelet disorder/acute myeloid leukemia patients with haploinsufficient RUNX1 allele. Blood Cancer J. 2016;6:e392.

Kanagal-Shamanna R, Singh RR, Routbort MJ, et al. Principles of analytical validation of next-generation sequencing based mutational analysis for hematologic neoplasms in a CLIA-certified laboratory. Expert Rev Mol Diagn. 2016;16(4):461-472.

haematologica | 2017; 102(10)

ARTICLE

Bone Marrow Failure

Endotoxemia shifts neutrophils with TIMP-free gelatinase B/MMP-9 from bone marrow to the periphery and induces systematic upregulation of TIMP-1

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Jennifer Vandooren, Wannes Swinnen, Estefania Ugarte-Berzal, Lise Boon, Daphne Dorst, Erik Martens and Ghislain Opdenakker Laboratory of Immunobiology, Department of Microbiology and Immunology, Rega Institute for Medical Research, University of Leuven, KU Leuven, Belgium.

Haematologica 2017 Volume 102(10):1671-1682

ABSTRACT

L

ipopolysaccharides or endotoxins elicit an excessive host inflammatory response and lead to life-threatening conditions such as endotoxemia and septic shock. Lipopolysaccharides trigger mobilization and stimulation of leukocytes and exaggerated production of pro-inflammatory molecules including cytokines and proteolytic enzymes. Matrix metalloproteinase-9 (MMP-9) or gelatinase B, a protease stored in the tertiary granules of polymorphonuclear leukocytes, has been implicated in such inflammatory reactions. Moreover, several studies even pinpointed MMP-9 as a potential target molecule to counter excessive inflammation in endotoxemia. Whereas the early effect of lipopolysaccharide-induced inflammation in vivo on the expression of MMP-9 in various peripheral organs has been described, the effects on the bone marrow and during late stage endotoxemia remain elusive. We demonstrate that TIMP-free MMP-9 is a major factor in bone marrow physiology and pathology. By using a mouse model for late-stage endotoxemia, we show that lipopolysaccharides elicited a depletion of neutrophil MMP-9 in the bone marrow and a shift of MMP-9 and MMP-9containing cells towards peripheral organs, a pattern which was primarily associated with a relocation of CD11bhighGr-1high cells. In contrast, analysis of the tissue inhibitors of metalloproteinases was in line with a natural, systematic upregulation of TIMP-1, the main tissue inhibitor of TIMP-free MMP-9, and a general shift toward control of matrix metalloproteinase activity by tissue inhibitors of metalloproteinases.

Correspondence: ghislain.opdenakker@kuleuven.be

Received: March 15, 2017.

Accepted: July 27, 2017.

Pre-published: August 3, 2017. doi:10.3324/haematol.2017.168799

Introduction The regulation of leukocytosis is a balance between the exit of leukocytes into the periphery and the production of these cells in the bone marrow. Within this compartment systemic cytokine levels orchestrate a regulated response to peripheral signals, leading to leukocyte expansion through the action of colony-stimulating factors. At the molecular level this is translated into oligosaccharide-lectin interactions, expression of cell adhesion molecules, balances between proteases and inhibitors, and chemokine-mediated leukocyte recruitment to specific body compartments.1 Interference at any of these levels could represent a possible treatment for acute inflammatory reactions such as shock syndromes.2 An often explored strategy is interference with balances between proteases and inhibitors, for example, matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs).3,4 MMPs are named after their ability to degrade extracellular matrix proteins, namely gelatinases (MMP-2/9), collagenases (MMP-1/8/13), stromelysins (MMP-3/10/11) and others, but are also able to degrade cell-surface and intracellular molecules.5 One particular MMP implicated in cell mobility in and out of the bone marrow is MMP-9 or gelatinase B.5 MMP-9 is primarily produced by neutrophils, which pre-store large quantities of proMMP-9 (zymogen) in their tertihaematologica | 2017; 102(10)

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1671 Š2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1671

J. Vandooren et al.

ary granules (also called gelatinase granules) for swift release upon encounter with an inflammatory stimulus.6 When released in the bone marrow, activated MMP-9 processes membrane-bound Kit-ligand into soluble Kit-ligand and thereby keeps progenitor cells locally and initiates recruitment of “escaped” hematopoietic stem cells to their proliferative niche.7 In the periphery, MMP-9 contributes to the recruitment of bone marrow-derived neutrophils to the site of an inflammatory stimulus.8 It is not, therefore, surprising that studies involving mouse models for sepsis show beneficial outcomes for MMP-9-deficient mice and/or mice treated with MMP inhibitors.9-12 In the body, MMP proteolytic activity is kept in check by natural tissue inhibitors of metalloproteinases (TIMPs).

While all four TIMPs (TIMP-1 to -4) have inhibitory activity on all MMPs, TIMP-1 preferentially binds to MMP-9 and TIMP-2 to MMP-2.13 A unique feature of TIMP-2 is that it can initiate the formation of a MT1-MMP/TIMP2/proMMP-2 cell surface complex that aids the conversion of the proMMP-2 zymogen into activated MMP-2.14,15 Similar to MMP-9, TIMP-1 is of importance in bone marrow physiology, in particular in the maintenance of hematopoietic stem cell quiescence16 and leukocytosis.17,18 Lipopolysaccharides (LPS) or endotoxins are glycolipids present in the outer membrane of Gram-negative bacteria, and are released upon lysis of bacteria. In the body, LPS are detected as an alarm signal by leukocytes, endothelial cells and parenchymal cells through their LPS-receptor complex composed of myeloid differentiation factor 2 (MD2) and toll-like receptor 4 (TLR4), a process which is facilitated by LPS-binding protein (LBP) and CD14.19 This interaction triggers cytokine production, leukocyte activation and inflammation,20 and excessive stimulation can lead to simultaneous activation of multiple parallel cascades that lead to adult respiratory distress syndrome and shock.21 In this study we used injection of LPS as an animal model to study the mechanisms behind acute inflammation (endotoxemia) caused by Gram-negative bacteria.

Whereas the early effects of LPS in vivo on the expression of MMPs in various peripheral organs are known (Online Supplementary Table S1), effects on bone marrow and during late stages remain elusive. This is remarkable, given the importance of MMP-9 in leukocytosis and hematopoietic recovery.7,8 The aim of this study was to evaluate the effects of endotoxemia on bone marrow gelatinases (MMP-2 and MMP-9) and, in parallel, the effect on their natural inhibitors. With the use of real-time polymerase chain reaction (qPCR), gelatin zymography and flow cytometry analysis, we demonstrated that endotoxemia results in depletion of MMP-9 from the bone marrow. In parallel, this MMP-9 shift is complemented by systematic upregulation of TIMP-1. Finally, immunohistochemical analysis confirmed the migration of MMP-9-containing cells from the bone marrow to blood and peripheral organs.

hours after injection, mice were sacrificed and organs were collected. Bone marrow cells were collected from femora by flushing the medullary cavity with PBS. Blood samples were collected by cardiac puncture with a heparin-coated needle and syringe, and immediately processed by centrifugation (2000 g for 10 min at 4°C). The supernatant (plasma) was collected and stored for protein analysis and the cell pellet was immediately processed for downstream analysis. Tissues were homogenized and proteins extracted with a Precellys lysing kit (Bertin Technologies), as described in the Online Supplementary Methods. All procedures were conducted in accordance with protocols approved by the local ethics committee (project number P201/2012, KU Leuven, Belgium).

Gelatin zymography Protein levels of proMMP-9, MMP-9, multimeric MMP-9, proMMP-2 and MMP-2 were determined by gelatin zymography on affinity-purified samples, as previously described.23 and detailed in the Online Supplementary Methods. Prior to zymography analysis, the total protein content was determined using a standard Bradford assay (Bio-Rad).

RNA expression analysis RNA was extracted and equal amounts of RNA were converted to cDNA. qPCR was performed using TaqMan® fast universal PCR master mix (Applied Biosystems), PrimeTime® predesigned qPCR assays (IDT) and a 7500 Fast Real-Time PCR System (Applied Biosystems). Details, including primer specifications, are provided in the Online Supplementary Methods. The housekeeping gene Tbp was used as a calibrator for the relative quantification of gene expression. Normalization for Tbp and calculation of the relative expression was performed using the ΔΔCT method.24

Immunohistochemistry Femora were placed in 6% formaldehyde (1 day), decalcified in 7% formic acid (2 days) and embedded in paraffin. Paraffinembedded bones were sliced into 5 µm sections and dried overnight at 50°C. For immunohistochemical staining the EnVisionTM FLEX kit (DAKO) was used. Goat anti-mouse MMP9 (R&D Systems) was used as the primary antibody. A detailed protocol is provided in the Online Supplementary Methods.

Flow cytometry analysis Spleens and bone marrow were passed through cell strainers to obtain single cell suspensions. Red blood cells of spleen, bone marrow and blood cell suspensions were lysed with 0.83% NH4Cl. Cells were incubated with Fc-receptor-blocking antibodies anti-CD16/anti-CD32 (BD Biosciences Pharmingen, San Diego, CA, USA), Zombie aqua BV510 (dead cell staining) (BioLegend, San Diego, CA, USA) and stained with anti-Gr1, anti-F4/80, anti-CD11b, anti-CD3 or anti-CD19 (eBioscience, San Diego, CA, USA). Cells were fixed and analyzed with a FACS Fortessa flow cytometer. Data were processed with the FlowJo software (Becton Dickinson Labware, Franklin Lakes, NJ, USA). A detailed protocol is provided in the Online Supplementary Methods.

Methods

Enzyme-linked immunosorbent assays and gelatin degradation assays

Mouse model of endotoxemia and sample collection

TIMP-1 and TIMP-2 concentrations were determined using mouse TIMP-1 (DY980) and TIMP-2 (DY6304-05) DuoSet enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems). To determine the gelatinolytic activity we used a previously described gelatin degradation assay.25 A detailed protocol is provided in the Online Supplementary Methods.

Endotoxemia was induced in female 8-week old C57BL/6 mice by intraperitoneal (i.p.) injection of LPS (E. coli 0111:B4, Sigma Aldrich, L4391) at a dose of 10 mg/kg.22 Control mice were injected with an equal volume of vehicle (pyrogen-free phosphatebuffered saline, PBS; LPS level below 12.5 pg/mL). Twenty-four 1672

haematologica | 2017; 102(10)

Gelatinases and their natural inhibitors in endotoxemia

Statistical analysis Data were analyzed using GraphPad Prism 7 software and are expressed as mean ± standard error of mean (SEM). Differences were determined using a Mann-Whitney test. Data were collected and confirmed over a course of five independent experiments with experimental groups ranging from two to six animals.

Results The bone marrow compartment contains considerable levels of MMP-9 RNA and protein Publicly available mRNA expression data26 of human and mouse tissues reveal that gene expression levels of MMP-9 are highest in bone and bone marrow of humans and mice (Figure 1A). We, therefore, first evaluated the baseline levels of MMP-9 protein in bone marrow, spleen, lungs, liver and plasma by using gelatin zymography. The main source of MMP-9 (here a combination of high-molecular weight proMMP-9 and MMP-9) was indeed the bone marrow (Figure 1B and Online Supplementary Figure S1), with bone marrow MMP-9 representing over 0.1% of the total bone marrow protein. Mean MMP-9 protein levels in bone marrow, spleen, lungs, liver and plasma were respectively 1131±212.8 ng/mg protein, 168.4±29.8 ng/mg protein, 49.1±4.9 ng/mg protein, 2.12±0.26 ng/mg protein and 0.6±0.09 ng/mg protein (mean±SEM, n = 5-10).

MMP-2 and MMP-9) or chemical modification of the propeptide cysteine thereby disturbing the pro-peptide/active site interaction. These processes are fine-tuned and require a local environment favoring MMP activation.5,28,29 A second level of complexity involves the formation of highmolecular weight disulfide linked MMP-9 multimers, including MMP-9 trimers.30 Detailed analysis of these forms provides better insights into the dynamics of protease activity and can be performed using gelatin zymography analysis (Figure 3A).23 In Table 1, protein data for

A

Endotoxemia results in a shift of the Mmp2/9 and Timp1/2/3 expression pattern Acute inflammation was induced by i.p. injection of endotoxin. To follow hematologic changes during late stages, samples were collected from bone marrow, blood and spleen 24 h after LPS/PBS injection. Liver and lungs, two organs prone to dysfunction during endotoxemia and sepsis syndromes were also collected.27 Analysis of RNA expression of Mmp9, Mmp2, Timp1, Timp2, Timp3 and Timp4 in these tissues revealed a shift in the expression pattern of both gelatinases and TIMPs (Figure 2). In the bone marrow, the basal expression level of Mmp9 RNA was reduced by endotoxemia, while Mmp9 expression in liver and lungs increased significantly. In blood cells a trend toward increased Mmp9 expression was seen, while Mmp9 mRNA levels in the spleen remained unaltered.

Interestingly, although the effect of endotoxemia on Mmp2 mRNA was less pronounced, it generally was opposing the expression pattern of Mmp9. Steady-state Mmp2 mRNA levels were increased in the bone marrow upon induction of endotoxemia, while expression in the spleen and lungs was decreased. While basal Timp1 expression was low, endotoxin triggered a considerable systematic induction.

For Timp2 the opposite effect occurred, namely, a general downregulation. A significant alteration in Timp3 expression was found in bone marrow (upregulation) and liver (downregulation) while Timp4 expression remained low and unaltered. All four TIMPs are thus differently regulated in response to LPS.

B

Endotoxemia shifts MMP-9 from bone marrow to the circulation and peripheral organs Biological samples contain complex mixtures of forms of MMP-2 and MMP-9 proteins. A first level of complexity lies in the fact that both proteases are produced as inactive pro-enzymes, proMMP-2 and proMMP-9. Activation of these pro-forms into active enzymes requires proteolytic removal of the pro-peptide by other proteases (to form haematologica | 2017; 102(10)

Figure 1. Organ-specific gene and protein expression of MMP-9. (A) Gene expression of MMP-9 in human and mouse tissue based on publicly available array data [human GeneAtlas U133A, gcrma and mouse GeneAtlas GNF1M, gcrma dataset26]. Array data from different locations in one organ were pooled in order to visualize the expression per organ (n = 2-28). (B) MMP-9 protein levels in C57BL/6 mice as determined by gelatin zymography (n = 5-10).

1673

J. Vandooren et al.

Table 1. Levels of MMP-2 and MMP-9 forms detected in bone marrow, spleen, liver, lungs and plasma.

Bone marrow PBS LPS P Spleen PBS LPS P Lungs PBS LPS P Liver PBS LPS P Plasma PBS LPS P

MMP-9 multimers

proMMP-9

activated MMP-9

proMMP-2

MMP-2

648±285 56.6±20.0 ****

578± 378 77.8± 39.4 ****

ND ND ND

ND ND ND

ND ND ND

54±9.1 42.4±6.6 *

59.3± 21.7 65.0± 21.0 ns

4.28±1.97 5.97±5.52 ns

8.44±2.78 6.63±2.44 ns

ND ND ND

26.4±7.4 40.7±17.8 ns

20.9±4.3 43.7±13.0 ****

9.06±1.10 15.99±4.65 **

16.44±2.51 15.11±2.87 ns

5.64±2.70 9.53±2.64 ns

1.1±0.24 2.3±0.4 ****

1.03±0.39 3.09±0.84 ****

0.21±0.18 0.59±0.19 ***

1.53±0.46 1.47±0.65 ns

0.27±0.02 0.30±0.04 ns

0.56±0.29 0.9±0.06 ns

0.49±0.10 1.18±0.48 *

ND ND ND

3.22±0.48 4.68±0.62 ns

ND ND ND

Data are shown as mean±SD and are expressed as nanograms protease per milligram of tissue protein (ng/mg) as determined by gelatin zymography. Control mice (administered PBS) were statistically compared with mice given i.p. LPS injection. ND; no signal detected, ns; non-significant, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001 as determined by the Mann-Whitney test. Data were pooled from three independent experiments; n = 4-10.

MMP-2 and MMP-9 forms are provided for bone marrow, liver, lungs, spleen and plasma from healthy mice and mice with endotoxemia. As expected, basal levels of all MMP-9 forms were highest in the bone marrow followed by, respectively, spleen, lungs, liver and plasma. Highest levels of (pro)MMP-2 were detected in lungs, followed by spleen and liver. MMP-2 was not detected in bone marrow, due to an overflow effect of the high quantities of MMP-9.

Similarly, proteolytically activated MMP-9 could not be detected in bone marrow due to the overflow effect of proMMP-9. While proteolytic activation of MMP-9 probably occurs in this compartment, we could not detect this due to the high levels of proMMP-9 within this compartment. In plasma, MMP-2 levels were below the detection limit but proMMP-2 was detected. Upon induction of endotoxemia, drastic changes occurred in MMP-9 protein levels. The most severe changes were observed in the bone marrow, where protein levels of proMMP-9 and multimeric MMP-9 dropped by approximately 90% or, respectively, around 500 and 600 ng/mL (Table 1 and Figure 3B). This major decrease of bone marrow (pro)MMP-9 coincided with an increase of (pro)MMP-9 in liver and lungs, while less pronounced effects were observed in the spleen.

Interestingly, mice with endotoxemia had a significant decrease in the percentage of multimeric MMP-9 in spleen, lungs and liver (Figure 3C). Indeed, downregulation of protein disulfide isomerase in sepsis was previously shown.31 While Mmp2 RNA expression analysis suggested an opposing effect to MMP-9, this could not be confirmed at the protein level, endorsing the existing notion that during inflammation MMP-2 is less prone to major stimulation, in comparison to the fine-tuned and controlled regulation of MMP-9.32 Nevertheless, one should take into account that non-secreted MMP-2 variants exist, which might shift the balance between intracellular and secreted MMP-2.33 Interestingly, although the total (pro)MMP-2 content remained similar, the percentage of proteolytically activat1674

ed MMP-2 was increased in the lungs of endotoxemic mice (Figure 3B,E). However, one needs to consider the possibility that proMMP-2 and proMMP-9 might also be activated by other non-proteolytic mechanisms. Many non-proteolytic activation mechanisms of MMPs have been discovered, for example, allosteric activation of proMMPs by substrate binding or ligand binding34,35 and chemical modification of the propeptide cysteine by reactive oxygen/nitrogen species such as peroxynitrite.28,29 To evaluate the possibility that LPS induce different MMP-9 modifications and charge variants, we evaluated the samples by two-dimensional zymography.36 MMP-9 appeared as a wide band of different charge variants, spanning almost the full length of the isoelectric point (pI) gradient (Online Supplementary Figure S2), likely caused by different posttranslational modifications. However, no major changes in this pattern were observed in the LPS-stimulated samples.

With the limitations of the resolution and sensitivity of this technique, we can conclude that no major charge modifications occurred upon induction of endotoxemia.

Systematic induction of TIMP-1 and reduction of TIMP-2 MMP activity is based on a balance between MMPs and their natural inhibitors. We, therefore, analyzed TIMP-1 and TIMP-2 protein levels. Basal levels of TIMP-1 and TIMP-2 were highest in the lungs (Table 2). Interestingly, basal levels of TIMP-1 in bone marrow were low while TIMP-2 was well represented. Next, LPS challenge resulted in a systematic change in TIMP-1 protein levels (Figure 4A). In agreement with RNA expression data, TIMP-2 was significantly decreased by LPS in bone marrow and spleen.

In sharp contrast, endotoxemia induced the production of TIMP-1 in all organs, an effect which was most pronounced in the bone marrow compartment. Taken together, with a decrease in MMP-9 of approximately 90% and an almost 8-fold increase of TIMP-1, the effect of endotoxhaematologica | 2017; 102(10)

Gelatinases and their natural inhibitors in endotoxemia

Figure 2. The effects of endotoxemia on Mmp9, Mmp2, Timp1, Timp2, Timp3 and Timp4 RNA expression. Relative RNA expression in bone marrow, spleen, lungs, liver and blood cells of LPS-treated mice in comparison with the control group given PBS. Data were normalized for the housekeeping gene Tbp. Histograms represent group medians and individual data points from single animals are shown by dots. *P<0.05, **P<0.01, as determined by the Mann-Whitney test (n = 3-6).

haematologica | 2017; 102(10)

1675

J. Vandooren et al.

A

B

C

D

E

Figure 3. Gelatin zymography analysis of the protein levels of MMP-9 and MMP-2 forms. (A) Representative gelatin zymography gels of bone marrow, spleen, lungs, liver and plasma of mice treated with LPS versus control mice (PBS). Each lane represents the analysis of the sample from a single mouse (n = 3-4). Each sample was spiked with an internal processing and loading control (MMP-9ΔHemOG – MMP-9 form lacking the C-terminal hemopexin and O-glycosylated domain) and each gel has three lanes of recombinant MMP-9 standard protein (RS), including multimeric, monomeric and MMP-9 ΔHemOG proteins, to serve as a molecular weight marker and standard (10, 5 and 3 pg). The loading quantity of the samples corresponded to, respectively, 0.4 mg (bone marrow), 2 mg (spleen), 4 mg (lungs), 60 µg (liver) and 30 mg (plasma) of total protein. (B) Detailed analysis of protein levels of MMP-9 multimers, proMMP-9 monomers, activated MMP-9 monomers, proMMP2 and activated MMP-2. Protein quantity was expressed as nanograms of MMP in one milligram of total protein. (C) Percentages of multimeric MMP-9 out of total MMP-9. (D) Percentages of proteolytically activated MMP-9 out of total MMP-9. (E) Percentages of proteolytically activated MMP-2. Histograms represent group medians and individual data points are shown by dots, each representing data from a single mouse. *P<0.05, **P<0.01, ***P<0.001, ***P<0.0001, as determined by the Mann-Whitney test.

1676

haematologica | 2017; 102(10)

Gelatinases and their natural inhibitors in endotoxemia

Table 2. Levels of TIMP-1 and TIMP-2 detected in bone marrow, spleen, liver and lungs.

Bone marrow PBS LPS P

TIMP-1

TIMP-2

MMP-9/TIMP-1

MMP-2/TIMP-2

0.003±0.006 0.23±0.07 **

2.48±0.49 0.41±0.07 **

102167 146

ND ND

Spleen PBS LPS P Lungs PBS LPS P

0.15±0.06 0.98±0.41 **

1.61±0.62 0.98±0.23 *

196 28.9

1.53 1.97

4.12±1.09 17.53±4.56 **

8.84±1.29 8.77±2.02 ns

3.42 1.43

0.73 0.82

Liver PBS LPS P

1.32±0.53 4.20±2.66 *

0.41±0.41 0.83±0.82 ns

0.44 0.36

1.28 0.62

Data are shown as mean±SD and are expressed as nanogram of TIMP per milligram of tissue protein (ng/mg) as determined by ELISA. Control mice (PBS) were statistically compared with mice given i.p. LPS injection. The two last columns represent the molar ratios of MMP-9/TIMP-1 and MMP-2/TIMP-2, calculated based on data in Table 1 and Table 2, and taking into account the molecular weight of each of the proteins.Values higher than one represent an excess of MMPs compared to TIMPs. ND; no signal detected, ns; nonsignificant, *P<0.05, **P<0.01 as determined by the Mann-Whitney test; n = 5-6.

emia on the protease/inhibitor balance in the bone marrow is substantial. Molar ratios of MMP-9/TIMP-1 indeed suggest a basal proteolysis-favoring environment in the bone marrow compartment, which decreases upon induction of endotoxemia (Table 2). It does, however, need to be emphasized that basal gelatinase-B in the bone marrow is primarily in the proMMP-9 form. To confirm this effect, net gelatinolytic activity in the bone marrow was measured using a gelatin degradation assay. As expected, bone marrow samples contained net-proteolytic activity which was significantly reduced upon treatment with LPS (Figure 4B). Moreover, by adding SB-3CT (an inhibitor of MMP2/MMP-9) and ElaV (an elastase inhibitor) we found that bone marrow proteolysis, here represented by gelatinolysis, is a combination of MMP-9 and elastase activity. MMP9 proteolysis accounted for approximately 30% of gelatin proteolysis while elastase accounted for the remaining 70%. An interesting observation is that organs with high levels of activated MMP-2/-9 (e.g. lungs and liver) also contained higher levels of TIMPs. Although being paradoxical, this observation might be explained by TIMP-assisted proteolytic activation, as previously shown for MMP-2/TIMP2.14,15

Changes in MMP-9 levels correlate with neutrophil migration patterns Neutrophils are the main producers of MMP-9 and they are unique because they do not produce TIMP-1.37,38 This concept was elaborated in a pioneering study on angiogenesis in tumor biology,38 but seemingly also has relevance in bone marrow physiology. We, therefore, hypothesized that the TIMP-1/MMP-9 shift during endotoxemia might be due to neutrophil mobilization into the circulation and into peripheral organs. To investigate this, we performed flow-cytometry analysis of bone marrow, spleen and blood cells to study the migration pattern of T cells (CD3+), B cells (CD19+), neutrophils (CD11b+Gr1+) and macrophages (CD11b+F4/80+) in animals with and without haematologica | 2017; 102(10)

LPS injection (Figure 5). Indeed, in the bone marrow a significant decrease (from 52% to 34%) in the CD11b+Gr-1+ population of neutrophils was seen with LPS (Figure 5A,B).

In addition, a shift of this population occurred towards the peripheral blood circulation, from 7% to 34% CD11b+Gr1+ cells. Interestingly, endotoxemia also caused a reduction in splenic macrophages (CD11b+F4/80+ cells: from 2% to 1%) and a decrease in circulating B cells (CD19+ cells: from 36% to 15%). In addition, we evaluated the RNA expression of myeloperoxidase (MPO) and neutrophil elastase (ELANE), two molecules most abundantly expressed by neutrophils and hence functioning as neutrophil markers, in bone marrow, spleen, blood, liver and lungs. Overall, the relative RNA expression of the neutrophil markers Mpo and Elane followed a similar pattern as that for MMP-9, except in the spleen, where a significant decrease was seen in both Mpo and Elane expression upon induction of endotoxemia.

Immunohistochemical analysis of MMP-9 Next, we examined the tissue location of MMP-9 by immunohistochemistry. In the bone marrow, total staining for MMP-9 was markedly reduced in the animals administered LPS (Figure 6A). In addition, the immunoreactive staining was clearly confined to single cells which appeared as polymorphonuclear cells. In control mice (given PBS) these cells were diffusely found across the bone marrow while in endotoxemic conditions more cells were associated with the vasculature. In lungs (Figure 6B) and liver (Figure 6C) the opposite phenomenon occurred.

While the baseline level of MMP-9-immunostained cells was low and predominantly restricted to blood vessels, LPS injection caused MMP-9-positive cells to migrate from blood vessels into lung alveoli, bronchioles and liver parenchyma. Again, staining was associated with polymophonuclear cells. Spleen immunohistochemistry revealed a similar staining pattern for both the PBS and LPS conditions (Figure 6D). Generally, MMP-9-positive cells 1677

J. Vandooren et al.

were located in red pulp, while no staining was seen in white pulp. In conclusion, immunohistochemical analysis of bone marrow, spleen, liver and lung sections further reinforced our data, showing decreased staining for MMP9 in bone marrow, increased staining in liver and lungs, and no changes in the MMP-9 staining in spleen upon systemic in vivo challenge with LPS.

Discussion Infection with Gram-negative bacteria accounts for about 60% of the bacterial infections causing sepsis, a disease associated with considerable lethality, even with the most sophisticated medical care.21 Previous studies suggest that MMP inhibition may become a treatment for endotoxin shock, although more in-depth preclinical work is necessary. In our model we used LPS, a Gram-negative bacterial

component, to trigger endotoxemia and to study the distribution patterns of the two key gelatinases (MMP-9 and MMP-2) and their inhibitors (TIMPs). Since the time-dependent increase of MMP-9 during early pathology (1-12 h after induction of the syndrome) is well documented9,39,40 (Online Supplementary Table S1), our study was focused at a later stage of endotoxemia (24 h) and involved an in-depth analysis thereof. Whereas the importance of MMP-9 in the development of acute inflammation, such as sepsis syndrome, is evident from the beneficial effects observed with the use of MMP inhibitors,9-12 little is known about bone marrow gelatinases (MMP-9 and MMP-2) or the effects on the balance between MMPs and their natural inhibitors. Here we provide several new insights into this topic (Figure 7).

A first striking observation is the finding that MMP-9 accounts for more than 0.1% of total bone marrow protein, making MMP-9 an important factor in bone marrow

A

B

Figure 4. Systematic induction of TIMP-1 and reduced bone marrow MMP gelatinolytic activity. (A) TIMP-1 and TIMP-2 protein content in bone marrow, spleen, lungs and liver of mice injected i.p. with LPS or control mice (PBS injection) as determined by ELISA. Protein levels detected by ELISA were corrected for the total protein concentration and are presented as nanograms of TIMP in one milligram of total protein. (B) Degradation of fluorogenic gelatin by gelatinases present in bone marrow samples from mice treated with LPS and control mice, in the presence or absence of an inhibitor of MMP-2 and MMP-9 (SB-3CT, 10 mM) or elastase inhibitor (ElaV, 10 mM). The velocity of the gelatin degradation reaction was expressed as fluorescence units (FU) per minute and is indicative of the net proteolytic activity present in the samples. Inhibition percentages are shown. *P<0.05, **P<0.01, as determined by the Mann-Whitney test.

1678

haematologica | 2017; 102(10)

Gelatinases and their natural inhibitors in endotoxemia

physiology and pathology, for example in hematopoiesis.

In addition, we show that during endotoxemia, the bone marrow is the major source of MMP-9 and that this MMP9 is associated with polymorphonuclear cells. A limitation of the present study is that the analysis was performed at the single time-point of 24 h. Previous studies have shown that a peak in plasma MMP-9 occurs as early as 1-12 h after induction of sepsis39,40 and that MMP-9 is predominantly associated with neutrophils6 and late-stage maturing neutrophils such as band cells and segmented cells, present in the bone marrow.32,41 In our model of late endotoxemia (24 h), a significant shift in MMP-9 was still evident. In particular, we observed depletion of bone marrow MMP-9 and significant increases in MMP-9 in lungs and liver, while the spleen remained unaffected. This is in line with previous

studies showing that patients suffering from sepsis have increased plasma levels of MMP-942 and that lungs and liver are indeed two highly affected organs during endotoxemia and sepsis syndromes.27 In parallel, we found that LPS induced mobilization of MMP-9-containing cells into the bloodstream. These cells predominantly migrated into lungs and liver, where clear infiltrations of MMP-9-positive cells were observed. We further confirmed this shift by showing a similar pattern for Gr-1high cells, which have been shown to contain large amounts of intracellular MMP-9.43 MMPs are indeed known to aid cell migration because of their ability to degrade extracellular matrix molecules.5,8 This effect relies on direct proteolytic activity which is the result of a fine-tuned balance of protease levels, protease activation (the conversion of proMMP-9 into active MMP-

A

C

B

Figure 5. MMP-9 levels relate with influx or efflux of a population of Gr-1high neutrophils. (A) Bone marrow, spleen and blood neutrophil content of mice not given an LPS injection (PBS, top) and given an LPS injection (bottom). Flow cytometry plots represent the distribution of live cells based on surface Gr-1 and CD11b staining. (B) Analysis of the T-cell (CD3+), B-cell (CD19+), neutrophil (CD11b+Gr1+) and macrophage (CD11b+F4/80+) populations in bone marrow, spleen and blood as determined by flow cytometry. (C) Relative MPO and Elane RNA expression in bone marrow, spleen, blood cells, liver and lungs of mice treated with PBS (control) or LPS. RNA expression data were normalized for the housekeeping gene Tbp. *P<0.05, **P<0.01, as determined by the Mann-Whitney test.

haematologica | 2017; 102(10)

1679

J. Vandooren et al.

A

B

C

D

Figure 6. Immunohistochemical analysis of MMP-9 in bone marrow, lungs, liver and spleen. (A) MMP-9 immunohistochemical analysis of bone marrow from mice with i.p. LPS injection and control mice (PBS). Arrowheads indicate immunoreactivity for MMP-9 in leukocytes associated with a blood vessel wall. (B) Lung MMP-9 immunohistochemical analysis. Increased infiltrations of MMP-9-positive cells from blood vessels (V) to surrounding alveoli and bronchioles (B) in lungs from mice with endotoxemia (LPS), indicated by the arrowheads. The insert shows the typical polymorphonuclear nature of the infiltrating MMP-9-containing cells. (C) Liver MMP-9 histology also shows clear infiltration of MMP-9-positive cells from blood vessels into the surrounding tissue. Infiltrating cells are indicated by arrowheads and the magnification shows the typical polymorphonuclear nature of these cells. (D) Immunohistochemical analysis of the spleens of mice subjected to i.p. LPS or control mice (PBS). MMP-9-positive cells were predominantly associated with the red pulp (RP) and not with white pulp (WP).

9 by proteolysis of the prodomain or chemical modification of the pro-peptide cysteine) and the presence of inhibitors (e.g. TIMP-1). An important new finding of our work is that, although MMP-9 increases systemically after LPS challenge, a natural anti-proteolytic response occurs by systematically increasing TIMP-1. This is evident from both tissue MMP-9/TIMP-1 ratios and proteolytic activity tests. In addition, the relative expression of Timp1 also considerably increased upon LPS challenge in almost all organs, including the bone marrow. These findings allowed us to address two questions: (i) whether inhibition of proteolysis by MMP-9 still has merit once acute inflammation has progressed (latestage) and (ii) whether the contribution of TIMP-1 is disease-promoting or disease-limiting. Most animal studies, showing beneficial effects for MMP-9 inhibitors, were performed with inhibitor injections immediately after the induction of endotoxemia10-12 or in MMP-9 knock-out animals.8,9 In addition, higher serum levels of TIMP-1 have been found in non-survivors compared to survivors of 1680

severe sepsis.44 This confirms that the fine-tuned regulation of MMPs, including their balance with TIMPs, is crucial for tissue homeostasis.18 A good approach to the treatment of acute inflammations might, therefore, rely on carefully restoring this balance depending on the proteolytic state.

So far, the protease/anti-protease balance has been predominantly investigated in lungs and is key to normal lung physiology. The most studied example in lungs is the elastase/anti-elastase balance. This is exemplified by mouse models in which both induction of elastase and mutations of antitrypsin (elastase inhibitor) lead to emphysema.45 We here show that endotoxemia also drastically affects the lungs by inducing the accumulation of MMP-9-positive cells. Nevertheless, this effect is counterbalanced by induction of TIMP-1. Indeed, it was previously suggested that the pulmonary accumulation of neutrophils in sepsis is due to shedding of platelet-derived surface-expressed CD40L by MMP-9.46 Although TIMP-1 increases in bone marrow, most MMPhaematologica | 2017; 102(10)

Gelatinases and their natural inhibitors in endotoxemia

Figure 7. Overview of the distribution pattern of MMP-9/TIMP-1 and MMP2/TIMP-2 balances and MMP-9-positive cells upon challenge with lipopolysaccharide. Induction of endotoxemia results in the migration of MMP-9-positive polymorphonuclear cells out of the bone marrow to accumulate primarily in lungs and liver. This release of MMP-9 is meanwhile counteracted by a systematic upregulation of TIMP-1. In contrast, MMP-2 protein levels remain similar while TIMP-2 is generally reduced. NE; neutrophil elastase.

9 (>99%) remains TIMP-free and thus potentially catalytically active. Indeed, neutrophils are unique in the fact that they do not secrete TIMP-1.37,38 Moreover, this TIMP-free MMP-9 was shown to have angiogenic capacity,38 in particular when neutrophils infiltrate peripheral tissue or tumor microenvironments.47 In general, we show that bone marrow functioning relies on a combination of proteolysis by MMP-9 and neutrophil elastase and that both activities are reduced during late-stage endotoxemia. Previously it was shown that neutrophil elastase can compensate for MMP-9 deficiency in a model of leukocyte infiltration in experimental peritonitis.48 Neutrophil elastase is a serine protease found in the azurophilic granules of neutrophils. In contrast to MMP-9, neutrophil elastase knock-out mice have impaired host defense against Gram-negative bacterial sepsis.49 Neutrophil elastase can, therefore, also be considered as an important factor in endotoxin shock and requires further investigation.

To date, the management of acute systemic inflammation, such as sepsis in patients, is mostly limited to supportive care. Although targeting MMP-9 in these conditions has been suggested, it has been shown that MMP-9 inhibition has a limited therapeutic time window.50 haematologica | 2017; 102(10)

During late-stage endotoxin shock, the host responds with systematic upregulation of TIMP-1 and shift in the protease/anti-protease balance towards reduced proteolysis. Together with reports showing higher levels of TIMP1 in non-survivors44 and the present data on profound effects of LPS on MMP-9-laden neutrophils in the bone marrow, successful treatment of sepsis with protease inhibitors might first require detailed analysis of the protease/anti-protease balance in humans. Nevertheless, our study provides new insights, techniques and data to probe critical enzymes and inhibitors. In addition, it shows the technical limitations encountered during indepth analysis of non-proteolytic MMP activation in a biological setting. A future aim for the treatment of endotoxin shock might be the containment of neutrophils in the bone marrow compartment for the purpose of prohibiting collateral damage in the tissues of peripheral organs.

Funding The authors would like to thank KU Leuven for C1 grant support (C16/17/010) and the Research Foundation of Flanders (FWO-vlaanderen).

1681

J. Vandooren et al.

References 1. Opdenakker G, Fibbe WE, Van Damme J. The molecular basis of leukocytosis.

Immunol Today. 1998;19(4):182-189.

2. Fullerton JN, Gilroy DW. Resolution of inflammation: a new therapeutic frontier.

Nat Rev Drug Discov. 2016;15(8):551-567.

3. Hu J, Van den Steen PE, Sang QX, Opdenakker G. Matrix metalloproteinase inhibitors as therapy for inflammatory and vascular diseases. Nat Rev Drug Discov.

2007;6(6):480-498.

4. Clark IM, Swingler TE, Sampieri CL, Edwards DR. The regulation of matrix metalloproteinases and their inhibitors. Int J Biochem Cell Biol. 2008;40(6-7):1362-1378.

5. Vandooren J, Van den Steen PE, Opdenakker G. Biochemistry and molecular biology of gelatinase B or matrix metalloproteinase-9 (MMP-9): the next decade. Crit Rev Biochem Mol Biol. 2013;48(3):222-272.

6. Masure S, Proost P, Van Damme J, Opdenakker G. Purification and identification of 91-kDa neutrophil gelatinase.

Release by the activating peptide interleukin-8. Eur J Biochem. 1991;198(2):391398.

7. Heissig B, Hattori K, Dias S, et al.

Recruitment of stem and progenitor cells from the bone marrow niche requires MMP9 mediated release of kit-ligand. Cell.

2002;109(5):625-637.

8. D'Haese A, Wuyts A, Dillen C, et al. In vivo neutrophil recruitment by granulocyte chemotactic protein-2 is assisted by gelatinase B/MMP-9 in the mouse. J Interferon Cytokine Res. 2000;20(7):667-674.

9. Dubois B, Starckx S, Pagenstecher A, Oord J, Arnold B, Opdenakker G. Gelatinase B deficiency protects against endotoxin shock. Eur J Immunol. 2002;32(8):2163-2171.

10. Steinberg J, Halter J, Schiller HJ, et al.

Metalloproteinase inhibition reduces lung injury and improves survival after cecal ligation and puncture in rats. J Surg Res. 2003;111(2):185-195.

11. Lalu MM, Gao CQ, Schulz R. Matrix metalloproteinase inhibitors attenuate endotoxemia induced cardiac dysfunction: a potential role for MMP-9. Mol Cell Biochem.

2003;251(1-2):61-66.

12. Hu J, Van den Steen PE, Dillen C, Opdenakker G. Targeting neutrophil collagenase/matrix metalloproteinase-8 and gelatinase B/matrix metalloproteinase-9 with a peptidomimetic inhibitor protects against endotoxin shock. Biochem Pharmacol. 2005;70(4):535-544.

13. Murphy G. Tissue inhibitors of metalloproteinases. Genome Biol. 2011;12(11):233.

14. Ward RV, Atkinson SJ, Reynolds JJ, Murphy G. Cell surface-mediated activation of progelatinase A: demonstration of the involvement of the C-terminal domain of progelatinase A in cell surface binding and activation of progelatinase A by primary fibroblasts.

Biochem J. 1994;304:263-269.

15. Strongin AY, Collier I, Bannikov G, Marmer BL, Grant GA, Goldberg GI. Mechanism of cell surface activation of 72-kDa type IV collagenase. Isolation of the activated form of the membrane metalloprotease. J Biol Chem. 1995;270(10):5331-5338.

16. Rossi L, Ergen AV, Goodell MA. TIMP-1 deficiency subverts cell-cycle dynamics in murine long-term HSCs. Blood. 2011;117 (24): 6479-6488.

17. Kim KH, Burkhart K, Chen P, et al. Tissue inhibitor of metalloproteinase-1 deficiency

1682

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

amplifies acute lung injury in bleomycinexposed mice. Am J Respir Cell Mol Biol.

2005;33(3):271-279.

Khokha R, Murthy A, Weiss A. Metalloproteinases and their natural inhibitors in inflammation and immunity.

Nat Rev Immunol. 2013;13(9):649-665.

Dauphinee SM, Karsan A. Lipopolysaccharide signaling in endothelial cells. Lab Invest. 2006;86(1):9-22.

Bohannon JK, Hernandez A, Enkhbaatar P, Adams WL, Sherwood ER. The immunobiology of toll-like receptor 4 agonists: from endotoxin tolerance to immunoadjuvants.

Shock. 2013;40(6):451-462.

Cohen J. The immunopathogenesis of sepsis. Nature. 2002;420(6917):885-891.

Zenker S, Panteleev-Ivlev J, Wirtz S, et al. A key regulatory role for Vav1 in controlling lipopolysaccharide endotoxemia via macrophage-derived IL-6. J Immunol. 2014; 192(6):2830-2836.

Vandooren J, Geurts N, Martens E, Van den Steen PE, Opdenakker G. Zymography methods for visualizing hydrolytic enzymes. Nat Methods. 2013;10(3):211-220.

Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods. 2001;25(4):402408.

Vandooren J, Geurts N, Martens E, et al.

Gelatin degradation assay reveals MMP-9 inhibitors and function of O-glycosylated domain. World J Biol Chem. 2011;2(1):1424.

Su AI, Wiltshire T, Batalov S, et al. A gene atlas of the mouse and human proteinencoding transcriptomes. Proc Natl Acad Sci USA. 2004;101(16):6062-6067.

Deutschman CS, Tracey KJ. Sepsis: current dogma and new perspectives. Immunity.

2014;40(4):463-475.

Okamoto T, Akaike T, Sawa T, Miyamoto Y, van der Vliet A, Maeda H. Activation of matrix metalloproteinases by peroxynitriteinduced protein S-glutathiolation via disulfide S-oxide formation. J Biol Chem.

2001;276(31):29596-29602.

Viappiani S, Nicolescu AC, Holt A, et al.

Activation and modulation of 72kDa matrix metalloproteinase-2 by peroxynitrite and glutathione.

Biochem Pharmacol.

2009;77(5):826-834.

Vandooren J, Born B, Solomonov I, et al.

Circular trimers of gelatinase B/matrix metalloproteinase-9 constitute a distinct population of functional enzyme molecules differentially regulated by tissue inhibitor of metalloproteinases-1. Biochem J. 2015;465(2): 259-270.

Zhou M, Jacob A, Ho N, et al.

Downregulation of protein disulfide isomerase in sepsis and its role in tumor necrosis factor-alpha release. Crit Care.

2008;12(4):R100.

Van den Steen PE, Dubois B, Nelissen I, Rudd PM, Dwek RA, Opdenakker G. Biochemistry and molecular biology of gelatinase B or matrix metalloproteinase-9 (MMP-9). Crit Rev Biochem Mol Biol.

2002;37(6):375-536.

Hughes BG, Schulz R. Targeting MMP-2 to treat ischemic heart injury. Basic Res Cardiol. 2014;109(4):424.

Bannikov GA, Karelina TV, Collier IE, Marmer BL, Goldberg GI. Substrate binding of gelatinase B induces its enzymatic activity in the presence of intact propeptide. J Biol Chem. 2002;277(18):16022-16027.

35. Geurts N, Martens E, Van Aelst I, Proost P, Opdenakker G, Van den Steen PE. Betahematin interaction with the hemopexin domain of gelatinase B/MMP-9 provokes autocatalytic processing of the propeptide, thereby priming activation by MMP-3.

Biochemistry. 2008;47(8):2689-2699.

36. Rossano R, Larocca M, Riviello L, et al.

Heterogeneity of serum gelatinases MMP-2 and MMP-9 isoforms and charge variants. J Cell Mol Med. 2014;18(2):242-252.

37. Opdenakker G, Van den Steen PE, Dubois B, et al. Gelatinase B functions as regulator and effector in leukocyte biology. J Leukoc Biol.

2001;69(6):851-859.

38. Ardi VC, Kupriyanova TA, Deryugina EI, Quigley JP. Human neutrophils uniquely release TIMP-free MMP-9 to provide a potent catalytic stimulator of angiogenesis.

Proc Natl Acad Sci USA. 2007;104(51): 20262-20267.

39. Pagenstecher A, Stalder AK, Kincaid CL, Volk B, Campbell IL. Regulation of matrix metalloproteinases and their inhibitor genes in lipopolysaccharide-induced endotoxemia in mice. Am J Pathol. 2000;157(1):197-210.

40. Lalu MM, Csont T, Schulz R. Matrix metalloproteinase activities are altered in the heart and plasma during endotoxemia. Crit Care Med. 2004;32(6):1332-1337.

41. Kjeldsen L, Sengelov H, Lollike K, Nielsen MH, Borregaard N. Isolation and characterization of gelatinase granules from human neutrophils. Blood. 1994;83(6):1640-1649.

42. Lauhio A, Hastbacka J, Pettila V, et al. Serum MMP-8, -9 and TIMP-1 in sepsis: high serum levels of MMP-8 and TIMP-1 are associated with fatal outcome in a multicentre, prospective cohort study. Hypothetical impact of tetracyclines. Pharmacol Res. 2011;64(6):590-594.

43. Gounko NV, Martens E, Opdenakker G, Rybakin V. Thymocyte development in the absence of matrix metalloproteinase9/gelatinase B. Sci Rep. 2016;6:29852.

44. Hoffmann U, Bertsch T, Dvortsak E, et al.

Matrix-metalloproteinases and their inhibitors are elevated in severe sepsis: prognostic value of TIMP-1 in severe sepsis.

Scand J Infect Dis. 2006;38(10):867-872.

45. Hautamaki RD, Kobayashi DK, Senior RM, Shapiro SD. Requirement for macrophage elastase for cigarette smoke-induced emphysema in mice. Science. 1997;277 (5334):2002-2004.

46. Rahman M, Zhang S, Chew M, Syk I, Jeppsson B, Thorlacius H. Platelet shedding of CD40L is regulated by matrix metalloproteinase-9 in abdominal sepsis. J Thromb Haemost. 2013;11(7):1385-1398.

47. Deryugina EI, Zajac E, Juncker-Jensen A, Kupriyanova TA, Welter L, Quigley JP. Tissue-infiltrating neutrophils constitute the major in vivo source of angiogenesis-inducing MMP-9 in the tumor microenvironment.

Neoplasia. 2014;16(10):771-788.

48. Kolaczkowska E, Grzybek W, van Rooijen N, et al. Neutrophil elastase activity compensates for a genetic lack of matrix metalloproteinase-9 (MMP-9) in leukocyte infiltration in a model of experimental peritonitis. J Leukoc Biol. 2009;85(3):374-381.

49. Belaaouaj A, McCarthy R, Baumann M, et al.

Mice lacking neutrophil elastase reveal impaired host defense against Gram negative bacterial sepsis. Nat Med. 1998;4(5):615-618.

50. Qiu Z, Chen J, Xu H, et al. Inhibition of neutrophil collagenase/MMP-8 and gelatinase B/MMP-9 and protection against endotoxin shock. J Immunol Res. 2014; 2014:747426.

haematologica | 2017; 102(10)

ARTICLE

Bone Marrow Failure

Incidence and outcome of acquired aplastic anemia: real-world data from patients diagnosed in Sweden from 2000–2011

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Krista Vaht,1,11 Magnus Göransson2 Kristina Carlson,3 Cecilia Isaksson,4 Stig Lenhoff,5 Anna Sandstedt,6 Bertil Uggla,7 Jacek Winiarski,8 Per Ljungman,9 Mats Brune1,11 and Per-Ola Andersson10,11

Section of Hematology and Coagulation, Sahlgrenska University Hospital, Gothenburg; Department of Pediatrics, The Queen Silvia Children's Hospital, Sahlgrenska University Hospital; 3Department of Hematology, Uppsala University Hospital; 4Department of Hematology, Cancer Centre, University Hospital, Umeå; 5Department of Hematology, Skåne University Hospital, Lund University; 6Department of Hematology, Linköping University Hospital; 7Section of Hematology Department of Medicine, Faculty of Medicine and Health, Örebro University; 8Astrid Lindgren Children's Hospital, Karolinska University Hospital, Stockholm; 9Centre of allogeneic stem cell transplantation (CAST), Karolinska University Hospital Huddinge, Stockholm; 10South Älvsborg Hospital Borås and 11Sahlgrenska Academy at Gothenburg University, Sweden 1 2

Haematologica 2017 Volume 102(10):1683-1690

ABSTRACT

A

plastic anemia is a rare life-threatening disease. However, since the introduction of immunosuppressive therapy and allogeneic stem cell transplantation, the outcome has improved considerably, and the 5-year survival is reported to be 70–80% in selected patient cohorts. Yet, contemporary population-based data on incidence and survival are lacking. We performed a national retrospective study to determine the incidence, treatment, and survival of patients with aplastic anemia diagnosed in Sweden from 2000–2011. Patients were included via the National Patient Registry, and diagnosed according to the Camitta criteria. In total, 257 confirmed cases were identified, with an overall incidence of 2.35 (95% CI: 2.06–2.64) cases per million inhabitants per year. Median age was 60 years (range: 2–92), and median follow up was 76 (0–193) months. Primary treatments included immunosuppressive therapy (63%), allogenic stem cell transplantation (10%), or single-agent cyclosporine/no specific therapy (27%). The 5-year survival was 90.7% in patients aged 0–18 years, 90.5% in patients aged 19–39 years, 70.7% in patients aged 40–59 years, and 38.1% in patients aged ≥60 years.

Multivariate analysis showed that age (both 40-59 and ≥60 age groups), very severe aplastic anemia and single-agent cyclosporine/no specific therapy were independent risk factors for inferior survival. In conclusion, younger aplastic anemia patients experience a very good long-term survival, while that of patients ≥60 years in particular remains poor.

Apparently, the challenge today is to improve the management of older aplastic anemia patients, and prospective studies to address this medical need are warranted.

Correspondence: krista.vaht@vgregion.se

Received: March 29, 2017.

Accepted: July 19, 2017.

Pre-published: July 27, 2017. doi:10.3324/haematol.2017.169862 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1683 ©2017 Ferrata Storti Foundation

Introduction Aplastic anemia (AA) is a rare life-threatening disease. The incidence and median age at diagnosis, which varies according to geography, ranges from 1.5 to about seven cases per million inhabitants/year, and from 25–60 years, respectively.1-7 Following the introduction of immunosuppressive therapy (IST) with antithymocyte globulin (ATG), and allogeneic stem cell transplantation (SCT) in the 1980–90s, several studies have reported an improved outcome with a 5-year overall survival of approximately 70–80%.8-16 However, most of these studies were performed in selected patient cohorts, and included randomized trials, transplantation registry studies, or single center experiences, in which the majority of included patients were younger. In addition, there are several epidemiological studies on AA from haematologica | 2017; 102(10)

Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1683

K. Vaht et al.

Methods

Europe, the United States, South America, and Asia. Most of them reported on patients diagnosed from 1970 until the 1990s, and only a few presented outcome data, with a relatively short follow up.2,4,6,17,18 The most recent epidemiological study in Scandinavia was published in 1996, and focused on AA in the pediatric population.19 Therefore, it appears as if contemporary population-based (or realworld) data on treatment and survival in patients with AA are missing. Genuine population-based studies are difficult to perform because of the lack of nationwide AA registries on incidence and outcome. Additionally, given that many patients with AA are referred to a regional or academic center, followed by a transplant unit, many cases are lost to follow up at the referring center, and their longterm outcome is not reported. The importance of complete longitudinal population-based data was highlighted at the International Working Group on Severe Aplastic Anemia meeting in 2010, where the establishment of a population-based registry for longitudinal collection of data on patients from diagnosis, during and after treatment was proposed.20 In Sweden, where disease codes of all patients in the community-based health care system are centrally registered, there is the unique possibility to identify patients with AA, and collect detailed data on incidence, treatment, and outcome of the entire population. Such a study, while waiting for mature data from the proposed population-based AA registry, could allow for the investigation into several important issues (e.g., possible changes in incidence, potential etiological factors, the proportion of patients undergoing potentially curative treatment, and if the reported improvement of outcome is translatable to a population-based cohort, especially in older patients).

Therefore, the aim of the present national retrospective population-based study was to determine the incidence, treatment type, and survival of patients with AA diagnosed in Sweden during the years 2000–2011.

A

Identification of patients We first aimed to identify all patients with AA (both children and adults) diagnosed in Sweden between January 2000 and December 2011. In the year 2000, Sweden had 8,882,792 inhabitants, while in 2011 the figure was 9,482,855. Patients were identified in the National Patient Registry held by the Swedish National Board of Health and Welfare, which collects patient, geographical, administrative, and medical data. The registry includes data on all patients treated as in- or outpatients in the national health care system regarding primary and secondary diagnoses and procedures. In the registry, patients are identified by a unique national social security number. For the registry search, the 10th version of International Classification of Diseases was used, and for a complete search, the disease codes D61.0–D61.9 were initially applied. For a detailed description of the identification of patients from the registry, see Online Supplementary Figure S1. The study was approved by the Regional Ethical Review Board in Gothenburg.

Most of the patients were treated at one of the seven university hospitals. Some patients were treated at the regional or county hospitals (29 hospitals in total) without referral to a university hospital.

The diagnosis of AA was confirmed according to the Camitta criteria21: bone marrow biopsy cellularity <25% (or 25–50% with <30% residual hematopoietic cells) together with two of the following three criteria: hemoglobin <100 g/L, reticulocytes <50×109/L, or <1%; platelets <50×109/L; and neutrophil leucocytes (ANC) <1.5×109/L. Disease severity was classified as follows: severe aplastic anemia (SAA), with two of the following three characteristics: ANC <0.5×109/L, platelet count <20×109/L, or reticulocyte count <20×109/L; very severe aplastic anemia (VSAA) had the same characteristics as SAA, but with ANC <0.2×109/L; and non-severe aplastic anemia (NSAA). Patients with congenital disease, pancytopenia without a marrow biopsy performed, marrow fibrosis, or other signs of malignancy or dysplasia were excluded. A mild dysplasia in erythropoiesis was accepted.

B

Figure 1. Overall survival. A. For all patients. B. In different age groups: 0–18 years, 19–39 years, 40–59 years, and ≥60 years.

1684

haematologica | 2017; 102(10)

Aplastic anemia in Sweden from 2000-2011

All data were collected in a case report form, and unclear cases were re-evaluated. Follow-up information was obtained from medical charts and/or through matching the social security number in the Swedish Cause of Death Registry.

Statistics Incidence and confidence intervals (CI) were calculated according to Rothmann.22 Rates and proportions were compared using Pearson’s χ2 test. The Kaplan-Meier method was used for estimation of overall survival, and comparisons were based on the logrank test. Analysis of risk factors for survival were calculated using Cox regression proportional hazards, including age, sex, type of treatment and disease severity. Relative survival ratios were calculated using the Ederer II23 method by dividing the observed survival of patients with AA with the expected survival in a general Swedish population with corresponding age, sex, and calendar year. The statistical analyses were performed either by SPSS version 23 or Stata for Macintosh, version 13.1, and relative survival ratio was calculated by use of the strs module.

Results Basic data Between 1st January 2000 and 31st December 2011, we identified 257 confirmed cases of acquired AA among 1,362 potential cases. The remaining cases had malignant diseases, secondary anemia/pancytopenia, or other benign hematological disorders (Online Supplementary Figure S1).

Clinical characteristics are shown in Table 1. The median age at diagnosis was 60 years (95% CI: 54–64, range: 2– 92). A total of 133 patients (52%) were female, and the median age was 60 years for both sexes (females: 95% CI: 51–66, range: 2–90; males: 95% CI: 51–65, range: 7–92).

At diagnosis, 38% had NSAA, 38% had SAA, and 24% had VSAA. There was no significant age-related distribution difference between patients with SAA and VSAA, however NSAA patients were older (P=0.028 and P=0.001, respectively).

Incidence The overall incidence was estimated to be 2.35 (95% CI: 2.06–2.64) cases per million inhabitants per year. For all patients, a biphasic age distribution was observed; one peak in patients aged 15–20 years, 2.87 (95% CI: 1.72– 4.03), and one in patients >60 years old, 4.36 (95% CI: 3.55–5.18). The incidence according to age groups and sex are shown in Online Supplementary Table S1. The biphasic

Table 1. Clinical characteristics.

Cases

257

Age (years) Median Range Sex (F/M), n (%) Follow up (months) Median Range Severity of AA, n (%) VSAA SAA NSAA Primary treatment, n (%) IST SCT Cyclosporine alone/No specific therapy: Cyclosporine Transfusions Spontaneous remission Second-line treatment, n SCT Second IST Third-line treatment, n SCT Third IST

60 2–92 133(52)/124(48) 76 0–193 62 (24) 97 (38) 98 (38) 161 (63)* 25 (10) 71 (27) 45** 20*** 6 28 55 13 7

IST-immunosuppressive therapy, SCT-allogeneic stem cell transplantation. *ATG (n=158), basiliximab (n=2), alemtuzumab (n=1); **Together with erythropoietin (EPO) (n=8), granulocyte colony-stimulating factor (G-CSF) (n=2), EPO and G-CSF (n=3), androgen (n=1); ***Together with EPO (n=3), G-CSF (n=2), androgen (n=1).

distribution was predominantly observed in male patients, while the incidence among females was more evenly distributed and increased steadily with a peak above the age of 60 years. The incidence in children <10 years old was lower: 1.8 per million per year. Furthermore, no difference in incidence was observed when grouping the patients according to two different time periods (2000–2005 and 2006–2011); 2.19 (95% CI: 1.80–2.29) and 2.5 (95% CI: 2.08–2.91), respectively.

Table 2. Primary treatment in different age groups.

Treatment modality IST SCT Cyclosporine alone Transfusions

n (%) n (%) n (%) n (%)

Spontaneous remission

n (%)

Total

n (%)

haematologica | 2017; 102(10)

0–18

19–39

29 (67.5) 13 (30.2) 0 (0) 0 (0) 1 (2.3) 43 (100)

30 (71.4) 7 (16.7) 2 (4.8) 0 (0) 3 (7.1) 42 (100)

Age group (years) 40–59 33 (80.5) 4 (9.8) 2 (4.9) 0 (0) 2 (4.9) 41 (100)

≥60

Total

69 (52.7) 1 (0.8) 41 (31.3) 0 (0) 20 15.2 131 (100)

161 25 45 20 6 257 1685

K. Vaht et al.

Treatment Primary treatment included IST (defined as treatment with ATG or other anti-T-cell monoclonal antibody together with cyclosporine, n=161, 63%), a total of 25 patients (10%) underwent allogeneic SCT, and 71 patients (27%) were treated with single-agent cyclosporine (CSA alone) or no specific therapy. Spontaneous remission was observed in six cases (five cases of NSAA and one of SAA).

Treatment data are shown in Table 1, and data on type of initial treatment in the different age groups are shown in Table 2. The median age in the ≥60 years group who received IST, CSA alone, and no specific therapy was 67 (range: 60–85), 79 (range: 60–90), and 82.5 (range: 62–92) years, respectively.

Survival The median follow up was 76 months (95% CI: 66–86, range: 0–193). During follow up, 121 (47.1%) patients died. Twenty-six died within 3 months, and a further 48 died within the first 24 months. The most common causes of death during the first 24 months were infections (n=41), bleeding (n=14), and unspecified conditions related to AA (n=8). The 5-year survival for all patients with AA was 60.7% (95% CI: 57.7–63.7) (Figure 1A), and median survival was 150 months. The 5-year survival, irrespective of treatment modality, varied according to the different age groups, and was significantly lower in patients aged 40–59 years and ≥60 years: 90.7% (95% CI: 77.1–96.4) in patients aged 0–18 years, 90.5% (76.6–96.3) in patients 19–39 years (P=0.95), 70.7% (54.3–82.2) in patients 40–59 years (P=0.029), and 38.1% (29.8–46.4) in patients ≥60 years (P=0.001) (Figure 1B). When dividing the ≥60 years group into two further groups, 60-69 years and ≥70 years, the 5-year survival was 57.5% (41.8-70.5) and 27.9 (18.937.6) (P=0.001), respectively. The age-related survival difference was obvious early after diagnosis: patients ≥60 years had a 3-month survival of 84.0% compared with 97.7% for patients aged 0–18 years, 97.6 % for patients 19–39 years, and 92.7% for patients 40–59 years (P=0.021, P=0.024, and P=0.155, respectively). When grouping patients according to disease severity, the 5-year survival was lower in patients with VSAA compared with those with SAA (P=0.025), but not compared with NSAA (P=0.13) (Figure 2). However, during follow up, almost 39% of patients with NSAA developed SAA or VSAA and the majority (84.7%) were treated: IST (n=61), CSA alone (n=18), or SCT (n=4). Early mortality rate (at 3 months) was significantly higher in VSAA compared with SAA and NSAA: 22.6% versus 8.2% and 4.1% (P=0.009 and P<0.0001), respectively.

The 5-year survival rate was 96.0% (74.8–99.4) in patients who underwent SCT, 68.9% (61.2–75.5) in the IST group (P=0.009), and 29.6% (19.5–40.4) in patients who received CSA alone or no specific therapy (P<0.0001) (Figure 3A). In younger patients (0–18 and 19–39 years), there was no significant difference in survival, whether they were primarily treated with IST or if they underwent SCT: 0–18 years, 86.2% vs. 100% (P=0.169); and 19–39 years, 90% vs. 100% (P=0.395). Furthermore, when grouping these patients together, the corresponding values were 88.1% vs. 100% (P=0.113). Regarding the group of patients treated with IST (n=161), there was no survival difference at 5 years between the age groups, 0–18 years: 86.2% (67.3–94.6), 19–39 years: 90% (72.1–96.7), and 40– 59 years: 69.7% (51.0–82.4) (P=0.67, P=0.12, and P=0.056, 1686

Figure 2. Overall survival according to disease severity at diagnosis (very severe, severe, and non-severe AA).

respectively). Patients ≥60 years had a significantly worse survival rate of 52.2% (39.8–63.2) compared with patients aged 0–18 years (P=0.003) and 19–39 years (P=0.001), but not with patients aged 40–59 years (P=0.15) (Figure 3B).

The Cox regression analysis revealed that age (40-59 and ≥60 years age groups), VSAA and treatment with CSA alone/no therapy were independent risk factors for inferior survival (Table 3). The hazard ratio for SCT compared to IST was 0.17, but due to the small number of patients this was not statistically significant.

Forty-three (27%) patients in the entire IST group were allografted (related donor, n=11; unrelated donor, n=32) after not responding to/relapsing following one or two cycles of IST. Only two (3%) patients ≥60 years underwent SCT compared with 14 (48%) patients in the group aged 0–18 years, 17 (57%) in the group aged 19–39 years, and 10 (30%) in the group aged 40–59 years.

For patients receiving CSA alone/no specific therapy, we observed no survival difference between patients with a more palliative approach (“Transfusions”; n=20) and CSA alone patients (n=45), 20.0% vs. 24.4% (P=0.172).

When grouping the entire patient cohort according to two different time-periods (from 2000–2005 or 2006– 2011), we found no difference in 5-year survival for all patients (61.0% vs. 61.5%), or in the different age groups (data not shown).

The relative 5-year survival (i.e., excess mortality: the difference between observed mortality and expected mortality) for all patients was 65.4% (95% CI: 58.6–71.5) (Figure 4A). When grouping all patients according to the median age at diagnosis, the relative 5-year survival was 84.6% (76.9–90.0) in patients less than 60 years, while the corresponding figure for patients ≥60 years was significantly worse, 45.3% (35.4–55.1) (Figure 4B). When grouping the patients into a younger group, 0–39 years, and a group ≥40 years, the relative survival at 5 years was significantly higher in the younger group: 90.8% (82.2–95.4) haematologica | 2017; 102(10)

Aplastic anemia in Sweden from 2000-2011

Table 3. Uni- and multivariable Cox-regression.

Variable Therapy IST SCT Cyclosporine alone/ No specific therapy Severity NSAA SAA VSAA Gender Male Female Age group (years) 0-18 19-39 40-59 60-69 70-

Univariable Cox-regression Deaths*/N (%) HR 95% CI

Multivariable Cox-regression P HR 95% CI

P

60/161 (37%) 1/25 (4%) 56/71 (79%)

1.00 0.09 3.28

0.01-0.63 2.27-4.73

0.016 <0.001

1.00 0.17 1.84

0.02-1.28 1.18-2.88

0.086 0.007

45/98 (46%) 39/97 (40%) 33/62 (53%)

1.00 0.84 1.35

0.55-1.29 0.86-2.11

0.422 0.195

1.00 1.24 3.47

0.80-1.91 2.16-5.59

0.329 <0.001

65/133 (49%) 52/124 (42%)

1.00 0.83

0.58-1.20

0.325

5/43 (12%) 5/42 (12%) 13/41 (32%) 23/45 (51%) 71/86 (83%)

1.00 1.09 3.32 5.84 13.7

0.31-3.75 1.18-9.31 2.22-15.4 5.50-34.1

0.896 0.023 <0.001 <0.001

1.00 1.22 4.38 7.06 12.8

0.35-4.24 1.52-12.6 2.60-19.2 4.78-34.3

0.771 0.007 <0.001 <0.001

* Within 10 years.

and 52.1% (43.4–60.3), respectively. When dividing the ≥60 years group into 60-69 years and ≥70 years, there was a numerical difference however without statistical significance: 60.7% (44.1-74.5) compared to 37.1% (25.1-49.9).

Discussion In this comprehensive population-based study on patients with AA patients in Sweden, diagnosed from 2000–2011, we found that the 5-year survival among younger patients (up to the age of 40) was about 90%, which is similar to the reported survival rates from modern clinical trials with ATG, or from SCT registry data.7,9,1316,24 Furthermore, there was no survival difference between IST and SCT as the primary treatment in the younger patients, which corroborate with recently published data from the European Society for Blood and Marrow Transplantation (EBMT) registry of the SAA Working Party.25 Additionally, patients from 40-59 years old experienced a 5-year survival of about 70%. Together, these figures appear to be superior in comparison to data from the latest population-based AA study, on Spanish patients diagnosed between 1980 and 2003, where the reported 2year survival was about 80% and 40% in the comparable patient groups.7 Thus, also in a population-based cohort, the outcome for patients with AA below the median age seems to have improved during the last decade. There could be several possible reasons for this: a higher percentage of these patients are treated with SCT, earlier onset of treatment, and better supportive care including infection prophylaxis.

However, even though AA patients 40-59 years also had an inferior survival, patients aged ≥60 years do not seem haematologica | 2017; 102(10)

to have gained any survival improvement in the last few decades. Less than 40% of these patients were alive after 5 years, and the relative 5-year survival was 45%, which indicated considerable excess mortality. In addition, patients ≥60 years had a significantly higher risk of early death compared with younger patients; 16% were deceased 3 months after diagnosis. On the other hand, patients ≥60 years old treated with IST (52.7 %) had a 5year survival of around 50%, which was similar to the results of IST reported from EBMT AA registry data.26 In their study, although survival was worse in older patients than in younger ones, the response to IST was independent of age. This finding, together with our results, may imply that more older patients should be treated with IST. In our cohort of patients ≥60 years old, a relatively large number (n=45) were treated with single-agent therapy of cyclosporine. In a phase III trial comparing cyclosporine alone versus the combination of ATG and cyclosporine in patients with NSAA, the overall response rate of cyclosporine alone was 46% compared with 74% in the combination arm.27 After a median follow up of 1 year, no difference in overall survival was observed. In contrast, our cohort of patients treated with cyclosporine had no survival benefit compared with patients treated with a stricter transfusion policy, i.e., with a palliative approach.

In our cohort that received cyclosporine alone, only 38% had NSAA, whereas 62% had SAA or VSAA. In retrospective studies, obtaining reliable information on therapeutic intent and endurance can be difficult. However, since patients starting treatment with cyclosporine alone likely require sufficient renal function to tolerate therapeutic doses, it is plausible to assume that at least some of the patients in this cohort would have been eligible for IST. Furthermore, only three patients ≥60 years received an 1687

K. Vaht et al.

allograft (one upfront, and two after failing IST). In the first comprehensive treatment guidelines for severe AA published in 1994,28 the Working Party on SAA recommended SCT from a HLA-identical sibling as upfront treatment for patients up to the age of 40. The first British AA guidelines (published in 2003) had similar recommendations, although patients from 30–40 years old could receive either IST or SCT.29 In the guidelines from 2009, the age limit was set at 40 years,30 and in the most recent guidelines from 2016,31 patients up to the age of 50 were considered eligible for upfront SCT. Regarding SCT as salvage for patients failing IST, the guidelines from 2003 advocated a matched unrelated donor SCT up to the age of 40. In the subsequent update, this age limit was changed to 50 years, but was also considered as an option for patients between 50–60 years if they had good performance status. Finally, in the 2016 version, the authors indicated that there is no strict upper age limit. At present, one of the most frequently allotransplanted groups, with a reduced intensity conditioning regimen, are those with acute myelogenous leukemia (AML) between 50–70 years old. This procedure is now considered standard of care for older patients with AML.32 Therefore, given our poor outcome data for patients ≥60 years old, this treatment option should likely be discussed more often. A possible alternative for older patients who are not eligible for IST with ATG, or for SCT, is treatment (preferably within a multicenter prospective clinical trial) with the thrombopoietin receptor antagonist, eltrombopag, which has shown promising results in refractory AA.33,34 In addition, very promising results with the combination of eltrombopag and IST was most recently reported,35 which could potentially become a future treatment option for all patients considered eligible for IST. We found that the incidence of AA during the study period was 2.35 per million inhabitants per year, and showed a biphasic age distribution. Previous epidemiolog-

A

ical studies have shown a broad variation in incidence depending on time and geographical location. Reports from Europe and the United States in the 1960–70s presented a very high incidence: from six to ten cases of AA per million per year. In addition, in some studies, there was an association with toxic agricultural substances.3,4,6,17 However, different diagnostic criteria were used, and many cases likely represented other diseases.4 The International Agranulocytosis and Aplastic Anemia Study was published in 1987, and established a well-accepted overall incidence of around two cases of AA per million.6 This was later confirmed by data from Spain published in 2008, with an overall incidence of 2.34 per million per year.7 In contrast, studies performed in Asia showed a higher incidence: over four patients with AA per million per year.5,36 A higher incidence in younger people in Asia has been suggested to be associated with environmental factors related to occupation.36 Our incidence data correspond to the aforementioned later reports from Europe, the United States, and South America. In some studies, the incidence has also been reported to be slightly higher among females,2,6 while data from Turkey and Bangkok have instead shown a male predominance.5,17 We found no differences between female and male incidence rate (1.07:1), which was consistent with data from the Spanish study from 2008.

The majority of patient deaths that occurred within two years from diagnosis were from serious infections (50%) or bleeding (15%), indicating the importance of early treatment and prevention of infections with antibacterial and antifungal therapy. It has been reported that despite the lack of progress in achieving higher response rates for patients with AA in the last two decades, both infectionrelated and overall mortality have been reduced in patients with SAA who are unresponsive to initial IST, mainly because of prompt empirical antifungal therapy.37 Patients in our cohort who died later had refractory dis-

B

Figure 3. Overall survival. A. According to primary treatment. B. For patients treated with IST divided into different age groups.

1688

haematologica | 2017; 102(10)

Aplastic anemia in Sweden from 2000-2011

A

B

Figure 4. Relative survival. A. For all patients. B. Divided into two age groups according to median age at diagnosis: < and ≥60 years.

ease, complications from transplantation, worsening of heart failure because of chronic anemia, or secondary cancer.

Although our study included all identifiable patients with AA diagnosed in Sweden between 2000 and 2011, and therefore reasonably reflects the true incidence and survival, it had limitations. First, there could be additional patients with AA who were not listed in the national diagnosis registry (e.g., patients with a mild form and subsequent spontaneous remission or patients who died before established diagnosis). However, based on the pilot study, when the search criteria were established, we estimated that the number of missing patients was going to be at most 2%, i.e., only four to five patients. Furthermore, other cases that could have been missed because of our search criteria of at least two medical contacts with D61.0–D61.9 would likely have been older individuals unsuited for active treatment, with a short life expectancy.

Such patients would worsen the survival data of the entire ≥60 years old patient group. Second, the patient data were collected retrospectively. However, all of the charts were thoroughly examined, and our follow up data are almost complete (only one patient was lost during follow up).

Third, we did not re-evaluate the bone marrow biopsies.

References 1. Camitta BM, Thomas ED, Nathan DG, et al. Severe aplastic anemia: a prospective study of the effect of early marrow transplantation on acute mortality. Blood.

1976;48(1):63-70.

2. Cartwright RA, McKinney PA, Williams L, et al. Aplastic anaemia incidence in parts of the United Kingdom in 1985. Leuk Res. 1988;12(6):459-463.

3. Mary JY, Baumelou E, Guiguet M. Epidemiology of aplastic anemia in France: a prospective multicentric study. The French Cooperative Group for Epidemiological Study of Aplastic Anemia.

Blood. 1990;75(8):1646-1653.

4. Szklo M, Sensenbrenner L, Markowitz J,

haematologica | 2017; 102(10)

5.

6.

7.

8.

Nevertheless, we scrutinized the biopsy reports from the hematopathologist for all patients, and even though the possibility of solitary cases of hypoplastic myelodysplastic syndromes cannot be ruled out, it seems unlikely that such cases would have significantly impacted our results.

In conclusion, apart from the limitations, incidence data obtained from this contemporary population-based study are consistent with previously established figures.

Furthermore, younger patients, regardless of initial therapy, experienced a very good long-term survival. However, for patients above the median age at diagnosis (≥60 years), excess mortality was still substantial. Therefore, the management of older patients with AA should be improved.

Prospective studies to address this medical need are warranted.

Funding This study was supported by grants from ALF Västra Götaland, Gothenburg Medical Society, and a scholarship from Alexion Sweden.

Acknowledgments We would also like to thank Erik Holmberg, Regional Cancer Centrum Väst, for expert statistical assistance.

Weida S, Warm S, Linet M. Incidence of aplastic anemia in metropolitan Baltimore: a population-based study. Blood. 1985;66 (1):115-119.

Issaragrisil S, Kaufman DW, Anderson T, et al. The epidemiology of aplastic anemia in Thailand. Blood. 2006;107(4):1299-1307.

IAAS. Incidence of aplastic anemia: the rel evance of diagnostic criteria. By the International Agranulocytosis and Aplastic Anemia Study. Blood. 1987;70(6):17181721.

Montane E, Ibanez L, Vidal X, et al.

Epidemiology of aplastic anemia: a prospective multicenter study.

Haematologica. 2008;93(4):518-523.

Speck B, Gratwohl A, Nissen C, et al.

Treatment of severe aplastic anaemia with antilymphocyte globulin or bone-marrow

transplantation. Br Med J (Clin Res Ed).

1981;282(6267):860-863.

9. Locasciulli A, Oneto R, Bacigalupo A, et al.

Outcome of patients with acquired aplastic anemia given first line bone marrow transplantation or immunosuppressive treatment in the last decade: a report from the European Group for Blood and Marrow Transplantation (EBMT). Haematologica.

2007;92(1):11-18.

10. Frickhofen N, Kaltwasser JP, Schrezenmeier H, et al. Treatment of aplastic anemia with antilymphocyte globulin and methylprednisolone with or without cyclosporine. The German Aplastic Anemia Study Group. N Engl J Med. 1991;324(19):1297-1304.

11. Bacigalupo A, Bruno B, Saracco P, et al.

Antilymphocyte globulin, cyclosporine, prednisolone, and granulocyte colony-

1689

K. Vaht et al.

12.

13.

14.

15.

16.

17.

18.

19.

1690

stimulating factor for severe aplastic anemia: an update of the GITMO/EBMT study on 100 patients. European Group for Blood and Marrow Transplantation (EBMT) Working Party on Severe Aplastic Anemia and the Gruppo Italiano Trapianti di Midolio Osseo (GITMO).

Blood.

2000;95(6):1931-1934.

Bacigalupo A, Brand R, Oneto R, et al.

Treatment of acquired severe aplastic anemia: bone marrow transplantation compared with immunosuppressive therapy-The European Group for Blood and Marrow Transplantation experience.

Semin Hematol. 2000;37(1):69-80.

Bacigalupo A, Socie G, Lanino E, et al.

Fludarabine, cyclophosphamide, antithymocyte globulin, with or without low dose total body irradiation, for alternative donor transplants, in acquired severe aplastic anemia: a retrospective study from the EBMTSAA Working Party. Haematologica.

2010;95(6):976-982.

Afable MG, 2nd, Shaik M, Sugimoto Y, et al.

Efficacy of rabbit anti-thymocyte globulin in severe aplastic anemia. Haematologica.

2011;96(9):1269-1275.

Marsh JC, Pearce RM, Koh MB, et al.

Retrospective study of alemtuzumab vs ATG-based conditioning without irradiation for unrelated and matched sibling donor transplants in acquired severe aplastic anemia: a study from the British Society for Blood and Marrow Transplantation. Bone Marrow Transplant. 2014;49(1):42-48.

Scheinberg P, Nunez O, Weinstein B, et al.

Horse versus rabbit antithymocyte globulin in acquired aplastic anemia. N Engl J Med. 2011;365(5):430-438.

Baslar Z, Aktuglu G, Bolaman Z, et al.

Incidence of aplastic anemia in Turkey: a hospital-based prospective multicentre study. Leuk Res. 1997;21(11-12):1135-1139.

Maluf E, Hamerschlak N, Cavalcanti AB, et al. Incidence and risk factors of aplastic anemia in Latin American countries: the LATIN case-control study.

Haematologica.

2009;94(9):1220-1226.

Clausen N, Kreuger A, Salmi T, Storm-

20.

21.

22.

23.

24.

25.

26.

27.

Mathisen I, Johannesson G. Severe aplastic anaemia in the Nordic countries: a population based study of incidence, presentation, course, and outcome. Arch Dis Child.

1996;74(4):319-322.

Pulsipher MA, Young NS, Tolar J, et al.

Optimization of therapy for severe aplastic anemia based on clinical, biologic, and treatment response parameters: conclusions of an international working group on severe aplastic anemia convened by the Blood and Marrow Transplant Clinical Trials Network, March 2010. Biol Blood Marrow Transplant.

2011;17(3):291-299.

Camitta BM, Rappeport JM, Parkman R, Nathan DG. Selection of patients for bone marrow transplantation in severe aplastic anemia. Blood. 1975;45(3):355-363.

KJ R. Modern Epidemiology. Boston: Little Brown; 1986.

Ederer F, Axtell LM, Cutler SJ. The relative survival rate: a statistical methodology. Natl Cancer Inst Monogr. 1961;6:101-121.

Marsh JC, Bacigalupo A, Schrezenmeier H, et al. Prospective study of rabbit antithymocyte globulin and cyclosporine for aplastic anemia from the EBMT Severe Aplastic Anaemia Working Party. Blood. 2012; 119(23):5391-5396.

Bacigalupo A, Giammarco S, Sica S. Bone marrow transplantation versus immunosuppressive therapy in patients with acquired severe aplastic anemia. Int J Hematol.

2016;104(2):168-174.

Tichelli A, Socie G, Henry-Amar M, et al.

Effectiveness of immunosuppressive therapy in older patients with aplastic anemia.

European Group for Blood and Marrow Transplantation Severe Aplastic Anaemia Working Party. Ann Intern Med. 1999; 130(3):193-201.

Marsh J, Schrezenmeier H, Marin P, et al.

Prospective randomized multicenter study comparing cyclosporin alone versus the combination of antithymocyte globulin and cyclosporin for treatment of patients with nonsevere aplastic anemia: a report from the European Blood and Marrow Transplant (EBMT) Severe Aplastic Anaemia Working

Party. Blood. 1999; 93(7):2191-2195.

28. Bacigalupo A. Guidelines for the treatment of severe aplastic anemia. Working Party on Severe Aplastic Anemia (WPSAA) of the European Group of Bone Marrow Transplantation (EBMT). Haematologica.

1994;79(5):438-444.

29. Marsh JC, Ball SE, Darbyshire P, et al.

Guidelines for the diagnosis and management of acquired aplastic anaemia. Br J Haematol. 2003;123(5):782-801.

30. Marsh JC, Ball SE, Cavenagh J, et al.

Guidelines for the diagnosis and management of aplastic anaemia. Br J Haematol.

2009;147(1):43-70.

31. Killick SB, Bown N, Cavenagh J, et al.

Guidelines for the diagnosis and management of adult aplastic anaemia. Br J Haematol. 2016;172(2):187-207.

32. Champlin R. Reduced intensity allogeneic hematopoietic transplantation is an established standard of care for treatment of older patients with acute myeloid leukemia. Best Pract Res Clin Haematol. 2013;26(3):297300.

33. Desmond R, Townsley DM, Dumitriu B, et al. Eltrombopag restores trilineage hematopoiesis in refractory severe aplastic anemia that can be sustained on discontinuation of drug. Blood. 2014;123(12):18181825.

34. Olnes MJ, Scheinberg P, Calvo KR, et al.

Eltrombopag and improved hematopoiesis in refractory aplastic anemia. N Engl J Med. 2012;367(1):11-19.

35. Townsley DM, Scheinberg P, Winkler T, et al. Eltrombopag added to standard immunosuppression for aplastic anemia. N Engl J Med. 2017;376(16):1540-1550.

36. Yong AS, Goh AS, Rahman M, Menon J, Purushothaman V. Epidemiology of aplastic anaemia in the state of Sabah, Malaysia.

Med J Malaysia. 1998;53(1):59-62.

37. Valdez JM, Scheinberg P, Nunez O, Wu CO, Young NS, Walsh TJ. Decreased infectionrelated mortality and improved survival in severe aplastic anemia in the past two decades. Clin Infect Dis. 2011; 52(6):726735.

haematologica | 2017; 102(10)

ARTICLE

Bone Marrow Failure

Rapamycin is highly effective in murine models of immune-mediated bone marrow failure Xingmin Feng,1 Zenghua Lin,1,2 Wanling Sun,1,3 Maile K. Hollinger,1 Marie J. Desierto,1 Keyvan Keyvanfar,1 Daniela Malide,4 Pawel Muranski,1 Jichun Chen,1 and Neal S. Young1

Hematology Branch, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD, USA; 2Department of Hematology, Affiliated Hospital of Nantong University, Nantong, Jiangsu, China; 3Department of Hematology, Xuanwu Hospital, Capital Medical University, Beijing, China and 4Light Microscopy Core Facility, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD, USA

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

1

Haematologica 2017 Volume 102(10):1691-1703

ABSTRACT

A

cquired aplastic anemia, the prototypical bone marrow failure disease, is characterized by pancytopenia and marrow hypoplasia. Most aplastic anemia patients respond to immunosuppressive therapy, usually with anti-thymocyte globulin and cyclosporine, but some relapse on cyclosporine withdrawal or require long-term administration of cyclosporine to maintain blood counts. In this study, we tested efficacy of rapamycin as a new or alternative treatment in mouse models of immune-mediated bone marrow failure. Rapamycin ameliorated pancytopenia, improved bone marrow cellularity, and extended animal survival in a manner comparable to the standard dose of cyclosporine.

Rapamycin effectively reduced Th1 inflammatory cytokines interferon-γ and tumor necrosis factor-α, increased the Th2 cytokine interleukin-10, stimulated expansion of functional regulatory T cells, eliminated effector CD8+ T cells (notably T cells specific to target cells bearing minor histocompatibility antigen H60), and preserved hematopoietic stem and progenitor cells. Rapamycin, but not cyclosporine, reduced the proportion of memory and effector T cells and maintained a pool of naïve T cells.

Cyclosporine increased cytoplasmic nuclear factor of activated T-cells-1 following T-cell receptor stimulation, whereas rapamycin suppressed phosphorylation of two key signaling molecules in the mammalian target of rapamycin pathway, S6 kinase and protein kinase B. In summary, rapamycin was an effective therapy in mouse models of immune-mediated bone marrow failure, acting through different mechanisms to cyclosporine. Its specific expansion of regulatory T cells and elimination of clonogenic CD8+ effectors support its potential clinical utility in the treatment of aplastic anemia.

Correspondence: fengx2@nhlbi.nih.gov

Received: January 13, 2017.

Accepted: July 12, 2017.

Pre-published: July 20, 2017. doi:10.3324/haematol.2017.163675 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1691

Introduction Aplastic anemia (AA) is a disease of bone marrow (BM) failure characterized by pancytopenia and marrow hypocellularity. In most patients, this is due to immune attack of hematopoietic stem and progenitor cells (HSPCs) by auto-reactive T cells.1 Standard immunosuppressive therapy (IST) with horse anti-thymocyte globulin (ATG) and cyclosporine A (CsA) is effective in 60-70% of AA patients, resulting in hematologic recovery. However, patients who have responded to IST often relapse after CsA withdrawal or are dependent on continued CsA administration in order to maintain blood counts.2 The overall and complete response rates to immunosuppressive therapy have increased to almost 100% with the addition of the thrombopoietin mimetic eltrombopag, but relapse on discontinuation of CsA may be especially problematic in these patients.3 ATG and CsA appear to partially eliminate and functionally suppress activation of expanded CD8+ effector T-cell clones.4 However, oligoclones are often not eliminated, and relapse is likely due to their reactivation and renewed destruction of HSPCs and precursors. In the clinic, therapeutic strategies to achieve tolerance are highly desirable in order to avoid complihaematologica | 2017; 102(10)

©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1691

X. Feng et al.

cations of recurrent pancytopenia that may require re-initiation of transfusions, hospitalizations for neutropenic fever, and control of chronic toxicity due to repeated interventions.

Human AA has been modeled in mice by adaptation of historic “runt disease” in which infusion of lymph node (LN) cells into recipients mismatched at MHC or minor histocompatibility (minor-H) antigen loci produced BM failure with severe pancytopenia and marrow hypoplasia that mimics human AA.5,6 Like human AA, treatment of murine BM failure in these models with CsA and other immunosuppressive agents ameliorates disease. These models have been used to test the plausibility of immune mechanisms suggested by the study of patients and patient cells.

In the search for an alternative and/or supplementary treatment for AA and BM failure, we turned our attention to rapamycin, an inhibitor of the mammalian target of rapamycin (mTOR) pathway that has been used in a variety of animal models of human diseases, such as murine experimental allergic encephalomyelitis, nephritis, lupus erythematosus, and inflammatory bowel disease.7-12 In the clinic, rapamycin is used to treat autoimmune hepatitis and uveitis,13,14 and to prevent rejection in solid organ and hematopoietic stem cell transplantation.15-17 In this study, we employed murine models to test efficacy of rapamycin as a therapy for immune-mediated BM failure, based on its well-characterized immunosuppressive activity and its tolerogenic role in organ transplantation, and aimed at its application as prophylaxis or salvage treatment of AA patients at risk of relapse. We were especially interested in comparing the mechanisms of action between rapamycin and CsA.

Methods Animals, induction of BM failure, and immunosuppressive therapies Inbred C57BL/6 (B6) and FVB/N (FVB), congenic C.B10H2b/LilMcd (C.B10) and B6-Cg-Tg(CAG-DsRed*MST)1Nagy/J (DsRed), and hybrid CByB6F1/J (CByB6F1) mice were all originally obtained from the Jackson Laboratory (Bar Harbor, ME, USA) and were bred and maintained in National Institutes of Health animal facilities (Bethesda, MD, USA) under standard care and nutrition. All animal studies were approved by the Animal Care and Use Committee at the National Heart, Lung, and Blood Institute.

Induction of immune-mediated BM failure was performed as previously reported.5,6 In brief, LN cells from B6 or DsRed donors were homogenized, washed, filtered and intravenously injected into sex-matched CByB6F1 or C.B10 recipients pre-irradiated with 5 Gy of total body irradiation (TBI) 4-6 hours (h) earlier (or LN cells from FVB donors were infused into 6.5-Gy-pre-irradiated B6 recipients). Mice were used at 8-12 weeks of age in all experiments. In most experiments, animals were bled and euthanized 12-18 days (d) later to obtain tissue for histology and cytology. Some mice were maintained long-term to measure survival under normal animal care conditions.

Rapamycin was obtained from LC Laboratories (Woburn, MA, USA), dissolved in pure ethyl alcohol (The Warner-Graham Company, Cockeysville, MD, USA) at 50 mg/mL for storage at -30˚C, diluted to 200 mg/mL with a 5% PEG-400 (polyethylene glycol MW 400, Sigma-Aldrich, St. Louis, MO, USA) and 5% Tween-80 (Sigma-Aldrich) solution before use,18 and administered through intraperitoneal injection (i.p.) once daily for 5-13 d at 2 1692

mg/kg. CsA (NDC 0078-0109-01, Novartis Pharmaceutical Corporation, East Hanover, NJ, USA) was diluted in Iscove’s Modified Dulbecco’s Medium (IMDM) to 5 mg/mL and injected (i.p.) into animals at 50 mg/kg body weight once daily for 5-10 d.

Both drugs were filtered through a 0.22 mM syringe-driven MillexGS filter (Millipore Corporation, Billerica, MA, USA) before injection.

Methods regarding flow cytometry analyses, regulatory T-cell isolation and functional analysis, cytokine measurement, transcriptome assay, cell culture, immunoblotting, and hematopoietic progenitor assays are described in the Online Supplementary Appendix.

Statistical analysis Data obtained from complete blood count, BM cell counting and flow cytometry were analyzed by unpaired t-test, variance analyses, and multiple comparisons using GraphPad Prism statistical software. Data are presented as means with standard errors.

The differences in survival among different groups of animals were evaluated by log rank test. P<0.05 was considered statistically significant.

Results Rapamycin ameliorates pancytopenia and BM hypoplasia in AA mice To evaluate the potential prophylactic effect of rapamycin in immune-mediated murine BM failure, we first infused LN cells from B6 donors into MHC-mismatched, pre-irradiated CByB6F1 recipients (Figure 1A) and induced severe BM failure in recipient animals with dramatically decreased white blood cells (WBCs), red blood cells (RBCs), platelets (PLT), and total BM cells on d13 after LN cell infusion (Figure 1B). Treatment with rapamycin at 2 mg/kg/day for 13 d (d0-12) preserved BM cellularity and ameliorated peripheral blood pancytopenia (Figure 1B). Relative to TBI and normal controls, BM failure mice had marked expansion of T cells in the BM, while rapamycin eliminated most BM-infiltrating T cells, especially CD8+ T cells (Figure 1C). Using DsRed mice on the B6 background as donors, we found that BM-infiltrating lymphocytes were essentially donor-derived (Figure 1D), consistent with previous observations.19 In TBI and normal control mice, BM had dense DAPI staining to show high cellularity with megakaryocytes scattered throughout the BM cavity; no DsRed LN cells were present (Figure 1D, upper panel). In untreated BM failure mice, the marrow cavity was infiltrated with DsRed LN cells with reduced BM cellularity and no visible megakaryocytes. In contrast, rapamycin-treated BM failure mice had very few DsRed LN cells in the BM, much higher BM cellularity, and abundant megakaryocytes (Figure 1D, lower panel).

To assess whether rapamycin has comparable efficacy to CsA, a standard treatment for AA patients and murine BM failure models (50 mg/kg, i.p. d0-9),5,6 we tested different regimens of rapamycin and found that all rapamycin treatments were effective, with the d0-12 treatment group produced optimal improvements in BM cellularity and peripheral blood counts relative to the standard CsA d0-9 treatment group (Figure 2A). In parallel survival experiments, untreated BM failure mice died within three weeks after LN cell infusion while short-term treatments with rapamycin d0-4 were ineffective and animals haematologica | 2017; 102(10)

Rapamycin in murine bone marrow failure

died within one month (Figure 2B). In contrast, rapamycin treatment regimens of d0-9, d0-12 and d3-12 were all effective and maintained animal survival to more than 100 days, comparable to CsA d0-9 treatment (P<0.0001 vs. untreated BMF) (Figure 2B). We measured complete blood counts (CBCs) in TBI, rapamycin d0-9 and d0-12 groups at 28, 42 and 100 d. RBC and PLT counts in rapamycin-treated mice recovered to similar levels as in TBI controls at all

three time points; recovery of WBC was slower until 42 days but reached TBI control levels at 100 d. BM cellularity in rapamycin-treated mice increased to similar levels as in TBI controls when the mice were euthanized and evaluated at 100 d (Figure 2C).

To mimic human disease treatment, we further delayed initiation of rapamycin injection until d5 (Rapa-D5) or d7 (Rapa-D7) post LN infusion and found that these delayed

A

B

C

D

Figure 1. Rapamycin ameliorates bone marrow (BM) failure (BMF). (A) CByB6F1 mice received 5 Gy total body irradiation (TBI) and infusion of 5x106 B6 lymph node (LN) cells to induce BMF. Some BMF mice received daily rapamycin injections at 2 mg/kg/day for 13 days (d). Mice were bled and euthanized at d13 for analyses.

(B) BM nucleated cell number, white blood cell (WBC), red blood cell (RBC) and platelets (PLT) counts in normal (n=5), TBI (n=6), BMF (n=5), and rapamycin-treated mice (n=9). (C) Flow cytometric analyses of CD8+ and CD4+ T cells infiltrating BM. (D) Confocal images of sternums. Specimens were fixed with 4% paraformaldehyde and subsequently stained for megakaryocytes with anti-mouse CD41-FITC (green) and for nuclei with DAPI (blue), respectively. DsRed LN cells are visible (red) without staining. Representative images from 3 mice/group are shown. Original magnification, x100. Rapa: rapamycin. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

haematologica | 2017; 102(10)

1693

X. Feng et al.

A

B

C

D

E

Figure 2. Efficacy of rapamycin in treatment of bone marrow (BM) failure. (A) Comparison of BM cell number, white blood cell (WBC) and platelet (PLT) count in different regimens of rapamycin at day (d)13. CByB6F1 BM failure (BMF) mice were treated with cyclosporine (CsA, 50 mg/kg/day) at the standard duration of d0-9, or with rapamycin (2 mg/kg/day) d0-9, d0-12, and d3-12, respectively. n=10 for each group, except n=5 for total body irradiation (TBI). (B) Survival curves using different treatment regimens. TBI, Rapa d0-9, Rapa d0-12, Rapa d3-12, CsA d0-9 versus BMF or Rapa d0-4; P<0.0001. n=8 for each group, except n=4 for TBI. (C) Recovery of complete blood counts in rapamycin-treated BMF mice at d28, d42 and d100 (n=8 for Rapa d0-9 and d0-12, respectively; n=4 for TBI control). Normal mice were used as a reference (n=3). BM cellularity was evaluated at 100 days. (D) Delayed rapamycin treatment to d5 (Rapa-D5, n=5) and d7 (Rapa-D7, n=5) following BMF induction (n=4) also preserved BM cells and alleviated pancytopenia when animals were evaluated at d18. n=5 for TBI. (E) Delayed rapamycin treatment to d5 (Rapa-D5, n=5) led to 100% animal survival with blood and BM cellularity comparable to that of TBI control (n=3) when animals were analyzed at ten weeks.

Rapa: rapamycin. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

1694

haematologica | 2017; 102(10)

Rapamycin in murine bone marrow failure

rapamycin treatments also ameliorated pancytopenia and BM failure, with a better response in the Rapa-D5 than the Rapa-D7 group (Figure 2D). Furthermore, the efficacy of the delayed treatment (Rapa-D5) was sustainable, with BM cellularity and CBCs reaching similar levels to TBI controls at ten weeks post-BM failure, although the recovery of WBC was slower than that of TBI controls (Figure 2E). We also tested the efficacy of rapamycin in another BM failure model by infusing LN cells from FVB donors into sublethally-irradiated B6 recipients20 (Online Supplementary Figure S1A). Rapamycin eliminated CD8+

A

T cells in BM, increased the CD4/CD8 T-cell ratio, and improved animal survival, comparable to the efficacy of CsA (Online Supplementary Figure S1B-E), indicating that rapamycin is effective in ameliorating BM failure independent of strains.

Rapamycin expands regulatory T cells and modulates T-cell function Based on the optimal efficacy and survival of rapamycin d0-12 mice, we chose this regimen as standard treatment relative to the previously-established CsA d0-9 treatment

E

B

C

D

Figure 3. Rapamycin expands regulatory T cells and suppresses effector T-cell function. (A) Both cyclosporine (CsA) and rapamycin were efficient in improving total bone marrow (BM) cell number and in decreasing frequencies of CD3+, CD8+, and CD4+ T cells in total BM cells (n=10 for each group, except n=5 for cyclosporine).

(B) Rapamycin increased regulatory T cells in the spleen of BM failure (BMF) mice compared with untreated and CsA-treated mice, but comparable with total body irradiation (TBI, n=3) and normal (n=3) mice. The regulatory T cells (CD4+CD25+) in rapamycin-treated mice were FACS-sorted, and were tested for functionality in suppressing proliferation of CFSE-labeled effector T cells upon T-cell receptor (TCR) stimulation on day (d)5. Red graph represents effector T cells alone; dashed open graph with regulatory T cells from normal mice; gray closed graph with regulatory T cells from rapamycin-treated mice. (C) Transcriptome analyses of BM-infiltrating CD8+ and CD4+ T cells using a PCR-based array. Each row represents one pooled population of CD8+ or CD4+ T cells from the same groups; each column represents one gene with more than 2-fold differences between any two groups. Blue: low expression level; red: high expression level. (D) Validation of CD11a (Itgal) and granzyme B expression in BM-infiltrating T cells in rapamycin-treated or CsA-treated mice by flow cytometry (n=3 for each group). (E) Plasma was collected from peripheral blood of BMF mice (n=10) and rapamycin-treated (n=10) or CsA-treated (n=5) mice at d13 post lymph node (LN) injection and analyzed for T-cell-related cytokines by Luminex. Rapa: rapamycin. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

haematologica | 2017; 102(10)

1695

X. Feng et al.

in follow-up studies. In a new experiment of B6==>CByB6F1 LN cell infusion-induced AA, both CsA and rapamycin mitigated BM failure by suppressing CD8+ and CD4+ T-cell infiltration in the BM. In the spleen, rapamycin exerted greater suppression of CD8+ T cells than did CsA, though both drugs preserved CD4+ T-cell numbers (Figure 3A). Strikingly, the spleens of rapamycintreated mice had a much higher proportion of CD4+CD25+FoxP3+ regulatory T (Treg) cells than in CsAtreated or untreated BM failure animals (Figure 3B), which was also observed in the BM and LN (Online Supplementary Figure S2). Isolated Tregs from the spleens of rapamycintreated mice were capable of suppressing the proliferation of effector cells reflected by decreased CFSE dilution after T-cell receptor (TCR) stimulation, similar to Tregs derived from normal mice (Figure 3B), suggesting that the expanded Tregs by rapamycin are functionally competent.

We sorted BM CD4+ and CD8+ T cells from BM failure mice with or without rapamycin or CsA treatment and performed transcriptome analyses focusing on genes related to T-cell function using a PCR-based array. Genes showing more than 2-fold changes in expression between any two groups (BM failure, CsA, and rapamycin) were

plotted in a heat map. In CD8+ T cells, rapamycin suppressed expression of Icam1 and Tnfsf14, whereas CsA suppressed expression of Lgals3 (Figure 3C, left). In CD4+ T cells, rapamycin suppressed expression of Cd27, Lgals3, Il10ra, Itgal, Tbx21, Gzmb, Tnfsf14, and Cd70 but increased expression of Il-2ra, Tnfrsf8, and Il-4, while CsA downregulated the expression of Cd70 only without affecting expression of other genes (Figure 3C, right). Thus, rapamycin and CsA appeared to affect different molecular pathways while modulating immune activity. Consistent with our transcriptome analyses, we found that rapamycin, but not CsA, reduced CD11a, a protein encoded by the Itgal gene, on the cell surface of both BM CD4+ and CD8+ T cells (Figure 3D, left). More strikingly, rapamycin abrogated intracellular granzyme B expression in residual CD4+ and CD8+ T cells in BM. In contrast, CsA increased the frequency of granzyme B-secreting T cells (Figure 3D, right).

Plasma cytokine levels were measured by Luminex assays to further evaluate differences in the T-cell secretory profiles of rapamycin-treated and CsA-treated mice (Figure 3E). Plasma MIP1β, Fas, Fas ligand, and TNF-α were reduced by both CsA and rapamycin, but the reduc-

A

B

C

Figure 4. Rapamycin preserves hematopoietic stem and progenitor cells in bone marrow (BM) failure mice. (A) Representative flow cytometry analyses of BM KSL (upper panel, c-kit+Scal1+ in Lin–) and KSLCD150 (lower panel, CD150+ in KSL cells) in normal mice, total body irradiation (TBI), untreated, cyclosporine (CsA)-treated, and rapamycin-treated BM failure (BMF) mice on day (d)13. (B) Frequencies and absolute numbers of total BM KSL and KSLCD150+ cells in normal mice, TBI, untreated, CsA-treated, and rapamycin-treated BMF mice. (C) Colony-forming unit (CFU) assay of BM cells from CsA-treated and rapamycin-treated BMF mice. 3x104 total BM cells were seeded into 1 mL methylcellulose culture medium; total colonies were counted after ten days. n=5 for each group. Rapa: rapamycin. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

1696

haematologica | 2017; 102(10)

Rapamycin in murine bone marrow failure

tion of these cytokines was greater with rapamycin than with CsA. Granzyme B levels were significantly reduced only by rapamycin, consistent with transcriptome and flow cytometry data (Figure 3C and D). Also, consistent with transcriptome data, CsA induced higher levels of

IFN-Îł, and IL-5, but lower levels of IL-10 and IL-4 than did rapamycin (Figure 3C). Plasma IL-2, RANTES, and sCD137 levels were similar in both CsA-treated and rapamycin-treated mice. Thus, rapamycin significantly down-regulated cytokines related to Th1 immune

A

B

C

D

E

Figure 5. Eradication of antigen-specific T cells by rapamycin in a minor histocompatibility antigen mismatched bone marrow (BM) failure (BMF) model. (A) C.B10 mice received 5 Gy total body irradiation (TBI) and infusion of 5x106 DsRed B6 lymph node (LN) cells to induce BM failure (BMF). Some BMF mice received daily rapamycin injections at 2 mg/kg/day or cyclosporine (CsA) at 50 mg/kg/day for ten days. Mice were bled and euthanized at day (d)12 for analyses. (B) BM cell number, white blood cell (WBC), red blood cell (RBC), and platelet (PLT) counts in TBI control (n=4), BMF (n=7), CsA-treated (n=7) and rapamycin-treated (n=9) mice. (C) Frequency and absolute number of DsRed CD3+ T cells in total BM. (D) Proportion of CD8+ and CD4+ T cells in DsRed T cells. (E) Frequency and absolute number of H60-specific CD8+ T cells in DsRed CD3+ T cells. Representative flow cytometry analyses are shown. Rapa: rapamycin. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

haematologica | 2017; 102(10)

1697

X. Feng et al.

responses, such as IFN-γ, and up-regulated cytokines related to Th2 immune responses, such as IL-10. CsA produced a different profile, more similar to that of Th1-like cytokine than rapamycin.

A

Rapamycin preserves HSPCs We examined the efficacy of rapamycin and CsA in preserving HSPCs as defined by KSL (c-Kit+Sca1+Lin–) and KSL-SLAM (c-Kit+Sca1+Lin–CD150+) markers. Rapamycin

C

B

Figure 6. Effects of rapamycin (Rapa) and cyclosporine on T-cell activation, proliferation and T-cell phenotype post-T-cell receptor (TCR) stimulation in vitro. (A) Suppression of CD25 expression by cyclosporine (CsA) and rapamycin upon TCR stimulation on day (d)1, and increase of CD25 expression by rapamycin on d3. (B) Representative flow cytometric figures and statistics of different memory phenotypes based on CD44 and CD62L expression in CsA-treated and rapamycin-treated lymph node (LN) T cells on d3. (C) Apoptosis in different T-cell phenotypes in CsA and rapamycin-treated LN T cells on d2. n=3 for each group. (D) Effect of rapamycin and CsA on T-cell proliferation in total CD4+ and CD8+ T cells. Rapa: rapamycin; unsti: unstimulated; stimu: stimulated. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

1698

haematologica | 2017; 102(10)

Rapamycin in murine bone marrow failure

increased the frequency of the c-Kit+Sca-1+ population in Lin– cells compared to untreated or CsA-treated BM failure mice (Figure 4A, upper panel). Frequencies of CD150+ cells in KSL cells were similar in each group (30-50%) but lower in the BM failure group (Figure 4A, lower panel).

Calculating total BM cells, rapamycin increased frequencies and absolute numbers of both KSL and KSLCD150+ cells as compared with other groups (Figure 4B). When BM cells were cultured in vitro in semisolid medium, BM cells from rapamycin-treated BM failure mice formed significantly more total colonies than did cells from CsAtreated or untreated BM failure mice (Figure 4C). Thus, rapamycin effectively preserved HSPCs in animals with immune-mediated BM failure, with an efficacy comparable to that of standard CsA treatment.

Rapamycin eradicates antigen-specific effector T cells The effectiveness of rapamycin was further evaluated in the B6-DsRed==>C.B10 LN cell infusion-induced BM failure model, in which C.B10 mice express the dominant minor-H antigen H60 and B6 mice carry a null allele at H60 locus, allowing us to detect antigen-specific T cells5 (Figure 5A). We treated BM failure C.B10 mice with CsA or rapamycin for ten days and evaluated their effects on H60-specific T cells. At d12, untreated BM failure mice had decreased BM cellularity and peripheral blood cell counts. Both CsA and rapamycin ameliorated pancytopenia and improved BM cellularity (Figure 5B). Flow cytom-

etry analyses show that BMs of untreated BM failure mice were infiltrated with large numbers of DsRed donor T cells, but both rapamycin and CsA markedly reduced the proportion of these cells in BM (Figure 5C). Most infiltrating DsRed cells in untreated BM failure mice were CD8+ T cells; rapamycin treatment eliminated almost all of the CD8+ T cells within DsRed population and more than 90% of the remaining DsRed cells were CD4+ T cells.

While suppressing donor DsRed cell infiltration to a similar extent as that achieved by rapamycin, CsA maintained a substantial proportion of CD8+ T cells in the residual DsRed cells (Figure 5D). As expected, a large clone of H60specific CD8+ T cells had infiltrated the BM of untreated AA mice. In rapamycin-treated mice, very few H60-specific CD8+ T cells in DsRed cells were present in recipient animals, and after CsA treatment, H60-specific CD8+ T cells were reduced in number but could be observed in host marrows (Figure 5E).

Rapamycin eliminates memory-like and effector T cells but retains naïve T cells in vitro To understand the mechanisms related to the different effects of CsA and rapamycin on T cells, we investigated T-cell activation, proliferation, and phenotypic differentiation. Stimulating mouse LN cells in vitro with anti-mouse CD3 and CD28 antibodies induced T-cell activation. Both CsA and rapamycin inhibited T-cell activation on d1, with reduced CD25 expression following TCR stimulation. On

A

B

Figure 7. Preservation of naïve T cells and elimination of active effector T cells by rapamycin after in vivo T-cell activation. (A) Representative flow cytometry figures and statistics of different cell phenotypes of spleen CD4+ and CD8+ T cells from CByB6F1 mice that received B6 lymph node (LN) injection (10x106) and subsequent treatment with rapamycin (Rapa, 2 mg/kg) or cyclosporine (CsA, 50 mg/kg) via i.p. for three days. Statistics of two major populations CD44–CD62L– (active effector) and CD44–CD62L+ (naïve) are shown. (B) CD62L and CD44 expression in spleen CD4+ and CD8+ T cells in CsA-treated and rapamycin-treated CByB6F1 mice after LN injection. n=5 for each group. Ctrl: LN injection without treatment. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

haematologica | 2017; 102(10)

1699

X. Feng et al.

d3, however, expression of CD25 was increased in CD4+ T cells following rapamycin exposure (Figure 6A), probably due to the Treg-stimulatory effect of rapamycin.

Activation generally results in T-cell differentiation into different phenotypes. We determined T-cell phenotypes after exposure to rapamycin or CsA in combination with anti-mouse CD3 and CD28 mAb in vitro. CD4+ and CD8+ T cells without stimulation retained a naïve phenotype (CD44–CD62L+), but effector memory (CD44+CD62L–) and central memory (CD44+CD62L+) phenotypes were induced by stimulation with CD3 and CD28 antibodies.

TCR-stimulated T cells exposed to CsA or rapamycin showed phenotypes similar to stimulated T cells on d1, but a substantial proportion of CD4+ and CD8+ T cells treated with rapamycin displayed markers of a naïve phenotype on d3 after stimulation. In the presence of CsA, however, T cells retained their central memory phenotype

(Figure 6B). Rapamycin induced more apoptosis in central memory and effector memory CD4+ and CD8+ T cells than did TCR stimulation and CsA treatment, especially in CD8+ T cells (Figure 6C).

To investigate the effects of each drug on T-cell proliferation, we measured Ki67 expression in TCR-stimulated T cells on d2 by flow cytometry. Both CsA and rapamycin suppressed Ki67 frequencies in CD4+ and CD8+ T cells with respect to TCR-stimulation controls (Figure 6D).

Rapamycin eliminates effector T cells and retains naive T cells in vivo To test if rapamycin and CsA have the same effect on Tcell differentiation in vivo as they do in vitro, we injected B6 LN cells into unirradiated CByB6F1 mice to activate the host immune system, with or without administration of rapamycin (2 mg/kg, i.p.) or CsA (50 mg/kg, i.p.) for three

A

B

Figure 8. Effects of rapamycin and cyclosporine on phosphorylation of mTOR pathway components after T-cell receptor (TCR) stimulation. B6 lymph node (LN) cells were stimulated with CD3/CD28 mAb in the presence or absence of different concentrations of cyclosporine (CsA) or rapamycin for 5 hours (h) (A) and 12 h (B), respectively, and proteins were extracted, immunoblotted to detect mTOR, p-mTOR, p-S6K, p-S6, S6, p-AKT, AKT, and NFAT1, with β-actin expression used as a loading control. The bands were quantified based on β-actin levels, the proteins differentially expressed between rapamycin and CsA are shown in bar graphs. Representative figures from 3 separate experiments are shown. Bands of p-S6K at 12 h were visible when exposure time was extended. Rapa: rapamycin; no stim: no stimulation; stim: stimulation. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

1700

haematologica | 2017; 102(10)

Rapamycin in murine bone marrow failure

days. On d4, spleen cells were collected and T-cell phenotypes were evaluated by flow cytometry. Active effector (CD44–CD62L–) T cells and naïve (CD44–CD62L+) T cells differed among the three groups: LN cell injection without drug treatment induced expansion of CD44–CD62L– active effector T cells in both CD4+ and CD8+ T-cell fractions, while rapamycin reduced the frequencies of this population in both CD4+ and CD8+ T cells, and CsA reduced the effector population in CD4+ but not in CD8+ T cells.

Rapamycin induced a marked increase in CD44–CD62L+ naïve CD4+ and CD8+ T cells; CsA only slightly increased naïve CD4+ T cells (Figure 7A), confirming our observations in vitro (Figure 6B). When we measured median fluorescence intensity of CD44 or CD62L in CD4+ and CD8+ T cells, rapamycin decreased CD44 but increased CD62L fluorescence intensity compared to CsA-treated or untreated mice (Figure 7B), suggesting suppression of effector T cells and retention of naïve T cells by rapamycin.

Rapamycin and CsA act on different signaling pathways We stimulated T cells with CD3/CD28 mAb in the presence or absence of rapamycin or CsA to investigate molecular changes in mTOR and nuclear factor of activated Tcell (NFAT) signaling pathways. By immunoblot, TCR stimulation increased total mTOR and ribosomal S6 levels compared to unstimulated T cells. CsA or rapamycin exposure post-TCR stimulation did not alter the increased levels of mTOR and ribosomal S6. In contrast, rapamycin inhibited mTOR phosphorylation and almost eliminated phosphorylation of S6 kinase (S6K) and its downstream target S6, effects not seen with CsA. At 5 h, similar levels of total and phosphorylated AKT were detected in all conditions. Phosphorylated AKT was greatly increased in stimulated T cells at 12 h; rapamycin, but not CsA, almost completely suppressed phosphorylated AKT (Figure 8).

Baseline NFAT1 was observed in unstimulated T cells at both 5 and 12 h. TCR stimulation reduced cytoplasmic NFAT1 levels at 5 h, which was further decreased at 12 h.

In contrast, CsA exposure post-TCR stimulation increased cytoplasmic NFAT1 levels at 5 h, with higher NFAT1 levels becoming even more pronounced at 12 h. Rapamycin exposure appeared to increase NFAT1 levels at 5 h, but returned to similar levels as TCR stimulation alone at 12 h, indicating that rapamycin might affect NFAT1 transiently (Figure 8). Thus, rapamycin appeared to exert its immunosuppressive function by modulating mTOR activity, while CsA suppresses immune activity by interfering with the NFAT signaling pathway.

Discussion In this study, we demonstrate that rapamycin effectively and reproducibly attenuated immune-mediated BM failure in AA mouse models, with efficacy similar to that of the standard dose of CsA. Rapamycin showed high efficacy in suppressing Th1 immune responses, eradicating pathogenic CD8+ T cells, stimulating immunosuppressive Treg cells, and protecting HSPCs. Our data indicate that modulation of mTOR activity and its downstream signaling molecules is key to the therapeutic efficacy of rapamycin, which differs from CsA immunosuppressive function through modulation of NFAT pathway.

Rapamycin, but not CsA, expanded Treg cells in BM haematologica | 2017; 102(10)

failure mice. This result is consistent with previous observations that activation of mTOR suppresses FoxP3 expression and that complete inhibition of mTOR activity in CD4+ T cells stimulates the generation of Treg cells even under activating conditions.21-25 There are two distinct mTOR complexes, mTOR complex 1 (mTORC1) and mTORC2, and both contribute to suppression of FOXP3 expression.26 In our in vitro experiments, rapamycin increased CD25 expression on CD4+ T cells at d3 post TCR stimulation. Increased plasma IL-10 concentration by rapamycin may also indicate enhanced immune regulatory function of Tregs, as IL-10 is a potent suppressor of effector T-cell proliferation.27 Tregs are central to the maintenance of self-tolerance and tissue homeostasis: Treg impairment had been reported in human autoimmune and immune-mediated conditions.28-30 In patients with AA, both the frequency and absolute number of Treg cells are reduced.31 Furthermore, the superiority of horse over rabbit ATG in AA treatment correlates to preservation of Treg cells,32 specific human Treg subpopulations defined using mass cytometry may be useful as predictive biomarkers for response to IST in AA,33 addition of rapamycin to anti-CD3/CD28 beads, high-dose IL-2, and all-trans retinoic acid led to more than 30-fold expansion of Treg from AA patients.33 Thus, modulation of Treg number and function becomes an important determinant of therapeutic efficacy in AA. Rapamycin selectively eradicated effector and memory T cells. TCR stimulation in vitro or MHC-mismatched LN injection in vivo augmented differentiation of T cells from naïve to central memory, effector memory, and active effector phenotypes. CsA maintained T cells with effector and memory phenotypes, which may provide a logical explanation for the frequent relapse of AA patients following withdrawal of CsA. In contrast, rapamycin augmented memory T-cell apoptosis, decreased effector and memory T cells, and increased naïve T cells. The efficacy of rapamycin on effector T cells was best demonstrated in the C.B10 BM failure mouse model in which rapamycin eradicated almost all H60-specific T cells. Our observations reflect an early report in a heart transplantation model in which rapamycin alone or in combination with co-stimulation blockade eradicated effector and memory T cells and induced immune tolerance.34 Alteration of mTOR activity by rapamycin affects T-cell differentiation, since deficiency of an important mTORC1 activator RHEB (Ras homolog enriched in brain) in CD4+ T cells leads to normal Th2 but impaired Th1 and Th17 effector lineages through reduced activation of STAT4 and STAT3.22 Loss of STAT4 and STAT3 is also correlated with diminished expression of T-bet, a Th1 master transcription factor, and RORγt, a Th17 master transcription factor.35 Our observation of reduced Tbet and Granzyme B in BM CD4+ T cells from rapamycin-treated mice is consistent with suppression of Th1 immune function. Thus, rapamycin not only induced Tregs and eradicated memory and effector T cells, but also skewed T-cell differentiation away from pro-inflammatory Th1 and Th17 subsets.

In T cells, mTOR bridges immune signals and metabolic cues to regulate T-cell responses. Active effector T cells preferentially utilize aerobic glycolysis to meet energy demands associated with expansion of cell numbers, cytolytic activity, and homing.36,37 Within the two mTOR complexes, the mTORC1-S6K1 axis is a crucial determinant of T-cell activation.38 Although AKT is not down1701

X. Feng et al.

stream of the mTORC1 complex, we observed AKT phosphorylation to be abrogated by rapamycin but not by CsA. This effect was observed at 12 h but not at 5 h postTCR stimulation, secondary to S6 phosphorylation, suggesting that mTORC2 inhibition requires prolonged treatment with rapamycin despite the fact that rapamycin primarily affects mTORC1.35,39 The PI3K-Akt-mTORC1S6K1/2 axis was reported to control Th17 cell differentiation via downregulation of Gfi1, a negative regulator of Th17 differentiation.40 There is an increased ratio of Th17 to regulatory T cells in AA,41 and the inhibitory effect of rapamycin on Th17 differentiation may be of benefit to AA patients treated with rapamycin.

The effectiveness of rapamycin in BM failure mouse models is relevant to the establishment of a clinical trial for AA patients. However, results from our previous randomized clinical trial suggested that treatment of AA patients with horse-ATG/CsA/rapamycin produced no improved beneficial effect than the standard horseATG/CsA treatment.42 CsA has been reported to prevent the induction of the cytosolic branched chain aminotransferase (BCATc). Blockade of BCATc increases phosphorylation of mTORC1 downstream targets, S6 and 4EBP-1,43 which counteracts the inhibitory effect of rapamycin on

References 1. Young NS. Current concepts in the pathophysiology and treatment of aplastic anemia. Hematology Am Soc Hematol Educ Program. 2013;2013:76-81.

2. Scheinberg P, Rios O, Scheinberg P, et al.

Prolonged cyclosporine administration after antithymocyte globulin delays but does not prevent relapse in severe aplastic anemia. Am J Hematol. 2014;89(6):571574.

3. Townsley DM, Scheinberg P, Winkler T, et al. Eltrombopag Added to Standard Immunosuppression for Aplastic Anemia.

N Engl J Med. 2017;376(16):1540-1550.

4. 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.

5. Chen J, Ellison FM, Eckhaus MA, et al.

Minor antigen h60-mediated aplastic anemia is ameliorated by immunosuppression and the infusion of regulatory T cells. J Immunol. 2007;178(7):4159-4168.

6. 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.

7. Donia M, Mangano K, Amoroso A, et al.

Treatment with rapamycin ameliorates clinical and histological signs of protracted relapsing experimental allergic encephalomyelitis in Dark Agouti rats and induces expansion of peripheral CD4+CD25+Foxp3+ regulatory T cells. J Autoimmun. 2009;33(2):135-140.

8. Zheng P, Chang X, Lu Q, Liu Y. Cytopenia and autoimmune diseases: a vicious cycle

1702

9.

10.

11.

12.

13.

14.

15.

16.

17.

mTORC1, possibly explaining the ineffectiveness of concurrent CsA/rapamycin treatment in AA. Data from our current animal study showed the importance of treatment duration, as short treatments with either CsA or rapamycin were not effective. We selected our CsA treatment duration based on its effectiveness from our previous experience, since longer-term CsA has produced adverse effects in mice, including hunching and scruffy appearance, weight loss, and death, probably due to nephrotoxicity and hepatotoxicity.44,45 In contrast, mice treated with rapamycin appeared healthy and active.

In conclusion, rapamycin treated immune-mediated murine BM failure as compared to standard dose CsA. Our observations support the potential utility of rapamycin in the clinic for the treatment of AA. A phase II clinical trial application is under development at our institution (clinicaltrials.gov identifier: 02979873) to test rapamycin as prophylaxis in severe AA patients who are at high risk of relapse on withdrawal of long-term CsA. Funding This work was supported by the Intramural Research Program of the National Heart, Lung and Blood Institute, National Institutes of Health.

fueled by mTOR dysregulation in hematopoietic stem cells. J Autoimmun.

2013;41:182-187.

Lui SL, Tsang R, Chan KW, et al.

Rapamycin attenuates the severity of established nephritis in lupus-prone NZB/W F1 mice. Nephrol Dial Transplant.

2008;23(9):2768-2776.

Ramos-Barron A, Pinera-Haces C, GomezAlamillo C, et al. Prevention of murine lupus disease in (NZBxNZW)F1 mice by sirolimus treatment. Lupus. 2007; 16(10):775-781.

Laragione T, Gulko PS. mTOR regulates the invasive properties of synovial fibroblasts in rheumatoid arthritis. Mol Med. 2010;16(9-10):352-358.

Guan Y, Zhang L, Li X, et al. Repression of Mammalian Target of Rapamycin Complex 1 Inhibits Intestinal Regeneration in Acute Inflammatory Bowel Disease Models. J Immunol. 2015;195(1):339-346.

Ytting H, Larsen FS. Everolimus treatment for patients with autoimmune hepatitis and poor response to standard therapy and drug alternatives in use. Scand J Gastroenterol. 2015;50(8):1025-1031.

Ibrahim MA, Sepah YJ, Watters A, et al.

One-Year Outcomes of the SAVE Study: Sirolimus as a Therapeutic Approach for UVEitis. Transl Vis Sci Technol. 2015;4(2):4.

Leventhal JR, Mathew JM, Salomon DR, et al. Nonchimeric HLA-Identical Renal Transplant Tolerance: Regulatory Immunophenotypic/Genomic Biomarkers.

Am J Transplant. 2016;16(1):221-234.

Hlavaty KA, Luo X, Shea LD, Miller SD. Cellular and molecular targeting for nanotherapeutics in transplantation tolerance.

Clin Immunol. 2015;160(1):14-23.

Hsieh MM, Kang EM, Fitzhugh CD, et al.

Allogeneic hematopoietic stem-cell trans-

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

plantation for sickle cell disease. N Engl J Med. 2009;361(24):2309-2317.

Huang J, Nguyen-McCarty M, Hexner EO, Danet-Desnoyers G, Klein PS. Maintenance of hematopoietic stem cells through regulation of Wnt and mTOR pathways. Nat Med. 2012;18(12):17781785.

Takaku T, Malide D, Chen J, et al.

Hematopoiesis in 3 dimensions: human and murine bone marrow architecture visualized by confocal microscopy. Blood.

2010;116(15):e41-55.

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.

Sauer S, Bruno L, Hertweck A, et al. T cell receptor signaling controls Foxp3 expression via PI3K, Akt, and mTOR. Proc Natl Acad Sci USA. 2008;105(22):7797-7802.

Haxhinasto S, Mathis D, Benoist C. The AKT-mTOR axis regulates de novo differentiation of CD4+Foxp3+ cells. J Exp Med. 2008;205(3):565-574.

Liu G, Burns S, Huang G, et al. The receptor S1P1 overrides regulatory T cell-mediated immune suppression through Akt-mTOR. Nat Immunol. 2009;10(7):769-777.

Liu G, Yang K, Burns S, Shrestha S, Chi H. The S1P(1)-mTOR axis directs the reciprocal differentiation of T(H)1 and T(reg) cells.

Nat Immunol. 2010;11(11):1047-1056.

Park Y, Jin HS, Lopez J, et al. TSC1 regulates the balance between effector and regulatory T cells. J Clin Invest. 2013; 123(12):5165-5178.

Delgoffe GM, Kole TP, Zheng Y, et al. The mTOR kinase differentially regulates effector and regulatory T cell lineage commitment. Immunity. 2009;30(6):832-844.

Taylor A, Verhagen J, Blaser K, Akdis M,

haematologica | 2017; 102(10)

Rapamycin in murine bone marrow failure

28.

29.

30.

31.

32.

33.

Akdis CA. Mechanisms of immune suppression by interleukin-10 and transforming growth factor-beta: the role of T regulatory cells. Immunology. 2006;117(4):433-442.

Venken K, Thewissen M, Hellings N, et al.

A CFSE based assay for measuring CD4+CD25+ regulatory T cell mediated suppression of auto-antigen specific and polyclonal T cell responses. J Immunol Methods. 2007;322(1-2):1-11.

Li Z, Arijs I, De Hertogh G, et al. Reciprocal changes of Foxp3 expression in blood and intestinal mucosa in IBD patients responding to infliximab. Inflamm Bowel Dis. 2010;16(8):1299-1310.

Longhi MS, Ma Y, Mitry RR, et al. Effect of CD4+ CD25+ regulatory T-cells on CD8 Tcell function in patients with autoimmune hepatitis. J Autoimmun. 2005;25(1):63-71.

Solomou EE, Rezvani K, Mielke S, et al.

Deficient CD4+ CD25+ FOXP3+ T regulatory cells in acquired aplastic anemia.

Blood. 2007;110(5):1603-1606.

Scheinberg P, Nunez O, Weinstein B, et al.

Horse versus rabbit antithymocyte globulin in acquired aplastic anemia. N Engl J Med. 2011;365(5):430-438.

Kordasti S, Costantini B, Seidl T, et al.

Deep-phenotyping of Tregs identifies an immune signature for idiopathic aplastic

haematologica | 2017; 102(10)

34.

35.

36.

37.

38.

39.

anemia and predicts response to treatment.

Blood. 2016;128(9):1193-1205.

Li Y, Li XC, Zheng XX, et al. Blocking both signal 1 and signal 2 of T-cell activation prevents apoptosis of alloreactive T cells and induction of peripheral allograft tolerance.

Nat Med. 1999;5(11):1298-1302.

Delgoffe GM, Pollizzi KN, Waickman AT, et al. The kinase mTOR regulates the differentiation of helper T cells through the selective activation of signaling by mTORC1 and mTORC2. Nat Immunol.

2011;12(4):295-303.

Jones RG, Thompson CB. Revving the engine: signal transduction fuels T cell activation. Immunity. 2007;27(2):173-178.

Wang R, Dillon CP, Shi LZ, et al. The transcription factor Myc controls metabolic reprogramming upon T lymphocyte activation. Immunity. 2011;35(6):871-882.

Salmond RJ, Brownlie RJ, Meyuhas O, Zamoyska R. Mechanistic Target of Rapamycin Complex 1/S6 Kinase 1 Signals Influence T Cell Activation Independently of Ribosomal Protein S6 Phosphorylation. J Immunol. 2015;195(10):4615-4622.

Sarbassov DD, Ali SM, Sengupta S, et al.

Prolonged rapamycin treatment inhibits mTORC2 assembly and Akt/PKB. Mol Cell. 2006;22(2):159-168.

40. Kurebayashi Y, Nagai S, Ikejiri A, et al.

PI3K-Akt-mTORC1-S6K1/2 axis controls Th17 differentiation by regulating Gfi1 expression and nuclear translocation of RORgamma. Cell Rep. 2012;1(4):360-373.

41. de Latour RP, Visconte V, Takaku T, et al.

Th17 immune responses contribute to the pathophysiology of aplastic anemia. Blood.

2010;116(20):4175-4184.

42. Scheinberg P, Wu CO, Nunez O, et al.

Treatment of severe aplastic anemia with a combination of horse antithymocyte globulin and cyclosporine, with or without sirolimus: a prospective randomized study.

Haematologica. 2009;94(3):348-354.

43. Ananieva EA, Patel CH, Drake CH, Powell JD, Hutson SM. Cytosolic branched chain aminotransferase (BCATc) regulates mTORC1 signaling and glycolytic metabolism in CD4+ T cells. J Biol Chem.

2014;289(27):18793-18804.

44. Naesens M, Kuypers DR, Sarwal M. Calcineurin inhibitor nephrotoxicity. Clin J Am Soc Nephrol. 2009;4(2):481-508.

45. Welz A, Reichart B, Uberfuhr P, Kemkes B, Klinner W. Cyclosporine as the main immunosuppressant in clinical heart transplantation: correlation of hepatotoxicity and nephrotoxicity. Transplant Proc.

1984;16(5):1212-1213.

1703

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Chronic Myeloid Leukemia

Ferrata Storti Foundation

Haematologica 2017 Volume 102(10):1704-1708

Prognostic discrimination based on the EUTOS long-term survival score within the International Registry for Chronic Myeloid Leukemia in children and adolescents Frédéric Millot,1 Joëlle Guilhot,1 Meinolf Suttorp,2 Adalet Meral Güneş,3 Petr Sedlacek,4 Eveline De Bont,5 Chi Kong Li,6 Krzysztof Kalwak,7 Birgitte Lausen,8 Srdjana Culic,9 Michael Dworzak,10 Emilia Kaiserova,11 Barbara De Moerloose,12 Farah Roula,13 Andrea Biondi14 and André Baruchel15

1 Inserm CIC 1402, University Hospital, Poitiers, France; 2Department of Pediatrics, University Hospital Carl Gustav Carus, Dresden, Germany; 3Department of Pediatric Hematology, Uludağ University Hospital, Görükle Bursa, Turkey; 4Department of Pediatric Hematology–Oncology, University Hospital Motol, Charles University, Prague, Czech Republic; 5Department of Pediatric Oncology/Hematology, University Medical Center Groningen, University of Groningen, the Netherlands, and Dutch Childhood Oncology Group, the Hague, the Netherlands; 6Department of Pediatrics, Prince of Wales Hospital, The Chinese University of Hong Kong, China; 7Department of Pediatric Hematology Oncology and Transplantation, Wroclaw Medical University, Poland; 8 Department of Pediatrics, Rigshospitalet, University Hospital, Copenhagen, Denmark; 9 Department of Pediatric Hematology Oncology Immunology and Medical Genetics, Clinical Hospital Split, Croatia; 10Children's Cancer Research Institute and St. Anna Children's Hospital, Vienna, Austria; 11Department of Pediatric Oncology of University Children's Hospital, Bratislava, Slovakia; 12Department of Pediatrics, Ghent University Hospital, Belgium; 13Department of Pediatrics, Saint George Hospital University Medical Center, Beirut, Lebanon; 14Department of Pediatrics, University of Milano-Bicocca, San Gerardo Hospital, Fondazione MBBM, Monza, Italy and 15Department of Pediatric Hematology, Robert Debré Hospital, Paris, France

ABSTRACT

Correspondence: f.millot@chu-poitiers.fr

Received: March 30, 2017.

Accepted: August 17, 2017.

Pre-published: August 24, 2017. doi:10.3324/haematol.2017.170035 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1704 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1704

T

he EUTOS Long-Term Survival score was tested in 350 children with chronic myeloid leukemia in first chronic phase treated with imatinib and registered in the International Registry for Childhood Chronic Myeloid Leukemia. With a median follow up of 3 years (range, 1 month to 6 years) progression and/or death (whichever came first) occurred in 23 patients. For the entire cohort of patients the 5-year progression-free survival rate was 92% (95% CI: 87%-94%) and the 5-year survival accounting for chronic myeloid leukemia deaths was 97% (95% CI: 94%-99%). Of the 309 patients allocated to low (n=199), intermediate (n=68) and high (n=42) risk groups by the EUTOS Long-Term Survival score, events (progression and/or death) occurred in 6.0%, 8.8% and 26.2%, respectively. Estimates of the 5-year progression-free survival rates according to these three risk groups were 96% (95% CI: 92%98%), 88% (95% CI: 76%-95%) and 67% (95% CI: 48%-81%), respectively. Differences in progression-free survival according to these risk groups were highly significant (P<0.0001, overall). The EUTOS LongTerm Survival score showed better differentiation of progression-free survival than the Sokal (<45 years), Euro and EUTOS scores in children and adolescents with chronic myeloid leukemia and should be considered in therapeutic algorithms. (Trial registered at: www.clinicaltrials.gov NCT01281735) Introduction Prognostic scores such as the Sokal score, the Euro score and the EUTOS score based on clinical and biological features at diagnosis have proven their usefulness in predicting the outcome of adults receiving defined treatment for chronic myeloid leukemia (CML).1-3 While the Sokal score for patients less than 45 years haematologica | 2017; 102(10)

Pertinence of ELTS score in children with CML Table 1. Probabilities of progression-free survival and survival accounting for competing events in children with chronic myeloid leukemia treated with imatinib.

Prognostic score

Number of cases n (%)

Sokal young score Low risk 54 (18%) Intermediate risk 118 (38%) High risk 137 (44%) Missing 41 Euro score Low risk 165 (55%) Intermediate risk 103 (34%) High risk 35 (12%) Missing 47 EUTOS score Low risk 238 (78%) High risk 68 (22%) Missing 44 EUTOS Long-Term Survival score Low risk 199 (64%) Intermediate risk 68 (22%) High risk 42 (14%) Missing 41

Progression or death (% risk group)

5-year PFS estimate (95% CI)

3 (5.5%) 6 (5%) 14 (10.2%)

93% (81-98) 94% (87-97) 87% (79-92)

9 (5.5%) 9 (8.7%) 5 (14.3%)

94% (88-97) 89% (79-94) 81% (60-92)

13 (5.5%) 10 (14.7%)

93% (88-96) 81% (67-89)

6 (3%) 6 (8.8%) 11 (26.2%)

96% (92-98) 88% (76-95) 67% (49-82)

P

CML Competing deaths events (% risk group) (% risk group)

5-year survival estimate (95% CI) accounting for CML deaths

P

P=0.279

0 2 (1.7%) 3 (2.2%)

2 (3.7%) 2 (1.7%) 3 (2.2%)

100% 97% (92-100) 96% (89-100)

P=0.576

P=0.211

2 (1.2%) 1 (1%) 2 (5.7%)

3 (1.8%) 4 (3.9%) 0

98% (94-100) 99% (93-100) 87% (75-98)

P=0.182

3 (1.3%) 2 (3%)

4 (1.7%) 3 (4.4%)

98% (93-99) 94% (83-99)

1 (0.5%) 2 (2.9%) 2 (4.8%)

2 (1%) 3 (4.4%) 2 (4.8%)

99% (95-100) 96% (88-99) 89% (70-98)

P=0.009

P<0.0001

P=0.340

P=0.107

CI: confidence interval; CML: chronic myeloid leukemia; PFS: progression-free survival. Because of some lacking data (spleen size n=31; platelet count n=1; eosinophil count n = 11; basophil count n=15 or blast and myeloblast percentage n=11) determination of at least one prognostic score was not possible in a total of 48 children (all scores and EUTOS Long-Term Survival score were missing in 38 and 41 of them, respectively). All patients with critical events (progression and/or deaths) were assessable for the calculation of the risk score.

old and the Euro score were defined in cohorts of patients including children, the usefulness of these prognostic scores has not been formally established in the pediatric population.4 Limited data are available regarding the utility of the EUTOS score in the pediatric population.5 Recently, a new EUTOS score, the EUTOS Long-Term Survival (ELTS) score was validated in the adult population and showed better discrimination of the probability of dying of CML than had previous prognostic scores.6 The International Registry for Chronic Myeloid Leukemia in children and adolescents (I-CML-Ped Study registered at www.clinicaltrials.gov as NCT01281735) gave us the opportunity to compare risk group allocations and outcome between these prognostic scores in the pediatric population.

Methods The I-CML-Ped Study was established to assess the epidemiology, management and outcome of CML in the pediatric population. Newly diagnosed children and adolescents less than 18 years old with Philadelphia chromosome-positive CML in chronic or advanced phase diagnosed later than January 2000 were eligible for this study. The calculations for the Sokal (for patients less than 45 years old), Euro, EUTOS and ELTS scores were performed using mathematical equations including the following parameters: sex, spleen size, hematocrit, platelets and blasts in blood for the Sokal score; age, spleen size, platelets, blasts, basophils and eosinophils in blood for the Euro score; spleen size and basophils in blood for the EUTOS score; and age, spleen size, platelets and haematologica | 2017; 102(10)

blasts in blood for the ELTS score, as previously reported.2-4,6 On the basis of the calculated scores, the children were categorized into low risk, intermediate risk or high risk groups for the Sokal (for patients less than 45 years), Euro and EUTOS scores and into low risk or high risk for the ELTS score. The phase of the disease was determined according to the European leukemiaNet (ELN) recommendations as previously reported.7 The study protocol was approved by the institutional review committee of the university hospital of Poitiers (France). Written informed consent was obtained from the children and/or their guardians. For analyses of progression-free survival, events of interest included progression to accelerated phase or blast crisis and death, irrespective of cause, whichever came first.8 For analysis of survival, the event of interest was death from CML disease, deaths from other causes being considered as competing events, as initially designed in the ELTS score model. The follow up of patients was not censored at the time of switching to other drugs or allogeneic hematopoietic stem cell transplantation (HSCT). Estimates of progression-free survival were calculated using the Kaplan-Meier method and comparisons were performed using the log-rank test. For the estimation of cause-specific death in a competing model, the Gray test was used for comparison.9 The level of statistical significance was 0.05.

Results Between January 2011 and June 2016, 350 patients with CML in chronic phase at diagnosis treated with standard dose (260 to 300 mg/m2 daily) imatinib front line were registered from 13 countries. The patientsâ&#x20AC;&#x2122; median age at diagnosis of CML was 12.2 years (range, 8 months to 18 1705

F. Millot et al.

A

B

C

D

Figure 1. Progression-free survival stratified according to risk categorization by the four scores. (A) Sokal score, (B) Euro score, (C) EUTOS score, (D) EUTOS LongTerm Survival (ELTS) score. Green represent low risk patients, orange represent intermediate risk patients and red represents high-risk patients.

years) and 56% were male; a palpable spleen was noted in 77% of the patients and the median spleen size was 5 cm (range, 0 to 32 cm) below the costal margin; the median white blood cell count and the median hemoglobin level were 228x109/L (range, 4.8x109/L to 1037x109/L) and 94 g/L (range, 31 g/L to 170 g/L), respectively.

The distribution of the children into the risk categories by the Sokal (for patients less than 45 years), Euro, EUTOS and ELTS scores is reported in Table 1. Discordant risk categorizations of the children were observed when comparing the four scores. Regarding the Sokal (for patients less than 45 years) and the ELTS scores, all the children categorized as low risk according to the Sokal system were allocated to the low-risk group according to the ELTS score.

By contrast, among the children in the intermediate-risk group according to the Sokal system, only 13% remained in the intermediate-risk group according to the ELTS score while 1% and 86% were allocated to the high-risk group and low-risk group, respectively. Among the children in the high-risk group according to the Sokal system, 30% remained in the high-risk group according to the ELTS score while 39% and 31% were allocated to the intermediate-risk group and low-risk group, respectively. The median follow up of the 350 patients in chronic phase treated with imatinib front line was 3 years (range, 1 month to 6 years). Imatinib was administered with a 1706

median observational time of 11 months (range, 1 to 131 months); 149 patients discontinued treatment with imatinib because of progression of their disease, toxicity, failure to achieve optimal response, or physicianâ&#x20AC;&#x2122;s choice (HSCT in optimal response). Progression and/or death (whichever came first) were recorded in 23 patients: progression occurred in 19 (5.4%) patients and death was recorded in 12 (3.4%) children. Among the 19 patients who progressed as first event, five patients progressed to accelerated phase and 14 to blastic phase at a median time of 12 months (range, 3 to 32 months) after diagnosis.

Eleven of these 19 children are alive including ten who were transplanted with a graft from a sibling donor (4 patients) or an unrelated donor (6 patients). The remaining 8/19 patients have died including five children who died of uncontrolled CML disease (2 children with recurrent disease after HSCT for disease progression of the disease) and three who died after HSCT because of graft-versushost disease (n=1) or infection (n=2). In addition, death occurred as the first event in four patients who were transplanted (unrelated donor 1 case, sibling donor 3 cases) in first chronic phase in accordance with the choice of the clinician. The causes of these four deaths were graft-versus-host disease (n=1) and infection (n=3). Overall, considering all 12 deaths, these occurred at a median time of 22 months (range, 12 to 56 months) after the diagnosis of haematologica | 2017; 102(10)

Pertinence of ELTS score in children with CML

CML and five were related to CML while the other seven deaths were due to post-transplant complications (graftversus-host disease 2 cases, infection 5 cases) and for this analysis were considered as non-CML-related deaths.

Overall, the 5-year overall survival rate was 94% (95% CI: 90%-97%), the 5-year progression-free survival rate was 92% (95% CI: 87%-94%) and the 5-year survival rate accounting for competing events was 97% (95% CI: 94%99%). Among the patients allocated to the low-, intermediate- and high-risk groups by the ELTS score, events (progression and/or death) occurred in 6.0%, 8.8% and 26.2%, respectively. When the patients were stratified according to the Sokal, Euro, EUTOS and ELTS scores, only the EUTOS and the ELTS scores were able to discriminate risk groups with significantly different progression-free survival (P=0.009 and P<0.0001, respectively) (Table 1, Figure 1). None of the Sokal, Euro, EUTOS and ELTS scores was able to discriminate risk groups with significant differences in survival based on CML deaths only (Table 1).

Discussion The prognosis of adult patients with CML can be predicted with established prognostic scores based on clinical (spleen size) and biological parameters. The characteristics of CML differ with age with larger spleen size and higher leukocyte count at diagnosis in the present population of children and adolescents than reported in adults with CML.10-12 Because of the rarity of CML in children, a specific prognostic score incorporating clinical, biological and molecular features has not been established for this population. The Sokal and Euro scores were developed in a cohort of patients including children with CML in the conventional chemotherapy (busulfan, hydroxyurea) and in the interferon eras, respectively.1,2 A Sokal score for young patients was established in a cohort of patients less than 45 years old and is still useful in the era of therapy with tyrosine kinase inhibitors.4,13 Subsequently, the EUTOS scoring system was introduced in adult patients treated with imatinib.3 The improved life expectancy of adults with CML treated with imatinib currently approaches that for the general population, with 41% to 44% of the deaths not directly related to CML but rather to comorbidities.6,14,15 Based on the concept of competing risks, the ELTS score was recently developed in order to consider disease-specific death in adults with CML.6 This new score differentiated the probability of dying of CML in the adult population better than did the Sokal, Euro and EUTOS scores.6 The aim of the present study was to test the relevance of the ELTS score in a large cohort of children and adolescents with CML. In the present cohort of 350 children treated with imatinib for CML in first chronic phase, the ELTS score identified a lower proportion of high-risk children than the Sokal score, as observed in adults, while the proportions of the children allocated to low-risk (64%), intermediate-risk (22%) and high-risk (14%) groups by the ELTS score were similar to the proportions reported in adults.6 The 5-year progression-free survival rate of 92% for the entire cohort of children is consistent with previous reports in children and adults with CML in chronic phase treated with imatinib front line.16-19 The recently developed ELTS score divided the children of the present study into three separate risk groups according to their progressionhaematologica | 2017; 102(10)

free survival with all risk groups differing significantly from each other. The ELTS score showed better differentiation of progression-free survival than the other scores in our cohort of children and could be used to predict the long-term outcome of children with CML in chronic phase. This finding suggests the establishment of new treatment policies with the incorporation of this score into the therapeutic algorithms of the current recommendations proposed for childhood CML.20 The high probability of progression for children allocated to the high-risk group could favor risk-adapted treatment with the use of second-generation tyrosine kinase inhibitors as first-line therapy in these patients.

The estimated 5-year overall survival rate reported in our non-selected cohort of children compares favorably with results reported in adults treated in trials with imatinib.21,22 Although children have more aggressive features at presentation compared to adults, probabilities of overall survival remain high and comparable in children, in adolescents and in young adults treated with imatinib.10,11,23 Because the improvement in the survival of patients with CML after introduction of imatinib has resulted in increased life expectancy, about half of adult patients now die of causes unrelated of CML. The main non-related CML deaths reported in adults in the tyrosine kinase inhibitor era are those due to secondary malignancies and cardiovascular events.6,24 Thus CML-related death could represent a better assessment of treatment efficacy. In the present study, the 5-year survival rate accounting for competing events of 97% corresponded to a 3% probability of death because of CML which is rather similar to the 4% probability reported in adults.6 However, in contrast to the adult study, the follow up was not censored at transplantation in the present study, consequently deaths from HSCT are competing events. The non-related CML deaths notified in the present study were due to post-transplant complications and were more common than CML as a cause of death. Thus HSCT should be reserved for cases of treatment failure in children in chronic phase, as proposed in the recommendation of the International BerlinFrankfurt-Munster study group.20 The ELTS score discriminates the probability of dying of CML better than do the Sokal, Euro and EUTOS scores in adults with CML. In the present study none of these scores was able to discriminate risk groups with significant differences in survival based on CML deaths only .The low number of events (only 5 CML-related deaths) is one of the possible explanations for these findings.

Moreover, because of the low number of comorbidities in the pediatric population, the risk of dying due to competing events is restricted to the complications of HSCT. In this pediatric cohort, the ELTS score demonstrated better differentiation of progression-free survival than did the Sokal (in patients less than 45 years old) and Euro scores in children and adolescents with CML in chronic phase treated with imatinib. We therefore propose that the ELTS score should be considered in therapeutic algorithms and clinical trials in children and adolescents.

Acknowledgments We gratefully acknowledge Professor Irene Roberts (Department of Paediatrics, University of Oxford, UK) and Professor Franรงois Guilhot (Inserm CIC 1402, Poitiers, France) for helpful comments on the manuscript. We also thank Violaine Goyeau for the data monitoring.

1707

F. Millot et al.

References 1. Sokal JE, Cox EB, Baccarani M, et al.

Prognostic discrimination in good-risk chronic granulocytic leukemia. Blood.

1984;63(4):789-799.

2. Hasford J, Pfirmann M, Hehlman R, et al. A new prognostic score for survival of patients with chronic myeloid leukemia treated with interferon alpha. J Natl Cancer Inst. 1998;90(11):850–858.

3. Hasford J, Baccarani M, Hoffman V, et al.

Predicting complete cytogenetic response and subsequent progression-free survival in 2060 patients with CML on imatinib treatment: the EUTOS score. Blood.

2011;118(3):686–692.

4. Sokal JE, Baccarani M, Tura S, et al.

Prognostic discrimination among younger patients with chronic granulocytic leukemia: relevance to bone marrow transplantation. Blood. 1985;66(6):1352-1357.

5. Gurrea Salas D, Glauche I, Tauer JT, et al.

Can prognostic scoring systems for chronic myeloid leukemia as established in adults be applied to pediatric patients? Ann Hematol. 2015;9(8):1363 -1371.

6. Pfirrmann M, Baccarani M, Saussele S, et al.

Prognosis of long-term survival considering disease-specific death in patients with chronic myeloid leukemia. Leukemia.

2016;30(1):48-56.

7. Baccarani M, Deininger MW, Rosti G, et al.

European LeukemiaNet recommendations for the management of chronic myeloid leukemia. Blood. 2013;122(6):872-884.

8. Guilhot J, Baccarani M, Clark RE, et al.

Definitions, methodological and statistical issues for phase 3 clinical trials in chronic myeloid leukemia: a proposal by the European LeukemiaNet. Blood. 2012;119 (25):5963-5971.

9. Gray RJ. A class of k-samples tests for com-

1708

10.

11.

12.

13.

14.

15.

16.

17.

paring the cumulative incidence of a competing risk. Ann Stat. 1988;16(3):1141-1154.

Castagnetti F, Gugliotta G, Baccarani M, et al. Differences among young adults, adults and elderly chronic myeloid leukemia patients. Ann Oncol. 2015;26(1):185-192.

Kalmanti L, Saussele S, Lauseker M, et al.

Younger patients with chronic myeloid leukemia do well in spite of poor prognostic indicators: results from the randomized CML study IV. Ann Hematol.

2014;93(1):71–80.

Hoffmann VS, Baccarani M, Hasford J, et al.

The EUTOS population-based registry: incidence and clinical characteristics of 2904 CML patients in 20 European countries. Leukemia. 2015;29(6):1336–1343.

Oyekunle AA, Osho PO, Aneke JC, et al.

The predictive value of the Sokal and Hasford scoring systems in chronic myeloid leukemia in the imatinib era. J Hemat Malign. 2012;2(2):25-32.

Bower H, Björkholm M, Dickman PW, Höglund M, Lambert PC, Andersson TM. Life expectancy of patients with chronic myeloid leukemia approaches the life expectancy of the general population. J Clin Oncol. 2016;34(24):2851-2857.

Hehlmann R, Müller MC, Lauseker M, et al. Deep molecular response is reached by the majority of patients treated with imatinib, predicts survival, and is achieved more quickly by optimized high-dose imatinib: results from the randomized CMLStudy IV. J Clin Oncol. 2014;32(5):415-423.

Millot F, Baruchel A, Guilhot J, et al.

Imatinib is effective in children with previously untreated chronic myelogenous leukemia in early chronic phase: results of the French national phase IV trial. J Clin Oncol. 2011;29(20):2827-2832.

Millot F, Guilhot J, Baruchel A, et al. Impact of early molecular response in children

18.

19.

20.

21.

22.

23.

24.

with chronic myeloid leukemia treated in the French Glivec phase 4 study. Blood.

2014;124(15):2408-2410.

Hehlmann R, Lauseker M, Jung-Munkwitz S, et al. Tolerability-adapted imatinib 800 mg/d versus 400 mg/d versus 400 mg/d plus interferon-α in newly diagnosed chronic myeloid leukemia. J Clin Oncol.

2011;29(12):1634-1642.

Hoffmann VS, Baccarani M, Hasford J, et al.

Treatment and outcome of 2904 CML patients from the EUTOS population-based registry. Leukemia. 2017;31(3):593–601.

De la Fuente J, Baruchel A, Biondi A, et al.

How I manage CML in children - guidelines for the management of chronic myeloid leukaemia in children and young people up to the age of 18 years. Br J Haematol. 2014;167(1):33-47.

Cortes JE, Saglio G, Kantarjian HM, et al.

Final 5-year study results of DASISION: the dasatinib versus imatinib study in treatment-naïve chronic myeloid leukemia patients trial.

J Clin Oncol.

2016;34(20):2333-2340.

Hochhaus A, Saglio G, Hughes TP, et al.

Long-term benefits and risks of frontline nilotinib vs imatinib for chronic myeloid leukemia in chronic phase: 5-year update of the randomized ENESTnd trial Leukemia.

2016;30(5):1044–1054.

Pemmaraju N, Kantarjian H, Shan J, et al.

Analysis of outcomes in adolescents and young adults with chronic myelogenous leukemia treated with upfront tyrosine kinase inhibitor therapy. Haematologica.

2012;97(7):1029–1035.

Sasaki K, Strom S, O'Brien S, et al. Relative survival in patients with chronic-phase chronic myeloid leukaemia in the tyrosinekinase inhibitor era: analysis of patient data from six prospective clinical trials. Lancet Haematol. 2015;2(5):e186–e193.

haematologica | 2017; 102(10)

ARTICLE

Acute Myeloid Leukemia

Vosaroxin in combination with decitabine in newly diagnosed older patients with acute myeloid leukemia or high-risk myelodysplastic syndrome

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Naval Daver,1 Hagop Kantarjian,1 Guillermo Garcia-Manero1, Elias Jabbour,1 Gautam Borthakur,1 Mark Brandt,1 Sherry Pierce,1 Kenneth Vaughan,1 Jing Ning,2 Graciela M. Nogueras González,2 Keyur Patel,3 Jeffery Jorgensen,3 Naveen Pemmaraju,1 Tapan Kadia,1 Marina Konopleva,1 Michael Andreeff,1 Courtney DiNardo,1 Jorge Cortes,1 Renee Ward,4 Adam Craig4 and Farhad Ravandi1

1 Department of Leukemia, 2Department of Biostatistics, and 3Department of Hematopathology, The University of Texas M.D. Anderson Cancer Center, Houston, TX, USA and 4Sunesis Pharmaceuticals Inc., South San Francisco, CA, USA

Haematologica 2017 Volume 102(10):1709-1717

ABSTRACT

V

osaroxin is an anti-cancer quinolone-derived DNA topoisomerase II inhibitor. We investigated vosaroxin with decitabine in patients ≥60 years of age with newly diagnosed acute myeloid leukemia (n=58) or myelodysplastic syndrome (≥10% blasts) (n=7) in a phase II non-randomized trial. The initial 22 patients received vosaroxin 90 mg/m2 on days 1 and 4 with decitabine 20 mg/m2 on days 1-5 every 4-6 weeks for up to seven cycles. Due to a high incidence of mucositis the subsequent 43 patients were given vosaroxin 70 mg/m2 on days 1 and 4. These 65 patients, with a median age of 69 years (range, 60-78), some of whom with secondary leukemia (22%), adverse karyotype (35%), or TP53 mutation (20%), are evaluable. The overall response rate was 74% including complete remission in 31 (48%), complete remission with incomplete platelet recovery in 11 (17%), and complete remission with incomplete count recovery in six (9%). The median number of cycles to response was one (range, 1-4). Grade 3/4 mucositis was noted in 17% of all patients.

The 70 mg/m2 induction dose of vosaroxin was associated with similar rates of overall response (74% versus 73%) and complete remission (51% versus 41%, P=0.44), reduced incidence of mucositis (30% versus 59%, P=0.02), reduced 8-week mortality (9% versus 23%; P=0.14), and improved median overall survival (14.6 months versus 5.5 months, P=0.007). Minimal residual disease-negative status by multiparametric flow-cytometry at response (± 3 months) was achieved in 21 of 39 (54%) evaluable responders and was associated with better median overall survival (34.0 months versus 8.3 months, P=0.023). In conclusion, the combination of vosaroxin with decitabine is effective and well tolerated at a dose of 70 mg/m2 and warrants randomized prospective evaluation.

ClinicalTrials.gov: NCT01893320

Introduction Over two-thirds of patients with newly diagnosed acute myeloid leukemia (AML) in the United States of America (USA) and Europe are aged 65 years or older.1-3 These older patients do not fare as well with intensive induction therapy, having complete remission (CR) rates <50%, a median survival of 4-9 months, and increased induction mortality (15-30%).4-6 Hypomethylating agents (decitabine and azacytidine) are commonly used in the treatment of less fit, older patients with AML in the USA and Europe.7 The pivotal DACO-016 study demonstrated a superior CR/CR without platelet recovery (CRp) rate with decitabine versus investigahaematologica | 2017; 102(10)

Correspondence: ndaver@mdanderson.org or fravandi@mdanderson.org Received: March 17, 2017.

Accepted: July 12, 2017.

Pre-published: July 20, 2017.

doi:10.3324/haematol.2017.168732 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1709 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1709

N. Daver et al.

tors’ choice of treatment (including low-dose cytarabine or best supportive care) in 485 older patients with AML (median age 73 years) who were ineligible for cytotoxic chemotherapy (17.8% versus 7.8%, P=0.001)7 and improved survival with decitabine, leading to the approval of decitabine for the treatment of AML in the elderly in Europe.7,8 Vosaroxin is a non-anthracycline anticancer quinolonederivative that intercalates DNA and inhibits topoisomerase II, causing site-selective DNA breaks, G2 arrest, and apoptosis.9 Vosaroxin is not a substrate of P-glycoprotein-mediated efflux and can induce apoptosis independently of P53 function.9-11 In a phase II dose regimen optimization study in patients with previously untreated, unfavorable prognosis AML ≥60 years of age, single-agent vosaroxin resulted in a CR/CRp rate of 32%, a 30-day mortality of 12%, and a median survival of 7.0 months.12 The 72 mg/m2 days 1 and 4 and 90 mg/m2 day 1 and 4 schedules of single-agent vosaroxin were well tolerated with the highest CR/CRp rates. The pivotal phase III, randomized, controlled, double-blind, multinational clinical study of the efficacy and safety of vosaroxin and cytarabine versus placebo and cytarabine in patients with first relapsed or refractory AML (VALOR) (n=711) demonstrated that vosaroxin in combination with intermediate-dose cytarabine produced a significantly superior remission rate (30% versus 16%; P<0.0001) and improved overall survival (OS) with equivalent 60-day mortality as that following cytarabine alone.13 The overall survival benefit with the combinations was most prominent in patients older than 60 years (7.1 months versus 5.0 months, P=0.003).

The non-confluent safety profile of vosaroxin and decitabine, their non-overlapping molecular mechanisms of action, and the encouraging data with vosaroxin alone, and vosaroxin in combination with cytarabine in the older AML population lent support to this phase II trial of vosaroxin with decitabine in untreated elderly patients (≥60 years) with AML or high-risk myelodysplastic syndrome (MDS) unsuitable for intensive induction. This study was designed to assess whether the addition of vosaroxin to decitabine can improve response rates and OS compared to established outcomes with decitabine

A

alone while maintaining an acceptable safety profile. The decitabine dose and schedule were those used in AML registration studies in the USA and Europe7 and in published phase II clinical studies with decitabine alone14 or in combination with idarubicin, amsacrine or daunorubicin.15,16 The vosaroxin dose and schedule were selected from the phase II study of frontline vosaroxin in older AML patients.12

Methods Patients’ eligibility

Eligible patients were subjects ≥60 years of age with untreated AML or untreated high-risk MDS (intermediate-2 or high according to the International Prognostic Scoring System and ≥10% blasts) who were unsuitable for standard induction in the opinion of the treating physician. Non-suitability for induction chemotherapy was based on the predictive prognostic model for outcome in older patients with AML published by Kantarjian et al.4 Patients with an Eastern Cooperative Oncology Group performance status ≤3; serum creatinine ≤2.0 mg/dL; serum bilirubin ≤2.0 mg/dL; serum transaminase ≤2.5 times the upper limit of the normal range or ≤5 times upper limit of the normal range if the transaminase elevation was deemed related to leukemic infiltration, were enrolled on the study. This was a single-center, open-label, non-randomized study. All patients signed an informed consent form approved by the University of Texas - M. D. Anderson Cancer Center (UT/MDACC) Institutional Review Board. The study was conducted in accordance with the Declaration of Helsinki.

(ClinicalTrials.gov identifier: NCT01893320)

Study design and objectives This study recruited patients between 15 September, 2013 and 23 May, 2016. A total of 65 patients were enrolled. The latest follow-up date was 20 September, 2016. The primary trial endpoint was to establish the safety and efficacy [overall response rate (ORR) = CR, CRp or CR with incomplete recovery of peripheral counts (CRi) assessed as the best response achieved on study] of the combination. Secondary endpoints included analysis of the OS, event-free survival, toxicities, and the correlation of outcomes to baseline cytogenetic and molecular profiles.

B

Figure 1. Survival and relapse-free survival in all patients on study. (A) Survival and relapse-free survival among all patients treated with vosaroxin in combination with decitabine on trial not censored and (B) censored for allogeneic stem cell transplant (SCT).

1710

haematologica | 2017; 102(10)

Vosaroxin and decitabine in older AML patients

Treatment regimen The induction regimen included 5 days of decitabine at a dose of 20 mg/m2 given intravenously (IV) over 60 to 90 min. The vosaroxin was initially administered at a dose of 90 mg/m2 to 22 patients (patients #1-22) on days 1 and 4 (Figure 1). Grade 3/4 mucositis was noted in five of these 22 (23%) patients, prompting a dose reduction of vosaroxin. The next 43 patients (patients #2365) received vosaroxin 70 mg/m2 on days 1 and 4. Patients underwent bone marrow aspiration on day 28 (±5) days. Patients whose day 28 bone marrow showed ≥5% blasts received re-induction with the same dose and schedule as the induction. Patients who did not achieve morphological remission (<5% blasts) at the end of course 1 had a repeat bone marrow examination at the end of course 2. Patients with a response or clinical benefit after one or two induction courses received post-induction therapy with up to five additional cycles of the combination with decitabine 20 mg/m2 on days 1-5 and either vosaroxin 70 mg/m2 on days 1 and 4 or a reduced dose of vosaroxin of 50 mg/m2 or lower, based on response including minimal residual disease (MRD) status, toxicity, and count recovery. Post-induction cycles were repeated every 4-6 weeks, depending on count recovery and resolution of other toxicity. Bone marrow aspirations were repeated every three to four courses while on therapy. Patients who maintained a response (CR or CRp or CRi) at the end of post-induction therapy could receive maintenance with decitabine alone every 4-6 weeks for up to 24 additional cycles.

Baseline assessments Pretreatment evaluations included complete history and physical examination, complete blood count with differential, a comprehensive biochemistry panel, pregnancy test and counseling, and bone marrow aspiration for histological, multiparametric flow-cytometric, cytogenetic analyses, and next-generation sequencing. Multiparametric flow-cytometry and cytogenetics were performed at our institution.17,18 A next-generation sequencing-based analysis for the detection of somatic mutations in the coding sequences of 28 genes was performed on DNA extracted from the bone marrow sample. The methodology of our mutation analysis panel and coverage by genes has been previously published19 (Online Supplementary Table S1).

Response criteria and definitions Responses were according to established criteria for AML and included the best response achieved on study.20,21

Toxicity assessment In the lead-in portion of the study, the safety and tolerable dose of the combination were assessed to identify the maximum tolerated doses. Six patients were to be treated in the lead-in portion.

If clinically significant, drug-related grade 3-4 toxicity was observed during the first 28 days on therapy in one or none of six patients, this would define a safe schedule and the study would proceed to expansion. If study drug-related grade 3-4 toxicity was observed in two or more of six patients during the first 28 days, this dose would exceed the maximum tolerated dose, and a lower dose schedule would be investigated. The dosing algorithm is presented in Table 1. The maximum tolerated dose was considered as the highest dose level at which fewer than two of six patients developed dose-limiting toxicity in the first 28 days on therapy.

In the phase II portion of the study, patients were monitored continuously for toxicity.22 We denoted the probability of toxicity by θE, where toxicity was defined as any clinically significant grade 3 or 4 non-hematologic toxic effect or death, according to the Common Terminology Criteria for Adverse Events version 4.0, attributable to the study drug. We assumed non-informative haematologica | 2017; 102(10)

toxicity prior of beta θE ~ beta (0.3, 1.7). The stopping rule was given by the following probability statement: P (θE >0.15 | data) >0.90. That is, we would stop the trial if, at any time during the study, we determined that there was a more than 90% chance that the toxicity rate would be greater than 15%.

Statistical methods The expected response rate with single-agent decitabine in this population of patients is 18-40%.7,14 Under a null hypothesis of 40% with decitabine alone, a sample size of 59 patients would have more than 80% power to detect a difference between a response rate of 60% for the combination of vosaroxin and decitabine and the null response rate using a one-sample exact binomial test with a two-sided alpha of 0.05.

In the expansion phase of the study, patients were monitored continuously for futility. The ORR was assumed to follow a noninformative prior of beta (1.2, 0.8). The stopping boundaries for ORR were that if, at any time during the study, we determined that there was a less than 2.5% chance that the ORR was greater than 60%, we would terminate the study.22 Differences among variables were evaluated by the chi-square test (or Fisher exact test for cell frequencies <5) for categorical variables and t-test or Wilcoxon-Mann-Whitney test for continuous variables. Survival distributions were estimated using Kaplan– Meier methods and compared using the log-rank test.23 All P values were two-sided and P<0.05 was considered statistically significant. Statistical analyses were carried out using IBM SPSS Statistics 21 for Windows (SPSS Inc., Chicago, IL, USA).

Results Patients’ characteristics The first six patients in the lead-in portion were treated at dose level 0 (Table 1), receiving decitabine 20 mg/m2 on days 1-5 and vosaroxin 90 mg/m2 on days 1 and 4. There were no documented dose-limiting toxicities during the first 28 days in the first six patients and the study opened broadly for expansion at this dose. In the expansion phase 16 additional patients received the combination with vosaroxin 90 mg/m2 on days 1 and 4 and grade 3/4 mucositis was noted in five of these 16 patients (31%).

The high incidence of mucositis prompted amendment of the protocol to reduce the vosaroxin dose to 70 mg/m2 on days 1 and 4. The subsequent 43 patients (patients #23-65) received decitabine 20 mg/m2 on days 1-5 and vosaroxin at a dose of 70 mg/m2 on days 1 and 4.

The median age of the patients was 69 years (range, 6078) and 40% of them were older than 70 years of age.

Their pre-treatment clinical characteristics are summarized in Table 2. Fifty-eight patients (89%) had AML and seven had MDS. One-third (35%) of the patients had complex cytogenetics. A clinically validated next-generation sequencing-based analysis was performed in 63 of the

Table 1. Protocol dosing algorithm.

Dose level 0 -1 -2 -3

Vosaroxin Days 1 and 4

Decitabine

90 mg/m2 70 mg/m2 50 mg/m2 50 mg/m2

20 mg/m2 (Days 1-5) 20 mg/m2 (Days 1-5) 20 mg/m2 (Days 1-5) 20 mg/m2 (Days 1-4) 1711

N. Daver et al.

65 (97%) patients. TP53 (20%), IDH2 (18%), TET2 (15%), and K/N-RAS (18%) were the most frequent mutations.

Response to therapy All 65 patients are evaluable for response. The ORR among the 65 patients was 74%, including 31 CR (48%), 11 CRp (17%), and six CRi (9%). The median number of cycles to response was one (range, 1-4). Ten patients (15%) were primary refractory. Deaths were documented in one (2%) and nine (14%) patients at 4 and 8 weeks, respectively. MRD was assessed by multiplanar flow cytometry at the time of response (± 3 months) in 39 of the 48 responders (81%) including 25 of the 32 (78%) responders at the 70 mg/m2 dose and 14 of the 16 (88%) responders at the 90 mg/m2 dose. MRD was not detectable in 21 of the 39 (54%) patients evaluated.

We assessed response by patient’s age at enrollment, baseline cytogenetics and molecular profile (Table 3). The response rates were not significantly different among patients 60-74 years of age and those ≥75 years of age.

Patients with an adverse karyotype (including complex

Table 2. Characteristics of the study population (n=65).

Characteristic

Category

Age (years)

Diagnosis

Prior Rx for AHD

60-69 ≥70 AML – de novo Secondary AML HR MDS Secondary MDS HMA Lenalidomide

BM blast % WBC x109/L Platelets x109/L Cytogenetics

Diploid Miscellaneous -5/-7/complex Insufficient Mutation status (n=65) TP53 IDH2 IDH1 TET2 RAS (K/N) DNMT3A CEBPA ASXL1 JAK2 FLT3 EZH2

N (%); Median [range] 69 [60-78] 38 (58) 27 (42) 44 (68) 14 (22) 7 (11) 0 (0) 3 (5) 1 (2) 36 [9-97] 3.6 [0.4-57.0] 36 [7-333] 24 (37) 15 (23) 23 (35) 3 (5) 13 (20) 12 (18) 9 (14) 10 (15) 12 (18) 8 (12) 8 (12) 8 (12) 3 (5) 4 (6) 2 (3)

karyotype and abnormalities of chromosome 5 and/or 7) at baseline had a response rate of 65%, as compared to 79% in patients with diploid or other non-adverse karyotype. The response rate among the 13 patients with mutated TP53 was 77%. Three patients had been given prior hypomethylating agent (HMA) therapy for MDS or MDS/myeloproliferative neoplasm and went on to receive decitabine with vosaroxin at the time of progression to AML. Two of the three achieved a response (1 CRp and 1 CRi) indicating that the response rate in patients previously treated with HMA was comparable to that in the entire study population.

We compared the outcomes of the 22 patients (patients 1-22) who received vosaroxin 90 mg/m2 in their induction course to the 43 patients (patients #23-65) who received vosaroxin 70 mg/m2 (Table 4). The 8-week mortality was lower with the lower dose of vosaroxin (23% versus 9%; P=0.14) and the response rates were similar (ORR = 73% versus 74%, CR = 41% versus 51%; P=0.435). Of note, a higher proportion of patients required more than one cycle of induction to achieve a response with the 70 mg/m2 dose, with 19 (59%) achieving a response after one course, nine (28%) after two courses, and four (13%) after three courses. Among the 16 responders at the 90 mg/m2 induction dose, 13 (81%) achieved a response after one course, one (6%) after two courses, and two (13%) after three courses. Individual patient’s response status, survival, and disposition of the 48 responders are provided in a swim plot (Online Supplementary Figure S1).

A number of patients who were not considered to be transplant candidates at the time of induction had improvement in their physical condition after achieving remission and could be considered for an allogeneic stem cell transplant (ASCT). Twelve patients proceeded to ASCT, including two of 16 (13%) responders given the vosaroxin 90 mg/m2 dose and ten of 32 (31%) responders on the vosaroxin 70 mg/m2 dose. The median time from the start of therapy to ASCT was 3.9 months (range, 1.8 –

Table 3. Response by baseline characteristics of the patients (n=65).

Parameter

Category

Age (years)

N/n: number; %: percentage; AML: acute myeloid leukemia; HR: high risk; MDS: myeloid dysplastic syndrome; Rx: treatment; AHD: antecedent hematologic disorder; HMA: hypomethylating agent; BM: bone marrow; WBC: white blood cell count.

60-74 ≥75 Cytogenetics Diploid -5/-7/other adverse Miscellaneous Mutation status IDH2 IDH1 TP53 RAS

N

Overall response (CR, CRp, CRi)

CR

52 13 24 23 18 12 9 13 12

75% 69% 79% 65% 78% 92% 33% 77% 58%

50% 38% 54% 35% 56% 75% 33% 46% 17%

N: number; CR: complete remission; CRp,: complete remission with incomplete platelet recovery; CRi,: complete remission with incomplete blood count recovery.

Table 4. Outcomes by induction dose of vosaroxin, N=65.

Induction dose (vosaroxin)

N

Median overall survival

8-week mortality

Overall response (CR, CRp, CRi)

Complete remission

Need >1 cycle to response

90 mg/m2 70 mg/m2

22 43

5.5 months 14.6 months

23% 9%

73% 74%

41% 51%

3 (19%) 13 (41%)

N: number; CR. complete remission; CRp: complete remission with incomplete platelet recovery; CRi: complete remission with incomplete recovery of blood counts.

1712

haematologica | 2017; 102(10)

Vosaroxin and decitabine in older AML patients

7.6). All patients were in CR/CRp/CRi at the time of transplantation. Six of the ASCT donors were matched siblings, the other six were matched, unrelated donors.

Remission duration and survival With a median follow up of 17.5 months (range, 3.634.2), 23 patients are alive and nine are in remission. The median OS for all 65 patients is 9.8 months (range, 0.7 – 34.2) with 1-year and 2-year OS rates of 39% and 32%, respectively (Figure 1A). The median OS censored for patients who underwent ASCT was 10.9 months (range, 0.7 – 34.2) with 1-year and 2-year censored OS rates of 42% and 32%, respectively (Figure 1B).

Of interest, patients treated with the 70 mg/m2 dose of vosaroxin at induction had a significantly longer median OS than those treated with the 90 mg/m2 dose (14.6 months versus 5.5 months, P=0.007) (Figure 2A). The 70 mg/m2 induction dose was associated with a significantly improved 1-year survival (51% versus 18%) and 2-year survival (44% versus 16%). Furthermore, patients treated with the 70 mg/m2 induction dose continued to fare better with censoring for ASCT (16.1 months versus 5.5 months, P=0.01) (Figure 2B). Patients 60-74 years of age had a significantly longer median OS with the vosaroxin 70 mg/m2 induction dose than with the vosaroxin 90 mg/m2 induction dose (16.1 months versus 7.6 months, P=0.025) (Figure 2C). Similarly, patients ≥75 years of age had a longer median OS with vosaroxin 70 mg/m2 induction as compared to vosaroxin 90 mg/m2 induction (5.7 months versus 1.6 months, P=0.06) (Figure 2D).

A

Patients with complex cytogenetics had a shorter median OS than those with diploid or miscellaneous cytogenetics (5.7 months versus 34.0 months versus 9.8 months, P=0.003) (Figure 3A). Patients treated with vosaroxin 70 mg/m2 with complex cytogenetics had a median OS of 6.6 months and this remained inferior to that of patients with diploid or miscellaneous cytogenetics treated at the same dose (6.6 months versus median not reached, P=0.011) (Figure 3B). The patients with TP53 mutations (n=13) had a median OS of 5.7 months as compared to a median OS of 11.2 months in patients without a TP53 mutation (n=51) (P=0.01). Patients with TP53 mutations who were treated with the 70 mg/m2 induction dose of vosaroxin (n=7) had a median OS of 7.2 months compared to the 1.8 months of those treated with the 90 mg/m2 dose (n=6) (P=0.03) (Figure 3C).

Three patients had received prior HMA therapy for MDS and were enrolled on decitabine and vosaroxin at progression to AML. The median OS of these three patients was 4.4 months which is significantly shorter than the median OS of 9.8 months achieved in the entire study population, although the number for comparison is small.

The median numbers of decitabine + vosaroxin cycles received by the 48 responders (CR/CRp/CRi) and the 17 non-responders were three (range, 1 – 7) and one (range, 1 – 3), respectively. The median numbers of cycles received by patients who achieved CR, CRp, and CRi were four (range, 2 – 7), three (range, 1 – 6), and two (range, 1 – 3), respectively.

B

C

Figure 2. Survival by vosaroxin induction dosage and age of the patients. Survival in patients treated with vosaroxin induction doses of 90 mg/m2 and 70 mg/m2 (A) not censored and (B) censored for allogeneic stem cell transplant (SCT), respectively. (C) Survival in patients 60-74 years and ≥75 years treated with the vosaroxin induction doses of 70 mg/m2 and 90 mg/m2.

haematologica | 2017; 102(10)

1713

N. Daver et al.

To evaluate the impact of the number of cycles of decitabine and vosaroxin among the 48 responders (CR/CRp/CRi) we analyzed OS according to whether the patients received more or less than the median number of cycles of the combination (median number of cycles of decitabine and vosaroxin in the 48 responders was 3). The median OS was significantly shorter in responders who received three or fewer cycles of the combination than in responders who received four or more cycles of the combination (6 months versus 34 months, P<0.001).

We performed the same analysis among the 31 patients who achieved a CR (median number of cycles of decitabine and vosaroxin in the 31 CR patients was 4).

The median OS was significantly shorter in patients with CR who received four or fewer cycles of the combination than in responders who received five or more cycles of the combination (6 months versus 34 months, P<0.001).

Minimal residual disease at response Responders who achieved MRD-negative status at the time of their response (Âą 3 months) had a significantly longer OS than those who remained MRD-positive at response (34.0 months versus 8.3 months, P=0.023).

Thirteen of 25 (52%) evaluable responders treated at the 70 mg/m2 dose achieved MRD-negative status at response and had a significantly improved survival as compared to those who remained MRD-positive (median not reached versus 7.6 months, P=0.006). Eight of 14 (57%) evaluable responders treated at the 90 mg/m2 dose achieved MRDnegative status at response but this was not associated with improved survival (6.9 versus 8.3 months, P=0.81)

A

(Figure 4A-C). We compared baseline clinical characteristics (age, vosaroxin induction dose, de novo/secondary AML, cytogenetic group, bone marrow blasts, platelet count, mutated/non-mutated TP53) among responders with MRD evaluable at response to see whether any influenced the achievement of MRD-negative status (Online Supplementary Table S2). None of the baseline features was predictive for achievement of MRD negativity.

The patients who achieved CR had a longer overall survival than the patients who achieved CRp or CRi (18.3 months versus 9.8 months versus 4.4 months, P=0.023) (Online Supplementary Figures S2 and S3).

Safety and tolerability The most common drug-related toxicity was mucositis

Table 5. Toxicities in the study patients (n=65).

Toxicities

Grade 1/2 N (%)

Mucositis 26 (40) Bilirubin 21 (32) Nausea/vomiting 8 (12) Diarrhea 2 (3) Major infections NA (pneumonia, sepsis) Other site infections NA Fungal infections NA

Grade 3/4 N (%)

Total N (%)

11 (17) 8 (12) 1 (2) 0 (0) 46 (71)

37 (57) 29 (45) 9 (14) 2 (3) 46 (71)

6 (9) 2 (3)

6 (9)

N/n: number; %: percentage.

B

C

Figure 3. Survival by baseline salient cytogenetic and molecular features. Survival by cytogenetic subgroups in (A) all patients on the study and in (B) patients treated with the vosaroxin induction dose of 70 mg/m2. (C) Survival of patients with TP53 mutations and those without TP53 mutations treated with the vosaroxin induction dose of 70 mg/m2.

1714

haematologica | 2017; 102(10)

Vosaroxin and decitabine in older AML patients

Discussion

(Table 5). Grade 3/4 mucositis was seen in 11 (17%) patients and grade 1/2 mucositis in 26 (40%). With regards to the frequency and severity of the mucositis according to induction dose, grade 3/4 mucositis occurred in 16% versus 23% of the patients given 70 mg/m2 or 90 mg/m2, respectively (P=0.85), whereas grade 1/2 mucositis occurred in 30% versus 59%, respectively (P=0.02). Other drug-related toxicities included elevated levels of bilirubin in 45% of the patients: 11 (26%) of those given 70 mg/m2 and 18 (82%) of those given 90 mg/m2 (P<0.001). Most of these were grade 1/2 events and resolved. The rate and distribution of infectious complications were as expected for elderly AML patients receiving an induction regimen.

Seven of the 12 (58%) patients who underwent ASCT have died from bone marrow relapse (n=3), central nervous system relapse (n=1) and post-transplant infections (n=3; all 3 died in CR). Forty-two of the 65 (65%) enrolled patients have died by the time of making this report. The causes/timing of death were induction-related (within 8 weeks) in nine (21%) cases, relapsed/refractory AML in 22 (52%), post-transplant in seven (17%), infection while in CR/CRp/CRi in two (5%), and unknown (death after leaving MD Anderson) in another two (5%). Of the 13 patients with mutated TP53 treated on the trial, one remains alive. The causes/timing of death in the remaining 12 were induction-related (within 8 weeks) in three (25%) cases, relapsed/refractory AML in six (50%), posttransplant in two (17%), and unknown in one.

The initial 22 patients on our study received vosaroxin at the induction dose of 90 mg/m2 on days 1 and 4. We observed a high incidence of mucositis and early mortality (8-week mortality rate, 23%) from neutropenic infections likely related to mucositis. The 70 mg/m2 dose was then evaluated in the subsequent 43 patients with a reduction in the incidence of overall mucositis and early mortality (8-week mortality rate, 9%). This led to an improved OS in the cohort treated with the 70 mg/m2 dose.

Older patients with AML more frequently have an antecedent hematologic disorder, unfavorable cytogenetics, and poorer performance status at presentation.4,5 As a result, the outcomes of elderly patients with AML has not improved significantly over the last four decades.3,24,25 Overall, standard induction regimens can achieve CR/CRp rates of 45-50%, median survival of 4-9 months with an 8week mortality rate of up to 30%.4-6,10 These dismal outcomes have resulted in a shift over the last decade to lower intensity strategies such as HMA or low-dose cytarabine in Europe and the USA. Phase II and III trials of decitabine in elderly patients with AML have shown CR/CRp rates of 18-25% with median survival of 7-8 months and 60-day mortality rates of 10-18%.7,14 In the DACO-016 trial the response rate, median survival, and 60-day mortality in the low-dose cytarabine or supportive care comparator arm were inferior to those in the decitabine arm. In an effort to

A

B

C

Figure 4. Survival by minimal residual disease status at time of response (A) Survival by minimal residual disease (MRD) at the time of response (Âą 3 months) as determined by multiplanar flow cytometry among the evaluable responders for all patients on study. (B, C) Survival

by MRD at response among responders treated with a vosaroxin induction dose of (B) 70 mg/m2 and (C) 90 mg/m2.

haematologica | 2017; 102(10)

1715

N. Daver et al.

further improve outcomes, decitabine was administered at a dose of 20 mg/m2 in an extended 10-day regimen with a reported CR/ CRi rate of 64%, 8-week mortality of 15%, and median OS of 13.7 months.26 However, these superior outcomes have not been reproduced. In a randomized trial of azacytidine versus standard of care in 488 older patients (age â&#x2030;Ľ65 years) with AML, a response rate of 27.8% and a median survival of 10.9 months were reported for azacytidine27 as compared to 25.1% and 6.5 months, respectively for the control arm. The 74% response rate among the patients treated on this trial does, therefore, compare favorably to those achieved with intensive chemotherapy, single-agent decitabine in a 5-day or 10-day dose regimen, or azacytidine in older patients with AML.4,25-28 Although the median OS of 9.8 months for all patients in this study is similar to that which has been achieved with intensive chemotherapy or HMA, the 70 mg/m2 dose of vosaroxin at induction produced a clearly improved, encouraging median survival (14.6 months).

The reported incidences of adverse karyotype and TP53 mutations in older patients with newly diagnosed AML are 20-25% and 5-10%, respectively.29-32 Typically, patients with these characteristics are resistant to standard cytotoxic therapy, having remission rates of 3236% and a median survival of 4 â&#x20AC;&#x201C; 7 months with standard therapies.33-35 Welch et al. treated 114 patients (88 with AML and 26 with MDS) with a 10-day regimen of decitabine in monthly cycles and reported high rates of morphological remission (46%). They specifically noted higher response rates among patients with an unfavorable cytogenetic profile than among those with intermediate- or favorable-cytogenetic profiles (67% versus 37%; P<0.001) and among patients with TP53 mutations as compared to those without TP53 mutations (100% versus 41%; P<0.001).36 The patients in our trial had poor prognostic factors with 35% having an adverse karyotype and 20% having TP53 mutations. A potential benefit of vosaroxin is its ability to induce apoptosis independently of TP53 function. In this trial, we noted ORR of 65% and 77% among patients with an adverse karyotype and mutated TP53, respectively. In contrast to the findings of Welch et al., the response rates in TP53 mutated patients were similar to but not better than the ORR of 74% for our entire group. In spite of the improved response rate, the median OS for TP53 mutated patients was <6 months with the major reason for poor OS among these patients being relapsed/refractory disease. In summary

References 1. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2006. CA Cancer J Clin.

2006;56(2):106-130.

2. Lowenberg B, Downing JR, Burnett A. Acute myeloid leukemia. N Engl J Med. 1999;341(14):1051-1062.

3. Juliusson G, Antunovic P, Derolf A, et al. Age and acute myeloid leukemia: real world data on decision to treat and outcomes from the Swedish Acute Leukemia Registry. Blood.

2009;113(18):4179-4187.

4. Kantarjian H, O'Brien S, Cortes J, et al.

Results of intensive chemotherapy in 998 patients age 65 years or older with acute myeloid leukemia or high-risk myelodysplas-

1716

5.

6.

7.

8.

we believe that both our data and those from Welch et al. support the use of HMA-based therapies for patients with complex cytogenetic abnormalities and TP53 mutations, at least to achieve an initial response. However, further approaches including incorporation of novel agents may be needed to significantly improve the survival in this high-risk population.

Three patients had received prior therapy with HMA. Two of these three patients had a response, but their OS was significantly shorter than that of the other patients.

In summary, as has been well described in the past, patients who progressed to AML after having received treatment with HMA had a worse outcome than those who had not received prior HMA therapy, although the numbers are small.

MRD status has emerged as a very important prognostic factor for long-term outcomes in AML patients treated with cytotoxic induction.17,37-39 This was one of the first trials to monitor MRD status and correlate it with outcome in patients treated with hypomethylator-based therapies.

A multicolor flow cytometric immunophenotype for detecting MRD in AML was used in this trial. The methodology of MRD detection used at our institution has been previously published.17,18 MRD at remission was evaluated in 81% of the patients who achieved remission.

Half of the patients became MRD negative at the time of achieving remission and achieving MRD negativity was associated with a significantly improved OS. In conclusion the combination of vosaroxin with decitabine achieves a higher response rate with an equivalent 8-week mortality to that expected with decitabine or azacytidine alone. Patients treated with the 70 mg/m2 regimen had a median survival of 14.6 months and 51% were alive at 1 year. Prospective randomized trials to compare vosaroxin with decitabine to existing regimens in newly diagnosed older patients with AML are encouraged.

Acknowledgments Funding support received from Sunesis Pharmaceuticals and the MD Anderson Cancer Centre Leukaemia Support Grant (CCSG) CA016672.

Funding Sunesis Pharmaceuticals and the MD Anderson Cancer Centre Leukaemia Support Grant (CCSG) CA016672.

Prior presentations: Oral presentation ASH 2014, Oral presentation ASH 2015, Oral presentation EHA 2016.

tic syndrome: predictive prognostic models for outcome. Cancer. 2006;106(5): 1090-1098.

Appelbaum FR, Gundacker H, Head DR, et al. Age and acute myeloid leukemia. Blood.

2006;107(9):3481-3485.

Burnett A, Wetzler M, Lowenberg B. Therapeutic advances in acute myeloid leukemia. J Clin Oncol. 2011;29(5):487-494.

Kantarjian HM, Thomas XG, Dmoszynska A, et al. Multicenter, randomized, openlabel, phase III trial of decitabine versus patient choice, with physician advice, of either supportive care or low-dose cytarabine for the treatment of older patients with newly diagnosed acute myeloid leukemia. J Clin Oncol. 2012;30(21):2670-2677.

Thomas XG, Arthur C, Delaunay J, Jones M,

Berrak E, Kantarjian HM. A post hoc sensitivity analysis of survival probabilities in a multinational phase III trial of decitabine in older patients with newly diagnosed acute myeloid leukemia. Clin Lymphoma Myeloma Leuk. 2014;14(1):68-72.

9. Hawtin RE, Stockett DE, Byl JA, et al.

Voreloxin is an anticancer quinolone derivative that intercalates DNA and poisons topoisomerase II. PLoS One. 2010;5(4): e10186.

10. Hoch U, Lynch J, Sato Y, et al. Voreloxin, formerly SNS-595, has potent activity against a broad panel of cancer cell lines and in vivo tumor models. Cancer Chemother Pharmacol. 2009;64(1):53-65.

11. Walsby EJ, Coles SJ, Knapper S, Burnett AK.

haematologica | 2017; 102(10)

Vosaroxin and decitabine in older AML patients

12.

13.

14.

15.

16.

17.

18.

19.

The topoisomerase II inhibitor voreloxin causes cell cycle arrest and apoptosis in myeloid leukemia cells and acts in synergy with cytarabine. Haematologica. 2011;96(3): 393-399.

Stuart RK, Cripe LD, Maris MB, et al.

REVEAL-1, a phase 2 dose regimen optimization study of vosaroxin in older poorrisk patients with previously untreated acute myeloid leukaemia. Br J Haematol. 2015;168 (6):796-805.

Ravandi F, Ritchie EK, Sayar H, et al.

Vosaroxin plus cytarabine versus placebo plus cytarabine in patients with first relapsed or refractory acute myeloid leukaemia (VALOR): a randomised, controlled, double-blind, multinational, phase 3 study. Lancet Oncol. 2015;16(9):1025-1036.

Cashen AF, Schiller GJ, O'Donnell MR, DiPersio JF. Multicenter, phase II study of decitabine for the first-line treatment of older patients with acute myeloid leukemia.

J Clin Oncol. 2010;28(4):556-561.

Willemze R, Suciu S, Archimbaud E, et al. A randomized phase II study on the effects of 5-Aza-2'-deoxycytidine combined with either amsacrine or idarubicin in patients with relapsed acute leukemia: an EORTC Leukemia Cooperative Group phase II study (06893). Leukemia. 1997;11(Suppl 1):S24-27.

Schwartsmann G, Fernandes MS, Schaan MD, et al. Decitabine (5-Aza-2'-deoxycytidine; DAC) plus daunorubicin as a first line treatment in patients with acute myeloid leukemia: preliminary observations.

Leukemia. 1997;11(Suppl 1):S28-31.

Ravandi F, Jorgensen J, Borthakur G, et al.

Persistence of minimal residual disease assessed by multiparameter flow cytometry is highly prognostic in younger patients with acute myeloid leukemia. Cancer.

2017;123(3):426-435.

Jaso JM, Wang SA, Jorgensen JL, Lin P. Multicolor flow cytometric immunophenotyping for detection of minimal residual disease in AML: past, present and future. Bone Marrow Transplant. 2014;49(9):1129-1138.

Luthra R, Patel KP, Reddy NG, et al. Nextgeneration sequencing-based multigene mutational screening for acute myeloid leukemia using MiSeq: applicability for diagnostics and disease monitoring.

haematologica | 2017; 102(10)

Haematologica. 2014;99(3):465-473.

20. Cheson BD, Bennett JM, Kopecky KJ, et al.

Revised recommendations of the International Working Group for Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia. J Clin Oncol.

2003;21(24):4642-4649.

21. Sievers EL, Larson RA, Stadtmauer EA, et al.

Efficacy and safety of gemtuzumab ozogamicin in patients with CD33-positive acute myeloid leukemia in first relapse. J Clin Oncol. 2001;19(13):3244-3254.

22. Thall PF, Simon RM, Estey EH. Bayesian sequential monitoring designs for single-arm clinical trials with multiple outcomes. Stat Med. 1995;14(4):357-379.

23. Kaplan EL, Meier P. Nonparametric-estimation from incomplete observations. J Am Stat Assoc. 1958;53(282):457-481.

24. Buchner T, Berdel WE, Haferlach C, et al.

Age-related risk profile and chemotherapy dose response in acute myeloid leukemia: a study by the German Acute Myeloid Leukemia Cooperative Group. J Clin Oncol.

2009;27(1):61-69.

25. Kantarjian H, Ravandi F, O'Brien S, et al.

Intensive chemotherapy does not benefit most older patients (age 70 years or older) with acute myeloid leukemia. Blood.

2010;116(22):4422-4429.

26. Blum W, Garzon R, Klisovic RB, et al.

Clinical response and miR-29b predictive significance in older AML patients treated with a 10-day schedule of decitabine. Proc Natl Acad Sci USA. 2010;107(16):7473-7478.

27. Dombret H, Seymour JF, Butrym A, et al.

International phase 3 study of azacitidine vs conventional care regimens in older patients with newly diagnosed AML with >30% blasts. Blood. 2015;126(3):291-299.

28. Tallman MS, Gilliland DG, Rowe JM. Drug therapy for acute myeloid leukemia. Blood.

2005;106(4):1154-1163.

29. Grimwade D, Walker H, Harrison G, et al.

The predictive value of hierarchical cytogenetic classification in older adults with acute myeloid leukemia (AML): analysis of 1065 patients entered into the United Kingdom Medical Research Council AML11 trial.

Blood. 2001;98(5):1312-1320.

30. Stirewalt DL, Kopecky KJ, Meshinchi S, et al. FLT3, RAS, and TP53 mutations in elderly patients with acute myeloid leukemia.

Blood. 2001;97(11):3589-3595.

31. Schoch C, Kern W, Schnittger S, Buchner T, Hiddemann W, Haferlach T. The influence of age on prognosis of de novo acute myeloid leukemia differs according to cytogenetic subgroups.

Haematologica.

20040;89(9):1082-1090.

32. Rucker FG, Schlenk RF, Bullinger L, et al.

TP53 alterations in acute myeloid leukemia with complex karyotype correlate with specific copy number alterations, monosomal karyotype, and dismal outcome. Blood.

2012;119(9):2114-2121.

33. Nahi H, Lehmann S, Bengtzen S, et al.

Chromosomal aberrations in 17p predict in vitro drug resistance and short overall survival in acute myeloid leukemia. Leuk Lymphoma. 2008;49(3):508-516.

34. Seifert H, Mohr B, Thiede C, et al. The prognostic impact of 17p (p53) deletion in 2272 adults with acute myeloid leukemia.

Leukemia. 2009;23(4):656-663.

35. Kadia TM, Jain P, Ravandi F, et al. TP53 mutations in newly diagnosed acute myeloid leukemia: clinicomolecular characteristics, response to therapy, and outcomes.

Cancer. 2016 Jul 26. [Epub ahead of Print].

36. Welch JS, Petti AA, Miller CA, et al. TP53 and decitabine in acute myeloid leukemia and myelodysplastic syndromes. N Engl J Med. 2016;375(21):2023-2036.

37. Freeman SD, Virgo P, Couzens S, et al.

Prognostic relevance of treatment response measured by flow cytometric residual disease detection in older patients with acute myeloid leukemia. J Clin Oncol. 2013;31 (32):4123-4131.

38. Ouyang J, Goswami M, Tang G, et al. The clinical significance of negative flow cytometry immunophenotypic results in a morphologically scored positive bone marrow in patients following treatment for acute myeloid leukemia. Am J Hematol. 2015;90 (6):504-510.

39. Chen X, Xie H, Wood BL, et al. Relation of clinical response and minimal residual disease and their prognostic impact on outcome in acute myeloid leukemia. J Clin Oncol. 2015;33(11):1258-1264.

1717

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Myeloid Leukemia

Ferrata Storti Foundation

Haematologica 2017 Volume 102(10):1718-1726

Long non-coding RNA expression profile in cytogenetically normal acute myeloid leukemia identifies a distinct signature and a new biomarker in NPM1-mutated patients Etienne De Clara,1 Morgane Gourvest,1 Hanjing Ma,2 François Vergez,1,3 Marie Tosolini,1 Sébastien Dejean,4 Cécile Demur,1,3 Eric Delabesse,1,3 Christian Recher,1,3 Christian Touriol,1 Maria Paola Martelli,5 Brunangelo Falini,5 Pierre Brousset1,6 and Marina Bousquet1 EDC, MG and HM contributed equally to this work.

Cancer Research Center of Toulouse (CRCT), UMR1037 Inserm/Université Toulouse III Paul Sabatier, ERL5294 CNRS, Laboratoire d’Excellence Toulouse Cancer (TOUCAN), France; 2BGI, Shenzhen, China; 3Laboratoire et Service d’Hématologie, Centre Hospitalier Universitaire de Toulouse, Institut Universitaire du Cancer, France; 4Institut de Mathématiques de Toulouse, UMR 5219 Université de Toulouse/CNRS Université Paul Sabatier, France; 5Institute of Hematology, University of Perugia, Ospedale S. Maria della Misericordia, Italy and 6Department of Pathology, Institut Universitaire du Cancer de Toulouse-Oncopole and Centre Hospitalier Universitaire de Toulouse, France 1

ABSTRACT

L

Correspondence: bousquetmarina@gmail.com

Received: April 25, 2017.

Accepted: June 22, 2017.

Pre-published: July 04, 2017. doi:10.3324/haematol.2017.171645 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1718 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1718

ong non-coding RNAs are defined as transcripts larger than 200 nucleotides but without protein-coding potential. There is growing evidence of the important role of long non-coding RNAs in cancer initiation, development and progression. In this study, we sought to evaluate the long non-coding RNA expression profile of patients with cytogenetically normal acute myeloid leukemia (AML). RNA-sequencing of 40 cytogenetically normal AML patients allowed us to quantify 11,036 long non-coding RNAs. Among these, more than 8000 were previously undescribed long non-coding RNAs. Using unsupervised analysis, we observed a specific long non-coding RNA expression profile dependent on the mutational status of the NPM1 gene. Statistical analysis allowed us to identify a minimal set of 12 long non-coding RNAs capable of discriminating NPM1-mutated from NPM1-wild-type patients. These results were validated by qRT-PCR on an independent cohort composed of 134 cytogenetically normal AML patients. Furthermore, we have identified one putative biomarker, the long non-coding RNA XLOC_109948 whose expression pattern predicts clinical outcome. Interestingly, low XLOC_109948 expression indicates a good prognosis especially for NPM1-mutated patients. Transient transfection of GapmeR against XLOC_109948 in NPM1-mutated OCI-AML3 cell line treated with AraC or ATRA enhances apoptosis suggesting XLOC_109948 plays a role in drug sensitivity. This study improves our knowledge of the long noncoding RNA transcriptome in cytogenetically normal AML patients. We observed a distinct long non-coding RNA expression profile in patients with the NPM1 mutation. The newly identified XLOC_109948 long non-coding RNA emerged as a strong prognostic factor able to better stratify NPM1-mutated patients.

Introduction Acute myeloid leukemia (AML) is a clinically and biologically heterogeneous disease with marked differences in survival following intensive chemotherapy. These differences are based on age, blast cell count, cytogenetic abnormalities, and gene mutations.1 Patients with cytogenetically normal AML (CN-AML) account for approximately 50% of all AML and are currently categorized in the intermediaterisk group.2 However, this large group of AML is clinically and molecularly heterohaematologica | 2017; 102(10)

Distinctive LncRNA profile in NPM1-mutated AML

Table 1. Disease and patients’ characteristics according to XLOC_109948 expression level.

Sex Median age at diagnosis Median WBC x109/L Median hemoglobin Median platelet Median blast % ELN Favorable Intermediate-I Complete response Mutations NPM1 FLT3-ITD CEBPa DNMT3a IDH1R132 IDH2R140

All patients n=174

Low 109948 n=71

High 109948 n=103

P

M 88 F 86 57.2 (16-87) 22.03 (0.9-250) 9.6 (5.5-14.2) 75 (8-535) 74 (10-99)

M 39 F 32 55 (20-76) 31.7 (0.9-250) 10 (5.5-14.1) 74 (10-225) 74.5 (10-99)

M 49 F 54 60 (16-87) 16.44 (0.9-234) 9. 5 (5.5-14.2) 76 (8-535) 73.5 (19-99)

0.359 0.002** 0.053 0.28 0.56 0.28 0.03*

54/174 120/174 133/171

29/71 42/71 64/70

25/103 78/103 69/101

0.0003***

94/174 71/174 22/137 30/126 15/123 25/123

57/71 36/71 11/59 13/50 3/47 13/47

37/103 35/103 11/78 17/76 12/76 12/76

6.01E-09*** 0.029* 0.49 0.673 0.16 0.166

n: number; M: male; F: female; WBC: white blood cell count; ELN: European LeukemiaNet; *P<0.05, **P<0.01, ***P<0.001.

geneous. In recent years, the identification of several gene mutations, deregulated coding genes, and the aberrant expression of several microRNAs have provided important prognostic tools and a more complete understanding of the molecular basis of AML.3 However, important questions about the molecular mechanisms underlying AML development and progression remain unanswered. To date, most efforts have been focused on genetic alterations that affect protein-coding genes. Therefore, the discovery of genes that code for long non-coding RNAs (lncRNAs) could reveal a new set of players that participate in AML development. LncRNAs are defined as RNA transcripts that are larger than 200nt but do not appear to have protein-coding potential.4 Recent studies have demonstrated that lncRNAs regulate many processes such as transcription,5 translation,6 epigenetic modification,5,7 cellular differentiation,8 and cell cycle regulation.9 There is growing evidence that also points towards an important role for lncRNAs in cancer initiation, development, and progression.10 Recently, Garzon et al. showed that some deregulated lncRNAs are associated with recurrent mutations and clinical outcome in AML.11 However, our overall knowledge of lncRNA expression patterns in hematologic malignancies remains very limited.12,13 The aim of our study was to better decipher the lncRNA transcriptome and to assess their prognostic role in patients with CN-AML.

Methods AML samples Cohort 1 was composed of 40 AML samples collected from patients registered at the HIMIP (Hémopathies INSERM MidiPyrénées, France) collection (Online Supplementary Table S1). For the validation cohort (Cohort 2, n=134), 34 AML samples were obtained from the HIMIP collection, 7 AML samples from the Hematology Institute of Perugia University, Italy, and 93 AML haematologica | 2017; 102(10)

samples from the FILOtheque AML, Paris, France (Online Supplementary Table S1). The study was approved by local ethics committees.

RNA-Sequencing Library Preparation, Read Generation and Mapping RNA sequencing was performed at the BGI, Hong Kong, for samples from Cohort 1 (n=40). rRNA depletion was performed from total RNA with Ribo-Zero™ rRNA Removal Kits (Epicentre, Madison, WI, USA). Paired-end, strand-specific reads of 91 nt were generated on an Illumina HiSeqTM2000. Alignment and mapping were performed using Tophat against the hg19 genome, and the mapped reads were assembled by Cufflinks v.2.0.2.

Fluidigm 96.96 dynamic array integrated fluidic circuits LncRNA expression analysis was performed using the BioMark 96x96 gene expression platform (Fluidigm) according to the manufacturer’s instructions. Online Supplementary Table S2 shows the list of primers used in the experiment. RNAseP, 5S rRNA and MLN51 were used as reference genes for RT-qPCR data normalization. The Biomark System was used to run the chips and data were collected with Fluidigm Real-Time PCR analysis software.

Statistical analysis Differential expression analysis was performed with the EdgeR package on R and cut offs were established at a fold change of 4 or more, and False Discovery Rate (FDR) less than 0.001. Fisher's exact test and the Mann Whitney test were used to assess the statistical significance of the differences between groups. Survival curves were generated by the Kaplan-Meier method and P values were calculated by the log-rank test.

Cell culture The OCI-AML3 cells were grown in RPMI 1640 medium supplemented with 10% (v/v) fetal bovine serum, 100 U/mL penicillin 1719

E. De Clara et al.

Figure 1. Overview of the study design. RNA-sequencing was performed on rRNA-depleted total bone marrow of 40 cytogenetically normal acute myeloid leukemia (CN-AML) patients (Cohort 1) to determine the lncRNA transcriptome. A validation set composed of 134 new CN-AML patients (Cohort 2) was used to validate the results. Downregulation of XLOC_109948 lncRNA was performed in OCI-AML3 cells transfected with GapmeRs. sPLS-DA: Sparse partial least squares discriminant analysis.

and 100 U/mL streptomycin at 37°C in humidified atmosphere containing 5% CO2.

Transient transfection of GapmeRs: the antisense LNA GapmeRs were designed and supplied by Exiqon (Online Supplementary Table S2). OCI-AML3 cells were transiently transfected with 50 nM (final concentration) of GapmeR mediated by Lipofectamine RNAiMax (Invitrogen) according to the manufacturer's instructions.

Apoptosis assay Apoptosis was induced one day post transfection with Cytarabine (Ara-C, Sigma) or Retinoic acid (ATRA, Sigma) at a final concentration of 10 mM and 1 mM, respectively. The cells were harvested at day 1 after induction with Ara-C and day 2 after induction with ATRA and stained using Pacific Blueâ&#x201E;˘ Annexin V Apoptosis Detection Kit with PI (Biolegend) according to the manufacturer's instructions. AnnexinV/PI stained samples were analyzed using a MACSQuant-10 flow cytometer (Miltenyi Biotec) and further analysis was performed using FlowJo. Additional methods are described in the Online Supplementary Materials and Methods.

Results Identification and quantification of lncRNAs expressed in CN-AML by RNA-sequencing In this study, we sought to determine the lncRNA expression profiles of CN-AML samples. Our first goal was to determine whether specific signatures could be associated with common AML molecular characteristics.

We focused our attention on the well-known AML mutations in NPM1,14 FLT3 (FLT3-ITD),15 CEBPa,16 DNMT3a,17 and IDH (IDH1R132 and IDH2R140).18 Forty CN-AML patients were selected such that around 30% of patients carried each mutation (Figure 1, Cohort 1, and Online Supplementary Table S1). We then used RNA sequencing to identify and determine the lncRNA transcriptome (method described in the Online Supplementary Appendix).

1720

RNA sequencing was performed on rRNA-depleted total bone marrow samples in order to evaluate the entire lncRNA landscape (polyadenylated and non-polyadenylated transcripts). This approach allowed us to identify 11036 lncRNAs expressed amongst our CN-AML patients.

Among them, 8526 were new lncRNAs that had not been previously reported in the RefSeq, UCSC or ENCODE databases (GRCh37/hg19).

Unsupervised analysis of the 11036 CN-AML lncRNAs using hierarchical clustering (Figure 2A) segregated the samples into two main groups. When different criteria [namely French-American-British classification (FAB) status, age and presence of mutations] were evaluated in order to define the two groups, the main observation was that NPM1-mutated patients were largely enriched in the first group (80%) compared to the second group (8%) (P=6x10-6). This result suggested there is an lncRNA expression profile that is specific to NPM1-mutated patients.

Specific lncRNA signature in CN-AML with NPM1 mutation To identify the differentially-expressed lncRNAs that were associated with NPM1 mutations in the 40 CN-AML patients, we compared lncRNA expression in NPM1mutated AML patients (n=14) with NPM1-wild type AML samples (n=26), using the EdgeR package. We focused our study on the highly-expressed lncRNAs by selecting those that had at least 10 cpm (counts per million) in at least 14 of the 40 patients. Among the 5333 lncRNAs selected, 107 were significantly (>4-fold change) differentiallyexpressed (FDR<10-3) between the two groups, with 26 of these up-regulated in NPM1-mutated patients and 81 down-regulated.

To validate our strategy we then applied the same approach to mRNAs and compared mRNA expression in NPM1-mutated and NPM1-wild-type AML patients. The 71 most differentially-expressed mRNAs included the HOX genes, which were up-regulated, and the CD34, haematologica | 2017; 102(10)

Distinctive LncRNA profile in NPM1-mutated AML

A

B

C

Figure 2. Specific lncRNA expression profile for NPM1-mutated AML patients with normal cytogenetics. (A) Unsupervised hierarchical clustering analysis of 40 patients with cytogenetically normal acute myeloid leukemia (CN-AML) (Cohort 1: NPM1+, n=14; NPM1wt, n=26) using 11065 lncRNAs (RNA-seq data). (B) Hierarchical clustering and associated heatmap of Fluidigm data from the first cohort of CN-AML patients (n=40) with 31 lncRNAs differentiallyexpressed between NPM1-mutated (n=14) and NPM1-wild-type patients (n=26). (C) Hierarchical clustering and associated heatmap of Fluidigm data from the second cohort of 134 CN-AML patients [NPM1mutated (n=80) and NPM1-wild type (n=54)].

The heatmap depicts high expression (red: +1) and low expression (blue: â&#x2C6;&#x2019;1).

BAALC and MN1 genes, which were down-regulated (Online Supplementary Table S3). These have already been described in several previous studies and validate our lncRNA approach.19 Among the 107 lncRNAs that were found to be differentially-expressed in NPM1-mutated patients, 74 are located within the introns or exons of protein-coding genes and 33 are intergenic or anti-sense lncRNAs. To avoid the quantification of pre-mRNA by RT-qPCR, we focused our attention on the 33 intergenic and anti-sense haematologica | 2017; 102(10)

lncRNAs (Online Supplementary Table S4). Hierarchical clustering of these clearly separated NPM1-mutated from NPM1-wild-type AML patients, as shown in Online Supplementary Figure S1A (P=1.16x10-9). In order to validate the RNA-seq data, we used RT-qPCR (Fluidigm), but were not able to amplify lncRNA XLOC_051554. A Pearsonâ&#x20AC;&#x2122;s correlation between the two techniques showed a large positive correlation for all the lncRNAs except XLOC_085385 (Online Supplementary Table S5).

Hierarchical clustering using the 31 lncRNAs validated by 1721

E. De Clara et al.

Fluidigm confirmed the separation of the NPM1-mutated and NPM1-wild-type groups (P=1.16x10-9) (Figure 2B). In order to validate this lncRNA signature, a validation set (n=134) composed of 80 new NPM1-mutated AML patients and 54 NPM1-wild-type patients was used (Figure 1, Cohort 2, and Online Supplementary Table S1).

Hierarchical clustering again allowed us to discriminate between NPM1-mutated and NPM1-wild-type patients (P=2.60x10-17) (Figure 2C).

Identification of a minimal signature composed of 12 lncRNAs able to discriminate NPM1-mutated patients from NPM1-wild-type patients In order to reduce the number of discriminating lncRNAs we then used the Sparse partial least squares discriminant analysis (Sparse PLS-DA) approach on the 31 lncRNAs identified from the RNA-seq data. This allowed us to identify a minimal set of 12 lncRNAs that were able to discriminate NPM1-mutated patients from NPM1-wild-type patients in both Cohort 1 (P=5.17x10-9)

A

and Cohort 2 (P=9.07x10-20) (Figure 3 and Online Supplementary Figures S1B and S2).

XLOC_109948 lncRNA expression serves as a prognostic biomarker in CN-AML Although the lncRNAs identified in the minimal signature were differentially expressed between NPM1-mutated and NPM1-wild-type patients, we noticed a heterogeneous expression among patients. We asked whether one of them could be used as a prognostic biomarker. To this end, we selected patients who matched the following criteria: 1) those who had received intensive chemotherapy; and 2) those for whom we had clinical data available. This accounted for 25 of the 40 patients in Cohort 1 and all 134 patients in Cohort 2 (Figure 1 and Online Supplementary Table S6). We used ROC curves to predict the best cut off (â&#x2C6;&#x2019;Î&#x201D;Ct value) for each lncRNA for event-free survival (EFS) in Cohort 1. This cut off was then used to define subgroups with a high or low expression of each lncRNA and the log-rank test was used to compare the survival distri-

B

C

D

1722

Figure 3. A minimal set of 12 lncRNAs is able to discriminate between NPM1-mutated and NPM1-wild-type acute myeloid leukemia (AML) patients. (A) Sparse partial least squares discriminant analysis (sPLSDA) plot of NPM1-mutated versus NPM1-wild-type patients based on 12 discriminating lncRNAs. (B) Variable plot of the 12 discriminative lncRNAs. (C) Hierarchical clustering of 40 cytogenetically normal acute myeloid leukemia (CN-AML) patients (Cohort 1) and associated heatmap of the 12 lncRNA signature identified by Sparse PLS-DA to compare NPM1-mutated (NPM1+) patients with NPM1-wild-type (NPM1wt) patients (Fluidigm data). (D) Hierarchical clustering of 134 CN-AML patients (Cohort 2: NPM1+, n=80; NPM1wt, n=54) and associated heatmap of the 12 lncRNA signature identified by Sparse-PLS-DA (Fluidigm data).

haematologica | 2017; 102(10)

Distinctive LncRNA profile in NPM1-mutated AML

bution in the two groups. Only XLOC_005798 and XLOC_109948 were found to be good putative biomarkers for EFS in Cohort 1 (Figure 4A). Using the previously identified cut off, both lncRNAs could also separate patients according to overall survival (OS) and disease-free survival (DFS) in Cohort 1 (Figure 4A). In order to validate these lncRNAs as good biomarkers, we tested EFS, OS and

DFS in the validation cohort (Cohort 2, n=134) using the cut off identified for Cohort 1 (Figure 4A). These analyses allowed us to validate XLOC_109948 as a good biomarker in CN-AML. Indeed, patients in Cohort 2 with high and low XLOC_109948 transcript levels had median EFS times of 474 and 1128 days, respectively, and estimated 5-year EFS rates of 22% and 45%, respectively (P=0.003) (Figure

A

B

C

D

Figure 4. LncRNA XLOC_109948 expression levels can predict clinical outcome. (A) Clinical outcomes in patient subgroups defined according to cut offs (â&#x2C6;&#x2019;Î&#x201D;Ct value) and identified with ROC curves for each lncRNA. (B) Prognostic value of XLOC_109948 lncRNA expression in a validation cohort (Cohort 2, n=134) of cytogenetically normal acute myeloid leukemia (CN-AML) patients. Kaplan-Meier plots show the event-free survival (EFS), overall survival (OS) and disease-free survival (DFS) of patient subgroups with high versus low transcript levels of XLOC_109948 lncRNA. (C) Multivariate analyses of clinical outcome in 134 patients (Cohort 2). (D) Risk stratification of patients with CN-AML according to NPM1 mutational status and XLOC_109948 expression level (Cohort 2). EFS, OS and DFS are shown for the four subgroups defined by NPM1 and XLOC_109948 status. P values are given for the overall comparison across all four groups. optcut: optimal cut off; HR: hazard ratio.

haematologica | 2017; 102(10)

1723

E. De Clara et al.

4B). Patients with high XLOC_109948 expression levels also had significantly worse OS (estimated 5-year OS values of 36% and 55%, respectively; P=0.021) and DFS (estimated 5-year DFS values of 28% and 48%, respectively; P=0.046) than those with low XLOC_109948 levels (Figure 4B). Interestingly, we also observed that XLOC_109948 is almost unexpressed in CD33+ bone marrow cells from healthy samples (Online Supplementary Figure S3). Table 1 lists the clinical characteristics of the entire study population (Cohort 1 and Cohort 2) and of the patient subgroups distinguished by high versus low expression of XLOC_109948 lncRNA. We found that high XLOC_109948 expression was associated with older patients (P=0.002) and that the patients with low XLOC_109948 expression were enriched in the favorable ELN group (P=0.03) and had higher complete remission (CR) rates (P=0.0003). Low expression of XLOC_109948 also strongly correlated with NPM1+ (P=6.01x10-9) and FLT3-ITD (P=0.029) mutations.

A

Given the correlations we had observed between the various molecular risk markers, we then performed multivariate analyses to identify the factors that independently predicted prognosis in CN-AML. Multivariate analysis was conducted for Cohort 2 (n=134). In a multivariate model, XLOC_109948 expression was a significant prognostic factor for EFS (P=0.008) and DFS (P=0.026). NPM1 mutational status, age and WBC count were also prognostic factors for EFS, and NPM1 mutational status and age for DFS (Figure 4C). Patients with high XLOC_109948 expression had shorter OS rates (P=0.042) than those with low expression after analysis adjustment for WBC count (Figure 4C).

Finally, we investigated patient stratification by using a combination of XLOC_109948 expression and NPM1 mutational status. Patients with low XLOC_109948 expression and mutated NPM1 constituted a favorable subset of patients in terms of EFS (P=0.0002), OS (P=0.007) and DFS (P=0.003) (Figure 4D). These results

B

C

D

Figure 5. Downregulation of XLOC_109948 lncRNA enhances drug sensitivity in OCI-AML3 cell line. (A) Subcellular localization of XLOC_109948 lncRNA. The RNA level of XLOC_109948 in nuclear and cytoplasmic fraction was evaluated by RT-qPCR after OCI-AML3 cell fractionation. GAPDH was a positive control for cytoplasmic fraction and Snord44 was a positive control for nuclear fraction. (B) Quantification of XLOC_109948 expression in OCI-AML3 cells transiently transfected with two different GapmeRs against XLOC_109948 (a and b), GapmeR Negative Control or water. The RNA expression levels were evaluated by quantitative real-time PCR, normalized to the expressions of TBP and ABL1, and presented as fold change [2-ΔΔCt]±Standard Deviation (SD) (n≥3) relative to cells transfected with the GapmeR Negative Control; ***P<0.0005. (C and D) Apoptosis assay. One day post transfection, cells were treated with (C) Ara-C (10 mM) or (D) ATRA (1 mM), and annexinV/PI staining was performed respectively 24 hours (h) or 48 h later. One representative flow cytometry plot is shown. The histogram represents the average of apoptotic cells (Annexin V+) from four independent experiments. *P<0.05, **P<0.01.

1724

haematologica | 2017; 102(10)

Distinctive LncRNA profile in NPM1-mutated AML

suggest that XLOC_109948 lncRNA may be a useful marker in the risk stratification of patients with CN-AML, especially among patients with the NPM1 mutation.

XLOC_109948 lncRNA downregulation induces apoptosis in OCI-AML3 cells As XLOC_109948 low expression was found to be of good prognosis for patients with NPM1 mutation, we decided to down-regulate it in the only available NPM1-mutated AML cell line, OCI-AML3.20 In order to adopt the best strategy to inactivate the XLOC_109948 lncRNA, we first evaluated its subcellular localization by OCI-AML3 cell fractionation.

XLOC_109948 is mainly located into the nucleus (Figure 5A). Since the efficiency of siRNAs in the nucleus remained controversial,21 we chose to use a GapmeR strategy. GapmeR are antisense LNA oligonucleotides which are able to induce the cleavage of the target RNA by endogenous RNase H, a ubiquitous enzyme cleaving the RNA part of RNA/DNA hybrids.22 Transient transfections of OCI-AML3 cells with two different GapmeR against XLOC_109948, a GapmeR Control or H2O were performed and the efficiency of XLOC_109948 downregulation was evaluated by RTqPCR 48 h post transfection. Both GapmeRs against XLOC_109948 were able to down-regulate it with respectively 30% and 70% of efficiency (Figure 5B).

In order to evaluate the role of XLOC_109948 in drug sensitivity, we treated OCI-AML3 transfected cells with Ara-C (10mM) or ATRA (1mM), two drugs used in the clinic for AML treatment, and apoptosis was measured by flow cytometry 24 h post treatment for Ara-C and 48 h post treatment for ATRA. Both GapmeRs against XLOC_109948 enhanced apoptosis in cells treated with Ara-C or ATRA compared to cells transfected with the GapmeR control or water (Figure 5C and D).

Discussion The analysis of 200 AML genomes as part of the Atlas Genome Project has highlighted the genetic and epigenetic architecture of CN-AML.3 It has also provided important new information on the genetic alterations involved in CNAML development and has identified potential tools for risk stratification of CN-AML, as well as proposing new diagnostic and therapeutic targets. However, the studies based on this data have so far mainly focused on the genetic alterations that impact either on protein coding genes or on microRNAs. The discovery of long non-coding RNAs (lncRNAs) has provided a unique opportunity to identify new biomarkers and key players in leukemogenesis.

In this study, we sought to determine lncRNA expression profiles in CN-AML and to correlate them with the most common mutations underlying CN-AML and outcome.

Using RNA sequencing approach, we detected a broad spectrum of lncRNA molecules (11036 lncRNAs).

Compared with the previous study of Garzon et al., who quantified lncRNA expression by microarray,11 our results improve the overall knowledge of lncRNAs, as our RNAseq approach has led to the identification of more than 8000 new lncRNAs. A major finding of this study is the discovery that NPM1-mutated AML patients show a distinct global lncRNA expression profile. NPM1 is a nucleo-cytoplasmic haematologica | 2017; 102(10)

shuttling protein mainly localized in the nucleolus that plays a key role in many biological processes, including ribosome biogenesis,23 histone assembly,24 and the maintenance of genomic stability.25,26 In addition, by regulating the activity and stability of crucial tumor suppressors such as ARF27 and p53,28 NPM1 can also affect cell proliferation and apoptosis. With rare exceptions, NPM1 mutations in AML occur in exon 12 and result in the loss of two key tryptophans, creating a new nuclear export signal leading to aberrant cytoplasmic NPM1 localization.14 NPM1 mutations have been reported to occur in up to 60% of CNAML adult cases14 and are related to distinct expression profiles of both mRNA (including the downregulation of CD34 and the upregulation of HOX genes) and microRNA (including the upregulation of miR-10a and miR-10b).19 When taken together with other features of NPM1 mutations, such as their mutual exclusivity with other AML recurrent cytogenetic abnormalities, their high specificity for AML, their stability at relapse, and their association with unique gene expression and microRNA profiles,29 our findings that NPM1-mutated AML displays a distinct lncRNA signature further supports the view that this leukemia represents a distinct disease entity, in accordance with the recent update of the WHO classification.30 By using differential and statistical analysis, we have identified a minimal signature of 12 lncRNAs that are able to discriminate between NPM1-mutated and NPM1-wildtype patients.

NPM1 mutations are associated with a favorable prognosis in the absence of an accompanying internal tandem duplication mutation in FLT3 (FLT3-ITD), and with an intermediate prognosis if FLT3-ITD co-exists.29 In our study, we showed that low expression of XLOC_109948 lncRNA is significantly associated with better prognosis, especially among NPM1-mutated patients, thus constituting a potentially useful new biomarker to better stratify the risk for NPM1-mutated CN-AML patients. In accordance with this observation, we demonstrated that inactivation of XLOC_109948 sensitizes NPM1-mutated OCIAML3 cell line to drug treatment, suggesting that XLOC_109948 quantification could be used to predict the response to chemotherapy.

Altogether, our data suggest that lncRNAs should be considered for use as biomarkers and could be used as therapeutic targets in the pathogenesis of AML. Determining the molecular mechanisms of these lncRNAs will be of great interest in the future.

Acknowledgments The authors would like to thank all participating investigators from the GOELAMS. We also thank V. De Mas and S. Bertoli for the update of the clinical information for patients collected at the HIMIP. We thank C. Daugrois for her help with multivariate analysis. English proofreading was performed by Scientific Scripts (http://scientificscripts.com).

Funding MB was supported by a fellowship from ARC (Association pour la Recherche sur le Cancer). EDC was supported by a fellowship from LABEX. This work was also supported by the Fondation ARC pour la Recherche sur le Cancer, the Association Laurette Fugain, the Ligue RĂŠgionale Contre le Cancer (ComitĂŠs du Gers et de la Haute-Garonne), the Institut Universitaire de France (PB) and by the Associazione Italiana per la Ricerca sul Cancro (AIRC, IG 2013 n.14595).

1725

E. De Clara et al.

References 1. Frรถhling S, Scholl C, Gilliland DG, Levine RL. Genetics of myeloid malignancies: pathogenetic and clinical implications. J Clin Oncol. 2005;23(26):6285-6295.

2. Mrรณzek K, Marcucci G, Paschka P, Whitman SP, Bloomfield CD. Clinical relevance of mutations and gene-expression changes in adult acute myeloid leukemia with normal cytogenetics: are we ready for a prognostically prioritized molecular classification? Blood. 2007;109(2):431-448.

3. The Cancer Genome Atlas Research Network. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med. 2013; 368(22):2059-2074.

4. Derrien T, Johnson R, Bussotti G, et al. The GENCODE v7 catalog of human long noncoding RNAs: analysis of their gene structure, evolution, and expression. Genome Res. 2012;22(9):1775-1789.

5. Vance KW, Ponting CP. Transcriptional regulatory functions of nuclear long noncoding RNAs. Trends Genet. 2014;30(8):348-355.

6. Carrieri C, Cimatti L, Biagioli M, et al. Long non-coding antisense RNA controls Uchl1 translation through an embedded SINEB2 repeat. Nature. 2012;491(7424):454-457.

7. Geisler S, Coller J. RNA in unexpected places: long non-coding RNA functions in diverse cellular contexts. Nat Rev Mol Cell Biol. 2013;14(11):699-712.

8. Hu W, Yuan B, Flygare J, Lodish HF. Long noncoding RNA-mediated anti-apoptotic activity in murine erythroid terminal differentiation. Genes Dev. 2011;25(24):25732578.

9. Kitagawa M, Kitagawa K, Kotake Y, Niida H, Ohhata T. Cell cycle regulation by long non-coding RNAs. Cell Mol Life Sci. 2013;70(24):4785-4794.

10. Yang G, Lu X, Yuan L. LncRNA: A link between RNA and cancer. Biochim Biophys Acta. 2014;1839(11):1097-1109.

1726

11. Garzon R, Volinia S, Papaioannou D, et al.

Expression and prognostic impact of lncRNAs in acute myeloid leukemia. Proc Natl Acad Sci USA. 2014;111(52):1867918684.

12. Morlando M, Ballarino M, Fatica A. Long Non-Coding RNAs: New Players in Hematopoiesis and Leukemia. Front Med. 2015;2:23.

13. Evans JR, Feng FY, Chinnaiyan AM, et al.

The bright side of dark matter: lncRNAs in cancer. J Clin Invest. 2016;126(8):2775-2782.

14. Falini B, Mecucci C, Tiacci E, et al.

Cytoplasmic nucleophosmin in acute myelogenous leukemia with a normal karyotype. N Engl J Med. 2005;352(3):254-266.

15. Frรถhling S, Schlenk RF, Breitruck J, et al.

Prognostic significance of activating FLT3 mutations in younger adults (16 to 60 years) with acute myeloid leukemia and normal cytogenetics: a study of the AML Study Group Ulm. Blood. 2002; 100(13):4372-4380.

16. Pabst T, Mueller BU, Zhang P, et al.

Dominant-negative mutations of CEBPA, encoding CCAAT/enhancer binding protein-alpha (C/EBPalpha), in acute myeloid leukemia. Nat Genet. 2001;27(3):263-270.

17. Ley TJ, Ding L, Walter MJ, et al. DNMT3A Mutations in Acute Myeloid Leukemia. N Engl J Med. 2010;363(25):2424-2433.

18. Marcucci G, Maharry K, Wu Y-Z, et al. IDH1 and IDH2 Gene Mutations Identify Novel Molecular Subsets Within De Novo Cytogenetically Normal Acute Myeloid Leukemia: A Cancer and Leukemia Group B Study. J Clin Oncol. 2010;28(14):2348-2355.

19. Becker H, Marcucci G, Maharry K, et al.

Favorable prognostic impact of NPM1 mutations in older patients with cytogenetically normal de novo acute myeloid leukemia and associated gene- and microRNA-expression signatures: a Cancer and Leukemia Group B study. J Clin Oncol.

2010;28(4):596-604.

20. Quentmeier H, Martelli MP, Dirks WG, et al. Cell line OCI/AML3 bears exon-12 NPM

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

gene mutation-A and cytoplasmic expression of nucleophosmin. Leukemia. 2005; 19(10):1760-1767.

Lennox KA, Behlke MA. Cellular localization of long non-coding RNAs affects silencing by RNAi more than by antisense oligonucleotides. Nucleic Acids Res. 2016; 44(2):863-877.

Kurreck J, Wyszko E, Gillen C, Erdmann VA. Design of antisense oligonucleotides stabilized by locked nucleic acids. Nucleic Acids Res. 2002;30(9):1911-1918.

Yu Y, Maggi LB, Brady SN, et al.

Nucleophosmin is essential for ribosomal protein L5 nuclear export. Mol Cell Biol.

2006;26(10):3798-3809.

Okuwaki M, Matsumoto K, Tsujimoto M, Nagata K. Function of nucleophosmin/B23, a nucleolar acidic protein, as a histone chaperone. FEBS Lett. 2001;506(3):272-276.

Okuda M, Horn HF, Tarapore P, et al.

Nucleophosmin/B23 is a target of CDK2/cyclin E in centrosome duplication.

Cell. 2000;103(1):127-140.

Ziv O, Zeisel A, Mirlas N, et al.

Identification of novel DNA-damage tolerance genes reveals regulation of translesion DNA synthesis by nucleophosmin. Nat Comm. 2014;25;5:5437.

Colombo E, Bonetti P, Lazzerini Denchi E, et al. Nucleophosmin is required for DNA integrity and p19Arf protein stability. Mol Cell Biol. 2005;25(20):8874-8886.

Kurki S, Peltonen K, Latonen L, et al.

Nucleolar protein NPM interacts with HDM2 and protects tumor suppressor protein p53 from HDM2-mediated degradation. Cancer Cell. 2004;5(5):465-475.

Falini B, Martelli MP, Bolli N, et al. Acute myeloid leukemia with mutated nucleophosmin (NPM1): is it a distinct entity?

Blood. 2011;117(4):1109-1120.

Arber DA, Orazi A, Hasserjian R, et al. The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia. Blood. 2016; 127(20):2391-2405.

haematologica | 2017; 102(10)

ARTICLE

Acute Lymphoblastic Leukemia

Prolonged versus standard native E. coli asparaginase therapy in childhood acute lymphoblastic leukemia and non-Hodgkin lymphoma: final results of the EORTC-CLG randomized phase III trial 58951

Veerle Mondelaers,1 Stefan Suciu, 2 Barbara De Moerloose,1 Alina Ferster,3 Françoise Mazingue,4 Geneviève Plat,5 Karima Yakouben,6 Anne Uyttebroeck,7 Patrick Lutz,8 Vitor Costa,9 Nicolas Sirvent,10 Emmanuel Plouvier,11 Martine Munzer,12 Maryline Poirée,13 Odile Minckes,14 Frédéric Millot,15 Dominique Plantaz,16 Philip Maes,17 Claire Hoyoux,18 Hélène Cavé,19,20 Pierre Rohrlich,13 Yves Bertrand21 and Yves Benoit 1 for the Children’s Leukemia Group (CLG) of the European Organization for Research and Treatment of Cancer (EORTC)

Department of Pediatric Hematology-Oncology and Stem Cell Transplantation, Ghent University Hospital, Ghent University, Belgium; 2EORTC Headquarters, Brussels, Belgium; Department of Pediatric Hematology-Oncology, Children's University Hospital Queen Fabiola, Université Libre de Bruxelles (ULB), Belgium; 4Department of Pediatric Hematology-Oncology, CHRU, Lille, France; 5Department of Pediatric HematologyOncology, CHU-Hopital Purpan, Toulouse, France; 6Department of Pediatric Hematology, Robert Debré Hospital, AP-HP, Paris, France; 7Department of Pediatric HematologyOncology, University Hospital Gasthuisberg, Leuven, Belgium; 8Department of Pediatric Hematology-Oncology, University Hospital Hautepierre, Strasbourg, France; 9Department of Pediatrics, Portuguese Oncology Institute, Porto, Portugal; 10Department of Pediatric Hematology-Oncology, CHU, Montpellier, France; 11Pediatric Hematology Unit, CHU Jean Minjoz Hospital, Besançon, France; 12Department of Pediatric Hematology-Oncology, American Memorial Hospital, Reims, France; 13Department of Pediatric HematologyOncology, CHU Lenval, Nice, France; 14Department of Pediatric Hematology-Oncology, CHU, Caen, France; 15Pediatric Oncology Unit, University Hospital, Poitiers, France; 16 Department of Pediatric Oncology, University Hospital, Grenoble, France; 17Department of Pediatrics, University Hospital Antwerp, Belgium; 18Department of Pediatrics, CHR de la Citadelle, Liège, Belgium; 19Department of Genetics, Assistance Publique des Hôpitaux de Paris (AP-HP), Robert Debré Hospital, Paris, France; 20INSERM UMR 1131, University Institute of Hematology, University Paris Diderot, Paris Sorbonne Cité, France and 21Institute of Pediatric Hematology and Oncology (IHOP), Hospices Civils de Lyon, and University Lyon 1, France

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2017 Volume 102(10):1727-1738

1 3

ABSTRACT

A

sparaginase is an essential component of combination chemotherapy for childhood acute lymphoblastic leukemia and non-Hodgkin lymphoma. The value of asparaginase was further addressed in a group of non-very high-risk patients by comparing prolonged (longasparaginase) versus standard (short-asparaginase) native E. coli asparaginase treatment in a randomized part of the phase III 58951 trial of the European Organization for Research and Treatment of Cancer Children’s Leukemia Group. The main endpoint was disease-free survival. Overall, 1,552 patients were randomly assigned to long-asparaginase (775 patients) or short-asparaginase (777 patients). Patients with grade ≥2 allergy to native E. coli asparaginase were switched to equivalent doses of Erwinia or pegylated E. coli asparaginase. The 8-year disease-free survival rate (±standard error) was 87.0±1.3% in the long-asparaginase group and 84.4±1.4% in the shortasparaginase group (hazard ratio: 0.87; P=0.33) and the 8-year overall survival rate was 92.6±1.0% and 91.3±1.2% respectively (hazard ratio: 0.89; P=0.53). An exploratory analysis suggested that the impact of long-asparaginase was beneficial in the National Cancer Institute standard-risk group with regards to disease-free survival (hazard ratio: 0.70; P=0.057), but far less so with regards to overall survival (hazard ratio: 0.89). The incidences of grade 3-4 infection during consolidation (25.2% versus 14.4%) and late intensification (22.6% versus 15.9%) and the incidence of grade 2-4 allergy were higher in the long-asparaginase arm (30% versus 21%). Prolonged native E. coli asparaginase therapy in consolidation and late intensification for our non-very high-risk patients did not improve overall outcome but led to an increase in infections and allergy. This trial was registered at www.clinicaltrials.gov as #NCT00003728. haematologica | 2017; 102(10)

Correspondence: veerle.mondelaers@uzgent.be

Received: February 6, 2017.

Accepted: July 21, 2017.

Pre-published: July 27, 2017. doi:10.3324/haematol.2017.165845 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1727 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1727

V. Mondelaers et al.

Introduction The rate of success in the treatment of children with acute lymphoblastic leukemia (ALL) and non-Hodgkin lymphoblastic lymphoma (NHL) has increased steadily since the 1970s. With contemporary risk-directed therapy the 5-year overall survival rate is nowadays nearly 90%.1 This major improvement in outcome results from a better understanding of the disease, more accurate stratification according to prognostic risk factors and optimized drug use.

One of the essential components in treatment protocols for childhood ALL and NHL is asparaginase (ASNase), an enzyme that catalyzes serum asparagine and glutamine deamination. In contrast to healthy cells, lymphoblasts are unable to produce endogenous asparagine because of the lack of the enzyme asparagine synthetase, and rely on plasma levels of this amino acid for protein synthesis. A depletion of plasma asparagine by ASNase results in selective apoptosis of lymphoblasts.2 The adverse effects of ASNase are mainly related to disturbances in normal protein synthesis and to hypersensitivity reactions.

Several study groups have demonstrated a benefit of intensive ASNase treatment compared to less intensive regimens.3-7 In the European Organization for Research and Treatment of Cancer Children’s Leukemia Group (EORTC-CLG) 58881 trial, our study group showed that the use of a more potent ASNase with a longer half-life improved patients’ outcome. Patients receiving native E. coli ASNase 10,000 IU/m2 twice weekly in induction and late intensification experienced longer event-free and overall survival than patients randomized to Erwinia ASNase at the same dosage and frequency.3,8 These results

were confirmed by the DFCI 95-01 trial for patients treated with 25,000 IU/m2 weekly of native E. coli or Erwinia ASNase4. Both studies made clear that the use of native E. coli ASNase was responsible for a reduction in relapse rate, albeit resulting in more toxicity. Based on these studies and data from pharmacokinetic and ASNase activity monitoring, it is known that higher doses of Erwinia ASNase and shorter dose intervals are required to achieve ASNase activity that is adequate in comparison with that provided

by E. coli ASNase. Several study groups now incorporate ASNase activity monitoring for the optimization of ASNase therapy.

Not only do the formulation and dose of ASNase seem to play a role, but the duration of the ASNase treatment also has an impact on survival. The Italian, Dutch and Hungarian IDH-ALL-91 trial demonstrated an improved outcome for patients who received extended high-dose native E. coli ASNase treatment in consolidation therapy (25,000 IU/m2 weekly for 20 weeks) compared to the same treatment based on a reduced-intensity BerlinFrankfurt-Münster (BFM)-backbone without extra ASNase.5 The benefit of prolonged native E. coli ASNase for T-cell ALL and NHL was also proven in the Pediatric Oncology Group (POG) study 8704 in which a high-dose ASNase regimen of 25,000 IU/m2 given weekly for 20 weeks starting from consolidation resulted in higher continuous complete remission rates.6 An increase in eventfree survival was also observed in the subsequent Dana Farber Cancer Institute (DFCI) trial (ALL 91-01) as a result of the prolongation of high-dose native E. coli ASNase (25,000 IU/m2 weekly) or pegylated (PEG)-ASNase (2,500 IU/m2 every other week) for 20 to 30 weeks during intensification therapy.7 However, patients were not random-

Figure 1. General scheme of the EORTC-CLG 58951 trial. The EORTC-CLG trial 58951 embedded three main randomized comparisons: [R1] the value of prednisolone (PRED, 60 mg/m²/day) versus dexamethasone (DEX, 6 mg/m²/day) in induction for all patients;12 [R2] the value of prolonged courses of ASNase throughout consolidation and late intensification for all non-very high risk (non-VHR) patients; and [R3] the value of vincristine (VCR) and corticosteroid pulses introduced in continuation therapy for average risk (AR) patients.13 IA: induction phase; IB: consolidation phase; II A+B: late intensification phase; VCR: vincristine; VLR: very low risk (group); AR: average risk (group); VHR: very high risk (group).

1728

haematologica | 2017; 102(10)

Prolonged asparaginase therapy in childhood ALL

ized to a standard versus prolonged schedule of ASNase, and the duration of ASNase was actually based on ASNase tolerance.

In contrast, the controversy of intensified ASNase treatment was highlighted in the Associazione Italiana Ematologia e Oncologia Pediatrica (AIEOP) ALL-91 trial, in which highdose E. coli ASNase treatment (25,000 IU/m2 weekly for 20 weeks during late intensification and early continuation) did not have an impact on disease-free survival compared to standard E. coli ASNase treatment (4 doses of 10,000 IU/m2 during late intensification).9 The mode of administration (continuous or discontinuous ASNase administration) may have different implications on the development of ASNase hypersensitivity and silent inactivation of ASNase. The probability of the appearance of hypersensitivity reactions does not only increase with the number of administrations within the same cycle but also in discontinuous administration schedules, in which ASNase is reintroduced after an ASNase-free interval.10 Furthermore, more intensive ASNase therapy goes along with an increase in ASNase-related toxicities. With the advantages and side effects of ASNase in mind, the optimal dosage and number of ASNase administrations remain subject of debate. The EORTC-CLG, therefore, conducted a randomized phase III trial 58951, comparing a conventional native E. coli ASNase regimen with a prolonged native E. coli ASNase therapy in a BFM-based treatment (Figure 1).

Methods Patients As previously described,12,13 the EORTC-CLG 58951 protocol included all children aged less than 18 years, with previously untreated ALL of French-American-British (FAB) L1 or L2 morphology whatever the immuno-phenotype, or precursor B- or Tlymphoblastic NHL. Patients with ALL of FAB L3 morphology and diffuse large cell B-cell lymphoma, Burkitt lymphoma or highgrade B-cell lymphoma Burkitt-like, were excluded. Infants (<1 year) and patients with Philadelphia-positive ALL were allocated to separate disease-specific protocols (Interfant protocol for infants and Esphall protocol for Philadelphia-positive patients).

As described in the Online Supplementary Appendix, patients were assigned to different risk groups: very low risk, average risk low, average risk high and very high risk.12,13

Before entering the study, informed consent was obtained from the parents or legal guardians according to the Declaration of Helsinki. Both the EORTC Protocol Review Committee and the local institutional ethical committees of each participating center approved the study design.

Treatment and study design The EORTC-CLG 58951 study was based on a BFM-like protocol with a prephase, four-drug induction (IA), consolidation phase (IB), interval phase with central nervous system-directed treatment (without cranial radiotherapy), late intensification (IIA+B) and maintenance therapy (Figure 1). Online Supplementary Table S1A,B outlines the study design of the EORTC-CLG 58951 study for the very low risk, average risk low and average risk high risk groups.12,13 All non-very high-risk patients in complete remission or good partial response at the end of induction were randomly assigned to receive either conventional native E. coli ASNase therapy (12 doses, short-ASNase) or prolonged native E. coli ASNase therapy (24 doses, long-ASNase) (Figure 1). Patients in the short-ASNase arm had to receive 8x10,000 U/m2 in induction (IA) and 4x10,000 U/m2 in late intensification (IIA). The patients in the long-ASNase arm had to receive 12 extra doses, 8x5,000 U/m2 native E.coli ASNase in consolidation (IB) and 4x5,000 U/m2 extra doses in IIA. All native E.coli ASNase doses were scheduled twice a week. Patients with grade â&#x2030;Ľ2 allergy to native E. coli ASNase were switched to equivalent doses of Erwinia ASNase: four doses of native E. coli ASNase (5,000 or 10,000 U/m2) were replaced by six doses of 20,000 U/m2 Erwinia ASNase (3 per week, in view of the shorter half-life). As Erwinia ASNase was not available in Europe from the end of 2002 until 2006, patients were, in that time period, switched to equivalent doses of pegylated E. coli ASNase, and six doses of 20,000 U/m2 Erwinia ASNase in 2 weeks substituted by one dose of 2,500 U/m2 of pegylated E. coli ASNase (Table 1).

Definitions Definitions of central nervous system disease and complete remission have been published previously12,13 and are summarized in the Online Supplementary Appendix. Treatment-related toxicity was graded according to the National Cancer Institute (NCI) Common Toxicity Criteria version 1994.15

Statistical analysis The primary endpoint of this study was disease-free survival, secondary endpoints were overall survival and toxicity. In order to detect an increase in the 5-year disease-free survival rate from

Table 1. Native E.coli asparaginase administrations in different treatment cycles and switch to Erwinia or pegylated E.coli asparaginase in case of grade 2 or more allergy.

Phase Induction (IA) Consolidation (IB) Short-ASNase Long-ASNase Late-intensification (IIA) Short-ASNase Long-ASNase Maintenance (only AR2)

Native E.coli ASNase* 2

Erwinia ASNase** 2

Pegylated E.coli ASNase*

10,000 U/m x 8

20,000 U/m x 12

2,500 U/m2 x 2

0 5,000 U/m2x 8

0 20,000 U/m2 x12

0 2,500 U/m2 x 2

10,000 U/m2 x 4 10,000 U/m2 x 4 + 5,000 U/m2 x 4 25,000 U/m2 x 1

20,000 U/m2 x 6 20,000 U/m2 x 12 25,000 U/m2 x 2

2,500 U/m2 x 1 2,500 U/m2 x 2 2,500 U/m2 x 1

* intravenously. **intramuscularly or intravenously (on discretion of the treating physician). AR2: average risk high.

haematologica | 2017; 102(10)

1729

V. Mondelaers et al.

84% (short-ASNase arm) to 89% (long-ASNase arm), corresponding to a treatment hazard ratio (HR) of 0.67, 1,500 patients had to be randomized, of whom 212 had to be followed until an event (log-rank test, 2-sided alpha=5%, power=80%).

Further information on endpoint definitions, randomization technique, stratification factors, and the statistical analysis16 is included in the Online Supplementary Appendix.

Results Patients’ characteristics Between December 1998 and August 2008, 1,552 nonvery high-risk patients (1,481 with ALL and 71 with NHL)

were randomized in the EORTC-CLG trial 58951 to receive either long-ASNase (n=775) or short-ASNase (n=777). Of those patients 14.7% were very low risk, 66.2% average risk low and 19.1% average risk high (Table 2). Twenty-nine (16 long-ASNase and 13 shortASNase) patients were considered a posteriori as being ineligible after randomization17,18 (Figure 2). The reasons for ineligibility were: no complete remission/good partial response (1 long-ASNase versus 3 short-ASNase), previous toxicity (4 versus 2), very high risk (1 versus 6), minimal residual disease ≥10-2 (10 versus 2). Nevertheless, those patients were included in the disease-free survival and overall survival analyses according to the intent-to-treat principle.

Table 2. Patient and disease characteristics according to the native E. coli ASNase randomization.

Type Immunophenotype First randomization

Third randomization

Sex Age (years)

White blood cell count (x109/L)

NCI risk group* EORTC risk group

Cerebrospinal fluid status

Response to prephase† MRD at end of IA

ALL NHL B-lineage T-lineage Prednisolone Dexamethasone Registered No pulses Pulses Not randomized Male Female <1 1-<2 2-<6 6 - < 10 > 10 < 20 20- < 50 50 - < 100 > 100 Standard risk High risk VLR AR1 AR2 CNS-1/TPLCNS-2/TLP+ CNS-3 Missing “Poor response” “Good response” ≥ 10-2 10-3 - < 10-2 < 10-3 Missing

Long-ASNase n=775 N. of pts(%)

Short-ASNase n=777 N. of pts (%)

745 (96.1) 30 (3.9) 678 (87.5) 97 (12.5) 389 (50.2) 383 (49.4) 3 (0.4) 99 (12.7) 99 (12.7) 579 (74.6) 416 (53.7) 359 (46.3) 4 (0.5) 59 (7.6) 409 (52.8) 144 (18.6) 159 (20.5) 568 (73.3) 107 (13.8) 52 (6.7) 48 (6.2) 532 (68.6) 243 (31.4) 113 (14.6) 514 (66.3) 148 (19.1) 721 (93.0) 39 (5.1) 11 (1.4) 4 (0.5) 0 (0.0) 775 (100) 22 (2.8) 371 (47.9) 255 (32.9) 127 (16.4)

736 (94.7) 41(5.3) 670 (86.2) 107 (13.8) 389 (50.1) 382 (49.2) 6 (0.8) 94 (12.1) 94 (12.1) 587 (75.8) 416 (53.5) 361 (46.5) 0 (0) 55 (7.1) 381 (49.0) 173 (22.3) 168 (21.6) 568 (73.1) 104 (13.4) 55 (7.1) 50 (6.4) 519 (66.8) 258 (33.2) 114 (14.7) 514 (66.1) 149 (19.2) 733 (94.3) 32 (4.1) 10 (1.3) 2 (0.3) 1 (0.1) 776 (99.9) 13 (1.7) 384 (49.5) 242 (31.1) 138 (17.7)

ALL: acute lymphoblastic leukemia; NHL: non-Hodgkin lymphoma; NCI: National Cancer Institute; TLP-, traumatic lumbar puncture without blasts; TLP+, traumatic lumbar puncture with blasts; CNS: central nervous system;VLR: very low risk; AR: average risk; AR1: averange risk low; AR2: average risk high;VHR: very high risk; MRD: minimal residual disease; IA: induction; N: number of patients. *NCI Standard Risk: NHL or ALL with white blood cell count <50x109/L and age 1 - <10 years; NCI High Risk: NHL or ALL with white blood cell count ≥50x109/L or age <1 or age ≥10 years. †Response to prephase (1 week prednisolone or dexamethasone and 1 intrathecal methotrexate): “poor response”: > 1x109 blasts/L,“good response”: < 1x109 blasts/L.

1730

haematologica | 2017; 102(10)

Prolonged asparaginase therapy in childhood ALL

The baseline characteristics of the patients and leukemia, first randomization, response to prephase and induction were evenly distributed in both treatment groups (Table 2).

Among the 775 patients allocated to the long-ASNase group, three patients never got the extra doses for unknown reasons and received the short-ASNase arm. In 58 patients the long-ASNase was interrupted for various reasons: relapse in 32, excessive toxicity in 17, parental refusal in two, and other reasons in seven patients.

Of the 777 patients allocated to the short-ASNase group, two patients erroneously received the long-ASNase treatment, and two did not receive the consolidation block. In 50 patients ASNase was interrupted prematurely: two because of failure to achieve remission, 34 because of relapse, eight because of excessive toxicity, two because of protocol violation and four for other reasons (Figure 2).

A total of 1,391 patients (641 treated with long-ASNase and 750 with short-ASNase) received the total number of administrations according to the protocol and randomization arm. In the long-ASNase group, 103 (13.3%) patients received an incomplete number of administrations and/or insufficient dose of ASNase, while 13 (1.7%) of the 777 patients in the short-ASNase arm were undertreated.

Most of the protocol violations were due to a substitution with insufficient amount and/or doses of Erwinia ASNase (58 in the long-ASNase arm and 7 in the short-ASNase arm) after allergic reactions to native E. coli ASNase.

notype, EORTC risk group and the steroid assigned during induction (prednisolone versus dexamethasone) did not reveal significant heterogeneity in the treatment differences regarding disease-free survival and overall survival (Figures 4 and 5, Online Supplementary Figures S1-S3). In contrast, in the NCI standard-risk ALL group, the longASNase treatment had a positive effect [8-year diseasefree survival: 89.7% (long-ASNase) versus 85.0% (shortASNase), HR: 0.70; 99% confidence interval (CI): 0.441.13; P=0.057] (Figure 4, Online Supplementary Figure S4A), whereas in the NCI high-risk ALL group, a trend in the opposite direction was observed (HR: 1.36; 99% CI: 0.772.41; P=0.17) (Online Supplementary Figure S4C) (test for heterogeneity: P=0.02). The favorable effect of longASNase in this subgroup of patients was due to a decreased number of bone marrow relapses (34 isolated or combined bone marrow relapses versus 53 in the shortASNase group). The incidence of isolated central nervous system (4 versus 5) and other isolated relapses (5 versus 6) was equally low in both arms. In contrast, regarding overall survival, no treatment difference was noted in NCI standard-risk (HR: 0.89) and high-risk ALL patients (Figure 5, Online Supplementary Figure S4B,D) (test for heterogeneity: P=0.52).

Restraining the treatment comparisons to eligible patients only (n=1,523), in particular to those receiving ASNase treatment according to the protocol guidelines (n=1,391) and to those patients without grade 2-4 ASNase allergy (n=1,154), similar results were obtained regarding disease-free survival and overall survival (data not shown).

Treatment results

Toxicity

At a 7-year median follow-up there were 97 events in the long-ASNase arm and 111 in the short-ASNase group (Table 3). The two patients in the short-ASNase group who failed to achieve complete remission were considered as having had events at time 0. Relapse occurred in 87 children (11%) in the long-ASNase arm and in 102 children (13%) in the short-ASNase arm. The site of relapse for the long-ASNase versus short-ASNase comparison was isolated bone marrow (54 versus 60), combined bone marrow (16 versus 23), isolated central nervous system (10 versus 11) and other isolated relapse (7 versus 8). Death in complete remission occurred in ten children in the longASNase arm and in seven children in the short-ASNase arm. In the long-ASNase arm, three patients died of organ toxicity, four due to an infection, one due to graft-versushost disease after hematopoietic stem cell transplantation for a secondary myelodysplasia, one patient died of a preexisting cardiomyopathy during treatment and there was one unexplained, sudden death during late intensification.

In the short-ASNase arm, one patient died due to organ toxicity, two due to an infection, two due to a secondary malignancy (1 central nervous system tumor and 1 medulloblastoma), one due to graft failure after hematopoietic stem cell transplantation for a secondary acute myeloid leukemia and one patient died of a pre-existing cardiomyopathy during treatment.

The 8-year disease-free survival rate (± standard error) was 87.0±1.3% in the long-ASNase and 84.4±1.4% in short-ASNase group (HR: 0.87; P=0.33) (Figure 3A). The 8year overall survival rate was comparable in the two treatment arms: 92.6±1.0% in the long-ASNase group and 91.3±1.2% in the short-ASNase group (HR: 0.89; P=0.53) (Figure 3B). Subgroup analyses according to immunophe-

The incidence of grade 3-4 infection was higher in the long-ASNase group than in the short-ASNase group during consolidation (25.2% versus 14.4%) and late intensification (22.6% versus 15.9%). This difference was more pronounced in patients who were randomly assigned to dexamethasone in induction (Table 3).

In both arms the incidence of grade 3-4 pancreatitis was low (<2%) with a trend to a slightly higher incidence in the long-ASNase group during consolidation (15 versus 4) and late intensification (12 versus 3). Grade 2-4 allergy to ASNase in the long-ASNase versus short-ASNase group was 22.6% versus 0.2% during the consolidation phase and 10.3% versus 21.4% in late intensification phase. In the long-ASNase group, the incidence of grade 2-4 allergic reactions was higher during consolidation (22.1%) than in late intensification (10.3%). The type of corticosteroid used in induction (prednisolone or dexamethasone) had no impact on the incidence of allergy grade during consolidation or late intensification. During the whole treatment period, the incidence of grade 2-4 allergy was 30.5% in the long-ASNase arm and 21.7% in the short-ASNase arm.

The incidences of other grade 3-4 toxicities reported during consolidation and late intensification were similar in the two randomization groups (Table 3).

Treatment applicability

haematologica | 2017; 102(10)

Discussion The prognosis of acute lymphoblastic malignancies has been improved steadily by the use of multidrug chemotherapy of which one of the key elements is ASNase. Several contemporary collaborative studies have demonstrated that intensification of ASNase therapy 1731

V. Mondelaers et al.

could offer an advantage in terms of outcome for the patients. Nevertheless, extrapolation of these findings is difficult as different dosages and sources of ASNase were applied in different chemotherapy backbones. This was the reason why the EORTC-CLG study group decided to investigate the effect of extended native E. coli ASNase therapy in consolidation and late intensification on top of a BFM-backbone for all non-very high risk patients as one of the randomized issues in the 58951 study.

We observed a slightly higher 8-year disease-free survival in the long-ASNase group (87.0% versus 84.4%; HR: 0.87; P=0.33) and a similar overall survival (92.6% versus

91.3%; HR: 0.89; P=0.53) rate in the long-ASNase group compared to the short-ASNase group. The analogous outcome rates were confirmed in subgroup analysis according to immunophenotype, EORTC risk group, and prednisolone/dexamethasone randomization. Exploratory analysis with an a posteriori classification of the ALL patients according to NCI risk criteria showed a trend towards higher 8-year disease-free survival rate for the long-ASNase patients in the NCI standard-risk group due to a lower number of bone marrow relapses. However, the 8-year overall survival rate was comparable in the two groups. Excluding patients who did not receive the

Figure 2. CONSORT statement in flow diagram.

1732

haematologica | 2017; 102(10)

Prolonged asparaginase therapy in childhood ALL

ASNase treatment as planned according to the allocated arm or excluding allergic patients did not affect the disease-free survival rates. Our data are in accordance with the results of the AIEOP ALL-91 study, in which no advantage could be demonstrated for intermediate-risk patients randomized to receive either standard treatment (4x10,000 IU/m2 ASNase during late intensification) or weekly high-dose ASNase 25,000 IU/m2 for 20 weeks during late intensification and early continuation.9 The International BFM study group showed in the IDH-ALL90 trial that extended high-dose native E. coli ASNase (25,000 IU/m2 weekly for 20 doses, starting from the beginning of continuation therapy) versus no ASNase could improve outcome in standard-risk ALL patients treated with a reduced-intensity BFM-based protocol. In this protocol no consolidation block IB was given and two doses of anthracyclines were omitted in late intensification.5 Pession et al.5 postulated that the benefit obtained from the extended native ASNase therapy in this study probably overcame the reduced leukemia control by the applied treatment reduction.

As in most BFM-based studies at that time, all EORTC-

CLG patients received classical doses of 10,000 IU/m2 of native E. coli ASNase. For the additional doses in the longASNase arm, patients received 5,000 IU/m2 of native E. coli ASNase. The dose reduction was mainly dictated by the fear of an excess in toxicity for those patients. Ahlke et al. studied the pharmacokinetic and pharmacodynamic properties of different dosages of native E. coli ASNase. They found that 96% of the patients had complete asparagine depletion and sufficient native E. coli ASNase activity with a dose of 5,000 IU/m2 every third day.19 This observation gave sufficient evidence to apply a lower dosage of the additional doses of native E. coli ASNase during consolidation and late intensification in the long-ASNase group. As previously reported by Domenech et al.,12 the number of infections during induction was similar for the patients treated with dexamethasone or prednisolone in the EORTC-CLG 58951 trial. The concern of an excess in toxicity due to extended native E. coli ASNase use turned out to be correct as patients in the long-ASNase arm had a trend towards more grade 3-4 infections during consolidation and late intensification, especially when they received dexamethasone in induction. Others have reported that

Table 3. Outcome, type of event, and toxicity according to the randomized arm.

Long ASNase n=775 N. of pts (%) Disease-free survival status

Survival status Toxicity

No CR CCR Relapse BM CNS Other isolated BM+CNS BM+other BM+other+CNS Death in CR Alive Dead Consolidation Grade 3-4 infection PRED (n=776) DEX (n=764) Grade 3-4 hyperglycemia Grade 3-4 pancreatitis Grade 3-4 thrombosis Grade 3-4 hemorrhage Grade 3-4 hepatotoxicity Grade 2-4 allergy Switch of asparaginase Late Intensification Grade 3-4 infection PRED (n=751) DEX (n=735) Grade 3-4 hyperglycemia Grade 3-4 pancreatitis Grade 3-4 thrombosis Grade 3-4 hemorrhage Grade 3-4 hepatotoxicity Grade 2-4 allergy Switch of asparaginase

0 (0) 678 (87.5) 87 (11.2) 54 (7.0) 10 (1.3) 7 (0.9) 11 (1.4) 5 (0.6) 0 (0) 10 (1.3) 723 (93.3) 52 (6.7) n=774 195 (25.2) 90 (11.6) 104 (13.6) 11 (1.4) 15 (1.9) 9 (1.7) 7 (0.9) 152 (19.6) 174 (22.6) 172 (22.2) n=738 167 (22.6) 81 (10.8) 86 (11.7) 30 (4.1) 12 (1.6) 6 (0.8) 4 (0.5) 75 (10.2) 76 (10.3) 66 (8.9)

Short ASNase n=777 N. of pts (%) 2 (0.3) 666 (85.7) 102 (13.1) 60 (7.7) 11 (1.4) 8 (1.0) 11 (1.4) 10 (1.3) 2 (0.3) 7 (0.9) 718 (92.4) 59 (7.6) n=775 112 (14.4) 67 (8.6) 44 (5.8) 6 (0.8) 4 (0.5) 7 (0.9) 5 (0.6) 136 (17.6) 2 (0.2) 4 (0.4) n=748 119 (15.9) 67 (8.9) 52 (7.0) 22 (2.9) 3 (0.4) 5 (0.7) 3 (0.4) 64 (8.6) 160 (21.4) 174 (23.3)

No CR: no complete remission; CCR: continuous complete remission; CNS: central nervous system; BM: bone marrow; PRED: prednisolone; DEX: dexamethasone; N: number of patients.

haematologica | 2017; 102(10)

1733

V. Mondelaers et al.

dexamethasone was associated with a higher risk of infection during consolidation treatment, particularly in older children and adolescents.10,20 The excess in severe infections during consolidation and late intensification could be the result of the additional myelosuppressive effect of prolonged administration of native E. coli ASNase in combination with intensive treatment10,21 and decreased immunoglobulin production resulting from decreased protein synthesis. Although moderate, hematologic toxicity from prolonged ASNase treatment might increase the immunosuppressive effect of dexamethasone, especially due to the fact that ASNase diminishes dexamethasone clearance.22 As expected due to the longer exposure to the drug, the cumulative incidence of grade 2-4 allergy was higher in the long-ASNase arm than in the short-ASNase arm (30.5% versus 21.7%). Although intermittent administra-

tion of native E. coli ASNase is considered to be a risk factor for allergy, our study demonstrated that the rate of grade 2-4 allergic reactions in the long-ASNase arm was higher in consolidation than in late intensification (22.1% versus 10.3%). Concomitant administration of corticosteroids in late intensification reduces the risk of clinical hypersensitivity reactions, whereas the protective effect of the corticosteroids is absent in consolidation.2,23 Although concurrent administration of corticosteroids reduces the clinical signs of allergic reactions, it does not influence the neutralizing antibodies and the subsequent inactivity of ASNase. It is known that not only the frequency of allergy, but also the incidence of silent inactivation increases with longer duration of native E. coli ASNase exposure. As native E. coli ASNase activity monitoring was not performed in the EORTC 58951 trial, the incidence of silent inactivation in this study remains unclear.

A

B

Figure 3. (A) Disease-free survival and (B) overall survival according to the randomized arm.

1734

haematologica | 2017; 102(10)

Prolonged asparaginase therapy in childhood ALL

However, it is likely that the long treatment arm had more silent inactivators who subsequently received insufficient ASNase in the late intensification phase. The EORTC 58951 study ran in a period in which therapeutic drug monitoring was not routinely performed to detect silent inactivation and subtherapeutic asparaginase activity.

Nowadays it is well known that therapeutic drug monitor-

ing is essential in order to optimize asparaginase treatment and therefore an increasing number of study groups are including such monitoring in their study protocols.

Despite the fact that grade 3-4 pancreatitis was rare in both treatment arms (<2%), there was a trend to a slightly higher incidence among patients randomized to the longASNase group. The incidences of other typical ASNase-

Figure 4. Forest plot regarding disease-free survival according to the randomization arm. ALL: acute lymphoblastic leukemia; NHL: non-Hodgkin lymphoma; VL: very low risk; AR1: average risk low; AR2: average risk high; DEX: dexamethasone; PRED: prednisolone; WBC: white blood cell count; NCI: National Cancer Institute; HR: hazard ratio; CI: confidence interval. *NCI Standard Risk: ALL with WBC <50x109/L and age 1 - <10 years; NCI High Risk: ALL with WBC â&#x2030;Ľ50x109/L or age <1 or age â&#x2030;Ľ10 years.

haematologica | 2017; 102(10)

1735

V. Mondelaers et al.

related grade 3-4 toxicities, such as hyperglycemia, thrombosis, hemorrhage and hepatotoxicity, were similar in the two treatment groups.

Two large, randomized studies3,4 showed superiority of native E. coli ASNase over Erwinia ASNase when given at the same dose and frequency. In the EORTC-CLG trial

58881, 700 patients were randomized to receive either native E. coli or Erwinia ASNase at a dosage of 10,000 IU/m2 twice weekly during induction and late intensification. The patients randomized to receive Erwinia ASNase had lower 6-year event-free survival (59.8% versus 73.4%; P=0.0004) and overall survival (75.1% versus 83.9%;

Figure 5. Forest plot regarding overall survival according to the randomization arm. ALL: acute lymphoblastic leukemia; NHL: non-Hodgkin lymphoma; VLR: very low risk; AR1: average risk low; AR2: average risk high; DEX: dexamethasone; PRED: prednisolone; WBC: white blood cell count; NCI: National Cancer Institute; HR: hazard ratio; CI: confidence interval. *NCI Standard Risk: ALL with WBC <50x109/L and age 1 - <10 years; NCI High Risk: ALL with WBC â&#x2030;Ľ50x109/L or age <1 or age â&#x2030;Ľ10 years.

1736

haematologica | 2017; 102(10)

Prolonged asparaginase therapy in childhood ALL

P=0.002) rates than those randomized to native E. coli ASNase.3 In the DFCI 95-01 trial, 286 standard and highrisk patients received either 25,000 IU/m2 weekly of native E. coli or Erwinia ASNase. Patients treated with Erwinia ASNase had an inferior 10-year event-free survival rate (75.2% versus 84.6%; P=0.02) and overall survival rate (85.3% versus 93.1%; P=0.04).4 In both studies, patients treated with Erwinia ASNase experienced less toxicities but more relapses,3,4 Although all patients in the EORTCCLG 58951 study received native E. coli ASNase as firstline treatment, a large proportion of the patients in the long-ASNase arm had to be switched to Erwinia ASNase due to allergic reactions. Based on the results of the EORTC-CLG 58881 study, the subsequent 58951 trial recommended higher dosages of Erwinia ASNase (20,000 IU/m2 3 times a week) for patients with an allergic reaction. Nevertheless, despite these dose recommendations 7.5% of patients in the long-ASNase group erroneously received insufficient doses of Erwinia ASNase, which was only the case in 0.9% of the patients assigned to shortASNase treatment.

As already mentioned, at the start of the study, therapeutic drug monitoring of ASNase was not common practice in childhood ALL protocols. An increasing number of study groups now incorporate the monitoring of ASNase activity to optimize treatment efficacy. It is widely accepted that ASNase activity is optimal if trough activity levels are >100U/L.2 A recent DCOG study highlighted the short activity of Erwinia ASNase by measuring trough levels of ASNase activity. Effective ASNase activity levels were found in 100% (at 48 h) and 33% (at 72 h) of patients treated with 20,000 IU Erwinia ASNase twice or three times a week.24 In our study, allergic patients received Erwinia ASNase three times a week, mostly on Monday, Wednesday and Friday. Although ASNase monitoring was not routinely performed in our study, extrapolating the data of Tong et al. we assume that at least some of our patients receiving Erwinia ASNase were undertreated during the weekend. 24This, together with the fact that fewer patients in the long-ASNase arm received all intended doses as planned in the protocol, could mitigate the potential benefit of prolonged ASNase treatment. Moreover, prolonged native E. coli ASNase therapy resulted in an increase in allergic reactions, pancreatitis and infections which can hamper the treatment efficacy by omission or delay of other essential chemotherapeutics.

It is important to note that ASNase therapy in our study consisted of treatment for 12 weeks in the long-ASNase arm versus 6 weeks in the short arm, given in a discontinuous way. Our extended arm is not comparable with the very prolonged and continuous use of ASNase (30 weeks) in the DCOG and DFCI protocols. In the DCOG ALL10 protocol, patients with medium risk according to minimal

References 1. Pui CH, Yang JJ, Hunger SP, et al.

Childhood acute lymphoblastic leukemia: progress through collaboration. J Clin Oncol. 2015;33(27):2938–2948.

2. Avramis VI, Tiwari PN. Asparaginase (native ASNase or pegylated ASNase) in the treatment of acute lymphoblastic

haematologica | 2017; 102(10)

residual disease received therapy intensification including 30 weeks of PEG-ASNase exposure and dexamethasone/ vincristine pulses. This resulted in a significantly higher 5year event-free survival rate compared with that in historical controls (88% versus 76%; P=0.056).25 An increase in event-free survival was also observed in the DFCI ALL 9101 trial for patients who tolerated prolonged high-dose native E. coli ASNase (25,000 IU/m2 weekly) or PEGASNase (2,500 IU/m2 every other week) for 30 weeks compared to patients who tolerated less than 20 weeks of extended ASNase therapy.7 The relatively short extension of ASNase treatment in our trial together with the increased risk of silent inactivation and allergy due to the discontinuous administrations, may have nullified the effect seen in the very prolonged, continuous schedules of the DCOG and DFCI protocols. Moreover, we used native E. coli ASNase, which is being increasingly replaced by PEG-ASNase in contemporary childhood ALL trials, due to the longer action and the lower incidence of allergy and silent inactivation.

In conclusion, this study demonstrates that a 6-week prolongation of native E. coli ASNase therapy in a discontinuous administration schedule (during consolidation and late intensification) does not significantly improve patients’ overall outcome. Our study underscores the hypothesis of Pession that the overall treatment intensity of this BFM-based regimen, with four-drug induction for all patients, already results in maximal therapeutic efficacy, whereas intensification with prolonged native E. coli ASNase therapy does not add any benefit for these patients.5 One has to be cautious to extrapolate our results to other studies especially if less intensive induction therapy, other asparaginase preparations or other administration schedules are used. Future studies should aim to optimize the efficacy of ASNase therapy, not only by prolonged administration but by the use of PEGASNase and by therapeutic drug monitoring in real time in order to promptly intercept silent inactivation and subtherapeutic asparaginase activity.

Acknowledgments The authors would like to thank the EORTC-CLG study group members for their participation in the study and the EORTC HQ Data Management Department members (Séraphine Rossi, Lies Meirlaen, Liv Meert, Aurélie Dubois, Christine Waterkeyn, Alessandra Busato, Isabel VandeVelde and Gabriel Solbu) for their support in this trial as well as Drs. Francisco Bautista (EORTC-CLG fellow) and Matthias Karrasch (former EORTC Clinical Research Physician).

A complete list of the members of the Children's Leukemia Group of the European Organization for Research and Treatment of Cancer appears in the Online Supplementary Data.

leukemia. Int J Nanomedicine. 2006; 1(3):241–254.

3. Duval M, Suciu S, Ferster A, et al.

Comparison of Escherichia coli-asparaginase with Erwinia-asparaginase in the treatment of childhood lymphoid malignancies: results of a randomized European Organisation for Research and Treatment of Cancer-Children's Leukemia Group phase 3 trial. Blood. 2002;99(8):2734–2739.

4. Moghrabi A, Levy DE, Asselin B et al.

Results of the Dana-Farber Cancer Institute ALL Consortium Protocol 95-01 for children with acute lymphoblastic leukemia.

Blood. 2007;109(3):896–904.

5. Pession A, Valsecchi MG, Masera G, et al.

Long-term results of a randomized trial on extended use of high dose L-asparaginase for standard risk childhood acute lymphoblastic leukemia. J Clin Oncol. 2005;

1737

V. Mondelaers et al.

23(28):7161–7167.

6. Amylon MD, Shuster J, Pullen J, et al.

Intensive high-dose asparaginase consolidation improves survival for pediatric patients with T cell acute lymphoblastic leukemia and advanced stage lymphoblastic lymphoma: a Pediatric Oncology Group study. Leukemia. 1999; 13(3):335–342.

7. Silverman LB, Gelber RD, Dalton VK, et al.

Improved outcome for children with acute lymphoblastic leukemia: results of DanaFarber Consortium Protocol 91-01. Blood.

2001;97(5):1211–1218.

8. Vilmer E, Suciu S, Ferster A, et al. Longterm results of three randomized trials (58831, 58832, 58881) in childhood acute lymphoblastic leukemia: a CLCG-EORTC report. Leukemia. 2000;14(12):2257-2266.

9. Rizzari C, Zucchetti M, Conter V, et al. Lasparagine depletion and L-asparaginase activity in children with acute lymphoblastic leukemia receiving i.m. or i.v. Erwinia C. or E.coli L-asparaginase as first exposure.

Ann Oncol. 2000;11(2):189–193.

10. Müller HJ, Boos J. Use of L-asparaginase in childhood ALL. Critical Rev Oncol Hematol. 1998;28(2):97–113.

11. Mondelaers V, Suciu S, De Moerloose B, et al. Prolonged E. Coli asparaginase therapy does not improve significantly the outcome for children with low and average risk acute lymphoblastic leukemia (ALL) and non hodgkin lymphoma (NHL): final report of the EORTC-CLG randomized phase III trial 58951 [abstract]. Blood 2012;120:134.

12. Domenech C, Suciu S, De Moerloose B, et al. Dexamethasone (6 mg/m²/day) and

1738

13.

14.

15.

16.

17.

18.

19.

prednisolone (60 mg/m²/day) were equally effective as induction therapy for childhood acute lymphoblastic leukemia in the EORTC CLG 58951 randomized trial.

Haematologica. 2014;99(7):1220-1227.

De Moerloose B, Suciu S, Bertrand Y, et al.

Improved outcome with pulses of vincristine and corticosteroids in continuation therapy of children with average risk acute lymphoblastic leukemia (ALL) and lymphoblastic non-Hodgkin lymphoma (NHL): report of the EORTC randomized phase 3 trial 58951. Blood. 2010;116(1):36–44.

Cavé H, Van Der Werff Ten Bosch J, Suciu S, et al. Clinical significance of minimal residual disease in childhood acute lymphoblastic leukemia. New Engl J Med. 1998;339(9):591-598. https://www.eortc.be/services/doc/ctc/ Kalbfleisch JD, Prentice RL. The Clinical Analysis of Failure Time Date. 2nd ed.

Hoboken NJ: Wiley InterScience; 2002.

Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. Trials. 2010;11:32.

Moher D, Hopewell S, Schulz KF, et al.

CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trial. BMJ. 2010;340:c869.

Ahlke E, Nowak-Gottl U, SchulzeWesthoff P, et al. Dose reduction of asparaginase under pharmacokinetic and pharmacodynamic control during induction therapy in children with acute lymphoblastic leukaemia. Br J Haematol.

1997;96(4):675–681.

20. Vrooman LM, Stevenson KE, Supko JG, et al. Consolidation dexamethasone and individualized dosing of Escherichia coli Lasparaginase each improve outcome of children and adolescents with newly diagnosed acute lymphoblastic leukemia: Results from a randomized study–DanaFarber Cancer Institute ALL Consortium Protocol 00–01.

J Clin Oncol.

2013;31(9):1202–1210.

21. Merryman R, Stevenson KE, Gostic WJ 2nd, et al. Asparaginase-associated myelosuppression and effects on dosing of other chemotherapeutic agents in childhood acute lymphoblastic leukemia. Pediatr Blood Cancer. 2012;59(5):925-927.

22. Kawedia JD, Liu C, Pei D, et al.

Dexamethasone exposure and asparaginase antibodies affect relapse risk in acute lymphoblastic leukemia. Blood. 2012; 119(7):1658–1664.

23. Pieters R, Hunger SP, Boos J, et al. Lasparaginase treatment in acute lymphoblastic leukemia: a focus on Erwinia asparaginase. Cancer. 2011;117(2):238-249.

24. Tong WH, Pieters R, de Groot-Kruseman HA, et al. The toxicity of very prolonged courses of PEGasparaginase or Erwinia asparaginase in relation to asparaginase activity, with a special focus on dyslipidemia.

Haematologica. 2014; 99(11):1716–1721.

25. Pieters R, de Groot-Kruseman H, Van der Velden V, et al. Successful therapy reduction and intensification for childhood acute lymphoblastic leukemia based on minimal residual disease monitoring: study ALL10 from the Dutch Childhood Oncology Group. J Clin Oncol. 2016;34(22):2591-2601.

haematologica | 2017; 102(10)

ARTICLE

Acute Lymphoblastic Leukemia

Loss-of-function but not dominant-negative intragenic IKZF1 deletions are associated with an adverse prognosis in adult BCR-ABL-negative acute lymphoblastic leukemia

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Benjamin Kobitzsch,1 Nicola Gökbuget,2 Stefan Schwartz,1 Richard Reinhardt,3 Monika Brüggemann,4 Andreas Viardot,5 Ralph Wäsch,6 Michael Starck,7 Eckhard Thiel,1 Dieter Hoelzer2 and Thomas Burmeister1

Department of Hematology, Oncology and Tumor Immunology, Charité Universitätsmedizin Berlin; 2Department of Medicine II, Hematology/Oncology, Goethe University, Frankfurt/Main; 3Max Planck Genome Center, Köln; 4Department of Hematology, University Hospital Schleswig-Holstein, Kiel; 5Department of Medicine III (Hematology, Oncology), Ulm University; 6Department of Hematology, Oncology and Stem Cell Transplantation, University of Freiburg Medical Center, and 7Department of Hematology, Klinikum München-Schwabing, Munich, Germany 1

Haematologica 2017 Volume 102(10):1739-1747

ABSTRACT

G

enetic alterations of the transcription factor IKZF1 ("IKAROS") are detected in around 15-30% of cases of BCR-ABL-negative B-cell precursor acute lymphoblastic leukemia. Different types of intragenic deletions have been observed, resulting in a functionally inactivated allele ("loss-of-function") or in "dominant-negative" isoforms.

The prognostic impact of these alterations especially in adult acute lymphoblastic leukemia is not well defined. We analyzed 482 well-characterized cases of adult BCR-ABL-negative B-precursor acute lymphoblastic leukemia uniformly treated in the framework of the GMALL studies and detected IKZF1 alterations in 128 cases (27%). In 20%, the IKZF1 alteration was present in a large fraction of leukemic cells ("high deletion load") while in 7% it was detected only in small subclones ("low deletion load"). Some patients showed more than one IKZF1 alteration (8%).

Patients exhibiting a loss-of-function isoform with high deletion load had a shorter overall survival (OS at 5 years 28% vs. 59%; P<0.0001), also significant in a subgroup analysis of standard risk patients according to GMALL classification (OS at 5 years 37% vs. 68%; P=0.0002). Low deletion load or dominant-negative IKZF1 alterations had no prognostic impact. The results thus suggest that there is a clear distinction between loss-of-function and dominant-negative IKZF1 deletions. Affected patients should thus be monitored for minimal residual disease carefully to detect incipient relapses at an early stage and they are potential candidates for alternative or intensified treatment regimes. (clinicaltrials.gov identifiers: 00199056 and 00198991).

Introduction IKAROS family transcription factors have been identified as key players in lymphopoiesis.1-5 Alterations of IKZF1 in acute lymphoblastic leukemia (ALL) were first described in isolated cases in the early 1990s6,7 but it took several years to recognize the important role of IKZF1 in ALL development.8,9 The crucial role of IKZF1 in ALL development has also recently been underlined by the finding that certain non-coding single nucleotide polymorphisms in IKZF1 predispose to B lineage ALL development in later life.10-12 The first larger studies on the incidence and role of IKZF1 alterations in ALL were exclusively conducted on pediatric patients and revealed a prevalence of 15-30% of haematologica | 2017; 102(10)

Correspondence: thomas.burmeister@charite.de

Received: January 12, 2017.

Accepted: July 18, 2017.

Pre-published: July 27, 2017. doi:10.3324/haematol.2016.161273 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1739 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1739

B. Kobitzsch et al.

IKZF1 alterations in BCR-ABL-negative ALL3,9 compared with a particularly large fraction in BCR-ABL-positive ALL (more than 60%).8,13 IKZF1-alterated BCR-ABL-negative pediatric ALL patients were reported to have an adverse prognosis9,14-17 although this is still a subject of dispute.18 The negative prognostic effect was even found within BCR-ABL-positive pediatric19 and adult13,20 patients.

In adult BCR-ABL-negative ALL patients, studies suggested a worse outcome for IKZF1-mutated patients, albeit there have been inconsistent results concerning the prognostic impact of different IKZF1 alterations (Online Supplementary Table S1).21-24 Furthermore, to the best of our knowledge, the effect of multiple IKZF1 alterations or the impact of mutation load25,26 has not been systematically studied in this population.

The IKZF1 gene comprises eight exons, of which the first is non-coding. Its gene product is a 519 amino acid protein with six zinc finger domains.4 The two carboxyterminal zinc fingers (exon 8) are responsible for dimerization with other IKAROS family members.27 The four amino-terminal zinc fingers (exons 4-6) mediate DNA binding. Besides point mutations and the loss of the complete IKZF1 gene, various intragenic types of deletions have been experimentally observed. Loss of two or more amino-terminal zinc fingers encoded by exons 4-6 with deletion of the binding domain but retention of the dimerization domain results in dominant-negative isoforms, i.e. an isoform able to suppress the function of wild-type protein.27 Loss of exon 2 with the ATG start codon abolishes gene transcription at all and loss of exon 8 removes the dimerization domain. The latter two have historically been called "haploinsufficient".3 Since this term implies that the other allele is still functional, which could only be proven with certainty by single cell analysis, we will use the term "loss-of-function" for these alterations.

In this study, we present an in-depth analysis of 482 BCR-ABL-negative patients with B-precursor ALL with regard to their IKZF1 status. Patients were treated uniformly in the framework of the German Multicenter ALL (GMALL) studies between 1999 and 2009. We present a detailed genetic analysis and an assessment of the prognostic impact of the various IKZF1 alterations.

committees of participating institutions, and were conducted according to the Declaration of Helsinki.

Immunophenotyping and molecular genetic analysis At the time of diagnosis, immunophenotyping and molecular genetic analysis were performed at the GMALL central laboratory in Berlin, Germany. For all BCP-ALL patients, BCR-ABL status was determined by RT-PCR. Other molecular targets (TCF3-PBX1, ETV6-RUNX1 and MLL fusion genes) were analyzed according to our diagnostic guidelines as outlined previously.29,30

Genomic PCR for Δ4-7, Δ2-7, Δ4-8, Δ2-8

For all patients, genomic PCR was performed using HotStarTaq Polymerase Mastermix (QIAGEN) with 40-200 ng DNA and 500 nM of each primer under the following conditions: 15 minutes (min) at 95°C, followed by 35 cycles of 30 seconds (sec) at 94°C, 30 sec at 65°C and 60 sec at 72°C. Primers were located in intron 1 (F2A ACTACAGAGACTTCAGCTCTATTCCATTTC, F2B TGATTTGGATGTGTGTGTTTCATGCGTGG), intron 3 (F4 CTTAGAAGTCTGGAGTCTGTGAAGGTC), intron 7 (R7 AGGGACTCTCTAGACAAAATGGCAGGA) and 3’UTR of IKZF1 (R8 CCTCCTGCTATTGCACGTCTCGGT). For primer combinations see Online Supplementary Table S4. In all PCRs, a fragment of intron 7 or 3'UTR was amplified as internal control with primer concentration of 100 nM (F7 ACCATCAAATACAGGTCAACAGGACTGA, product 1,257 bp) or 50 nM (F8 CCCACTGCACAGATGAACAGAGCA, product 1,229 bp).

Primers were manufactured by metabion (Munich, Germany) or TIB Molbiol (Berlin, Germany) and HPLC-purified.

Methods Patients’ samples Originally, 507 patients with BCR-ABL-negative B-cell precursor (BCP) ALL were studied (Figure 1). Four were excluded because of irreproducible results, and 21 for missing follow-up data (of these only breakpoint sequences are presented).

Of the remaining 482 patients who were treated within the GMALL protocols 06/99 (n=84; clinicaltrials.gov identifier: 00199056) or 07/03 (n=398; clinicaltrials.gov identifier: 00198991), we analyzed bone marrow (n=330) or peripheral blood with peripheral blasts (n=132; bone marrow or peripheral blood not specified in n=20) obtained at the time of diagnosis between 1999 and 2009 (for blast count see Online Supplementary Tables S2 and S3). Matched samples from the time of relapse were available for 16 out of 482 patients

GMALL studies Detailed information on treatment has been published previously.28 The GMALL studies were approved by the ethics committee of the University of Frankfurt, Germany, and by local ethics 1740

Figure 1. Flowchart of the analysis.

haematologica | 2017; 102(10)

Prognosis of intragenic IKZF1 deletions in adult BCR-ABL-negative ALL

Reverse transcriptase PCR RT-PCR was performed with 2 µl cDNA, 500 nM of each primer and the HotStarTaq Polymerase Mastermix (QIAGEN) using the following conditions: 15 min at 95°C, followed by 35 cycles of 30 sec at 94°C, 30 sec at 64°C, and 60 sec at 72°C. Primers were located in exons 1 and 8 (RT-PCR ex1/8, primers ex1FA AAAGCGCGACGCACAAATCCA and ex8R CGTTGTTGATGGCTTGGTCCATCAC) or in exon 1 and exon 4 for detection of Δ2-3 (RT-PCR ex1/4, primers ex1FB CGAGGATCAGTCTTGGCCCCAA and ex4R GAATGCCTCCAACTCCCGACAAAG). Long IKZF1 isoforms were used as internal control. Bands of unexpected sizes were excised from the gel and sequenced.

In cases where RNA was not available for RT-PCR, we used our own and the PCR described by Meyer et al.31 as genomic screening PCR.

Quantitative PCR for Δ4-7, Δ2-7, Δ4-8

Quantitative PCR was performed in duplicates either for all patients (Δ4-7) or for patients positive in genomic PCR (Δ2-7 and Δ4-8) using a Rotorgene 6000 cycler (Corbett, Concorde,

Australia), the Thermo Scientific ABsolute QPCR Mix (Life Technologies, Darmstadt, Germany) with 200-250 ng DNA per PCR and the following conditions: 15 min at 95°C, followed by 55 cycles for 15 sec at 95°C, and 60 sec at 60°C. As DNA standard, we used the cell-line BV-173 for Δ4-7 (DSMZ, Braunschweig, Germany)32 or patient DNA (#100 for Δ27, #101 for Δ4-8). A PCR for the HCK gene served as internal control as described earlier.33 Oligonucleotides are given in Online Supplementary Table S4. Deletions were considered to be present in a large fraction of leukemic cells ("high deletion load", "highdel") when the relative PCR signal was >10-1, otherwise they were considered having a "low deletion load" ("lowdel"). The cut-off value was chosen a priori since this threshold appeared to separate samples with a high and low mutation load (Online Supplementary Figure S1). We used MLPA (SALSA MLPA P335 ALL-IKZF1 kit, MRC Holland, Amsterdam, the Netherlands) to correlelate the cut-off values of our quantitative PCRs with MLPA deletion values. We investigated a subset of patients with qPCR signals that we expected to yield a MLPA reduction of 0.3 or more (i.e. qPCR signal of 0.6 or higher). The chosen thresholds distinguishing highdel and lowdel corresponded to 5% deleted alleles in case of Δ2-7

A

D

B

E

C

F

Figure 2. Detection of IKZF1 deletions by RT-PCR and PCR screening. (A-C) RT-PCR ex1/8, PCR ∆4-7 and PCR ∆2-7 of the same 9 patients. (A) RT-PCR with primers in exon 1/8. Increased Ik6 expression in lanes 4-6 and increased Ik10 expression in lanes 6-8. Reduced full length isoform expression in lanes 1 and 7 is attributed to an additional deletion ∆2-3 in these 2 patients detected by another RT-PCR (see Online Supplementary Figure S2). (B) PCR ∆4-7. In lanes 1-3, ∆4-7 is present with a low deletion load; in lanes 4-6, the deletion is present with a high deletion load. Corresponding qPCR results are given below. Control band of 1257bp. (C) PCR ∆2-7 with low deletion load in lanes 3-4 and high deletion load in lanes 6-8. Control band of 1257bp. (D) Structure of the IKZF1 transcript isoforms Ik1 (full-length), Ik6 (loss of exons 4-7) and Ik10 (loss of exons 2-7). (E and F) PCR ∆4-8 and PCR ∆2-8 of the identical patients in lanes 10-17. Control band of 1229 bp. (E) PCR ∆48. See double bands in lanes 10 and 11. (F) PCR ∆2-8. See variant breakpoint in lane 17.

haematologica | 2017; 102(10)

1741

B. Kobitzsch et al. and Δ4-7, and 10% in Δ4-8, but the latter could equally well have been placed at 5%, since there were no samples between 5% and 10%.

In cases negative for Δ4-7 by conventional PCR but positive by qPCR, qPCR measurements were repeated and were considered positive when at least 3 out of 4 measurements were positive.

Gel densitometry When no quantification by qPCR was possible (n=41), we assessed the relative amount of cells with IKZF1 deletions (high vs. low deletion load) by gel band densitometry using the AlphaEaseFC v.4.0 software (Alpha Innotech, San Leandro, CA, USA). In deletions Δ2 (n=1) and Δ2-3 (n=17, missing values n=2), we compared deleted isoforms to full-length isoforms on RT-PCR images with a cut-off value of 0.60. In deletions Δ2-7 (n=5), Δ4-7 (n=3) and Δ5-7 (n=1) we compared deleted with long bands on RT-PCR images using a cut-off value of 1.20. In Δ2-8 (n=10) and Δ4-8 (n=2) we calculated the ratio of short PCR products to the long PCR control band with a cut-off value of 1.20.

Supplementary methods Nucleic acid preparation, identification of rare genomic breakpoints (primer sequences specified in Online Supplementary Table S5),31 DNA sequencing, bioinformatic analysis,34 and statistical analysis are all described in the Online Supplementary Methods.

Results Patients’ characteristics All 482 patients were aged between 16 and 65 years at diagnosis (Online Supplementary Table S6). The median age was 32 years [interquartile range (IQR) 22-47]. Two hundred and eighty-five patients (59%) were male. The distribution of immunophenotypes was 111 pre-B ALL (cyIg+; 23%), 314 common ALL (cyIg–,CD10+; 65%) and 57 pro-B ALL (CD10–; 12%). Two hundred and fourteen patients (44%) were considered high risk, the remaining standard risk. All patients were BCR-ABL-negative and a MLL rearrangement was detected in 44 patients (39 MLL-AF4, 4 MLL-ENL, 1 MLL-AF9), a TCF3-PBX1 fusion in 30, and an ETV6-RUNX1 fusion in 3 cases.

Frequency of IKZF1 deletions Two RT-PCRs were used to detect short IKZF1 isoforms (Figure 2A and Online Supplementary Figure S2A-C) and four separate PCRs to detect the Δ2-7, Δ2-8, Δ4-7 and Δ4-8 isoforms (Figure 2B-F). Deletions were then quantified using quantitative PCR or gel densitometry.

Dominant-negative deletions (Δ4-7, Δ5-7) were compared to loss-of-function deletions (Δ2, Δ2-3, Δ2-7, Δ2-8, Δ4-8).

Overall, 128 of 482 (27%) patients carried an IKZF1

A

B

Figure 3. Prevalence of IKZF1 deletions at the time of diagnosis. (A) Frequency of all deletions as detected by PCR (∆2-7, ∆2-8, ∆4-7, ∆4-8) and RT-PCR (exon 1/4, exon 1/8). (B) Only deletions classified as high deletion load by quantitative PCR and densitometry.

1742

haematologica | 2017; 102(10)

Prognosis of intragenic IKZF1 deletions in adult BCR-ABL-negative ALL

deletion (Figure 3A). Among these patients, we detected 175 different IKZF1 deletions. While 91 (19%) patients expressed only one deletion, in 37 (8%) patients more than one IKZF1 deletion was detected: 2 (n=28), 3 (n=8) or 4 (n=1) deletions (Online Supplementary Table S7; for an example, see lanes 3, 4 and 6 in Figure 2).

Among the 175 IKZF1 deletions, Δ4-7 was the most frequent (n=71). Δ2-7 was found in 47, Δ4-8 in 26, Δ2-3 in 19 and Δ2-8 in 10 patients. Rare deletions were Δ5-7 (n=1) and Δ2 (n=1). In summary, 56 patients (12%) carried only loss-of-function deletions, 50 (10%) had only dominantnegative deletions while 22 patients exhibited both types of deletions (5%).

We then quantified the amount of cells with IKZF1 deletions, as a variable deletion load was apparent from gel images (Figure 2B and C). We avoided the simple terminology "clonal" and "subclonal" since we did not prove clonality in a strict sense and did not investigate clonal relationships. Instead, we adopted the terms "high deletion load" (highdel) and "low deletion load" (lowdel) for IKZF1 aberrations present either in the vast majority of leukemic cells or only in a small fraction.

Out of 173 quantifiable deletions (n=2 not quantified), 106 (61%) were considered to have a high deletion load.

At least one highdel IKZF1 deletion could be found in 98 of 482 (20%) patients (Figure 3B). Among these, 50 had a highdel loss-of-function deletion only, 44 patients had a highdel dominant-negative deletion only, and there was a group of 4 patients expressing both deletions with a high deletion load level. qPCR screening revealed 50 additional cases positive for Δ4-7 with a low deletion load not detectable by our conventional PCR. In 41 of these cases, the lowdel Δ4-7 was the only IKZF1 deletion, while in 9 cases a loss-of-function deletion had been detected by conventional PCR. Patients with a lowdel Δ4-7 detected by qPCR only were considered IKZF1 wild-type.

Prognostic impact of IKZF1 deletions Four hundred and twenty-eight (89%) patients reached a complete remission, 31 patients (6%) died during induction, and 23 patients (5%) had a treatment failure after induction. The overall survival was 55% at five years.

We first calculated the effect of any IKZF1 deletion (n=128 vs. wild-type n=354) and then analyzed loss-offunction (n=78 vs. negative n=404) and dominant-negative deletions (n=72 vs. negative n=410) separately. We com-

pared the effect of high to low deletion load and no deletion in the group of loss-of-function (n=54/23/404, missing value n=1) and dominant-negative deletions (n=48/24/410).

There was a non-significant trend towards inferior overall survival (OS) for patients with any IKZF1 deletion (0.46 vs. 0.59; P=0.06) (Online Supplementary Figure S3A).

Patients carrying a loss-of-function IKZF1 deletion had a reduced OS (0.37 vs. 0.59; P=0.0012) (Figure 4A) while dominant-negative deletions had no effect on OS (0.54 vs.

0.56; P=0.95) (Figure 4B). Patients with both dominantnegative and loss-of-function deletions showed a clinical course comparable to loss-of-function deletions only (Online Supplementary Figure S3B). Analysis of the amount of IKZF1-deleted cells showed that the inferior survival in loss-of-function deletions was an effect of highdel loss-offunction deletions only (Figure 4C). Lowdel loss-of-function deletions did not influence the clinical course. In dominant-negative deletions, OS was not associated with the relative amount of IKZF1-deleted cells (Figure 4D).

Patients with highdel loss-of-function deletions showed a reduced OS (0.28 vs. 0.59; P<0.0001) (Table 1). In subgroups according to risk stratification, highdel loss-offunction IKZF1 deletions conferred a negative prognostic effect on standard-risk patients (0.37 vs. 0.68; P=0.0002), while in high-risk patients, the trend towards inferior OS narrowly missed statistical significance (0.26 vs. 0.46; P=0.06).

Clinico-biological characteristics of patients with IKZF1 deletions Patients with IKZF1 deletion showed a common immunophenotype significantly more often than patients without IKZF1 deletions (98 in 128, 77%, vs. 216 in 354, 61%; P=0.0064). The former were also significantly more likely to be CD34-positive (112 in 127, 88%, vs. 209 of 353, 59%; P<0.0001; n=2 CD34 N/A). The occurence of IKZF1 deletions was not associated with patients' age, gender, WBC or GMALL risk group, neither for all deletions (Online Supplementary Table S8) nor for different types of deletion (Online Supplementary Table S9).

TCF3-PBX1 and IKZF1 deletions were mutually exclusive (0 of 30 TCF3-PBX1+ vs. 64 of 250 TCF3-PBX1−; P=0.0004).

One in 3 ETV6-RUNX1-positive patients showed an IKZF1 deletion. There was a trend towards a lower frequency of IKZF1 deletions in MLL-positive patients (7 of 44 MLL+, 16% vs. 7 of 26 MLL-, 26%; P=0.3556).

Table 1. Effect of IKZF1 deletions on overall survival.

Type of IKZF1 deletion Any mutation Loss-of-function Dominant-negative High deletion load loss-of-function

Patient group

Cases pos/neg

Overall survival positive negative

all patients all patients all patients all patients SR HR

128/354 78/404 72/410 54/427 24/243 30/184

0.46±0.05 0.37±0.06 0.54±0.06 0.28±0.06 0.37±0.10 0.26±0.08

0.59±0.03 0.59±0.02 0.56±0.02 0.59±0.02 0.68±0.03 0.46±0.04

P ns (0.06) 0.0012 ns (0.95) <0.0001 0.0002 ns (0.06)

pos: positive; neg: negative; ns: not significant; SR: standard risk according to the German Multicenter Acute Lymphoblastic Leukemia (GMALL) studies; HR: high risk according to GMALL.

haematologica | 2017; 102(10)

1743

B. Kobitzsch et al.

Oligoclonality is more common in loss-of-function deletions Some patients showed more than one IKZF1 deletion (e.g. Δ2-7 and Δ4-7). Forty out of 175 deletions (23%) showed more than one chromosomal breakpoint resulting in the same type of RNA transcript. This oligoclonality may arise from multiple alterations in a single hyperdipoid clone or from alterations in different clones. This was evident either by gel electrophoresis (9 patients; see lanes 910 in Figure 2E and F) or by multiple sequences in chromatograms (2 breakpoints in 5 patients, Figure 5A; more than two breakpoints in 26 patients, Figure 5B). This kind of oligoclonal pattern occurred more often in loss-of-function deletions (31 of 103 deletions, 30%) compared with dominant-negative (9 of 72, 13%; P=0.0064).

Breakpoint sequences Sequencing of 193 breakpoints revealed four clusters (Figure 5C; for all breakpoints see Online Supplementary Table S10). In intron 1, 66 of 83 were located within 30bp.

In intron 3, 106 of 108 proximal breakpoints were located within 40bp. All 132 distal breakpoints in intron 7 clustered within 43bp. Thirty-six of 42 breakpoints in the 3'UTR region were located in a 27bp region, and an additional 5 breakpoints clustered around 500bp proximally.

The remaining 17 breakpoints in intron 1 were more diverse, covering a region of 7kb. Distal (3') breakpoints in intron 3 (Δ2-3) were scattered all over the 40kb intron. In

183 of 193 (95%) molecularly characterized breakpoints, putative cryptic recombination signal sequences, either with 23bp or 12bp spacer, were identified at both breakpoint sites (5' and 3'). This was the case for the four major breakpoint clusters (Figure 5 and Online Supplementary Table S11) but also true for the majority of the atypical breakpoints in intron 1 and 3. In 10 of 25 atypical breakpoints, only one cRSS could be identified (8 only on the 3' site, 2 only on the 5' site) (Online Supplementary Table S11).

There was no evidence of somatic hypermutation near the break sites.

Detection of deletions by RT-PCR

In 13 of 17 patients positive for Δ2-3 in RT-PCR ex1/4, a genomic breakpoint could be identified by eyer et al.'s PCR (Online Supplementary Figure S2A).31 In the remaining 4 patients, breakpoints were identified by a newly developed PCR (Online Supplementary Figure S2B). We also identified Δ2 once by RT-PCR ex1/4 and confirmed the genomic deletion. One patient expressed isoform Δ2-4 in RT-PCR ex1/8 but we could only find a deletion Δ2-3 on the genomic level and no deletion Δ2-4 or Δ4.

RT-PCR revealed 3 patients positive for Ik10 (lacking exons 2-7) but negative for Δ2-7 by genomic PCR due to a more proximal 5' breakpoint (Online Supplementary Figure S4A). In all 70 cases of RT-PCR positive for Ik6 (lacking exons 4-7) and negative for Ik6Δ (lacking exons 4-7 but with an additional 60 bp cryptic exon 3b),7,35 genomic PCR

B

A

P=0.0012

P=0.95 D

C

P=0.0002

P=0.62

Figure 4. Overall survival (OS) depending on IKZF1 deletions. (A) OS of patients with loss-of-function IKZF1 deletions. (B) OS of patients with dominant-negative deletions. (C) OS of patients with high or low deletion load loss-of-function IKZF1 deletions. (D) OS of patients with high or low deletion load dominant-negative IKZF1 deletions.

1744

haematologica | 2017; 102(10)

Prognosis of intragenic IKZF1 deletions in adult BCR-ABL-negative ALL was positive for deletion Δ4-7. In one patient with Ik6 and Ik6Δ we found two deletions Δ4-7, one with common breakpoints, one with a 5' breakpoint distal to the 60bp insert (Online Supplementary Figure S4B). The second patient with Ik6/Ik6Δ showed only a deletion Δ5-7 that was supposedly the reason for overexpression of Ik6 and Ik6Δ (Online Supplementary Figure S4C).

Comparison between diagnosis and relapse DNA at the time of relapse was available from 16 patients carrying 20 IKZF1 deletions. Four in 7 (57%) Δ47 and 9 in 13 (69%) loss-of-function deletions were conserved (P=0.65) (Online Supplementary Table S12). Eleven in 15 (73%) highdel and 1 in 4 lowdel deletions were conserved (P=0.12; 1 deletion not quantified). All genomic breakpoints were identical at the time of diagnosis and relapse. No newly acquired deletion Δ2-7, Δ2-7, Δ4-7 or Δ4-8 could be detected in relapse samples. We also inves-

A

tigated 5 relapse samples from patients who had shown a lowdel Δ4-7 IKZF1 deletion at diagnosis, detectable only by quantitative PCR. None of these cases evolved into a major clone, i.e. with high deletion load at relapse.

Discussion IKZF1 alterations have been recognized as recurrent aberrations in B-precursor ALL but their prognostic impact in adult ALL is still not well defined. Two major studies involving more than 200 patients have focused on the prognostic impact in BCR-ABL-negative adult BCP ALL. Moorman et al.21 investigated 304 patients and found IKZF1 deleted patients (29%) to have a lower OS, but this was only seen in a univariate analysis. The authors stated cautiously that "there was evidence to suggest that the poor outcome was not linked to the expression of the IK6

B

C

Figure 5. Distribution of IKZF1 breakpoints and clonality of deletions. (A) Chromatogram of patient #189 showing two distinguishable clones (sequenced sense and antisense reverse complement). (B) Chromatogram of patient #395 showing oligoclonality at the breakpoint junction in both sequencing directions. (C) Distribution of breakpoints in the IKZF1 gene locus. Proximal breakpoints are shown in black, distal breakpoints in blue. There are four major breakpoint clusters within intron 1, 3, 7 and 3’UTR of IKZF1.

haematologica | 2017; 102(10)

1745

B. Kobitzsch et al. isoform but rather to other types of IKZF1 deletions".21 Beldjord et al.22 investigated 216 younger adults and observed a significantly higher cumulative incidence of relapse in patients with focal IKZF1 alterations (25%) but not with whole gene deletion. No statistically significant difference between patients with different focal alterations was observed.

Our present study included 482 homogenously treated patients and revealed IKZF1 alterations in 128 cases. The incidence of focal deletions (27%) was comparable to both studies mentioned above. Our study is the first to systematically address the issue of IKZF1 mutation load and its implications for prognosis on a larger scale. This is of diagnostic interest if IKZF1 alterations are to be used as molecular markers for risk stratification and/or for detecting minimal residual disease.15,26 Ninety-eight patients revealed a high deletion load IKZF1 aberration while 29 patients showed low deletion load IKZF1 alterations only (n=1 not quantified). Regarding clinical implications, only high deletion load loss-of-function IKZF1 alterations were of prognostic relevance and conferred an adverse prognosis while low deletion load IKZF1 alterations or dominantnegative IKZF1 alterations did not have a prognostic effect.

In animal studies, double IKZF1 knock-out mice show a total absence of B cells.36 Mice with only IKZF1 deletions did not develop BCP ALL, but haploinsufficiency of IKZF1 in BCR-ABL-transgenic mice significantly accelerated the development of BCP ALL.37 Current evidence suggests that IKZF1 alterations alone are not sufficient to cause leukemia in humans but are an important co-factor or secondary event in the development and acceleration of ALL disease.

It may seem unexpected that the loss of one IKZF1 allele without apparent functional alteration of the other allele should have such a significant prognostic effect. However, this is supported by the above mentioned mouse model of Virely et al.37 The observation that loss-of-function IKZF1 deletions frequently occur in a small fraction of cells, but only seem to have an impact on prognosis if they are found in a large fraction, requires some explanation. A hypothetical explanation is the assumption that RAGmediated IKZF1 deletions occur sporadically during all stages of B-cell maturation because of the ongoing process of VDJ recombination.38,39 However, only those IKZF1 aberrations occurring at a very early maturation stage are thought to result in a cell phenotype with the full capacity of self-renewal, i.e. a "leukemia stem cell phenotype".40 IKZF1 alterations occurring at later stages of B-cell matu-

References 1. Georgopoulos K, Bigby M, Wang JH, et al.

The Ikaros gene is required for the development of all lymphoid lineages. Cell.

1994;79(1):143-156.

2. Georgopoulos K. Haematopoietic cell-fate decisions, chromatin regulation and ikaros.

Nat Rev Immunol. 2002;2(3):162-174.

3. Kastner P, Dupuis A, Gaub MP, Herbrecht R, Lutz P, Chan S. Function of Ikaros as a tumor suppressor in B cell acute lymphoblastic leukemia. Am J Blood Res. 2013;3(1):1-13.

1746

ration should result in low deletion load aberrations.

The extremely narrow clustering of breakpoints in regions comprising only a few nucleotides strongly argues in favor of a specific mechanism. The analysis of the breakpoint junctions revealed four breakpoint clusters in the vicinity of recombination signal sequences suggestive of a break mechanism involving the immunoglobulin VDJ recombination enzyme complex. RAG1 and RAG2 and other genes involved in VDJ rearrangement are not expressed at a very early stage of differentiation but only after lymphoid committment,41 which would be in line with the assumption that IKZF1 deletions are a later event in the path towards the malignant phenotype. The fact that cRSS could not be identified in 10 out of 193 breakpoints may be explained by limitations of the RSSsite software, since some of these breaks occurred in near vicinity, suggesting a specific mechanism.

The PCR method used in this study has the advantage that it can also detect IKZF1 alterations in a small fraction of leukemic cells, which is not possible when using MLPA.26 Since we analyzed the final IKZF1 cDNA transcript, we were in principle also able to detect deletions or aberrant splice isoforms arising from alterations involving only a few nucleotides that would escape detection by MLPA. However, MLPA has the advantage of also detecting whole gene deletions that are not detectable with our PCR-based approach. As long as there are no reliable PCRbased detection methods for the former, and given the fact that low deletion load alterations are prognostically irrelevant, we consider MLPA to be a suitable detection method.

To summarize, we detected partial IKZF1 gene deletions in approximately 27% of cases of adult BCR-ABL-negative adult ALL. Only high deletion load loss-of-function IKZF1 alterations, but not dominant-negative IKZF1 alterations, had negative prognostic implications and should thus be monitored closely, while those that were found in a small fraction of cells did not influence prognosis. We report extensive molecular data on these alterations which should help to establish suitable diagnostic methods for their detection and which shed additional light on the molecular pathogenesis.

Acknowledgments The authors are grateful for the excellent technical work of D. Grรถger, R. Lippoldt and colleagues and the members of the MPI sequencing team in Cologne. They thank all involved patients and physicians for participating in the GMALL studies. TB was supported by DFG grant BU 2453/1-1.

4. Olsson L, Johansson B. Ikaros and leukaemia. Br J Haematol. 2015;169(4):479491.

5. John LB, Ward AC. The Ikaros gene family: Transcriptional regulators of hematopoiesis and immunity. Mol Immunol. 2011;48(910):1272-1278.

6. Sun L, Heerema N, Crotty L, et al.

Expression of dominant-negative and mutant isoforms of the antileukemic transcription factor Ikaros in infant acute lymphoblastic leukemia. Proc Natl Acad Sci USA. 1999;96(2):680-685.

7. Sun L, Crotty ML, Sensel M, et al.

Expression of dominant-negative Ikaros

isoforms in T-cell acute lymphoblastic leukemia. Clin Cancer Res. 1999;5(8):21122120.

8. Mullighan CG, Miller CB, Radtke I, et al.

BCR-ABL1 lymphoblastic leukaemia is characterized by the deletion of Ikaros.

Nature. 2008;453(7191):110-114.

9. Mullighan CG, Su X, Zhang J, et al.

Deletion of IKZF1 and prognosis in acute lymphoblastic leukemia. N Engl J Med. 2009;360(5):470-480.

10. Papaemmanuil E, Hosking FJ, Vijayakrishnan J, et al. Loci on 7p12.2, 10q21.2 and 14q11.2 are associated with risk of childhood acute lymphoblastic

haematologica | 2017; 102(10)

Prognosis of intragenic IKZF1 deletions in adult BCR-ABL-negative ALL

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

leukemia. Nat Genet. 2009;41(9):10061010.

Treviño LR, Yang W, French D, et al.

Germline genomic variants associated with childhood acute lymphoblastic leukemia.

Nat Genet. 2009;41(9):1001-1005.

Burmeister T, Bartels G, Gröger D, et al.

Germline variants in IKZF1, ARID5B, and CEBPE as risk factors for adult-onset acute lymphoblastic leukemia: an analysis from the GMALL study group. Haematologica.

2014;99(2):e23-5.

Martinelli G, Iacobucci I, Storlazzi CT, et al. IKZF1 (Ikaros) deletions in BCR-ABL1positive acute lymphoblastic leukemia are associated with short disease-free survival and high rate of cumulative incidence of relapse: a GIMEMA AL WP report. J Clin Oncol. 2009;27(31):5202-5207.

Kuiper RP, Waanders E, van der Velden VH, et al. IKZF1 deletions predict relapse in uniformly treated pediatric precursor B-ALL. Leukemia. 2010;24(7):1258-1264.

Waanders E, van der Velden VH, van der Schoot CE, et al. Integrated use of minimal residual disease classification and IKZF1 alteration status accurately predicts 79% of relapses in pediatric acute lymphoblastic leukemia. Leukemia. 2011;25(2):254-258.

Dörge P, Meissner B, Zimmermann M, et al. IKZF1 deletion is an independent predictor of outcome in pediatric acute lymphoblastic leukemia treated according to the ALL-BFM 2000 protocol.

Haematologica. 2013;98(3):428-432.

Clappier E, Grardel N, Bakkus M, et al.

IKZF1 deletion is an independent prognostic marker in childhood B-cell precursor acute lymphoblastic leukemia, and distinguishes patients benefiting from pulses during maintenance therapy: results of the EORTC Children’s Leukemia Group study 58951. Leukemia. 2015;29(11):2154-2161.

Palmi C, Valsecchi MG, Longinotti G, et al.

What is the relevance of Ikaros gene deletions as a prognostic marker in pediatric Philadelphia-negative B-cell precursor acute lymphoblastic leukemia. Haematologica.

2013;98(8):1226-1231. van der Veer A, Zaliova M, Mottadelli F, et al. IKZF1 status as a prognostic feature in BCR-ABL1-positive childhood ALL. Blood.

2014;123(11):1691-1698.

DeBoer R, Koval G, Mulkey F, et al. Clinical impact of ABL1 kinase domain mutations and IKZF1 deletion in adults under age 60

haematologica | 2017; 102(10)

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

with Philadelphia chromosome-positive (Ph+) acute lymphoblastic leukemia (ALL): molecular analysis of CALGB (Alliance) 10001 and 9665. Leuk Lymphoma.

2016;57(10):2298-2306.

Moorman AV, Schwab C, Ensor HM, et al.

IGH@ translocations, CRLF2 deregulation, and microdeletions in adolescents and adults with acute lymphoblastic leukemia. J Clin Oncol. 2012;30(25):3100-3108.

Beldjord K, Chevret S, Asnafi V, et al.

Oncogenetics and minimal residual disease are independent outcome predictors in adult patients with acute lymphoblastic leukemia. Blood. 2014;123(24):3739-3749.

Mi JQ, Wang X, Yao Y, et al. Newly diagnosed acute lymphoblastic leukemia in China (II): prognosis related to genetic abnormalities in a series of 1091 cases.

Leukemia. 2012;26(7):1507-1516.

Dhédin N, Huynh A, Maury S, et al. Role of allogeneic stem cell transplantation in adult patients with Ph-negative acute lymphoblastic leukemia. Blood. 2015; 125(16):2486-2496.

Dupuis A, Gaub MP, Legrain M, et al.

Biclonal and biallelic deletions occur in 20% of B-ALL cases with IKZF1 mutations.

Leukemia. 2013;27(2):503-507.

Caye A, Beldjord K, Mass-Malo K, et al.

Breakpoint-specific multiplex polymerase chain reaction allows the detection of IKZF1 intragenic deletions and minimal residual disease monitoring in B-cell precursor acute lymphoblastic leukemia.

Haematologica. 2013;98(4):597-601.

Sun L, Liu A, Georgopoulos K. Zinc fingermediated protein interactions modulate Ikaros activity, a molecular control of lymphocyte development. EMBO J. 1996; 15(19):5358-5369.

Brüggemann M, Raff T, Flohr T, et al.

Clinical significance of minimal residual disease quantification in adult patients with standard-risk acute lymphoblastic leukemia. Blood. 2006;107(3):1116-1123.

Burmeister T, Meyer C, Schwartz S, et al.

The MLL recombinome of adult CD10-negative B-cell precursor acute lymphoblastic leukemia: results from the GMALL study group. Blood. 2009;113(17):4011-4015.

Burmeister T, Gökbuget N, Schwartz S, et al. Clinical features and prognostic implications of TCF3-PBX1 and ETV6-RUNX1 in adult acute lymphoblastic leukemia.

Haematologica. 2010;95(2):241-246.

31. Meyer C, zur Stadt U, Escherich G, et al.

Refinement of IKZF1 recombination hotspots in pediatric BCP-ALL patients.

Am J Blood Res. 2013;3(2):165-173.

32. Nakayama M, Suzuki H, YamamotoNagamatsu N, et al. HDAC2 controls IgM H- and L-chain gene expressions via EBF1, Pax5, Ikaros, Aiolos and E2A gene expressions. Genes Cells. 2007;12(3):359-373.

33. Burmeister T, Marschalek R, Schneider B, et al. Monitoring minimal residual disease by quantification of genomic chromosomal breakpoint sequences in acute leukemias with MLL aberrations. Leukemia.

2006;20(3):451-457.

34. Merelli I, Guffanti A, Fabbri M, et al.

RSSsite: a reference database and prediction tool for the identification of cryptic Recombination Signal Sequences in human and murine genomes. Nucleic Acids Res. 2010;38 (Web Server Issue):W262-267.

35. Payne KJ, Dovat S. Ikaros and tumor suppression in acute lymphoblastic leukemia.

Crit Rev Oncog. 2011;16(1-2):3-12.

36. Wang JH, Nichogiannopoulou A, Wu L, et al. Selective defects in the development of the fetal and adult lymphoid system in mice with an Ikaros null mutation.

Immunity. 1996;5(6):537-549.

37. Virely C, Moulin S, Cobaleda C, et al.

Haploinsufficiency of the IKZF1 (IKAROS) tumor suppressor gene cooperates with BCR-ABL in a transgenic model of acute lymphoblastic leukemia.[letter]. Leukemia.

2010;24(6):1200-1204.

38. Iacobucci I, Storlazzi CT, Cilloni D, et al.

Identification and molecular characterization of recurrent genomic deletions on 7p12 in the IKZF1 gene in a large cohort of BCRABL1-positive acute lymphoblastic leukemia patients: on behalf of Gruppo Italiano Malattie Ematologiche dell’Adulto Acute Leukemia Working Party (GIMEMA AL WP). Blood. 2009;114(10):2159-2167.

39. Yu W, Nagaoka H, Jankovic M, et al.

Continued RAG expression in late stages of B cell development and no apparent reinduction after immunization. Nature.

1999;400(6745):682-687.

40. Warner JK, Wang JC, Hope KJ, Jin L, Dick JE. Concepts of human leukemic development. Oncogene. 2004;23(43):7164-7177.

41. Nagaoka H, Yu W, Nussenzweig MC. Regulation of RAG expression in developing lymphocytes. Curr Opin Immunol.

2000;12(2):187-190.

1747

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Hodgkin Lymphoma

Ferrata Storti Foundation

Haematologica 2017 Volume 102(10):1748-1757

Secondary malignant neoplasms, progression-free survival and overall survival in patients treated for Hodgkin lymphoma: a systematic review and meta-analysis of randomized clinical trials

Dennis A. Eichenauer,1 Ingrid Becker,2 Ina Monsef,3 Nicholas Chadwick,4 Vitaliana de Sanctis,5 Massimo Federico,6 Catherine Fortpied,7 Alessandro M. Gianni,8 Michel Henry-Amar,9 Peter Hoskin,10 Peter Johnson,11 Stefano Luminari,6 Monica Bellei,6 Alessandro Pulsoni,12 Matthew R. Sydes,13 Pinuccia Valagussa,8 Simonetta Viviani,8 Andreas Engert1 and Jeremy Franklin2

First Department of Internal Medicine and German Hodgkin Study Group (GHSG), University Hospital Cologne, Germany; 2Institute of Medical Statistics, Informatics and Epidemiology, University of Cologne, Germany; 3Cochrane Haematological Malignancies Group, First Department of Internal Medicine, University Hospital Cologne, Germany; 4 University College London (UCL) Cancer Trials Centre, UK; 5Department of Radiotherapy, University “La Sapienza”, Rome, Italy; 6University of Modena and Reggio Emilia, Modena, Italy; 7European Organisation of Research and Treatment of Cancer (EORTC), Brussels, Belgium; 8Istituto Nazionale dei Tumori, Milan, Italy; 9Centre de Traitement des Données du Cancéropôle Nord-Ouest, Centre François Baclesse, Caen, France; 10Mount Vernon Cancer Centre, Northwood, UK; 11Cancer Research UK Centre, University of Southampton, UK; 12Cellular Biotechnology and Hematology Department, University “La Sapienza”, Rome, Italy and 13Medical Research Council (MRC), Clinical Trials Unit at University College London (UCL), UK 1

ABSTRACT

Correspondence: jeremy.franklin@uni-koeln.de

Received: February 23, 2017.

Accepted: July 19, 2017.

Pre-published: September 14, 2017. doi:10.3324/haematol.2017.167478 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1748 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1748

T

reatment intensification to maximize disease control and reduced intensity approaches to minimize the risk of late sequelae have been evaluated in newly diagnosed Hodgkin lymphoma. The influence of these interventions on the risk of secondary malignant neoplasms, progression-free survival and overall survival is reported in the meta-analysis herein, based on individual patient data from 9498 patients treated within 16 randomized controlled trials for newly diagnosed Hodgkin lymphoma between 1984 and 2007. Secondary malignant neoplasms were meta-analyzed using Peto’s method as time-toevent outcomes. For progression-free and overall survival, hazard ratios derived from each trial using Cox regression were combined by inversevariance weighting. Five study questions (combined-modality treatment vs. chemotherapy alone; more extended vs. involved-field radiotherapy; radiation at higher doses vs. radiation at 20 Gy; more vs. fewer cycles of the same chemotherapy protocol; standard-dose chemotherapy vs. intensified chemotherapy) were investigated. After a median follow-up of 7.4 years, dose-intensified chemotherapy resulted in better progression-free survival rates (P=0.007) as compared with standard-dose chemotherapy, but was associated with an increased risk of therapyrelated acute myeloid leukemia/myelodysplastic syndromes (P=0.0028).

No progression-free or overall survival differences were observed between combined-modality treatment and chemotherapy alone, but more secondary malignant neoplasms were seen after combined-modality treatment (P=0.010). For the remaining three study questions, outcomes and secondary malignancy rates did not differ significantly between treatment strategies. The results of this meta-analysis help to weigh up efficacy and secondary malignancy risk for the choice of firstline treatment for Hodgkin lymphoma patients. However, final conclusions regarding secondary solid tumors require longer follow-up. haematologica | 2017; 102(10)

Secondary malignant neoplasms after HL treatment

Introduction Hodgkin lymphoma (HL) is a lymphoid malignancy with an incidence of 3-4/100 000/year. Young adults are most often affected.1 At present, about 80% of patients achieve long-term remission after treatment with multiagent chemotherapy optionally followed by radiotherapy (RT).2 Given the mostly young age at diagnosis and the excellent long-term prognosis, therapy-related late effects including secondary malignant neoplasms (SMN), cardiovascular disease and infertility have become increasingly important.3-6 Several recent clinical trials evaluated the possibility of reducing toxicity by limiting chemotherapy and RT without compromising efficacy.2 Conversely, some studies for patients with newly diagnosed advanced HL investigated intensified chemotherapy protocols to improve the clinical outcome of high-risk patients.2 SMN are divided into secondary hematological malignancies including therapy-related acute myeloid leukemia/myelodysplastic syndromes (t-AML/MDS) and secondary non-Hodgkin lymphomas (NHL) and the heterogeneous group of secondary solid tumors. An association between the use of alkylating agents and topoisomerase II inhibitors and the development of t-AML/MDS has been demonstrated.7-9 Both drug classes are included in first-line chemotherapy protocols such as adriamycin, bleomycin, vinblastine and dacarbazine (ABVD), bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone escalated (BEACOPP) and vinblastine, doxorubicin, vincristine, bleomycin, mustard, etoposide, and prednisone (Stanford V).10-12 For secondary NHL, such associations have not been identified.13 The time interval between HL treatment and the occurrence of secondary hematological malignancies is usually short, ranging between four and ten years in most cases.9,13 In contrast, the risk of secondary solid tumors remains significantly increased for up to 25 years and more.3,4 In the meta-analysis herein, using individual patient data (IPD), SMN, progression-free survival (PFS) and overall survival (OS) of patients treated in randomized clinical trials comparing different treatment approaches (combined-modality treatment (CMT) vs. chemotherapy alone; more extended radiotherapy (RT) vs. involved-field RT (IFRT); RT at higher doses vs. RT at 20 Gy; more vs. fewer cycles of the same chemotherapy protocol; standard-dose chemotherapy vs. intensified chemotherapy) were investigated. Acceptable chemotherapy regimens were ABVD or similar (e.g., mechlorethamine, vincristine, procarbazine and prednisone/doxorubicin, bleomycin and vincristine [MOPP/ABV]) or (for the last study question above) any dose-intensified chemotherapy randomly compared with a standard dose ABVD-like regimen.

Methods Searches for randomized clinical trials including patients with newly diagnosed HL that compared treatment approaches according to the mentioned study questions and published from 1984 onwards were performed in March 2010 using the electronic literature databases Medline and Cochrane Central. Reference lists of all relevant retrieved publications and previous meta-analyses were searched. All identified articles were screened independently by two authors. Trials had to have enrolled at least 50 patients per haematologica | 2017; 102(10)

treatment group and to have finished recruitment before or during 2007, to avoid trials with inadequate follow-up. Searches were repeated in April 2015.

IPD were requested from the investigators of the eligible trials, including birth date, sex, HL diagnosis date, stage at diagnosis, presence of B symptoms, randomization date, allocated treatment, remission status after first-line treatment, relapse date, date and type of SMN, death date and last follow-up date concerning clinical outcome and vital status.

Statistical methods For quality control, each trial included in the meta-analysis was initially analyzed separately, comparing the treatment arms with respect to recruitment times, patient characteristics, complete remission rates, follow-up duration, PFS, OS and time to SMN. Results were compared with previous trial publications and inconsistencies were queried. Risk of bias was assessed for each trial according to the Cochrane recommendations.14 To assess completeness of follow-up, the median observation time was calculated using the Kaplan-Meier method with reverse censoring at death. The distribution of last information dates was quantified using the interquartile range of the dates of last information (IQRDLI). This range includes the central 50% of last information dates and thus represents the extent to which patients in a given study are lost to follow up over a broad time interval. Large IQR-DLI values (absolute or relative to median follow-up) suggest poorer quality of follow up.

Randomized comparisons for each study question were combined across the appropriate trials to obtain a pooled Petoâ&#x20AC;&#x2122;s odds ratio (OR) for SMN rates, with 95% confidence intervals (CI).15,16 Three types of SMN, i.e., t-AML/MDS, secondary NHL and secondary solid tumors were also analyzed separately.

Subgroup analyses were performed to investigate whether the SMN rates depended upon the stage according to Ann Arbor classification (early stages I and II vs. advanced stages III and IV), age (â&#x2030;¤50 years vs. >50 years) and sex. Treatment subgroups were defined for the intensified chemotherapy question only (escalated BEACOPP vs. Stanford V vs. epidoxirubicin, bleomycin, and vinblastine/lomustine, doxorubicin, and vindesine (EBV/CAD)-based vs. chlorambucil, vinblastine, prednisolone and procarbazine (ChlVPP)-based). Results were displayed chronologically by recruitment period in order to reveal time period effects.

Sensitivity analyses were conducted excluding SMN that had occurred after HL recurrence: follow-up times were censored at HL recurrence. Further sensitivity analyses were performed as follows: firstly, SMN data were analyzed using a one-step Cox proportional hazards regression, stratified by trial, secondly, analyses were repeated with the exclusion of the less complete follow-up periods in each trial, censoring at the date at which 75% of surviving patients in the particular trial were still being followed, thirdly, overall SMN and separate secondary solid tumor analyses were repeated excluding non-melanoma skin cancers (NMSC), and finally, the cumulative incidence method, which considers nonSMN death as a competing risk, was employed.17,18 All analyses were performed in SAS (version 9.3) and RevMan (version 5.2).

Results Results of the search A total of 3515 references (excluding duplicates) published after 1984 were identified in 2010 and reviewed for eligibility. The majority did not meet the predefined criteria and were excluded for the following reasons: 1419 ref1749

D.A. Eichenauer et al.

erences did not concern HL patients, 1162 did not report a clinical trial, 161 reported on patients in the second-line setting, 98 reported non-randomized trials, 97 were review articles, and 53 were duplicates. Hence, 578 references fulfilled the predefined general eligibility criteria and were reassessed concerning the exact treatment comparison. A total of 21 randomized clinical trials for the firstline treatment of HL were identified, which included at least 50 patients per study arm and compared treatment modalities that matched with at least one of the five study questions. Data were received for 16 trials conducted between 1984 and 2007 (Figure 1).19-33 No data were received for four studies.34-37 One additional trial first published in 2013 was only found in the 2015 search, so IPD

were not sought.38 One trial was split for analysis since the participating centers could choose between two alternative intensified chemotherapy regimens.32

Characteristics of the included studies All included studies, grouped according to the study questions, are described in Table 1.

Three aspects of the risk of bias (randomization, allocation concealment, attrition bias) according to the Cochrane scheme were judged to be low in 13 out of 16 trials, while randomization and allocation concealment were uncertain in two trials and considered high in one trial. High risk of bias due to lack of blinding applied to all trials with respect to SMN and PFS, OS was presumed to

Table 1. Description of included studies.

Comparison (standard vs. experimental)

Study

Stage Recruitment

Follow up Start Length Median (y) (y) (y)

Treatment arm Standard

CMT vs.

EORTC 20884 (19) IIIA-IVB chemotherapy GHSG HD3 (24) IIIB-IVB alone EORTC-GELA H9-F (21) IA-IIB

1989 1984 1998

10.7 4.2 5.75

9.0 6-8MOPP/ABV + IF 12.9 3(COPP+ABVD) + IF 6.6 6EBVP+ 20/36Gy IF

Extended vs.

GHSG HD8 (25) involved-field RT Milan (30) (after CT) EORTC-GELA H8-U (20) HD94 Rome (31)

IA-IIIA IA-IIA IA-IIB IA-IIIA

1993 1990 1993 1994

5 6.5 6 4

10.3 17.5 8.8 10.5

Higher dose vs. 20 Gy RT (after CT)

GHSG HD10 (22) GHSG HD11 (23)

IA-IIB IA-IIB

1998 1998

4.75 4.75

7.5 7.4

EORTC H9-F (21)

IA-IIB

1998

5.75

More vs. fewer CT EORTC H8U (20) cycles EORTC H9-U (21) GHSG HD10 (22)

IA-IIB IA-IIB IA-IIB

1993 1998 1998

Standard-dose GHSG HD9 (26) IIB-IVB vs. intensified CT IIL HD9601 (28) IIB - IVB (regimen +/- RT) GISL HD2000 (27) IIB-IVB GITIL-IIL NCT IB*, IIB-IVB 01251107 (29) UKLG LY09 Hyb (32) IA-IVB UKLG LY09 Alt (32) IA-IVB UKLG ISRCTN64141244 (33) I-IV**

N

Experimental

Total Standard Experimental

6-8MOPP/ABV 4(COPP+ABVD) 6EBVP

333 100 578

172 51 448

161 49 130

4COPP/ABVD + EF 4COPP/ABVD + IF 4ABVD + STNI 4ABVD + IF 4MOPP/ABV + STNI 4/6MOPP/ABV + IF 4ABVD + EF 4ABVD + IF

1064 140 984 209

532 68 324 102

532 72 660 107

2/4ABVD + 20Gy IF 4ABVD/4BEACOPP +20Gy IF 6EBVP+20Gy IF

1163 1351

575 675

588 676

6.6

2/4ABVD + 30Gy IF 4ABVD/4BEACOPP +30Gy IF 6EBVP+36Gy IF

448

239

209

6 4 4.75

8.8 7.0 7.5

6MOPP/ABV + IF 4MOPP/ABV + IF 6ABVD +RT 4ABVD +RT 4ABVD + 20/30Gy IF 2ABVD + 20/30Gy IF

669 553 1190

336 276 596

333 277 594

1993 1996 2000 2000

5.2 4.3 7.25 7.0

8.6 6.9 3.5 4.6

8COPP/ABVD 6ABVD 6ABVD 6-8ABVD

727 335 295 331

261 122 99 168

466 213 196 163

1998 1998

3.7 3.7

7.9 8.1

6-8ABVD

569 219

287 107

282 112

1998

3.7

5.5

ABVD

520

261

259

8BEACOPPesc Stanford V or 6MEC 6BEACOPP or 6CEC 4BEACOPPesc + 4BEACOPPbas 6-8CHLVPP/EVA Hybrid 6/8ChlVPP/PABLOE alternating Stanford V

*only 1 patient, ** symptoms not specified. CMT: combined-modality treatment; RT: radiotherapy; CT: chemotherapy; y: year; IF: involved field; EF: extended field; STNI: subtotal nodal irradiation; EORTC: European Organization for Research and Treatment of Cancer; GELA: Groupe d’Etude des Lymphomes de l’Adulte; GHSG: German Hodgkins' Study Group; GISL: Gruppo Italiano per lo Studio dei Linfomi; GITIL: Gruppo Italiano di Terapie Innovative nei Linfomi; IIL: Intergruppo Italiano Linfomi; UKLG: United Kingdom National Cancer Research Institute Lymphoma Group; MOPP/ABV: mechlorethamine, vincristine, procarbazine, prednisone/doxorubicin, bleomycin, vinblastine; EBVP: epirubicin, bleomycin, vinblastine, prednisone; ABVD: doxorubicin, bleomycin, vinblastine, dacarbazine; BEACOPP: bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, prednisone; COPP/ABVD: cyclophosphamide, vincristine, procarbazine, prednisone/doxorubicin, bleomycin, vinblastine, dacarbazine; CEC: COPPEBVCAD: cyclophosphamide, lomustine, vindesine, melphalan, prednisone, epidoxorubicin, vincristine, procarbazine, vinblastine, bleomycin; MEC: MOPP/EBV/CAD: mechlorethamine, vincristine, procarbazine, prednisone, epidoxorubicin, bleomycin, vinblastine, lomustine, doxorubicin, vindesine; Stanford V: adriamycin, vinblastine, mechlorethamine, vincristine, bleomycin, etoposide, prednisone; ChlVPP/PABlOE: chlorambucil, vinblastine, procarbazine, prednisolone/prednisolone, doxorubicin, bleomycin, vincristine, etoposide; ChlVPP/EVA: chlorambucil, vinblastine, procarbazine, prednisolone/etoposide, vincristine, doxorubicin.

1750

haematologica | 2017; 102(10)

Secondary malignant neoplasms after HL treatment

be entirely objective. Sixteen patients from two studies with no evaluable data after randomization were excluded from the meta-analysis.20,26 Those patients had also been excluded from the analyses of the respective study groups.

The median follow-up within the trials ranged between 3.5 and 17.6 years (overall median follow-up: 7.4 years). A histogram of follow-up times to SMN or last information is displayed in the Online Supplementary Figure S1; although 75% of patients had more than 5 years of follow-up, only 16% were followed beyond ten years. A comparison of the distribution of follow-up times between the treatment arms of the included studies yielded a significant difference according to the log-rank test (P=0.036) in only one out of 16 cases.

The IQR-DLI varied among trials from 0.4 to 6.6 years (median: 3.1 years). Studies with longer follow-up tended to have a wider scatter. The ratio between the IQR-DLI and the median follow-up varied between 0.05 and 0.59 (median: 0.34). Thus, in half of the included studies the central 50% of last information dates stretch over a time interval of at least three years or one third of the median follow-up duration.

Patient characteristics IPD from 9498 patients treated within 16 randomized clinical trials for newly diagnosed HL were included. At the time of HL treatment, patients were aged between 14 and 75 according to inclusion criteria of the included studies (eight exceptions between ten and 87 years), median age was 33 years. During the course of follow-up, an SMN was reported for 438/9498 patients (4.6%), including 63 tAML/MDS (0.7%), 86 secondary NHL (0.9%) and 276 (2.9%) secondary solid tumors. The sites most often affected were the breast (39/276), lung (35/276), skin (29/276) and bowel (colon, rectum) (23/276). In 13 patients (0.1%) diagnosed with an SMN, information on the tumor entity was lacking. Cumulative incidences of SMN at five, ten, 15 and 20 years (regarding death as a competing risk) were 2.4%, 5.8%, 13% and 23%, respectively.

Results of the treatment comparisons (1) CMT vs. chemotherapy alone A total of 1011 patients treated within the European Organisation for Research and Treatment of Cancer (EORTC) 20884 (advanced stages), EORTC-Groupe D'Etude des Lymphomes de L'Adulte (GELA) H9-F (early stages) and the German Hodgkin Study Group (GHSG) HD3 (advanced stages) trials were analyzed. After a median follow-up of 7.8 years, 30/671 patients (4.5%) treated with CMT and 10/340 patients (2.9%) treated with chemotherapy alone had been diagnosed with an SMN. This difference was significant, favoring patients who had received chemotherapy only (P=0.010; Petoâ&#x20AC;&#x2122;s OR: 0.43, 95%-CI: 0.23-0.82) (Figures 2 and 3; Table 2). In particular, patients aged 50 and younger (P=0.04), female patients (P=0.01) and those with advanced stages (P=0.01) had a significantly reduced risk of developing an SMN when treated with chemotherapy alone. No reduced SMN rate with chemotherapy alone was observed among patients diagnosed with early stages (P=0.68). An analysis separately evaluating the incidence rates for t-AML/MDS, secondary NHL and secondary solid tumors revealed a risk reduction after chemotherapy alone solely for the development of t-AML/MDS (P=0.037), but not for secondary NHL and solid tumors (Table 3). The PFS and OS rates did not significantly differ between treatment groups (Table 2), however, there was some evidence of inferior tumor control with chemotherapy alone (HR: 1.31, 95%-CI: 0.99-1.73, P=0.06). Subgroup analyses indicated that stage and age were significant effect modifiers (interaction P-values were <0.0001 (stage) and 0.02 (age)). PFS was significantly impaired after chemotherapy alone in comparison with CMT for early-stage patients and patients aged â&#x2030;¤50 (Online Supplementary Tables S1, S2 and S3).

(2) More extended RT vs. involved field RT

A total of 2397 early-stage patients treated within the EORTC-GELA H8-U, the GHSG HD8, the Italian HD94 and the Milan trial were analyzed. After a median followup of 10.8 years, 91/1026 patients (8.9%) who had

Figure 1. Search results (combined for both searches in 2010 and 2015). HL: Hodgkin lymphoma; IPD: individual patient data.

haematologica | 2017; 102(10)

1751

D.A. Eichenauer et al.

received RT to more extended fields and 96/1371 patients (7.0%) who had received IF-RT had been diagnosed with an SMN. This difference was not statistically significant (P=0.32; Peto’s OR: 0.86, 95%-CI: 0.64-1.16; Online Supplementary Figures S2 and S3). In addition, when separately analyzing the incidence rates of t-AML/MDS, secondary NHL and secondary solid tumors, there were no significant differences between the treatment groups. The same is true for the PFS and OS rates (Table 2). Subgroup

analyses according to sex and age did not identify any heterogeneity of the treatment effect across any subgroups.

(3) RT at higher doses vs. RT at 20 Gy A total of 2962 early-stage patients treated within the EORTC-GELA H9-F, GHSG HD10 and GHSG HD11 trials were analyzed. After a median follow-up of 7.4 years, 54/1489 patients (3.6%) who had RT at a dose of 30 Gy or 36 Gy and 56/1473 patients (3.8%) who had RT at a dose of 20 Gy had developed an SMN. Thus, the rate of SMN

(P=0.88) (P=0.01)

(P=0.68)

(P=0.88) (P=0.01)

Favors chemo. alone Favors chemo-radio

(P=0.63)

Figure 2. Additional radiotherapy, cumulative incidence of SMN (Peto meta-analysis). CI: confidence interval; O-E: observed minus expected; V: variance; I2 = measure of heterogeneity: EORTC: European Organization for Research and Treatment of Cancer; GHSG: German Hodgkins' Study Group.

Table 2. Treatment effect and heterogeneity for secondary malignant neoplasms (SMN), overall survival (OS) and progression-free survival (PFS).

Comparison (standard vs. experimental) CMT vs. chemotherapy alone

OR (95%-CI) 0.433 (0.28-0.82)

Extended vs. involved-field RT (after CT)

0.862 (0.64-1.16)

Higher dose vs. 20 Gy RT (after CT)

1.032 (0.71-1.50)

More vs. fewer CT cycles

1.096 (0.74-1.62)

Standard-dose vs. intensified CT (regimen +/- RT)

1.37 (0.89%2.10)

SMN I2

0%

N

Summary of main results OS HR I2 (95%-CI)

30 (4.5%) vs. 10 (2.9%)

92 67% (9.0%) vs. 96 (7.0%) 54 72% (3.6%) vs. 56 (3.8%) 48 0% (4.0%) vs. 53 (4.4%) 31 11% (2.4%) vs. 60 (3.6%)

0.71 (0.46- 1.11)

0.89 (0.70-1.12)

0.91 (0.65-1.28)

0.99 (0.73-1.34)

0.85 (0.70 – 1.04)

39%

N 53 (7.9%) vs. 31 (9.1%)

132 0% (12.9%) vs.

155 (11.3%) 72 0% (4.8%) vs.

66 (4.5%) 82 (6.8%) 0% vs.

82 (6.8%) 191 63% (14.6%) vs. 213 (12.6%)

HR (95%-CI)

PFS I2

1.31 (0.99-1.73)

89%

0.99 (0.81-1.21)

0%

1.20 (0.97-1.48)

1%

1.15 (0.91-1.45)

31%

0.82 (0.70 – 0.95)

85%

N

123 (18.3%) vs. 83 (24.4%) 175 (17.1%) vs. 224 (16.3%) 165 (11.1%) vs. 190 (12.9%) 133 (11.0%) vs. 152 (12.6%) 350 (26.8%) vs. 372 (22.0%)

CMT: combined-modality treatment; RT: radiotherapy; CT: chemotherapy; OR: odds ratio by Peto method, HR: hazard ratio by Cox regression; CI: confidence interval; I2: heterogeneity; N: number of events.

1752

haematologica | 2017; 102(10)

Secondary malignant neoplasms after HL treatment

did not differ between the treatment groups (P=0.87; Peto’s OR: 1.03, 95%-CI: 0.71-1.50; Online Supplementary Figures S4 and S5). There were also no differences when the rates for t-AML/MDS, secondary NHL and secondary solid tumors were separately considered. The clinical outcome was similar in both treatment groups, with no differences in PFS and OS. Additional subgroup analyses according to sex and age did not identify differences between any subgroups.

(4) More vs. fewer cycles of the same chemotherapy protocol A total of 2403 early-stage patients treated within the EORTC-GELA H8-U, EORTC-GELA H9-U and GHSG HD10 trials were analyzed. After a median follow-up of 7.8 years, 48/1201 patients (4.0%) who had been treated with more cycles and 53/1202 patients (4.4%) who had received fewer cycles of the same chemotherapy protocol had developed an SMN. This difference was not statistically significant (P=0.65; Peto’s OR: 1.10, 95%-CI: 0.741.62) (Online Supplementary Figures S6 and S7).

Additionally, the SMN incidences did not differ between the patient groups when t-AML/MDS, secondary NHL and secondary solid tumors were considered separately.

The clinical outcome after treatment with more or fewer cycles of the same chemotherapy was comparable, without any significant PFS and OS differences. Additional subgroup analyses according to sex and age did not identify differences between any subgroups.

(5) Standard-dose chemotherapy (ABVD or cyclophosphamide vincristine, procarbazine, and prednisone (COPP)/ABVD) vs. intensified chemotherapy A total of 2996 advanced-stage patients treated within a Gruppo Italiano Terapie Innovative nei Linfomi

(GITIL)-Intergruppo Italiano Linfomi (IIL) trial, the Gruppo Italiano per lo Studio dei Linfomi (GISL) HD2000 trial, the GHSG HD9 trial, the HD9601 trial from Italy and the British LY09 and ISRCTN64141244 trials were analyzed. The rates of RT were comparable in both treatment groups of the included studies with the exception of the ISRCTN64141244 and HD9601 trials, administering Stanford V as an intensified regimen. After a median follow-up of 6.7 years, 31/1305 patients (2.8%) who had received standard-dose chemotherapy (ABVD, COPP/ABVD) and 60/1691 patients (3.5%) treated with intensified chemotherapy protocols (escalated BEACOPP, Stanford V, ChlVPP/prednisolone, doxorubicin, bleomycin, vincristine, and etoposide (PABIOE), ChlVPP/etoposide, vincristine, and doxorubicin (EVA), MOPP/EBV/CAD, COPP/EBV/CAD) had developed an SMN. This difference was not statistically significant (P=0.15; Peto’s OR: 1.37, 95%-CI: 0.89-2.10; Figures 4 and 5). When considering t-AML/MDS, secondary NHL and secondary solid tumors separately, an increased risk to develop t-AML/MDS was seen for patients treated with intensified chemotherapy protocols (P=0.0028), whereas the incidence rates for secondary NHL and secondary solid tumors did not differ between the treatment groups (Table 3). Overall, tumor control was significantly better with intensified chemotherapy regimens as compared with standard-dose protocols (P=0.007; HR: 0.82, 95%-CI: (0.70 – 0.95)) while there were no significant differences in OS (P=0.12; HR: 0.85, 95%-CI: (0.70 – 1.04)). However, subgroup analyses revealed that patients aged ≤50, in particular, appear to benefit from more aggressive chemotherapy approaches as the improved PFS also translated into a

Figure 3. Additional radiotherapy, cumulative incidence of SMN (Peto meta-analysis). Vertical bars depict approximate 95% confidence intervals (CI) for cumulative incidence rates. CT: chemotherapy; RT: radiotherapy.

haematologica | 2017; 102(10)

1753

D.A. Eichenauer et al.

better OS in these patients (P=0.02). In contrast with patients older than 50 years, patients aged â&#x2030;¤50 also had a significantly greater SMN risk with intensified chemotherapy protocols than with standard-dose chemotherapy (interaction P=0.02; treatment effect in younger subgroup: P=0.01). Additional subgroup analyses revealed differences between the different intensified protocols for PFS (interaction P<0.00001) and OS (interaction P=0.006).

Patients treated with escalated BEACOPP had superior PFS and OS rates in comparison with those receiving standard-dose chemotherapy. For Stanford V, PFS was worse than with standard-dose chemotherapy. No PFS and OS differences in comparison with standard-dose chemotherapy were observed for ChlVPP/PABIOE, ChlVPP/EVA, MOPP/EBV/CAD and COPP/EBV/CAD. For SMN no interaction in chemotherapy subgroups were found (interaction P=0.06; Online Supplementary Table S3).

All of the main meta-analytic results are summarized in Table 2.

Sensitivity analyses agreed with the described main analyses. The results of sensitivity analyses concerning censoring of incomplete follow-up periods and exclusion of NMSC are summarized in Online Supplementary Tables S4 and S5.

Discussion The meta-analysis herein, including 9498 patients treated within 16 randomized clinical trials for newly diagnosed HL between 1984 and 2007, represents one of the

largest analyses of SMN, PFS and OS of HL patients based on randomized comparisons. The major findings were as follows: (1) after a median follow-up of 7.4 years, the overall SMN rate was 4.6%, (2) compared with patients receiving chemotherapy alone, an increased SMN rate was observed in patients receiving CMT, (3) patients with early-stage HL treated with CMT had a better PFS than patients treated with chemotherapy alone, (4) compared with patients receiving standard-dose chemotherapy, those receiving intensified chemotherapy protocols developed t-AML/MDS more often, and (5) compared with ABVD-like protocols, PFS and OS in advanced-stage patients were improved with escalated BEACOPP, but higher rates of SMN were observed.

The overall SMN rate of 4.6% in the meta-analysis herein was lower than in previous reports. A meta-analysis from our group included a total of 9312 patients treated between 1962 and 2000. After median follow-up times ranging between four and 32 years for the considered trials, the overall SMN rate was 7.6%.39,40 A British analysis including 5798 patients treated between 1963 and 2001 reported an SMN rate of 7.9%.4 According to a Dutch study with a median follow-up of 19.1 years, the overall SMN rate was 23% and the risk for the development of an SMN was still increased 30 years after HL treatment.3 Two reasons likely contribute to the higher SMN rates in these previous analyses: (1) a relevant proportion of the expected secondary solid tumors that are often diagnosed ten or more years after HL treatment has not yet occurred in the patients included in the meta-analysis herein due to the limited median follow-up of 7.4 years, and (2) the SMN

(P=0.35) (P=0.15)

Favors dose-intensified

Favors ABVD-like

Figure 4. Intensified chemotherapy, secondary malignant neoplasms, forest plot for Peto Odds Ratios. CI: confidence interval; O-E: observed minus expected; V: variance, I2: measure of heterogeneity; ABVD: doxorubicin, bleomycin, vinblastine, dacarbazine.

Table 3. Summary of SMN results for each SMN type.

Comparison (standard vs. experimental) CMT vs. chemotherapy alone Extended vs. involved-field RT (after CT) Higher dose vs. 20 Gy RT (after CT) More vs. fewer CT cycles Standard-dose vs. intensified CT (regimen +/- RT)

Solid tumor

t-AML/MDS

NHL

OR

P

OR

P

OR

P

0.627 0.851 1.20 1.15 1.00

0.29 0.37 0.43 0.56 1.0

0.293 0.517 0.662 0.261 4.51

0.037 0.14 0.65 0.10 0.0028

0.325 1.18 0.845 1.94 0.61

0.21 0.66 0.67 0.13 0.26

t-AML/MDS: therapy-related acute myeloid leukemia/myelodysplastic syndromes; NHL: non- Hodgkin lymphomas; CMT: combined-modality treatment; RT: radiotherapy; CT: chemotherapy; OR: Peto odds ratio; P: P-value.

1754

haematologica | 2017; 102(10)

Secondary malignant neoplasms after HL treatment

rate observed within the present meta-analysis is probably truly lower than the rates seen in the older analyses, since in recent years RT fields and doses were reduced and chemotherapy protocols were modified with the aim of decreasing the SMN risk.

In the meta-analysis herein, patients receiving CMT for HL had a significantly increased risk to develop an SMN when compared with chemotherapy alone. Figure 3 shows that at ten years after first-line treatment, the absolute cumulative SMN risks are approximately 3% and 10%, a risk difference of 7%. This result is in line with previous analyses such as the above mentioned British study, in which patients treated with chemotherapy alone had a significantly lower relative risk for the development of an SMN than patients receiving CMT.4 Subgroup analyses of the analysis herein considering t-AML/MDS, secondary NHL and secondary solid tumors separately, detected a significantly increased risk after CMT for tAML/MDS only, but not for secondary NHL and secondary solid tumors. This finding is consistent with older reports, including a study from Italy comprising 1659 patients treated with RT alone, CMT or chemotherapy alone. At 15 years, the t-AML/MDS rate after CMT was significantly higher than after chemotherapy alone (P=0.05).41 According to the meta-analysis herein, patients diagnosed with early-stage HL had a better PFS after CMT than after chemotherapy alone. However, this finding has to be interpreted with caution as it derives from only one of the included trials. Nonetheless, similar data also came from the randomized EORTC/Lymphoma Study Association (LYSA)/Fondazione Italiana Linfomi (FIL) H10 study and the RAPID trial conducted in the UK. These studies evaluated the positron emission tomography (PET)-guided omission of consolidating RT after chemotherapy in patients with early-stage HL. Both studies additionally revealed a significantly increased event rate after chemotherapy alone in patients with a good response to chemotherapy resulting in a negative interim

PET.42-44 Additional data supporting the use of CMT in early-stage HL come from a previous Cochrane systematic review including 1245 patients from five randomized studies and an analysis comprising 20600 patients registered in the U.S. National Cancer Data Base, both of which have not only demonstrated a better PFS but also an improved OS among patients treated with CMT compared with chemotherapy alone.45,46 In the meta-analysis herein, an increased t-AML/MDS rate was seen in patients receiving intensified chemotherapy compared with patients treated with standard-dose chemotherapy. This finding is consistent with other reports on t-AML/MDS after HL treatment. An analysis by the GHSG, including 11952 patients treated within prospective studies for newly diagnosed HL, demonstrated that patients receiving no BEACOPP or up to four cycles of escalated BEACOPP had significantly lower cumulative t-AML/MDS rates than patients treated with four or more cycles of escalated BEACOPP (0.3% vs. 0.7% vs. 1.7%; P<0.0001).7 For escalated BEACOPP, the increased t-AML/MDS risk contrasts with an improved clinical outcome. PFS (P<0.00001) and OS (P=0.0005) rates were better than those seen with standard-dose protocols, i.e., ABVD or COPP/ABVD. This is in line with the results of a network meta-analysis on the effect of the initial treatment strategy on the survival of patients with advanced HL. That analysis, which included a total of 9993 patients, revealed a survival advantage of 10% at five years for escalated BEACOPP in comparison with ABVD.47 Generally, this meta-analysis provides high-quality evidence on SMN, PFS and OS among patients treated for HL, as the used data are from participants of large randomized trials for the first-line treatment of HL. However, the analysis has some limitations. With a median overall follow-up of 7.4 years, valid estimates are only possible for secondary hematological malignancies, whilst final conclusions regarding secondary solid tumors that often occur more than ten years after HL cannot be drawn. The

Figure 5. Intensified chemotherapy, cumulative incidence of SMN (Peto meta-analysis). Vertical bars depict approximate 95% confidence intervals (CI) for cumulative incidence rates. ABVD: doxorubicin, bleomycin, vinblastine, dacarbazine.

haematologica | 2017; 102(10)

1755

D.A. Eichenauer et al.

long-term data of HL patients treated within clinical trials are therefore necessary but often difficult to obtain due to different factors, including a limited duration of insurance for study participants and a lack of funding sources. There is also some uncertainty about the completeness of SMN reporting which is of particular importance due to the small number of SMN events. Finally, for certain outcomes and study questions there was a considerable heterogeneity of up to 89% between the included trials, which signifies that the overall meta-analytic results may not apply in all situations.

Nonetheless, given the relevant proportion of HL patients that have already developed an SMN after a median observation of 7.4 years, the present report underscores the need for treatment approaches allowing a more accurate allocation to defined risk groups, in order to prevent overtreatment and reduce the risk of the development of potentially fatal SMN. At present, interim PET is considered the most promising tool to stratify treat-

References 1. Eichenauer DA, Engert A, Andre M, et al.

Hodgkin's lymphoma: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2014;25 Suppl 3:iii70-75.

2. Borchmann P, Eichenauer DA, Engert A. State of the art in the treatment of Hodgkin lymphoma. Nat Rev Clin Oncol. 2012; 9(8):450-459.

3. Schaapveld M, Aleman BM, van Eggermond AM, et al. Second cancer risk up to 40 years after treatment for Hodgkin's lymphoma. N Engl J Med. 2015; 373(26):2499-2511.

4. Swerdlow AJ, Higgins CD, Smith P, et al.

Second cancer risk after chemotherapy for Hodgkin's lymphoma: a collaborative British cohort study. J Clin Oncol. 2011; 29(31):4096-4104.

5. Aleman BM, van den Belt-Dusebout AW, De Bruin ML, et al. Late cardiotoxicity after treatment for Hodgkin lymphoma. Blood.

2007;109(5):1878-1886.

6. Behringer K, Mueller H, Goergen H, et al.

Gonadal function and fertility in survivors after Hodgkin lymphoma treatment within the German Hodgkin Study Group HD13 to HD15 trials. J Clin Oncol. 2013; 31(2):231-239.

7. Eichenauer DA, Thielen I, Haverkamp H, et al. Therapy-related acute myeloid leukemia and myelodysplastic syndromes in patients with Hodgkin lymphoma: a report from the German Hodgkin Study Group. Blood.

2014;123(11):1658-1664.

8. Koontz MZ, Horning SJ, Balise R, et al. Risk of therapy-related secondary leukemia in Hodgkin lymphoma: the Stanford University experience over three generations of clinical trials. J Clin Oncol. 2013; 31(5):592-598.

9. Leone G, Voso MT, Sica S, Morosetti R, Pagano L. Therapy related leukemias: susceptibility, prevention and treatment. Leuk Lymphoma. 2001;41(3-4):255-276.

10. Bonadonna G, Zucali R, Monfardini S, De Lena M, Uslenghi C. Combination chemotherapy of Hodgkin's disease with adriamycin, bleomycin, vinblastine, and

1756

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

ment. Some mature results addressing this issue are already available and additional studies will be analyzed in the near future.48,49 The replacement of conventional chemotherapy by novel agents such as the CD30-directed antibody-drug conjugate brentuximab vedotin and antibodies targeting the programmed death cell protein 1 (PD-1) may also reduce the risk for the development of SMN.50,51 Funding The authors would like to thank the Bundesministerium fĂźr Bildung und Forschung (German Federal Ministry of Education and Research, grant no. 01KG0921) for financial support.

Acknowledgments Further thanks go to the Cochrane Collaboration and in particular to the Cochrane Haematological Malignancies Review Group (coordinating editor: Dr. Nicole Skoetz, M.D.) for reviewing the protocol and draft of the Cochrane systematic review.

imidazole carboxamide versus MOPP. Cancer. 1975;36(1):252-259.

Diehl V, Sieber M, Ruffer U, et al. BEACOPP: an intensified chemotherapy regimen in advanced Hodgkin's disease. The German Hodgkin's Lymphoma Study Group. Ann Oncol. 1997;8(2):143-148.

Bartlett NL, Rosenberg SA, Hoppe RT, Hancock SL, Horning SJ. Brief chemotherapy, Stanford V, and adjuvant radiotherapy for bulky or advanced-stage Hodgkin's disease: a preliminary report. J Clin Oncol.

1995;13(5):1080-1088.

Rueffer U, Josting A, Franklin J, et al. NonHodgkin's lymphoma after primary Hodgkin's disease in the German Hodgkin's Lymphoma Study Group: incidence, treatment, and prognosis. J Clin Oncol. 2001;19(7):2026-2032.

Higgins JPT, Altman DG, Sterne JAC. Chapter 8: Assessing risk of bias in included studies. Cochrane Handbook for Systematic Reviews of Interventions (Editors: JPT Higgins, S Green). 2011.

Peto R, Pike MC, Armitage P, et al. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. I. Introduction and design. Br J Cancer. 1976;34(6):585-612.

Peto R, Pike LC, Armitage P, et al. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. II. analysis and examples. Bri J Cancer. 1977;35(1):1-39.

Pepe MS, Mori M. Kaplan-Meier, marginal or conditional probability curves in summarizing competing risks failure time data?

Stat Med. 1993;12(8):737-751.

Tai BC, Machin D, White I, Gebski V. Competing risks analysis of patients with osteosarcoma: a comparison of four different approaches. Stat Med. 2001;20(5):661684.

Aleman BM, Raemaekers JM, Tirelli U, et al. Involved-field radiotherapy for advanced Hodgkin's lymphoma. N Engl J Med. 2003;348(24):2396-2406.

Ferme C, Eghbali H, Meerwaldt JH, et al.

Chemotherapy plus involved-field radiation in early-stage Hodgkin's disease. N Engl J Med. 2007;357(19):1916-1927.

21. Thomas J, Ferme C, Noordijk EM, et al.

Results of the EORTC-GELA H9 randomized trials: The H9-F trial (comparing 3 radiation dose levels) and H9-U trial (comparing 3 chemotherapy schemes) in patients with favorable or unfavorable early stage Hodgkin´s lymphoma (HL).

Haematologica. 2007; 92(Supplement 5): abstract C010.

22. Engert A, Plutschow A, Eich HT, et al.

Reduced treatment intensity in patients with early-stage Hodgkin's lymphoma. N Engl J Med. 2010;363(7):640-652.

23. Eich HT, Diehl V, Gorgen H, et al.

Intensified chemotherapy and dosereduced involved-field radiotherapy in patients with early unfavorable Hodgkin's lymphoma: final analysis of the German Hodgkin Study Group HD11 trial. J Clin Oncol. 2010;28(27):4199-4206.

24. Diehl V, Loeffler M, Pfreundschuh M, et al.

Further chemotherapy versus low-dose involved-field radiotherapy as consolidation of complete remission after six cycles of alternating chemotherapy in patients with advance Hodgkin's disease. German Hodgkins' Study Group (GHSG). Ann Oncol. 1995;6(9):901-910.

25. Engert A, Schiller P, Josting A, et al.

Involved-field radiotherapy is equally effective and less toxic compared with extended-field radiotherapy after four cycles of chemotherapy in patients with early-stage unfavorable Hodgkin's lymphoma: results of the HD8 trial of the German Hodgkin's Lymphoma Study Group. J Clin Oncol.

2003;21(19):3601-3608.

26. Diehl V, Franklin J, Pfreundschuh M, et al.

Standard and increased-dose BEACOPP chemotherapy compared with COPPABVD for advanced Hodgkin's disease. N Engl J Med. 2003;348(24):2386-2395.

27. Federico M, Luminari S, Iannitto E, et al.

ABVD compared with BEACOPP compared with CEC for the initial treatment of patients with advanced Hodgkin's lymphoma: results from the HD2000 Gruppo Italiano per lo Studio dei Linfomi Trial. J Clin Oncol. 2009;27(5):805-811.

28. Gobbi PG, Levis A, Chisesi T, et al. ABVD versus modified Stanford V versus MOPPE-

haematologica | 2017; 102(10)

Secondary malignant neoplasms after HL treatment

29.

30.

31.

32.

33.

34.

35.

BVCAD with optional and limited radiotherapy in intermediate- and advancedstage Hodgkin's lymphoma: final results of a multicenter randomized trial by the Intergruppo Italiano Linfomi. J Clin Oncol.

2005;23(36):9198-9207.

Viviani S, Zinzani PL, Rambaldi A, et al.

ABVD versus BEACOPP for Hodgkin's lymphoma when high-dose salvage is planned. N Engl J Med. 2011;365(3):203212.

Bonadonna G, Bonfante V, Viviani S, Di Russo A, Villani F, Valagussa P. ABVD plus subtotal nodal versus involved-field radiotherapy in early-stage Hodgkin's disease: long-term results. J Clin Oncol. 2004; 22(14):2835-2841.

Anselmo AP, Cavalieri E, Osti FM, et al.

Intermediate stage Hodgkin's disease: preliminary results on 210 patients treated with four ABVD chemotherapy cycles plus extended versus involved field radiotherapy. Anticancer Res. 2004;24(6):4045-4050.

Johnson PW, Radford JA, Cullen MH, et al.

Comparison of ABVD and alternating or hybrid multidrug regimens for the treatment of advanced Hodgkin's lymphoma: results of the United Kingdom Lymphoma Group LY09 Trial (ISRCTN97144519). J Clin Oncol. 2005;23(36):9208-9218.

Hoskin PJ, Lowry L, Horwich A, et al.

Randomized comparison of the stanford V regimen and ABVD in the treatment of advanced Hodgkin's Lymphoma: United Kingdom National Cancer Research Institute Lymphoma Group Study ISRCTN 64141244. J Clin Oncol. 2009;27(32):53905396.

Nachman JB, Sposto R, Herzog P, et al.

Randomized comparison of low-dose involved-field radiotherapy and no radiotherapy for children with Hodgkin's disease who achieve a complete response to chemotherapy.

J Clin Oncol.

2002;20(18):3765-3771.

Kung FH, Schwartz CL, Ferree CR, et al.

POG 8625: a randomized trial comparing chemotherapy with chemoradiotherapy for children and adolescents with Stages I, IIA, IIIA1 Hodgkin Disease: a report from

haematologica | 2017; 102(10)

36.

37.

38.

39.

40.

41.

42.

the Children's Oncology Group. J Pediatr Hematol Oncol. 2006;28(6):362-368.

Laskar S, Gupta T, Vimal S, et al.

Consolidation radiation after complete remission in Hodgkin's disease following six cycles of doxorubicin, bleomycin, vinblastine, and dacarbazine chemotherapy: is there a need? J Clin Oncol. 2004;22(1):62-68.

Gerhartz HH, Schwenke H, Bazarbashi S, et al. Randomised comparison of COPP/ABVD versus dose- and time-escalated COPP/ABVD with GM-CSF support for advanced Hodgkin s disease. Proc Am Soc Clin Oncol. 1997;16:8a.

Gordon LI, Hong F, Fisher RI, et al.

Randomized phase III trial of ABVD versus Stanford V with or without radiation therapy in locally extensive and advanced-stage Hodgkin lymphoma: an intergroup study coordinated by the Eastern Cooperative Oncology Group (E2496). J Clin Oncol.

2013;31(6):684-691.

Franklin JG, Paus MD, Pluetschow A, Specht L. Chemotherapy, radiotherapy and combined modality for Hodgkin's disease, with emphasis on second cancer risk.

Cochrane Database Syst Rev. 2005; (4):CD003187.

Franklin J, Pluetschow A, Paus M, et al.

Second malignancy risk associated with treatment of Hodgkin's lymphoma: metaanalysis of the randomised trials. Ann Oncol. 2006;17(12):1749-1760.

Brusamolino E, Anselmo AP, Klersy C, et al.

The risk of acute leukemia in patients treated for Hodgkin's disease is significantly higher aft [see bined modality programs than after chemotherapy alone and is correlated with the extent of radiotherapy and type and duration of chemotherapy: a casecontrol study. Haematologica. 1998; 83(9):812-823.

Raemaekers JM, Andre MP, Federico M, et al. Omitting radiotherapy in early positron emission tomography-negative stage I/II Hodgkin lymphoma is associated with an increased risk of early relapse: Clinical results of the preplanned interim analysis of the randomized EORTC/LYSA/FIL H10 trial. J Clin Oncol. 2014;32(12):1188-1194.

43. Radford J, Illidge T, Counsell N, et al.

Results of a trial of PET-directed therapy for early-stage Hodgkin's lymphoma. N Engl J Med. 2015;372(17):1598-1607.

44. Andre MPE, Girinsky T, Federico M, et al.

Early positron emission tomography response-adapted treatment in stage I and II Hodgkin Lymphoma: final results of the randomized EORTC/LYSA/FIL H10 Trial. J Clin Oncol. 2017;35(16):1786-1794.

45. Herbst C, Rehan FA, Skoetz N, et al.

Chemotherapy alone versus chemotherapy plus radiotherapy for early stage Hodgkin lymphoma. Cochrane Database Syst Rev. 2011;(2):CD007110.

46. Olszewski AJ, Shrestha R, Castillo JJ. Treatment selection and outcomes in earlystage classical Hodgkin lymphoma: analysis of the National Cancer Data Base. J Clin Oncol. 2015;33(6):625-633.

47. Skoetz N, Trelle S, Rancea M, et al. Effect of initial treatment strategy on survival of patients with advanced-stage Hodgkin's lymphoma: a systematic review and network meta-analysis. Lancet Oncol. 2013; 14(10):943-952.

48. Engert A, Haverkamp H, Kobe C, et al.

Reduced-intensity chemotherapy and PETguided radiotherapy in patients with advanced stage Hodgkin's lymphoma (HD15 trial): a randomised, open-label, phase 3 non-inferiority trial. Lancet. 2012; 379(9828):1791-1799.

49. Johnson P, Federico M, Kirkwood A, et al.

Adapted Treatment Guided by Interim PET-CT Scan in Advanced Hodgkin's Lymphoma. N Engl J Med. 2016; 374(25):2419-2429.

50. Younes A, Gopal AK, Smith SE, et al.

Results of a pivotal phase II study of brentuximab vedotin for patients with relapsed or refractory Hodgkin's lymphoma. J Clin Oncol. 2012;30(18):2183-2189.

51. Younes A, Santoro A, Shipp M, et al.

Nivolumab for classical Hodgkin's lymphoma after failure of both autologous stem-cell transplantation and brentuximab vedotin: a multicentre, multicohort, singlearm phase 2 trial. Lancet Oncol. 2016; 17(9):1283-1294.

1757

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Non-Hodgkin Lymphoma

Ferrata Storti Foundation

Haematologica 2017 Volume 102(10):1758-1766

Exome sequencing identifies recurrent BCOR alterations and the absence of KLF2, TNFAIP3 and MYD88 mutations in splenic diffuse red pulp small B-cell lymphoma Laurent Jallades,1,2 Lucile Baseggio,1,2 Pierre Sujobert,1,2,3 Sarah Huet,1,2,3 Kaddour Chabane,1,2 Evelyne Callet-Bauchu,1,2,3 Aurélie Verney,2,3 Sandrine Hayette,1,2 Jean-Pierre Desvignes,4,5 David Salgado,4,5 Nicolas Levy,4,5,6 Christophe Béroud,4,5,6 Pascale Felman,1,2 Françoise Berger,2,3,7 Jean-Pierre Magaud,1,2,3 Laurent Genestier,2 Gilles Salles2,3,8 and Alexandra TraverseGlehen2,3,7

Hospices Civils de Lyon, Centre Hospitalier Lyon Sud, Laboratoire d’Hématologie, Pierre-Bénite; 2Cancer Research Center of Lyon, INSERM 1052 CNRS 5286, Team “Clinical and Experimental Models of Lymphomagenesis”, Faculté de Médecine et de Maïeutique Lyon-Sud Charles Mérieux, Oulins; 3Université Claude Bernard Lyon-1; 4 Aix-Marseille Université, GMGF, 13385, Marseillee; 5INSERM, UMR_S 910, 13385, Marseille; 6APHM, Hôpital TIMONE Enfants, Laboratoire de Génétique Moléculaire, 13385, Marseille; 7Hospices Civils de Lyon, Centre Hospitalier Lyon Sud, Laboratoire d’Anatomie Pathologique, Pierre-Bénite and 8Hospices Civils de Lyon, Centre Hospitalier Lyon Sud, Service d’Hématologie, Pierre-Bénite, France 1

ABSTRACT

S

Correspondence: gilles.salles@chu-lyon.fr

Received: November 24, 2016.

Accepted: July 12, 2017.

Pre-published: July 27, 2017.

doi:10.3324/haematol.2016.160192 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1758 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1758

plenic diffuse red pulp lymphoma is an indolent small B-cell lymphoma recognized as a provisional entity in the World Health Organization 2008 classification. Its precise relationship to other related splenic B-cell lymphomas with frequent leukemic involvement or other lymphoproliferative disorders remains undetermined. We performed whole-exome sequencing to explore the genetic landscape of ten cases of splenic diffuse red pulp lymphoma using paired tumor and normal samples. A selection of 109 somatic mutations was then evaluated in a cohort including 42 samples of splenic diffuse red pulp lymphoma and compared to those identified in 46 samples of splenic marginal zone lymphoma and eight samples of hairy-cell leukemia. Recurrent mutations or losses in BCOR (the gene encoding the BCL6 corepressor) – frameshift (n=3), nonsense (n=2), splicing site (n=1), and copy number loss (n=4) – were identified in 10/42 samples of splenic diffuse red pulp lymphoma (24%), whereas only one frameshift mutation was identified in 46 cases of splenic marginal zone lymphoma (2%). Inversely, KLF2, TNFAIP3 and MYD88, common mutations in splenic marginal zone lymphoma, were rare (one KLF2 mutant in 42 samples; 2%) or absent (TNFAIP3 and MYD88) in splenic diffuse red pulp lymphoma. These findings define an original genetic profile of splenic diffuse red pulp lymphoma and suggest that the mechanisms of pathogenesis of this lymphoma are distinct from those of splenic marginal zone lymphoma and hairy-cell leukemia.

Introduction Splenic diffuse red pulp lymphoma (SDRPL) with circulating villous lymphocytes is a rare indolent B-cell lymphoma involving the spleen, bone marrow and peripheral blood and is characterized by various clinical, morphological and immunological features.1-3 However, SDRPL was defined as a provisional entity in the World Health Organization 2008 classification and in its recent release in 2016, and assigned to the unclassifiable splenic B-cell lymphomas/leukemias.4,5 Indeed, SDRPL haematologica | 2017; 102(10)

Whole-exome sequencing of SDRPL

may present some overlapping features with other splenic B-cell lymphomas or small B-cell leukemias such as splenic marginal zone lymphoma (SMZL), hairy cell leukemia (HCL) and, especially, its variant form (HCL-v). The differential diagnosis may be difficult because of the absence of pathognomonic diagnostic markers. Recurrent mutations have been reported in HCL (BRAF V600E), HCL-v (MAP2K1) and in SMZL (KLF2, NOTCH2), indicating characteristic mutational patterns and distinctive oncogenic pathways in each of these entities.6-13 Some mutations in NOTCH1, NOTCH2, MYD88, TP53, MAP2K1, and CCND3 have recently been described in SDRPL, though the studies were non-exhaustive and lacked detailed comparisons with other B-cell malignancies.14-16 In the present study, we explored the genetic landscape of SDRPL using whole-exome sequencing of paired tumor and normal samples. We confirmed and extended our findings through the targeted sequencing of 109 mutations in a validation series of SDRPL and compared these data with those obtained for SMZL and HCL.

Methods

Germany). The purity of the B-cell fractions was determined by flow cytometry and systematically exceeded 90%. Non-B-cell fractions contained less than 5% CD20+ cells.

Genomic DNA was enriched in protein-coding sequences using the in-solution exome capture SureSelect Human All Exon 50-Mb kit (Agilent Technologies) according to the manufacturer’s protocol. The captured targets were subjected to sequencing using the Illumina HiSEQ2000 analyzer (Illumina) with the paired-end 2 x 75 bp read option. Exome capture, massively parallel sequencing and quality controls were performed at IntegraGen (Evry, France).

Paired-end reads obtained by high-throughput sequencing were aligned with the human genome reference hg19/NCBI GRCh37, and differences from the reference sequence were identified separately for tumor and normal samples using the CASAVA pipeline (Illumina) (IntegraGen), as well as another pipeline based on BWAMEM, SAMBAMBA, and GATK (INSERM UMR S910, Marseille, France). Merge analysis, data mining and manual review were performed using VarAFT (http://varaft.eu) and ALAMUT software (Interactive Biosoftware, Rouen, France). To investigate genomic copy number aberrations (e.g., copy number gains and copy number losses), we used the Bioconductor DNACopy package (DNAcopy 1.32.0), comparing the DNA exome data with the paired reference sample data and used the circular binary segmentation algorithm to segment DNA copy number data. All changes

Case selection Diagnoses of HCL, SDRPL and SMZL were established by histological analyses of spleen (62 cases) or peripheral blood (38 cases) (Table 1). Given the frequent policy of watchful waiting and the low rate of splenectomy in SDRPL patients, 31 samples were included in the study after a diagnostic procedure based on thorough cytological examination of peripheral blood smears completed with bone marrow analyses, extensive flow cytometry immunophenotyping, and cytogenetic analyses. The immunophenotypic characteristics of SDRPL were previously shown to discriminate this type of lymphoma from other lymphoid malignancies.2,3 In our experience, SDRPL can be clearly distinguished from SMZL using a scoring system based on five membrane markers (CD11c, CD22, CD76, CD27 and CD38) and from HCL given that SDRPL does not co-express typical HCL markers such as CD25, CD103 and CD123.2 However, in some cases, a partial expression of CD103 may be observed in SDRPL. The criteria used to recognize each entity were in accordance with the World Health Organization 2008 classification, completed with recent published updates.2,17-23 The clinico-pathological characteristics of the HCL, SDRPL and SMZL series are detailed in Table 1. Informed consent to participation in this study was obtained from patients, and the procedures were conducted in accordance with the Helsinki Declaration and the Biological Resource Center policy of the Hospices Civils de Lyon. Furthermore, the institutional review board of the Hospices Civils de Lyon approved the research protocol (DC-2015-2566).

Whole-exome sequencing Whole-exome sequencing was performed on a discovery cohort (flowchart in Online Supplementary Figure S1), which included ten cases of SDRPL (namely, SDRPL #1, 3, 5, 7, 8, 9, 10, 12, 13, and 15).

DNA from tumor cells was obtained from seven frozen spleen samples and three positively immunoselected villous lymphoma B-cell (CD19+) samples isolated from peripheral blood (to verify the origin of the samples, see Figure 1). Paired DNA from nonmalignant cells was purified from spleen fibroblasts obtained after culture of the original tissue biopsy or from the non-B-cell fraction obtained after immunoselection using a human anti-CD19 antibody-conjugated magnetic microbead kit according to the manufacturer’s instructions (Miltenyi Biotec, Bergisch Gladbach, haematologica | 2017; 102(10)

Table 1. Clinico-pathological comparison between the HCL, SDRPL and SMZL cases of our series

HCL Total samples (n=96) 8 Spleen 3 Peripheral blood 5 Sex ratio (F:M) 2:6 (0.33) Median age (years) 47 Immunophenotype# k:l ratio 4:3 (1.33) CD11c 7/7 (100%) CD25 7/7 (100%) CD27 0/6 (0%) CD103 6/6 (100%) CD123 6/6 (100%) CD76 4/4 (100%) CD38 1/7 (14%) CD23 0/7 (0%) ANXA1 3/3 (100%) Cytogenetic n.a.

Deletion 7q Trisomy 3 Trisomy 12 Trisomy 18 Complex karyotype IGHV gene status 4/8 IGHV mutated (<100%) a 4/4 IGHV1–2 0/4 IGHV3–23 0/4 IGHV3–30 1/4 IGHV4–34 1/4 IGHV4-59 0/4

SDRPL 42 11 31 15:27 (0.56) 80 14:28 (0.50) 35/41 (85%) 0/41 (0%) 8/33 (24%) 9/40 (23%) 0/31 (0%) 20/23 (87%) 2/40 (5%) 2/42 (5%) 0/11 (0%) 24/42 (57%) 6/24 (25%) 1/24 (4%) 3/24 (13%) 1/24 (4%) 3/24 (13%) 36/42 (86%) 32/36 (89%) 1/36 (3%) 5/36 (14%) 2/36 (6%) 8/36 (22%) 4/36 (11%)

SMZL 46 46 0 *25:21 (1.19) *68 *24:17 (1.41) *14/26 (54%) 0/26 (0%) *32/35 (91%) *0/27 (0%) 1/19 (5%) *2/13 (15%) *15/39 (39%) *19/41 (46%) 0/46 (0%) 41/46 (89%) 13/41 (32%) *9/41 (22%) *9/41 (22%) *4/41 (10%) *13/41 (32%) 40/46 (87%) 36/40 (90%) *15/40 (38%) *1/40 (3%) 1/40 (3%) *3/40 (8%) 5/40 (13%)

Marker (*) indicates feature that discriminates between SDRPL and SMZL entities. #All antigen expressions were studied by flow cytometry except CD76, which was assessed by immunocytochemistry. aPercentage of IGHV mutated cases considering a case ‘mutated’ when it had ≤ 100% of homology with the germline sequence. n.a.: not available.

1759

Figure 1. Distribution of mutations in hairy-cell leukemia, splenic diffuse red pulp lymphoma and splenic marginal zone lymphoma. Each column represents a type of lymphoma and corresponding tissue (HCL: hairy cell lymphoma; SDRPL: splenic diffuse red pulp lymphoma; SMZL: splenic marginal zone lymphoma; PB: peripheral blood; SP: spleen; B+: CD19+ immunoselected sample). The orange colored boxes indicate the ten cases investigated by whole-exome sequencing. Six of them were also reanalyzed by targeted sequencing as internal controls, whereas the four remaining were not (NR: not resequenced). Each row (top) represents a gene linked to the corresponding RAS/MAPK, NF-kB or NOTCH signaling pathway. Three types of mutations (missense, nonsense, frameshift indel) are highlighted in different colors. Details of the mutations are available in Online Supplementary Table S3. Gene copy number (CN) was only reported as a gain (↑) or loss (↓) for relevant genes (), namely BCOR and TNFAIP3. The last row (bottom) represents the IGHV gene status, homology with germline sequence percentage, and cytogenetic findings (complex karyotype, trisomy 3, 12, 18 and deletion 7q) for each case when available.

L. Jallades et al.

1760

haematologica | 2017; 102(10)

Whole-exome sequencing of SDRPL

were compared to the catalogue of the Database of Genomic Variants to provide a comprehensive summary of structural variations in the human genome.

Targeted gene sequencing A panel of 109 genes was selected from the exploratory exome data for subsequent analyses; these genes were selected according to their previous description in mutational landscape studies of other B-cell malignancies, to their well-established role in B-cell physiology or to the type of mutation they produce (predicted to be deleterious or damaging according to SIFT and PolyPhen2score). The genes, exon positions and coverage are listed in Online Supplementary Table S2. High-throughput sequencing was performed on a validation series of 38 SDRPL samples, consisting of six samples from the discovery cohort used as positive controls (namely, SDRPL #1, 3, 5, 8, 9 and 12) and 32 new SDRPL samples, as well as 46 SMZL and eight HCL samples. The origin of the samples is shown in Figure 1. DNA was extracted from frozen spleen samples, from CD19+ immunoselected cells, or from nonimmunoselected peripheral blood mononuclear cells after Ficollisolation using patients’ samples with a total lymphocyte count exceeding 5x109/L (mean 14x109/L; minimum: 5.8x109/L; maximum: 92x109/L). All of the spleen tissue samples were infiltrated with over 70% B-cell lymphoma tumor cells.

Library preparation, capture, sequencing, variant detection, and annotation were performed by IntegraGen. Exons of genomic DNA samples were captured using the Agilent in-solution enrichment methodology with the biotinylated oligonucleotide probe library, followed by paired-end 75-bp massively parallel sequencing on an Illumina HiSEQ2000. Image analysis and base calling were performed using the Illumina Real Time Analysis Pipeline version 1.14 and default parameters. Bioinformatics analysis of the sequencing data was based on the Illumina pipeline (CASAVA1.8.2). Only positions included in the bait coordinates were conserved. Genetic variation annotation was performed using the IntegraGen in-house pipeline, which consists of gene annotation (RefSeq) and detection of known polymorphisms (dbSNP 132, 1000Genome) followed by characterization of the mutations (exonic, intronic, silent, nonsense, etc.). For each position, exomic frequencies were determined using the IntegraGen Exome database, and exome results were provided by HapMap. The procedure for detecting copy number variation was adapted from previously published methods; read depths of each exon from all of the target genes were obtained for each sample using DeCovA.24,25 Exon read depths (ERD) were then normalized against the sum of all ERD from the same sample. A copy number ratio (CNR) was obtained by dividing the ERD of the sample by the median ERD of all of the samples as a control: CNRx=(ERDx/∑ERD1>n)sample/(ERDx/∑ERD1>n)control. For the two genes located on chromosome Xp (BCOR and KDM6A), a 2-fold extrapolation was applied to the normalized ERD for males. Copy number ratios <0.7 and >1.25 were considered to represent a loss and gain, respectively. Chromosomal abnormalities (chromosome X loss or gain) were confirmed by copy number variation detection by whole-exome sequencing using a circular binary segmentation algorithm and compared to that of the paired reference genome or karyotype analysis when available.

Microarray-based comparative genomic hybridization A 60K oligonucleotide microarray (Agilent Technologies) was used according to the manufacturer’s instructions to confirm the microdeletion of the BCOR locus, identified by the detection, by whole-exome sequencing, of copy number variation (SDRPL case: VL_218). haematologica | 2017; 102(10)

Results Whole-exome sequencing of the discovery cohort of SDRPL resulted in the identification of more than 300 unique somatic mutations among the ten different tumor samples (Online Supplementary Table S1). These included exon missense (83%), frameshift (5%), nonsense (4%), untranslated regions (5%), and intron splicing site (3%) alterations. Some of the mutated genes were previously associated with various other malignancies. These mutated genes encode proteins in various pathways, including cell cycle regulation (CCND3, MGA, MYC), epigenetic regulation (BCOR, EZH1, HIST1H1D, HIST1H2AD, HIST4H4, KAT6A, KDM6A, NCOA6), the RAS-MAPK pathway (HRAS, KIF26A, NRAS, RAF1, WNK1), NF-kB signaling (IKBKB, TBK1, TRAF3), NOTCH signaling (NOTCH1, NOTCH2, DTX3L, DTX1, SPEN), and cytoskeleton and cell-matrix interactions (ARHGAP20, ARHGEF15, ARHGEF17, ROCK1, MYLK, DOCK6, DNAH5, DNAH7, DNAI1, RAPGEF2, RFTN1, DSP, DTNB). In the SDRPL discovery cohort, we detected only three recurrently mutated genes [CCND3 (cyclin D3), HIST4H4 (histone cluster 4, H4), and RFTN1 (raftlin, lipid raft linker 1)] in two of the ten SDRPL samples (Online Supplementary Table S1).

Given the characteristic mutational profiles previously identified in other B-cell malignancies and after extensive manual review and data mining of the SDRPL discovery cohort, we selected a panel of 109 target genes the functions of which may be relevant during lymphomagenesis and the types of mutations of which were predicted by the SIFT and PolyPhen2-score to be deleterious or damaging (Online Supplementary Table S2). This panel was used to further explore the validation cohort, including SDRPL, SMZL and HCL samples, by high-throughput sequencing and to assess the frequency of mutations in each of these lymphoma/leukemia entities. All of the mutations detected by whole-exome sequencing were confirmed by targeted gene sequencing for SDRPL cases, sequenced using both methodologies. All of the samples of the validation cohort, except one, displayed at least one of the 109 target gene mutations, with the number of mutations ranging from 1-13 in some samples. Only a few recurrent and discriminative mutations were identified and are described hereafter, with a promising candidate being mutations or losses in BCOR since these affected 10/42 SDRPL patients compared to 1/46 patients with SMZL and 0/8 of those with HCL (Figure 1).

Recurrent BCOR mutations and deletions in splenic diffuse red pulp lymphoma A single mutation in a splicing site of BCOR was initially detected in the discovery cohort and was confirmed to be somatic (Figure 1 and Online Supplementary Table S3). This mutation was identified in the same splicing site position as another mutation reported in the COSMIC database as being COSM521431. Additional BCOR mutations were identified in five SDRPL cases of the validation set. These mutations were distributed along the gene (Figure 2) within exons 4, 5 and 11 (BCOR gene coverage: 100%; mean depth analysis: 345 reads). These mutations were characterized by splicing site (1/6), nonsense (2/6) and frameshift (3/6) alterations. The mutations exhibited a high variant allele frequency, with the exception of two frameshift mutations that exhibited a lower variant allele frequency.

1761

L. Jallades et al.

The tumor cell content was elevated in these two cases, which also harbored some mutations in other genes with a higher variant allele frequency, possibly indicating a subclonal change (Online Supplementary Table S3). Finally, the pattern of these mutations strongly suggests that they would result in loss of function of BCOR. Since such loss of function may arise from the deletion of BCOR on chromosome Xp11.4, we further analyzed our data to look for copy number variations. A single BCOR allele is supposed to be functional in both males (one copy on the single chromosome X) and females (because of X-inactivation). Copy number losses of the BCOR locus were identified by analysis of exon read depths in four SDRPL female patients displaying no BCOR mutation within the single remaining allele (Figure 3A). Three of these losses involved the whole chromosome X, whereas the last deletion was a microdeletion of approximately 669 kb assessed by microarray-comparative genomic hybridization [arr[GRCh37] Xp11.4(39576746_40245183)x1], only targeting an uncharacterized long non-coding RNA (LOC101927476) and the BCOR gene (Figure 3B). Two copy number gains of the BCOR locus were also observed in two SDRPL cases. In one male patient, who also had a BCOR frameshift mutation, the copy number gain explained the observed variant allele frequency of

40% in that individual (VL#208) (Figures 1 and 3A, Online Supplementary Table S3). Some of these abnormalities were confirmed by copy number variation detection from whole-exome sequencing or by karyotype analysis when available (Figure 3). Taken together, these findings demonstrated an alteration in BCOR in 11/42 SDRPL cases, most of them (10/11) being potentially pathogenic (mutations or loss). Furthermore, the percentage of BCOR alterations in cases with spleen histology (2/11 cases, 18%) was similar to that of the cases without spleen histology (8/31, 26%).

A single BCOR frameshift mutation (distinct from the previous ones identified in SDRPL samples) was detected in the 46 SMZL samples (2%) (Figure 1 and Online Supplementary Table S3). Three copy number variations of the BCOR locus were also observed in three SMZL patients, and in all three cases entailed a gain. No BCOR alteration (mutation, gain or deletion) was observed in the eight HCL samples.

MAP2K1 and BRAF mutations While no MAP2K1 (MEK1) mutation was detected in the SMZL samples (Figure 1), three mutated cases were observed in SDRPL samples (3/42; 7%). Two mutations (MAP2K1 p.I103N and p.C121S) were previously reported in HCL-v.7 Of the three patients with MAP2K1 mutations

Figure 2. BCOR mutations in splenic diffuse red pulp lymphoma, lymphoid and hematologic malignancies. The distribution of mutations along the BCOR sequence was drawn using cBioPortal (http://www.cbioportal.org). Mutation diagram circles are colored with respect to the corresponding mutation types. Six distinct mutations were detected in our SDRPL series (top), compared to mutations reported in lymphoid (middle) or hematologic (bottom) malignancies.43,44 BCOR: BCL-6 corepressor, non-ankyrin-repeat region (1202 - 1414); Ank_2: Ankyrin repeats (3 copies) (1467 - 1560); PUFD: BCORL-PCGF1-binding domain (1634 - 1747).

1762

haematologica | 2017; 102(10)

Whole-exome sequencing of SDRPL

A

B

C

D

E

F

G Figure 3. Copy number evaluation at the BCOR Xp11.4 locus in hairy-cell leukemia, splenic diffuse red pulp lymphoma and splenic marginal zone lymphoma. (A) Copy number (CN) variation was detected from exon sequencing of the target genes after normalization using DeCovA.25,24 Each row represents a lymphoma case (HCL: hairy-cell lymphoma; SDRPL: splenic diffuse red pulp lymphoma; SMZL: splenic marginal zone lymphoma). Each column represents an exon of two different genes (BCOR and KDM6A) located at locus Xp11. A progressive gradation of CN variation distinguishes loss (red) and gain (green). Right: mean copy number ratio of the BCOR locus, with some references in brackets corresponding to the next inserts (B-G). (B) Microarray-based comparative genomic hybridization confirmed a monoallelic microdeletion of approximately 669 kb (arr[GRCh37] Xp11.4(39576746_40245183)x1), encompassing the BCOR locus (SDRPL female patient VL_218), whereas the KDM6A gene was unaffected. (C-D) Copy number variation of chromosome X detected by whole-exome sequencing using a circular binary segmentation algorithm and compared to a paired reference genome (log2 ratio). These two female patients with SDRPL acquired complete monosomy X. (E-G) Detailed karyotype results confirmed monosomy X (E), tetrasomy X (F) or trisomy X (G).

haematologica | 2017; 102(10)

1763

L. Jallades et al.

in our series, two also harbored a BCOR mutation (VL#201, VL#208). In contrast to HCL-v cases with MAP2K1 mutations, these two SDRPL cases did not use the IGHV4-34 gene, as indicated in Online Supplementary Table S3 (comments column).

Whereas BRAF p.V600E is the hallmark of HCL, we identified one distinct BRAF mutation (p.G469A) in 1/42 SDRPL samples (2%). This case was characterized by an unmutated immunoglobulin IGHV4-34 sequence and a non-HCL immunological profile (CD25-/CD103+/CD123-) and harbored both BRAF p.G469A and MAP2K1 p.I103N mutations but no BCOR mutation.

A single sample of the 46 SMZL (2%) cases harbored both a BRAF p.V600E mutation and a frameshift mutation of KLF2 (MZ#104). The diagnosis of SMZL was confirmed after further review of the spleen histology in this patient, who also had a typical IGHV1-2*04 rearrangement and trisomy 3q.

KLF2, TNFAIP3 and MYD88 mutations Mutations in KLF2, TNFAIP3 and MYD88, which are known to activate the NF-kB pathway, were observed in 14/46 (30%), 9/46 (20%) and 4/46 (9%) of SMZL patients, respectively (Figure 1 and Online Supplementary Table S3).4,8,11,26 TNFAIP3 and MYD88 mutations were not detected in either SDRPL or HCL samples.

Only one KLF2 mutant (1/42; 2%) was observed in an SDRPL patient. The diagnosis of SDRPL in this case was based on cytological features of a peripheral blood smear with more than 60% villous lymphoma cells among the lymphoid cells, an immunological SDRPL score of 0/5 (dimCD11c+/dimCD22+/CD38-/CD27-/CD76+) and the fact that IGHV3-23 usage is more frequently observed in SDRPL than SMZL.2 One other KLF2 mutation was detected in a HCL sample, which also displayed a BRAF V600E mutation.

Other mutations found in both splenic diffuse red pulp lymphoma and splenic marginal zone lymphoma Finally, different mutations (already reported in other B-cell malignancies) likely implicated in the activation of the Notch pathway (i.e., mutations in NOTCH2, NOTCH1, and SPEN) were identified in 7/42 SDRPL (17%) and 14/46 SMZL (30%) samples (Figure 1 and Online Supplementary Table S3).6,8,9 Other mutations were recurrently observed but were distributed across both SDRPL and SMZL samples, in particular CCND3 [in 9 (21%) of 42 SDRPL and 6 (13%) of 46 SMZL samples], BIRC3, TP53, MYC and CXCR4. No additional recurrent mutations were observed in HIST4H4 and RFTN1, previously identified in two cases of the SDRPL discovery cohort. These findings show that, in addition to the original somatic mutation pattern described above, SDRPL can also share some mutations with its closest Bcell malignancy.

Discussion We explored the mutational landscape of SDRPL with circulating villous lymphocytes in an attempt to identify diagnostic pathognomonic markers associated with this rare unclassifiable (World Health Organization 2016 classification) splenic B-cell lymphoma. One of the limitations of our study was the restricted number of patients having 1764

undergone histological spleen analyses, and we thus partially relied on stringent immunophenotypic and cytological criteria to select SDRPL cases from peripheral blood samples. This methodology is consistent with the current medical management of patients suffering from splenic lymphoma, who are preferentially administered rituximab rather than being submitted to splenectomy. However, irrespective of the origin of the SDRPL tissues analyzed, their genetic landscape was clearly distinct from the SMZL and HCL samples used herein as comparative B-cell malignancy cases. Indeed, we identified recurrent BCOR mutations or losses in 10/42 SDRPL cases (24%), while these remained rare in SMZL (1/46) and absent in HCL (0/8).

Most BCOR mutations (identified in 6 cases), as well as BCOR deletions (in 4 cases including a microdeletion), were speculated to result in inactivation of gene function.

These mutations were not reported in the COSMIC database, except for one case with a BCOR p.S340Vfs*41 mutation similar to the reported pS340* mutation (COSM5945498). There were no particular associations between BCOR alterations and cytogenetic features, IGHV gene usage, or the mutational pattern of SDRPL cases.

Germline BCOR mutations have been detected in patients with inherited oculofaciocardiodental and Lenz microphthalmia syndromes.27 Recently, massively parallel sequencing has identified inactivating somatic BCOR mutations at a very low frequency in patients with various types of solid neoplasia but also in patients with hematologic malignancies, such as acute myeloid leukemia, myelodysplastic syndrome, T-cell prolymphocytic leukemia, and extranodal NK/T-cell lymphoma, nasal type.28-36 Altogether, these data underline the critical role of BCOR in cell differentiation and oncogenesis.

BCOR was first identified as a corepressor whose product interacts specifically with BCL6.37,38 Through epigenetic modifications, the enzymatic activity of the BCOR complex provides a mechanism for silencing BCL6 targets.

Although BCOR and BCL6 play key roles in germinal center formation, and BCL6 alterations are involved in the transformation of germinal center B cells, their functions in other B cells remain elusive.39,40 The predicted inactivating mutations of BCOR or acquired hemizygosity observed in SDRPL may lead to the loss of BCL6 repression, but BCL6 expression is not usually detected by immunohistochemistry in SDRPL. Recent data also indicate that BCL6/BCOR inhibits Notch-activated target genes during embryonic development.40 Given the frequency of mutations of genes involved in the Notch pathway in SDRPL (17%), it would be interesting to investigate whether BCOR loss-of-function might represent an alternative mechanism for activating this pathway in B cells. Overall, how BCOR mutations or deletions might participate in SDRPL oncogenesis remains unknown.

Recently, recurrent CCND3 mutations have been described in six of 25 (24%) of another series of SDRPL cases.16 In our series, we also observed recurrent CCND3 mutations in nine of 42 SDRPL cases (21%), but CCND3 mutations were also detected in six of 46 SMZL cases (13%). Case selection based on diagnostics, obtained either following spleen histology or through morphological and immunological examination of peripheral blood, may potentially account for these distinct findings. Further studies are, therefore, required to identify the spectrum of CCND3 mutations among splenic lymphomas more precisely. A nonsense mutation in CDKN1B, the gene that haematologica | 2017; 102(10)

Whole-exome sequencing of SDRPL

encodes the cyclin-dependent kinase inhibitor and that interacts with cyclin D3, was found in another case of SDRPL.16 The immunohistochemical expression of CCND3 showed a selective expression of cyclin D3 by most neoplastic cells in SDRPL spleen tissues. Considering that cyclin D3 works downstream of BCL6 in germinal center development, it could be hypothesized that dysregulation of the BCL6/BCOR complex on the one hand or CCND3 dysregulation on the other hand may represent two aspects of the same pathophysiology in SDRPL; this remains to be explored more thoroughly.41 Some MAP2K1 mutations were also reported in the present series of SDRPL cases but at a low frequency (7%).

In contrast, a high prevalence (48%) of MAP2K1 mutations was previously reported in HCL-v and IGHV4-34expressing HCL, suggesting a possible degree of overlap between HCL-v and SDRPL.7 Among the 12 lymphoma samples in the present series sharing the IGHV4-34 pattern (8 SDRPL, 3 SMZL and 1 HCL that was BRAF V600Epositive), we identified only one case of IGHV4-34 in a SDRPL patient with a MAP2K1 mutation, also harboring a BRAF G469A mutation (VL#203), but lacking the BCOR mutation. These previously unknown molecular findings may result in tools for the diagnosis of an atypical form of HCL rather than a true case of SDRPL. The two other MAP2K1 mutants were identified in samples with distinct IGHV rearrangements that also harbored a BCOR mutation, which was the most frequent abnormality in our SDRPL series. Further screening of BCOR mutations in the

References 1. Traverse-Glehen A, Baseggio L, Bauchu EC, et al. Splenic red pulp lymphoma with numerous basophilic villous lymphocytes: a distinct clinicopathologic and molecular entity? Blood. 2008;111(4):2253-2260.

2. Baseggio L, Traverse-Glehen A, CalletBauchu E, et al. Relevance of a scoring system including CD11c expression in the identification of splenic diffuse red pulp small B-cell lymphoma (SRPL). Hematol Oncol. 2011;29(1):47-51.

3. Miguet L, Lennon S, Baseggio L, et al. Cellsurface expression of the TLR homolog CD180 in circulating cells from splenic and nodal marginal zone lymphomas.

Leukemia. 2013;27(8):1748-1750.

4. Piris MA, Foucar K, Mollejo M, Campo E, Falini B. Splenic B-cell lymphoma/ leukaemia unclassifiable.

WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues, Fourth Edition. IARC Press Lyon. 2008;191193.

5. Swerdlow SH, Campo E, Pileri SA, et al.

The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood. 2016;127(20):2375-2390.

6. Tiacci E, Trifonov V, Schiavoni G, et al.

BRAF mutations in hairy-cell leukemia. N Engl J Med. 2011;364(24):2305-2315.

7. Waterfall JJ, Arons E, Walker RL, et al. High prevalence of MAP2K1 mutations in variant and IGHV4-34–expressing hairy-cell leukemias. Nat Genet. 2013;46(1):8-10.

8. Rossi D, Trifonov V, Fangazio M, et al. The coding genome of splenic marginal zone lymphoma: activation of NOTCH2 and

haematologica | 2017; 102(10)

9.

10.

11.

12.

13.

14.

15.

16.

HCL-v series may, therefore, refine our understanding of the relationship between HCL-v and SDRPL. Finally, the frequent BCOR mutations and the absence of alterations in genes regulating the NF-kB pathway (triplenegative for KLF2, TNFAIP3 and MYD88 mutations) or the absence of a BRAF mutation appear to delineate a specific genetic pattern of SDRPL, which is distinct from that already identified in SMZL, HCL or HCL-v. Exploration of these mutational patterns should improve the differential diagnosis of splenic B-cell lymphoma/leukemia and other lymphoproliferative disorders, in combination with morphological and immunological criteria currently used for their diagnosis. This will be essential for pursuing the biological and clinical characterization of this range of B-cell lymphoproliferative disorders. Our findings also pave the way for additional functional studies to characterize the oncogenic mechanism by which BCOR may promote lymphomagenesis and how this may potentially lead to the development of novel therapeutics.

Acknowledgments We acknowledge Prof Damien Sanlaville, Dr Nicolas Chatron, Audrey Labalme and Jessica Michel for assistance with the microarray-CGH experiment and the Centre de Ressources Biologiques Sudbiothèque of the Hospices Civils de Lyon. We thank Dr Brigitte Manship for improving the use of English in the manuscript and for her critical reading.

Funding

other pathways regulating marginal zone development.

J Exp Med. 2012;209(9):1537-1551.

Kiel MJ, Velusamy T, Betz BL, et al. Wholegenome sequencing identifies recurrent somatic NOTCH2 mutations in splenic marginal zone lymphoma. J Exp Med. 2012;209(9):1553-1565.

Martínez N, Almaraz C, Vaqué JP, et al.

Whole-exome sequencing in splenic marginal zone lymphoma reveals mutations in genes involved in marginal zone differentiation. Leukemia. 2013;28(6):1334-1340.

Clipson A, Wang M, de Leval L, et al. KLF2 mutation is the most frequent somatic change in splenic marginal zone lymphoma and identifies a subset with distinct genotype. Leukemia. 2014;29(5):1177-1185.

Piva R, Deaglio S, Fama R, et al. The Krüppel-like factor 2 transcription factor gene is recurrently mutated in splenic marginal zone lymphoma. Leukemia. 2014;29 (2):503-507.

Parry M, Rose-Zerilli MJ, Ljungstrom V, et al. Genetics and prognostication in splenic marginal zone lymphoma: revelations from deep sequencing. Clin Cancer Res. 2015;21 (18):4174-4183.

Martinez D, Navarro A, Martinez-Trillos A, et al. NOTCH1, TP53, and MAP2K1 mutations in splenic diffuse red pulp small B-cell lymphoma are associated with progressive disease. Am J Surg Pathol 2015;40(2):192201.

Traverse-Glehen A, Verney A, Gazzo S, et al. Splenic diffuse red pulp lymphoma has a distinct pattern of somatic mutations amongst B-cell malignancies. Leuk Lymphoma. 2016;58(3):666-675.

Curiel-Olmo S, Mondejar R, Almaraz C, et

17.

18.

19.

20.

21.

22.

23.

24.

al. Splenic diffuse red pulp small B-cell lymphoma displays increased expression of cyclin D3 and recurrent CCND3 mutations. Blood. 2017;129(8):1042-1045.

Traverse-Glehen A, Baseggio L, CalletBauchu E, et al. Hairy cell leukaemia-variant and splenic red pulp lymphoma: a single entity? Br J Haematol. 2010;150(1):113116.

Traverse-Glehen A, Baseggio L, Salles G, Felman P, Berger F. Splenic marginal zone Bcell lymphoma. Curr Opin Oncol.

2011;23(5):441-448.

Traverse-Glehen A, Baseggio L, Salles G, Coiffier B, Felman P, Berger F. Splenic diffuse red pulp small-B cell lymphoma: toward the emergence of a new lymphoma entity. Discov Med. 2012;13(71):253-265.

Wotherspoon AC. Extranodal and splenic small B-cell lymphoma. Mod Pathol.

2013;26:S29-S41.

Kanellis G, Mollejo M, Montes-Moreno S, et al. Splenic diffuse red pulp small B-cell lymphoma: revision of a series of cases reveals characteristic clinico-pathological features.

Haematologica. 2010;95(7):1122-1129.

Baseggio L, Traverse-Glehen A, Petinataud F, et al. CD5 expression identifies a subset of splenic marginal zone lymphomas with higher lymphocytosis: a clinico-pathological, cytogenetic and molecular study of 24 cases. Haematologica. 2010;95(4):604-612.

Ponzoni M, Kanellis G, Pouliou E, et al.

Bone marrow histopathology in the diagnostic evaluation of splenic marginal-zone and splenic diffuse red pulp small B-cell lymphoma. Am J Surg Pathol. 2012;36 (11):1609-1618.

Pasmant E, Parfait B, Luscan A, et al.

Neurofibromatosis type 1 molecular diag-

1765

L. Jallades et al.

25.

26.

27.

28.

29.

30.

1766

nosis: what can NGS do for you when you have a large gene with loss of function mutations?

Eur J Hum Genet.

2014;23(5):596-601.

Dimassi S, Simonet T, Labalme A, et al.

Comparison of two next-generation sequencing kits for diagnosis of epileptic disorders with a user-friendly tool for displaying gene coverage, DeCovA. Appl Transl Genom. 2015;7:19-25.

Rossi D, Deaglio S, Dominguez-Sola D, et al. Alteration of BIRC3 and multiple other NF-kappaB pathway genes in splenic marginal zone lymphoma.

Blood.

2011;118(18):4930-4934.

Ng D, Thakker N, Corcoran CM, et al.

Oculofaciocardiodental and Lenz microphthalmia syndromes result from distinct classes of mutations in BCOR. Nat Genet.

2004;36(4):411-416.

Pugh TJ, Weeraratne SD, Archer TC, et al.

Medulloblastoma exome sequencing uncovers subtype-specific somatic mutations. Nature. 2012;488(7409):106-110.

Yamamoto Y, Abe A, Emi N. Clarifying the impact of polycomb complex component disruption in human cancers. Mol Cancer Res. 2014;12(4):479â&#x20AC;&#x201C;484.

Moreira AL, Won HH, McMillan R, et al.

Massively parallel sequencing identifies recurrent mutations in TP53 in thymic car-

31.

32.

33.

34.

35.

36.

37.

cinoma associated with poor prognosis. J Thorac Oncol. 2015;10(2):373-380.

Abe H, Kaneda A, Fukayama M. EpsteinBarr virus-associated gastric carcinoma: use of host cell machineries and somatic gene mutations. Pathobiology. 2015;82(5):212223.

Grossmann V, Tiacci E, Holmes AB, et al.

Whole-exome sequencing identifies somatic mutations of BCOR in acute myeloid leukemia with normal karyotype. Blood.

2011;118(23):6153-6163.

Damm F, Chesnais V, Nagata Y, et al.

BCOR and BCORL1 mutations in myelodysplastic syndromes and related disorders. Blood. 2013;122(18):3169-3177.

Yoshizato T, Dumitriu B, Hosokawa K, et al. Somatic mutations and clonal hematopoiesis in aplastic anemia. N Eng J Med. 2015;373(1):35-47.

Stengel A, Kern W, Zenger M, et al. Genetic characterization of T-PLL reveals two major biologic subgroups and JAK3 mutations as prognostic marker. Genes Chromosomes Cancer. 2016;55(1):82-94.

Lee S, Park HY, Kang SY, et al. Genetic alterations of JAK/STAT cascade and histone modification in extranodal NK/T-cell lymphoma nasal type. Oncotarget. 2015;6(19): 17764-17776.

Huynh KD, Fischle W, Verdin E, Bardwell

38.

39.

40.

41.

42.

43.

VJ. BCoR, a novel corepressor involved in BCL-6 repression. Gen Dev. 2000;14(14): 1810-1823.

Hatzi K, Melnick A. Breaking bad in the germinal center: how deregulation of BCL6 contributes to lymphomagenesis. Trends Mol Med. 2014;20(6):343-352.

Polo JM, Dell'Oso T, Ranuncolo SM, et al.

Specific peptide interference reveals BCL6 transcriptional and oncogenic mechanisms in B-cell lymphoma cells. Nat Med. 2004;10(12):1329-1335.

Sakano D, Kato A, Parikh N, et al. BCL6 canalizes notch-dependent transcription, excluding mastermind-like1 from selected target genes during left-right patterning.

Dev Cell. 2010;18(3):450-462.

Cato MH, Chintalapati SK, Yau IW, Omori SA, Rickert RC. Cyclin D3 is selectively required for proliferative expansion of germinal center B cells. Mol Cell Biol.

2010;31(1):127-137.

Gao J, Aksoy BA, Dogrusoz U, et al.

Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6(269):pl1-pl1.

Cerami E, Gao J, Dogrusoz U, et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2012;2(5): 401-404.

haematologica | 2017; 102(10)

ARTICLE

Plasma Cell Disorders

Impact of prior therapy on the efficacy and safety of oral ixazomib-lenalidomide-dexamethasone vs. placebo-lenalidomide-dexamethasone in patients with relapsed/refractory multiple myeloma in TOURMALINE-MM1

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

María-Victoria Mateos,1 Tamas Masszi,2 Norbert Grzasko,3 Markus Hansson,4 Irwindeep Sandhu,5 Ludek Pour,6 Luísa Viterbo,7 Sharon R. Jackson,8 Anne-Marie Stoppa,9 Peter Gimsing,10 Mehdi Hamadani,11 Gabriela Borsaru,12 Deborah Berg,13 Jianchang Lin,13 Alessandra Di Bacco,13 Helgi van de Velde,13 Paul G. Richardson14 and Philippe Moreau15

Hospital Universitario de Salamanca, Instituto Biosanitario de Salamanca (IBSAL), Spain; 2St. István, St. László Hospital, 3rd Department of Internal Medicine, Semmelweis University, Budapest, Hungary; 3Medical University of Lublin and St John’s Cancer Center, Lublin, Poland; 4Skåne University Hospital, Lund University, Sweden; 5University of Alberta Edmonton, Canada; 6University Hospital Brno, Czech Republic; 7Instituto Português de Oncologia do Porto Francisco Gentil, Entidade Pública Empresarial (IPOPFG, EPE), Portugal; 8Middlemore Hospital, Auckland, New Zealand; 9Institut PaoliCalmettes, Marseille, France; 10University Hospital Rigshospitalet, Copenhagen, Denmark; 11Medical College of Wisconsin, Milwaukee, WI, USA; 12Spitalul Clinic Coltea, Bucharest, Romania; 13Millennium Pharmaceuticals Inc., Cambridge, MA, a wholly owned subsidiary of Takeda Pharmaceutical Company Limited, Cambridge, MA, USA; 14 Dana-Farber Cancer Institute, Boston, MA, USA and 15University Hospital Hôtel Dieu, Nantes, France 1

Haematologica 2017 Volume 102(10):1767-1775

ABSTRACT

P

rior treatment exposure in patients with relapsed/refractory multiple myeloma may affect outcomes with subsequent therapies. We analyzed efficacy and safety according to prior treatment in the phase 3 TOURMALINE-MM1 study of ixazomib-lenalidomide-dexamethasone (ixazomib-Rd) versus placebo-Rd. Patients with relapsed/refractory multiple myeloma received ixazomib-Rd or placebo-Rd. Efficacy and safety were evaluated in subgroups defined according to type (proteasome inhibitor [PI] and immunomodulatory drug) and number (1 vs.

2 or 3) of prior therapies received. Of 722 patients, 503 (70%) had received a prior PI, and 397 (55%) prior lenalidomide/thalidomide; 425 patients had received 1 prior therapy, and 297 received 2 or 3 prior therapies. At a median follow up of ~15 months, PFS was prolonged with ixazomib-Rd vs. placebo-Rd regardless of type of prior therapy received; HR 0.739 and 0.749 in PI-exposed and –naïve patients, HR 0.744 and 0.700 in immunomodulatory-drug-exposed and -naïve patients, respectively. PFS benefit with ixazomib-Rd vs. placebo-Rd appeared greater in patients with 2 or 3 prior therapies (HR 0.58) and in those with 1 prior therapy without prior transplant (HR 0.60) versus those with 1 prior therapy and transplant (HR 1.23). Across all subgroups, toxicity was consistent with that seen in the intent-to-treat population. In patients with relapsed/refractory multiple myeloma, ixazomib-Rd was associated with a consistent clinical benefit vs. placebo-Rd regardless of prior treatment with bortezomib or immunomodulatory drugs. Patients with 2 or 3 prior therapies, or 1 prior therapy without transplant seemed to have greater benefit than patients with 1 prior therapy and transplant.

TOURMALINE-MM1 registered at clinicaltrials.gov identifier: 01564537.

Introduction Novel agents such as proteasome inhibitors (PIs) and immunomodulatory drugs have revolutionized multiple myeloma (MM) treatment, with significant improvements in overall survival (OS) evident over the past 15 years.1-5 Despite the use of haematologica | 2017; 102(10)

Correspondence: mvmateos@usal.es

Received: March 31, 2017.

Accepted: July 19, 2017.

Pre-published: July 27, 2017. doi:10.3324/haematol.2017.170118 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1767 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1767

M.-V. Mateos et al.

Table 1. Number and type of prior therapies received by patients in TOURMALINE-MM1.

these novel agents, MM follows a relapsing course, with many patients receiving multiple lines of therapy and ultimately becoming refractory to some agents,6 possibly due to the development and selection of increasingly treatment-resistant clones.7 Long-term outcomes, including progression-free survival (PFS) and OS, also become progressively shorter with increasing number of prior therapies,6,8-11 as rates of medical comorbidities and complications increase.12 Prior therapies are therefore often considered when selecting a therapy at relapse, with prior therapies shown to affect the outcomes of subsequent lines of treatment. For example, outcomes for thalidomideexposed patients have been shown to be worse than for thalidomide-naïve patients following treatment with bortezomib12 and with lenalidomide-dexamethasone.13 Until 2012, bortezomib was the only PI available so subsequent treatment with other drugs of the same class was not possible outside of a clinical trial. However, retreatment with bortezomib has been shown to be effective14-17 and, following the introduction of carfilzomib, the feasibility of retreatment with a different agent of the same class and with a similar mechanism of action has been demonstrated.17 Similarly, for the immunomodulatory drugs, lenalidomide plus dexamethasone improved responses, time to progression (TTP), and PFS compared with dexamethasone alone in patients with or without prior thalidomide exposure.13 The phase 3, randomized, placebo-controlled, doubleblind TOURMALINE-MM1 study in 722 patients with relapsed/refractory MM (RRMM) demonstrated a significant 35% improvement in PFS with the all-oral combination of ixazomib plus lenalidomide-dexamethasone (Rd) compared with placebo-Rd (median PFS 20.6 vs. 14.7 months; hazard ratio 0.74; P=0.01).18 On the basis of these data, ixazomib, in combination with lenalidomide and dexamethasone (ixazomib-Rd), was approved in 2015 by the US Food and Drug Administration, and in 2016 by the European Medicines Agency, for the treatment of patients with MM who have received at least one prior line of therapy. Given the widespread use of PIs and immunomodulatory drugs as first-line therapy, it is important to determine their impact on the overall and relative efficacy of new agents for the treatment of RRMM. The TOURMALINE-MM1 study included patients with prior exposure to PIs and the immunomodulatory drugs thalidomide and lenalidomide, and patients with and without prior transplant. Here we present a subgroup analysis of efficacy and safety data for ixazomib-Rd compared with placebo-Rd according to the number and type of prior therapies received.

performed in accordance with the International Conference on Harmonisation Good Clinical Practice guidelines and appropriate regulatory requirements, and with approval of Institutional Review Boards at individual enrolling institutions. All patients provided written informed consent. A total of 722 patients were randomized 1:1 to receive oral ixazomib 4 mg (ixazomib-Rd arm, N=360) or placebo (placebo-Rd arm, N=362) on days 1, 8, and 15 of 28-day cycles, with oral lenalidomide 25 mg on days 1–21 and oral dexamethasone 40 mg on days 1, 8, 15 and 22, until disease progression or unacceptable toxicity. Stratification factors were number of prior therapies per investigator assessment (1 vs. 2 or 3), International Staging System disease stage (I or II vs. III), and prior PI exposure (yes vs. no); patients were not stratified by prior thal/R exposure or thalidomide-refractoriness. A prior line of therapy was defined as 1 or more cycles of a planned treatment program, as determined by the investigator. Overall patient baseline demographics and disease characteristics were well balanced between ixazomib-Rd and placebo-Rd arms.18 Responses were assessed per International Myeloma Working Group 2011 criteria19 every cycle until disease progression, using a central laboratory. Adverse events (AEs) were assessed per National Cancer Institute’s Common Terminology Criteria for Adverse Events version 4.03 during treatment and until 30 days after the last dose of study medication was administered.

Methods

Analyses by prior treatment exposure

Study design and participants Adult patients with measurable relapsed, refractory, or relapsed and refractory MM who had received 1-3 prior lines of therapy were eligible. Full eligibility criteria have been reported previously.18 Patients who had received prior PI- and thalidomide/lenalidomide (thal/R)-based regimens were eligible, as were primary refractory patients and patients refractory to thalidomide; patients who were refractory to prior PI- or lenalidomide-based therapy were not eligible. Study endpoints have been reported previously.18 The primary endpoint was PFS as assessed by a blinded independent review committee (IRC). The study was 1768

Number of prior therapies, n (%) 1 2–3 Prior therapy type, n (%) PI naïve Bortezomib naïve PI exposed Bortezomib exposed Carfilzomib exposed Immunomodulatory drug naïve Thalidomide naïve Lenalidomide naïve Immunomodulatory drug exposed Thalidomide exposed Lenalidomide exposed Thalidomide refractory

Ixazomib-Rd (N=360)

Placebo-Rd (N=362)

212 (59) 148 (41)

213 (59) 149 (41)

110 (31) 112 (31) 250 (69) 248 (69) 1 (<1) 167 (46) 203 (56) 316 (88) 193 (54) 157 (44) 44 (12) 40 (11)

109 (30) 112 (31) 253 (70) 250 (69) 4 (1) 158 (44) 192 (53) 318 (88) 204 (56) 170 (47) 44 (12) 49 (14)

PI: proteasome inhibitor; Rd: lenalidomide-dexamethasone.

Subgroup analyses were performed for efficacy and safety outcomes relative to type of prior regimen. Patient subgroups were defined according to prior exposure to the PIs bortezomib and carfilzomib, and the immunomodulatory drugs lenalidomide and thalidomide. Outcomes were also assessed according to number of prior lines of therapy (1 vs. 2/3, per study stratification) and, within those subgroups, according to components of prior therapies, including transplant.

Statistical analysis At a pre-planned analysis (median follow up of ~15 months), the study met the primary endpoint of a significant PFS benefit haematologica | 2017; 102(10)

Ixazomib-Rd in RRMM patients: impact of prior therapies

with ixazomib-Rd vs. placebo-Rd. Consistent with the statistical methodology, this was therefore the final statistical analysis for PFS. Per protocol, the study continued in a double-blind, placebocontrolled manner to gain more mature OS data; a second preplanned analysis (median follow up of ~23 months) was conducted for safety and survival. Time-to-event distributions were estimated using Kaplan-Meier methodology, with stratified log-rank tests and Cox models (alpha=0.05, two-sided) used for comparisons of time-to-event endpoints. A stratified Cochran-MantelHaenszel Ď&#x2021;2 test was used to assess inter-arm differences in response rates. The subgroup analyses were not powered for formal statistical testing.

253 [70%] in the placebo arm). The majority of PIexposed patients had received bortezomib (1 patient in the ixazomib arm and 4 patients in the placebo arm had received prior carfilzomib). Over half (55%) had received prior thalidomide or lenalidomide (193 [54%] in the ixazomib arm and 204 [56%] in the placebo arm) (Table 1).

Of these, in the ixazomib and placebo arms, respectively, 157 (44%) and 170 (47%) patients had received prior thalidomide, and 44 (12%) and 44 (12%) patients had received prior lenalidomide; there was no prior pomalidomide therapy. A total of 425 patients had received 1 prior therapy (212 in the ixazomib arm and 213 in the placebo arm) and 297 had received 2 or 3 prior therapies (148 in the ixazomib arm and 149 in the placebo arm).

Results Efficacy according to type of prior therapies received Patients Of the 722 patients in the ITT population, 70% had received a prior PI (250 [69%] in the ixazomib arm, and

At a median follow up of ~15 months (14.8 months in the ixazomib-Rd group and 14.6 months in the placeboRd group), there was a clinical benefit in terms of pro-

A

B

Figure 1. Forest plot of progression-free survival (PFS) according to number and type of prior therapies (A), and forest plot of PFS according to type of prior therapy in patients who have received 1 versus 2 or 3 prior therapies (B). CI: confidence interval; HR: hazard ratio; PI: proteasome inhibitor; Rd: lenalidomide-dexamethasone.

haematologica | 2017; 102(10)

1769

M.-V. Mateos et al.

longed PFS with ixazomib-Rd vs. placebo-Rd regardless of prior therapy received (Figure 1A, Figure 2); median PFS was 18.4 vs. 13.6 months (HR 0.74) in PI-exposed patients, not reached vs. 15.7 months (HR 0.749) in PI-naïve patients, not reached vs. 17.5 months (HR 0.744) in thal/Rexposed, and 20.6 vs. 13.6 months (HR 0.700) in thal/Rnaïve patients. PFS was also prolonged with ixazomib-Rd versus placebo-Rd in patients refractory to thalidomide (HR 0.726; median PFS 16.6 vs. 13.0 months).

TTP was also longer with ixazomib-Rd than placebo-Rd regardless of type of prior therapy received. When analyzed by prior PI exposure, median TTP with ixazomibRd vs. placebo-Rd was 18.5 vs. 13.9 months (HR 0.702, 95% CI 0.526, 0.936) in PI-exposed patients, and not

estimable vs. 17.5 months (HR 0.741, 95% CI 0.456, 1.203) in PI-naïve patients (Figure 3). For immunomodulatory drug exposure, median TTP was not estimable vs. 18.3 months (HR 0.727, 95% CI 0.515, 1.026) in exposed patients, and 20.6 vs. 13.6 months (HR 0.651, 95% CI 0.449, 0.945) in naïve patients.

Overall response rates (ORR) with ixazomib-Rd and placebo-Rd appeared generally similar across most subgroups (PI-naïve: 81% vs. 74%; PI-exposed: 77% vs. 70%; thal/R-naïve: 80% vs. 77%; R-naïve: 78% vs. 73%; Table 2) but were slightly lower in thalidomide-refractory patients (70% vs. 57%). Complete response plus very good partial response (CR+VGPR) rates with ixazomib-Rd vs. placebo-Rd by patient subgroup are shown in Table 2;

A

B

C

D

E

F

Figure 2. Kaplan-Meier analysis of progression-free survival (PFS) with ixazomib-Rd vs. placebo-Rd according to prior therapy. (A) PI-exposed patients; B) PI-naïve patients; C) immunomodulatory drug-exposed patients; D) immunomodulatory drug-naïve patients; E) patients with 1 prior therapy; F) patients with 2/3 prior therapies. CI: confidence interval; Rd: lenalidomide-dexamethasone.

1770

haematologica | 2017; 102(10)

Ixazomib-Rd in RRMM patients: impact of prior therapies

again, there was a consistent benefit with ixazomib-Rd vs. placebo-Rd.

Efficacy according to number of prior therapies The benefit of ixazomib-Rd vs. placebo-Rd was seen when assessed by number of prior therapies, with prolonged PFS and TTP and improved response rates seen with ixazomib-Rd versus placebo-Rd in patients with 1 prior therapy and in those with 2 or 3 prior therapies (Figures 1A and 2, Table 2). However, the benefit seemed less pronounced in patients with 1 prior therapy versus those with 2 or 3 prior therapies: the hazard ratio for PFS was 0.88 (95% CI: 0.65–1.20) vs. 0.58 (95% CI: 0.40–0.84) in patients with 1 vs. 2 or 3 prior therapies, respectively, and the hazard ratio for TTP was 0.842 (95% CI 0.614, 1.156) vs. 0.550 (95% CI 0.370, 0.819) in patients with 1 vs.

2 or 3 prior therapies, respectively (Figures 1A and 3).

To investigate this further, PFS was analyzed in patients with 1 prior therapy and patients with 2 or 3 prior therapies according to type of prior therapy received and other clinical characteristics (Figure 1B). In patients with 2 or 3

prior therapies, the PFS benefit was consistent across all subgroups, regardless of type of prior therapy received or cytogenetic risk status. In patients with 1 prior therapy, the magnitude of PFS benefit was consistent regardless of prior exposure to PIs or immunomodulatory drugs (HR ~0.7 across all subgroups; Figure 1A), but was greater in patients with high-risk cytogenetics (HR for PFS 0.64 vs. 0.81 for those with standard-risk cytogenetics) and those who did not have a prior transplant (HR for PFS 0.60 vs. 1.23 for those who did have a prior transplant) (Figure 1B).

Safety Of the 722 patients randomized, 720 received at least one dose of study drug and were included in the safety population (ixazomib-Rd N=361, placebo-Rd N=359). Per the primary study report,20 safety data are reported from a pre-specified analysis at a median follow up of approximately 23 months. Rates of all-grade AEs, grade ≥3 AEs, and serious AEs for the overall population and by patient subgroup are shown in Table 3. Rates of all-grade AEs,

Table 2. Response with ixazomib-Rd vs. placebo-Rd by type and number of prior therapies.

Ixazomib-Rd (N=360)

Placebo-Rd (N=362)

≥VGPR Ixazomib-Rd Placebo-Rd (N=360) (N=362)

282 (78) 193/250 (77) 89/110 (81) 149/193 (77) 133/167 (80) 122/157 (78) 160/203 (79) 34/44 (77) 248/316 (78) 28/40 (70) 254/320 (79) 163/212 (77) 119/148 (80)

259 (72) 178/253 (70) 81/109 (74) 137/204 (67) 122/158 (77) 114/170 (67) 145/192 (76) 26/44 (59) 233/318 (73) 28/49 (57) 231/313 (74) 159/213 (75) 100/149 (67)

173 (48) 114/250 (46) 59/110 (54) 87/193 (45) 86/167 (51) 73/157 (46) 100/203 (49) 20/44 (45) 153/316 (48) 12/40 (30) 161/320 (50) 95/212 (45) 78/148 (53)

ORR

Overall population Prior PI Prior immunomodulatory drug Prior thalidomide Prior lenalidomide Thalidomide-refractory Number of prior therapies

Exposed Naive Exposed Naïve Exposed Naïve Exposed Naïve Yes No 1 2 or 3

141 (39) 101/253 (40) 40/109 (37) 71/204 (35) 70/158 (44) 58/170 (34) 83/192 (43) 16/44 (36) 125/318 (39) 13/49 (27) 128/313 (41) 93/213 (44) 48/149 (32)

Ixazomib-Rd

≥CR

42 (12) 22/250 (9) 20/110 (18) 22/193 (11) 20/167 (12) 20/157 (13) 22/203 (11) 4/44 (9) 38/316 (12) 2/40 (5) 40/320 (13) 19/212 (9) 23/148 (16)

Placebo-Rd 24 (7) 15/253 (6) 9/109 (8) 13/204 (6) 11/158 (7) 10/170 (6) 14/192 (7) 3/44 (7) 21/318 (7) 2/49 (4) 22/313 (7) 17/213 (8) 7/149 (5)

CR: complete response; ORR: overall response rate; PI: proteasome inhibitor; Rd: lenalidomide-dexamethasone; VGPR: very good partial response.

Table 3. Overall summary of adverse events (AEs) according to number and type of prior therapies.

n / N (%)

All-grade AEs Ixazomib-Rd Placebo-Rd

Grade ≥3 AEs Ixazomib-Rd Placebo-Rd

Serious AEs Ixazomib-Rd Placebo-Rd

Overall population PI-naive PI-exposed Immuno-modulatory drug-naive Immuno-modulatory drug-exposed 1 prior therapy 2-3 prior therapies

355/361 (98) 108/109 (99) 247/252 (98) 161/166 (97)

357/359 (99) 109/109 (100) 248/250 (99) 156/158 (99)

267/361 (74) 86/109 (79) 181/252 (72) 125/166 (75)

247/359 (69) 72/109 (66) 175/250 (70) 112/158 (71)

168/361 (47) 54/109 (50) 114/252 (45) 80/166 (48)

177/359 (49) 47/109 (43) 130/250 (52) 80/158 (51)

15/361 (4) 6/109 (6) 9/252 (4) 8/166 (5)

23/359 (6) 8/109 (7) 15/250 (6) 12/158 (8)

194/195 (99)

201/201 (100)

142/195 (73)

135/201 (67)

88/195 (45)

97/201 (48)

7/195 (4)

11/201 (5)

208/212 (98) 147/149 (99)

209/211 (99) 148/148 (100)

153/212 (72) 114/149 (77)

134/211 (64) 113/149 (76)

99/212 (47) 69/149 (46)

94/211 (45) 83/149 (56)

10/212 (5) 5/149 (3)

10/211 (5) 13/149 (9)

On-study deaths Ixazomib-Rd Placebo-Rd

PI: proteasome inhibitor; Rd: lenalidomide-dexamethasone.

haematologica | 2017; 102(10)

1771

M.-V. Mateos et al. grade ≥3 AEs, and serious AEs by patient subgroup were largely consistent with those seen for the overall population, the only exception being slightly higher rates of grade ≥3 AEs and serious AEs with placebo-Rd in patients with 2-3 prior therapies (76% and 56%, vs. 69% and 49% in the overall population, respectively).

Rates of AEs of clinical interest, including neutropenia, thrombocytopenia, peripheral neuropathy, rash, diarrhea, nausea, and vomiting, are shown in Table 4; common grade ≥3 AEs are shown in Online Supplementary Table S1.

Rates of AEs were largely consistent across patient subgroups (Online Supplementary Table S1).

Across patient subgroups, the incidence of peripheral neuropathy, a known side effect of the first-in-class PI bortezomib, was largely consistent with the overall population (27% vs. 22% for ixazomib-Rd vs. placebo-Rd), including in PI-naïve (29% vs. 23%) and PI-exposed (26% vs. 21%) patients. Rates of grade ≥3 peripheral neuropathy with ixazomib-Rd vs. placebo-Rd were also similar across patient subgroups: 3% vs. <1% of PI-naïve, 2% vs. 2% of PI-exposed, 1% vs. 3% of immunomodulatory drug-naïve, 4% vs. <1% of immunomodulatory drugexposed patients, 2% vs. 2% of patients with 1 prior therapy, and 3% vs. 1% of patients with 2-3 prior therapies (Online Supplementary Table S1). As with the overall population, the incidence of cardiac, thromboembolism, and renal failure toxicities were consistently low and similar in both treatment groups regardless of prior therapy (Table 4).

Discussion This subgroup analysis demonstrated that, as with the overall TOURMALINE-MM1 study population,20 the addition of ixazomib to Rd was associated with prolonged PFS versus placebo-Rd across the patient subgroups analyzed, regardless of prior bortezomib or immunomodulatory drug exposure or number of prior therapies received.

This PFS benefit was accompanied by improved response rates and a prolonged TTP versus placebo-Rd across all prior therapy subgroups. Reflecting the findings in the overall study population, the addition of ixazomib to lenalidomide-dexamethasone was consistently associated

with limited additional toxicity regardless of prior therapy subgroup.18 These efficacy and safety data are particularly important given both the widespread use of PIs and immunomodulatory drugs as front-line therapy in MM and the relapsing nature of the disease.20 Retreatment with bortezomib has previously been shown to be feasible,14-17 as has the benefit of carfilzomibdexamethasone in patients with prior bortezomib exposure.21 However, the median PFS with carfilzomib-dexamethasone in bortezomib-exposed patients was less than that in bortezomib-naïve patients (15.6 months vs. not estimable), suggesting some effect of prior PI exposure on the efficacy of carfilzomib-dexamethasone.21 In the present study, ixazomib-Rd was associated with prolonged PFS and TTP and improved response rates vs. placebo-Rd in bortezomib-naïve and –exposed patients. Median PFS with ixazomib-Rd appeared longer in bortezomib-naïve vs. bortezomib-exposed patients (not estimable vs.18.5 months), but the associated hazard ratios vs. placebo-Rd were similar (0.746 vs. 0.747), suggesting a similar PFS benefit with ixazomib-Rd in bortezomib-naïve and -exposed patients. Although no conclusions can be drawn regarding patients refractory to bortezomib, these similar hazard ratios also suggest that the adverse impact of prior bortezomib exposure on PFS and OS seen in a previous study of Rd22 may not be the case when ixazomib is added to the Rd regimen.

The clinical benefit of ixazomib-Rd versus placebo-Rd was also consistent regardless of prior exposure to immunomodulatory drugs. Ixazomib-Rd was associated with prolonged PFS vs. placebo-Rd in both immunomodulatory drug-naïve and –exposed patients (with HR of approximately 0.7 for both subgroups). Although only 12% of patients in each arm had received prior lenalidomide, ixazomib-Rd was associated with a clinical benefit versus placebo-Rd in patients with prior lenalidomide exposure (median PFS, not estimable vs. 17.5 months; HR 0.582), highlighting the benefit of adding a drug with a different mechanism of action for these patients. Of note, the clinical benefit of ixazomib-Rd was also seen in thalidomide-refractory patients; as lenalidomide-refractory patients were not eligible for the study, no conclusions can be drawn regarding these patients.

Patients with MM who have received multiple prior

Figure 3. Forest plot of time to progression (TTP) according to number and type of prior therapies. CI: confidence interval; PI: proteasome inhibitor; Rd: lenalidomide-dexamethasone.

1772

haematologica | 2017; 102(10)

Ixazomib-Rd in RRMM patients: impact of prior therapies

therapies are a particularly difficult-to-treat population, with patient outcomes becoming progressively worse with increasing prior therapies.6,9,20 This may be in part due to clonal evolution, with multiple rounds of treatment with different agents exerting selection pressure on mutant plasma cells, leading to both the development of increasingly treatment-resistant clones and the re-emer-

gence of original clones not completely suppressed.11 It is therefore important that effective and tolerable treatments are available for this heterogeneous patient population.

Importantly, ixazomib-Rd was associated with a particular clinical benefit vs. placebo-Rd in patients with multiple prior therapies (HR 0.580; median PFS not estimable with ixazomib-Rd vs. 12.9 months with placebo-Rd), and this

Table 4. Adverse events (AEs) of clinical interest according to number and type of prior therapies.

Overall population Common AEs of clinical interest Neutropenia* Ixazomib-Rd 118/361 (33) Placebo-Rd 111/359 (31) Thrombocytopenia† Ixazomib-Rd 112/361 (31) Placebo-Rd 57/359 (16) ‡ Peripheral neuropathy Ixazomib-Rd 97/361 (27) Placebo-Rd 78/359 (22) Diarrhea Ixazomib-Rd 164/361 (45) Placebo-Rd 139/359 (39) Rash§ Ixazomib-Rd 72/361 (20) Placebo-Rd 45/359 (13) Nausea Ixazomib-Rd 104/361 (29) Placebo-Rd 79/359 (22) Vomiting Ixazomib-Rd 84/361 (23) Placebo-Rd 42/359 (12) Other AEs of clinical interest Acute renal failure# Ixazomib-Rd 31/361 (9) Placebo-Rd 41/359 (11) Venous embolic and thrombotic events# Ixazomib-Rd 29/361 (8) Placebo-Rd 38/359 (11) Heart failure# Ixazomib-Rd 16/361 (4) Placebo-Rd 14/359 (4) Myocardial infarction# Ixazomib-Rd 5/361 (1) Placebo-Rd 8/359 (2)

PI-naïve

PI-exposed

ImmunoImmunomodulatory drug- modulatory drugnaïve exposed

1 prior therapy

2-3 prior therapies

39/109 (36) 31/109 (28)

79/252 (31) 80/250 (32)

52/166 (31) 46/158 (29)

66/195 (34) 65/201 (32)

62/212 (29) 53/211 (25)

41/149 (28) 39/148 (26)

37/109 (34) 13/109 (12)

75/252 (30) 44/250 (18)

49/166 (30) 31/158 (20)

63/195 (32) 26/201 (13)

49/212 (23) 21/211 (10)

37/149 (25) 20/148 (14)

32/109 (29) 25/109 (23)

65/252 (26) 53/250 (21)

40/166 (24) 30/158 (19)

57/195 (29) 48/201 (24)

61/212 (29) 44/211 (21)

36/149 (24) 34/148 (23)

53/109 (49) 47/109 (43)

111/252 (44) 92/250 (37)

74/166 (45) 56/158 (35)

90/195 (46) 83/201 (41)

96/212 (45) 92/211 (44)

68/149 (46) 47/148 (32)

25/109 (23) 13/109 (12)

47/252 (19) 32/250 (13)

33/166 (20) 23/158 (15)

39/195 (20) 22/201 (11)

34/212 (16) 28/211 (13)

38/149 (26) 17/148 (11)

29/109 (27) 24/109 (22)

75/252 (30) 55/250 (22)

46/166 (28) 29/158 (18)

58/195 (30) 50/201 (25)

53/212 (25) 45/211 (21)

51/149 (34) 34/148 (23)

20/109 (18) 12/109 (11)

64/252 (25) 30/250 (12)

42/166 (25) 20/158 (13)

42/195 (22) 22/201 (11)

47/212 (22) 21/211 (10)

37/149 (25) 21/148 (14)

9/109 (8) 10/109 (9)

22/252 (9) 31/250 (12)

17/166 (10) 18/158 (11)

14/195 (7) 23/201 (11)

5/212 (2) 8/211 (4)

4/149 (3) 6/148 (4)

10/109 (9) 11/109 (10)

19/252 (8) 27/250 (11)

16/166 (10) 19/158 (12)

13/195 (7) 19/201 (9)

17/212 (8) 22/211 (10)

12/149 (8) 16/148 (11)

8/109 (7) 3/109 (3)

8/252 (3) 11/250(4)

7/166 (4) 6/158 (4)

9/195 (5) 8/201 (4)

12/212 (6) 7/211 (3)

4/149 (3) 7/148 (5)

3/109 (3) 3/109 (3)

2/252 (<1) 5/250 (2)

1/166 (<1) 4/158 (3)

4/195 (2) 4/201 (2)

5/212 (2) 3/211 (1)

0 5/211 (3)

*Data based upon standardized MedDRA query, including neutropenia and neutrophil count decreased. †Data based upon standardized MedDRA query, including thrombocytopenia and platelet count decreased. ‡High-level term including peripheral neuropathy, peripheral sensory neuropathy, peripheral sensorimotor neuropathy, and peripheral motor neuropathy. §High-level term including acute febrile neutrophilic dermatosis, acneiform dermatitis, allergic dermatitis, drug eruption, erythema multiforme, exfoliative rash, interstitial granulomatous dermatitis, pruritus, generalised pruritus, purpura, rash, erythematous rash, follicular rash, generalised rash, macular rash, maculo-papular rash, maculovesicular rash, morbilliform rash, papular rash, pruritic rash, pustular rash, vesicular rash, red man syndrome, Stevens-Johnson syndrome, Toxic epidermal necrolysis, urticaria, urticarial papular, and vasculitic rash. #Data based upon standardized MedDRA query, incorporating pooled preferred terms, or multiple preferred terms. PI, proteasome inhibitor; Rd, lenalidomide-dexamethasone.

haematologica | 2017; 102(10)

1773

M.-V. Mateos et al.

was seen regardless of the type of prior therapies received.

While there was a clinical benefit with ixazomib-Rd vs. placebo-Rd in patients with 1 prior therapy, the magnitude of benefit appeared reduced when compared to that in patients with multiple prior therapies (HR vs. placeboRd 0.88). These results appear to differ from those seen with carfilzomib-Rd versus Rd alone, in which there was a consistent benefit in patients with 1 prior therapy and in those with ≥2 prior therapies (HR 0.694 and 0.688).23 The further analysis of patients with 1 prior therapy in TOURMALINE-MM1 suggests this difference may, in part, be driven by effects in the subgroup of patients with prior transplant (HR 1.232, vs. 0.604 in those with no prior transplant). Across other subgroups of patients with 1 prior therapy there was a clear PFS benefit with ixazomibRd vs. placebo-Rd, including those with high-risk cytogenetics. One possibility suggested by preliminary findings is that tumors relapsed post-transplant may have a distinct biology with a less differentiated phenotype and lower expression of c-myc.24 The benefit (HR 0.44) in patients with no prior transplant but prior melphalan-containing therapy suggests that the difference is not due to prior alkylator therapy but possibly due to the transplant itself or, although speculative, due to the myeloablative dose of melphalan administered before the transplant.

Several published data have previously suggested a link between c-myc levels and the sensitivity to proteasome inhibitors.25-27 Immunomodulatory drugs and proteasome inhibitors appear to target different clones (less versus more differentiated phenotypes, respectively), which might explain in part their synergistic action and the increased benefit observed with ixazomib-Rd.24 As the study was not powered to detect a statistical difference between the subgroups, and the transplant vs. non-transplant analysis was retrospective and post-hoc rather than a prespecified subgroup analysis, this finding is hypothesis-generating and further investigations to characterize the tumor biology are ongoing.24 As seen in the overall population,18 the addition of ixazomib was associated with limited additional toxicity when compared with placebo-Rd across all patient subgroups. Overall, the safety profile of ixazomib-Rd was similar regardless of number and type of prior therapies

References 1. Brenner H, Gondos A, Pulte D. Recent major improvement in long-term survival of younger patients with multiple myeloma. Blood. 2008;111(5):2521-2526.

2. Kastritis E, Zervas K, Symeonidis A, et al.

Improved survival of patients with multiple myeloma after the introduction of novel agents and the applicability of the International Staging System (ISS): an analysis of the Greek Myeloma Study Group (GMSG).

Leukemia.

2009;23(6):1152-1157.

3. Kristinsson SY, Landgren O, Dickman PW, Derolf AR, Bjorkholm M. Patterns of survival in multiple myeloma: a populationbased study of patients diagnosed in Sweden from 1973 to 2003. J Clin Oncol.

2007;25(15):1993-1999.

1774

and was consistent with that reported for the overall patient population. Rates of all-grade AEs, grade ≥3 AEs and SAEs were similar between subgroups and were aligned with the rates seen in the overall study population.

Peripheral neuropathy and the hematologic AEs neutropenia and thrombocytopenia are known side effects of PIs. There were no consistent differences in all-grade or grade ≥3 AEs in patients with 1 vs. 2-3 prior therapies. This is in contrast to results with carfilzomib-dexamethasone, where rates of AEs were generally higher in patients with 2-3 prior therapies vs. 1 prior therapy.21 There are a number of limitations associated with subgroup analyses of this type. The subgroup analyses were not powered for formal statistical testing, some were not prespecified, and analyses did not use a multivariate approach, hence there may be confounding factors, such as an imbalance between some subgroups in terms of other prognostic factors.

In conclusion, ixazomib plus lenalidomide-dexamethasone demonstrated a clear PFS, TTP, and response rate benefit compared to lenalidomide-dexamethasone alone, with limited additional toxicity, in patients with RRMM, regardless of prior therapy received. The findings in patients with 1 prior therapy and transplant are hypothesis-generating and further investigations are ongoing.

Together, these findings support the results from the primary analysis of TOURMALINE-MM1, further demonstrating that the all-oral regimen of ixazomib, lenalidomide, and dexamethasone represents an effective and tolerable treatment option for patients with RRMM. Acknowledgments The authors would also like to acknowledge writing support from Jane Saunders of FireKite, an Ashfield company, part of UDG Healthcare plc, during the development of this manuscript, which was funded by Millennium Pharmaceuticals, Inc., and complied with Good Publication Practice 3 ethical guidelines (Battisti et al., Ann Intern Med 2015;163:461–4).

Funding Millennium Pharmaceuticals, Inc., a wholly owned subsidiary of Takeda Pharmaceutical Company Limited.

4. Kumar SK, Dispenzieri A, Lacy MQ, et al.

Continued improvement in survival in multiple myeloma: changes in early mortality and outcomes in older patients. Leukemia.

2014;28(5):1122-1128.

5. Kumar SK, Rajkumar SV, Dispenzieri A, et al. Improved survival in multiple myeloma and the impact of novel therapies. Blood.

2008;111(5):2516-2520.

6. Kumar SK, Therneau TM, Gertz MA, et al.

Clinical course of patients with relapsed multiple myeloma. Mayo Clin Proc.

2004;79(7):867-874.

7. Brioli A, Melchor L, Cavo M, Morgan GJ. The impact of intra-clonal heterogeneity on the treatment of multiple myeloma. Br J Haematol. 2014;165(4):441-454.

8. Jagannath S, Roy A, Kish J, et al. Real-world treatment patterns and associated progression-free survival in relapsed/refractory

multiple myeloma among US community oncology practices. Expert Rev Hematol.

2016;9(7):707-717.

9. Kumar SK, Lee JH, Lahuerta JJ, et al. Risk of progression and survival in multiple myeloma relapsing after therapy with IMiDs and bortezomib: a multicenter international myeloma working group study. Leukemia.

2012;26(1):149-157.

10. San Miguel J, Weisel K, Moreau P, et al.

Pomalidomide plus low-dose dexamethasone versus high-dose dexamethasone alone for patients with relapsed and refractory multiple myeloma (MM-003): a randomised, open-label, phase 3 trial. Lancet Oncol. 2013;14(11):1055-1066.

11. Willenbacher E, Weger R, Rochau U, Siebert U, Willenbacher W. Real-world use of 3rd line therapy for multiple myeloma in Austria: an Austrian Myeloma Registry

haematologica | 2017; 102(10)

Ixazomib-Rd in RRMM patients: impact of prior therapies

12.

13.

14.

15.

16.

17.

(AMR) analysis of the therapeutic landscape and clinical outcomes prior to the use of next generation myeloma therapeutics.

PLoS One. 2016;11(3):e0147381.

Vogl DT, Stadtmauer EA, Richardson PG, et al. Impact of prior therapies on the relative efficacy of bortezomib compared with dexamethasone in patients with relapsed/refractory multiple myeloma. Br J Haematol. 2009;147(4):531-534.

Wang M, Dimopoulos MA, Chen C, et al.

Lenalidomide plus dexamethasone is more effective than dexamethasone alone in patients with relapsed or refractory multiple myeloma regardless of prior thalidomide exposure. Blood. 2008;112(12):44454451.

Conner TM, Doan QD, Walters IB, LeBlanc AL, Beveridge RA. An observational, retrospective analysis of retreatment with bortezomib for multiple myeloma. Clin Lymphoma Myeloma. 2008;8(3):140-145.

Oriol A, Giraldo P, Kotsianidis I, et al.

Efficacy and safety of bortezomib-based retreatment at the first relapse in multiple myeloma patients: A retrospective study.

Hematology. 2015;20(7):405-409.

Petrucci MT, Giraldo P, Corradini P, et al. A prospective, international phase 2 study of bortezomib retreatment in patients with relapsed multiple myeloma. Br J Haematol.

2013;160(5):649-659.

Wolf J, Richardson PG, Schuster M, LeBlanc

haematologica | 2017; 102(10)

18.

19.

20.

21.

22.

23.

A, Walters IB, Battleman DS. Utility of bortezomib retreatment in relapsed or refractory multiple myeloma patients: a multicenter case series. Clin Adv Hematol Oncol. 2008;6(10):755-760.

Moreau P, Masszi T, Grzasko N, et al. Oral ixazomib, lenalidomide, and dexamethasone for multiple myeloma. N Engl J Med. 2016;374(17):1621-1634.

Rajkumar SV, Harousseau JL, Durie B, et al.

Consensus recommendations for the uniform reporting of clinical trials: report of the International Myeloma Workshop Consensus Panel 1. Blood. 2011; 117(18):4691-4695.

Sonneveld P, Broijl A. Treatment of relapsed and refractory multiple myeloma.

Haematologica. 2016;101(4):396-406.

Moreau P, Joshua D, Chng WJ, et al. Impact of prior treatment on patients with relapsed multiple myeloma treated with carfilzomib and dexamethasone vs bortezomib and dexamethasone in the phase 3 ENDEAVOR study. Leukemia. 2016;31(1):115-122.

Avet-Loiseau H, Soulier J, Fermand JP, et al.

Impact of high-risk cytogenetics and prior therapy on outcomes in patients with advanced relapsed or refractory multiple myeloma treated with lenalidomide plus dexamethasone. Leukemia. 2010;24(3):623628.

Dimopoulos MA, Stewart AK, Rajkumar SV, et al. Effect of carfilzomib, lenalido-

24.

25.

26.

27.

mide, and dexamethasone (KRd) vs lenalidomide and dexamethasone (Rd) in patients with relapsed multiple myeloma (RMM) by line of therapy: Secondary analysis from an interim analysis of the phase III study ASPIRE (NCT01080391). J Clin Oncol. 2015;33(15_suppl):abstract 8525.

Di Bacco A, Bahlis NJ, Munshi NC, et al.

Higher c-MYC expression is associated with ixazomib-lenalidomide-dexamethasone (IRd) progression-free survival (PFS) benefit versus placebo-Rd: biomarker analysis of the phase 3 tourmaline-MM1 study in relapsed/refractory multiple myeloma (RRMM).

Blood.

2016; 128(22):abstract 243 Chen S, Blank JL, Peters T, et al. Genomewide siRNA screen for modulators of cell death induced by proteasome inhibitor bortezomib. Cancer Res. 2010;70(11):43184326.

Chng WJ, Gertz MA, Chung TH, et al.

Correlation between array-comparative genomic hybridization-defined genomic gains and losses and survival: identification of 1p31-32 deletion as a prognostic factor in myeloma. Leukemia. 2010;24(4):833-842.

Ravi D, Beheshti A, Abermil N, et al.

Proteasomal inhibition by ixazomib induces CHK1 and MYC-dependent cell death in T-cell and hodgkin lymphoma.

Cancer Res. 2016;76(11):3319-333.

1775

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Plasma Cell Disorders

Ferrata Storti Foundation

Haematologica 2017 Volume 102(10):1776-1784

The BET bromodomain inhibitor CPI203 improves lenalidomide and dexamethasone activity in in vitro and in vivo models of multiple myeloma by blockade of Ikaros and MYC signaling

Tania Díaz,1,2 Vanina Rodríguez,2 Ester Lozano,1,2 Mari-Pau Mena,1,2 Marcos Calderón,1,2 Laura Rosiñol,1,2 Antonio Martínez,3 Natalia Tovar,1,2 Patricia Pérez-Galán,2 Joan Bladé,1,2 Gaël Roué2,4 and Carlos Fernández de Larrea1,2

Amyloidosis and Myeloma Unit, Department of Hematology, Hospital Clinic, University of Barcelona, Barcelona; 2Division of Hematology and Oncology, Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), CIBERONC, Barcelona; 3Hematopathology Unit, Department of Pathology, Hospital Clinic, IDIBAPS, Barcelona and 4Laboratory of Experimental Hematology, Department of Hematology, Vall d'Hebron Institute of Oncology, Vall d’Hebron University Hospital, Barcelona, Spain 1

GR and CFdL share the senior authorship of this article

ABSTRACT

M

Correspondence: cfernan1@clinic.ub.es or groue@clinic.ub.es

Received: January 20, 2017.

Accepted: July 19, 2017.

Pre-published: July 27, 2017. doi:10.3324/haematol.2017.164632 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1776 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1776

ost patients with multiple myeloma treated with current therapies, including immunomodulatory drugs, eventually develop relapsed/refractory disease. Clinical activity of lenalidomide relies on degradation of Ikaros and the consequent reduction in IRF4 expression, both required for myeloma cell survival and involved in the regulation of MYC transcription. Thus, we sought to determine the combinational effect of an MYC-interfering therapy with lenalidomide/dexamethasone. We analyzed the potential therapeutic effect of the combination of the BET bromodomain inhibitor CPI203 with the lenalidomide/dexamethasone regimen in myeloma cell lines. CPI203 exerted a dose-dependent cell growth inhibition in cell lines, indeed in lenalidomide/dexamethasone-resistant cells (median response at 0.5 mM: 65.4%), characterized by G1 cell cycle blockade and a concomitant inhibition of MYC and Ikaros signaling. These effects were potentiated by the addition of lenalidomide/dexamethasone. Results were validated in primary plasma cells from patients with multiple myeloma co-cultured with the mesenchymal stromal cell line stromaNKtert. Consistently, the drug combination evoked a 50% reduction in cell proliferation and correlated with basal Ikaros mRNA expression levels (P=0.04). Finally, in a SCID mouse xenotransplant model of myeloma, addition of CPI203 to lenalidomide/dexamethasone decreased tumor burden, evidenced by a lower glucose uptake and increase in the growth arrest marker GADD45B, with simultaneous downregulation of key transcription factors such as MYC, Ikaros and IRF4. Taken together, our data show that the combination of a BET bromodomain inhibitor with a lenalidomidebased regimen may represent a therapeutic approach to improve the response in relapsed/refractory patients with multiple myeloma, even in cases with suboptimal prior response to immunomodulatory drugs.

Introduction Multiple myeloma (MM) is a hematologic malignancy characterized by neoplastic growth of bone marrow plasma cells. It constitutes almost 15% of all hematologic malignancies.1,2 Virtually, all cases of MM are preceded by a pre-malignant state of clonal plasma cell known as monoclonal gammopathy of undetermined significance (MGUS).3 The introduction of novel drugs, such as thalidomide, bortehaematologica | 2017; 102(10)

Combinational effect of CPI203 with Len/Dex in myeloma

zomib and lenalidomide, has resulted in an improved response rate and progression-free survival (PFS).4-6 Lenalidomide, like other immunomodulatory drugs (IMiDs), acts by modulating the substrate specificity of the CRL4-CRBN E3 ubiquitin ligase complex. This drug enhances the binding of Ikaros (IKZF1) and Aiolos (IKZF3) to the ubiquitin ligase complex, leading to a degradation of these two factors by the proteasome, thus inducing a reduction in their protein levels.7-9 This, in turn, reduces MM cell survival.

Although the improvement achieved with new therapeutic approaches is clinically relevant, it is far from satisfactory as MM remains incurable, with a significant shortening of life-expectancy.10 For this reason, research into new drugs to treat relapse or refractory MM patients constitutes a field of intense investigation. In this regard, bromodomain and extra-terminal (BET) inhibitors have emerged as promising molecules for the treatment of hematologic malignancies. BET family proteins (BRD2, BRD3 and BRD4) are chromatin adaptors, functionally linked to important pathways for cellular viability and cancer signaling. In particular, BRD4 has a direct role in the regulation of transcription of different genes involved in the cell cycle and transcription of oncogenes.11 It has been demonstrated that BRD4 is highly expressed in dividing cells and has an important role in cell growth regulation.12,13 Therefore, it is conceivable that its deregulation can influence cancer cell biology and the inhibition of BRD4 could effectively disrupt tumor growth.14 Thus, BRD4 has been recently described as a therapeutic target for MM, among other hematologic diseases. The BET inhibitor (BETi) (+)-JQ1 selectively inhibits BRD4 by competitively binding to the acetyl-lysine recognition pocket of BET bromodomains from chromatin.15,16 This displacement of BRD4 from chromatin leads to the inhibition of MYC transcription in a dose- and time-dependent manner.

Although gene expression changes observed after BETbromodomain inhibition are mainly dominated by the MYC transcriptome, BET inhibitors influence the expression of a more extensive assortment of nearly 3000 genes.17,18 CPI203 is an analog of (+)-JQ1 with superior bioavailability via oral or intraperitoneal administration.11 Moreover, the antitumoral effects of CPI203 are comparable, and even higher in some cases, than the effects of (+)JQ1, both in vitro and in vivo.11,19 Previous studies in MM showed the capacity of CPI203 to inhibit cell growth, even in cells resistant to bortezomib and melphalan.18 CPI203 and bortezomib had a synergistic antiproliferative effect in vitro, where CPI203 causes a decrease in MYC expression levels sufficient to reduce proliferation and aggresome-mediated survival, yet permitting enough NOXA expression for bortezomib to potentiate apoptosis.18 Accordingly, mantle cell lymphoma cells are also notably sensitive to CPI203 in bortezomib-resistant cells with increased MYC basal expression. In this scenario, CPI203 and lenalidomide synergistically inhibit the growth of bortezomib-resistant tumors.20 Malignant plasma cells in MM require Ikaros family zinc finger factor 1 (IKZF1) for their survival, which is therapeutically targeted by lenalidomide. As this protein is involved in the regulation of MYC transcription, our aim was to explore the activity of therapy with CPI203 targeting MYC in combination with a lenalidomide-based regimen in both in vitro and in vivo models of myeloma. haematologica | 2017; 102(10)

Table 1. Drug sensitivity and combination index (CI) of CPI203 and lenalidomide in MM cell lines.

Cell line

ARP-1 JJN.3 U266 MM.1S MM.1R RPMI-8226 KMM.1

Lenalidomide sensitivity

CPI203 GI50 at 48h (mM)

CI values (CPI203 0.1 mM/Len 5 mM/Dex 0.1 mM)

Sensitive Sensitive Sensitive Sensitive Sensitive Resistant Resistant

0.16 0.08 >1 >1 0.12 >1 0.23

0.082 0.514 0.331 0.284 0.283 0.090 0.379

MM: myltiple myeloma; h: hours; Len: lenalidomide; Dex: dexamethasone.

Methods Cell lines and patient samples Human myeloma cell lines ARP-1, JJN-3, U266, MM.1S, MM.1R, RMPI-8226 and KMM.1 were maintained in 10-15% FCS-supplemented RPMI-1640 medium (Thermo Fisher, Waltham, MA, USA).

Primary mononuclear cells from bone marrow aspiration of 9 patients [4 male (M)/5 female (F), median age: 63 years (range: 5189)] with symptomatic MM were isolated by Ficoll/Hypaque sedimentation (GE Healthcare, Chalfont St Giles, UK). Ethical approval for this project, including patient informed consent, were granted following the guidelines of the Hospital Clinic Ethics Committee (IRB).

More detailed information is provided in the Online Supplementary Methods.

Cell proliferation assays Myeloma cell lines (5x104 per well) were incubated with CPI203 (kindly provided by Constellation Pharmaceuticals, Cambridge, MA, USA) and/or lenalidomide (Selleck Chemicals LLC, Houston, TX, USA) plus dexamethasone (Merck, S.L., Darmstadt, Germany) at indicated doses in triplicates. MTT [3-(4,5dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide] assay (Sigma-Aldrich, St Louis, MO, USA) was used to evaluate the effect of the drugs on cell proliferation.

Primary cells were labeled with CellTrackerTM Red CMPTX dye (Thermo Fisher) following the manufacturerâ&#x20AC;&#x2122;s protocol and co-cultured with the mesenchymal stromal cell line stromaNKtert in the presence of 10 ng/mL IL-6 (RnD Systems, Minneapolis, MN, USA). Cell proliferation was analyzed in an Attune acoustic focusing cytometer using Attune software (Thermo Fisher).

Gene expression profiling and gene set enrichment analysis RNA was analyzed on Affymetrix Human Genome U219 arrays. Gene set enrichment analysis (GSEA) v.2.0 (Broad Institute at MIT, Boston, USA; http://www.broadinstitute.org/gsea/) was used to identify gene signatures, interrogating C2CP and C3TFT gene sets from the Molecular Signature Database v.2.5 and experimentally-derived custom gene sets related to Ikaros.21 The microarray data have been deposited in the NCBIâ&#x20AC;&#x2122;s Gene Expression Omnibus and are accessible through GEO series accession number GSE87403.

1777

T. Diaz et al.

Xenograft mouse model SCID mice (Charles River Laboratories, L’Arbresle, France) were inoculated subcutaneously with 1.2x107 cells of RPMI-8226 cell line. Mice were randomly assigned into cohorts of 5 mice each

and received either a twice-daily dose of CPI203 (2.5 mg/kg) for two weeks, or a daily dose of lenalidomide (25 mg/kg) plus twice weekly dexamethasone (1 mg/kg), or the combination of both, or an equal volume of vehicle. Twenty-one days post-cell inocula-

A

B

C

D

E

F

Figure 1. Characterization of CPI203 effect in multiple myeloma cell lines. (A) A set of seven myeloma cell lines were exposed to increasing concentrations of CPI203 for 48 hours (h). The relative number of proliferating cells was analyzed by MTT assay. Results are represented as mean±Standard Error of Mean (SEM) of triplicate assays. (B) JJN-3, RPMI-8226 and ARP-1 cells were treated for 24 h with 0.1 mM CPI203 and cell cycle fractions were determined by flow cytometry of propidium iodide-labeled nuclei. (C) MYC protein levels were analyzed by Western blot after 0.1 mM CPI203 treatment (48 h) in the 7 cell lines; β-actin was used as loading control. (D) Heatmaps of the leading edges of IKAROS-related gene sets identified as enriched by GSEA in cells treated with CPI203 (6 h) versus control. Threshold FDR=0.01 and NES=1.80 and FDR=0.09 and NES=-1.62 for gene sets “Ikaros del down-regulated” and “Ikaros del up-regulated”, respectively. (E) Changes in the expression of the selected genes (Ikaros and GADD45B) were confirmed by Western blot in three representative myeloma cell lines after 24 h of treatment. (F) IKZF1 and MYC mRNA expression in three representative cell lines tested after 6 h of treatment with CPI203 (0.1 mM). Results are referred to the untreated control and GUSB was used as endogenous control. Data are shown as mean±SEM. t-test was performed with reference to the control. *P<0.05, **P<0.01.

1778

haematologica | 2017; 102(10)

Combinational effect of CPI203 with Len/Dex in myeloma

tion, intratumoral glucose uptake was evaluated with an Odyssey infra-red scanner (Li-Cor, Lincoln, NE, USA) in mice previously injected with an IRDye 800CW 2-deoxyglucose probe (Li-Cor).

Animals were then sacrificed according to institutional guidelines and tumor xenografts were isolated. Paraffin-embedded tumor

samples were subjected to immunohistochemical staining using primary antibodies against Ikaros, MYC, IRF4, GADD45B and pH3 and evaluated with an Olympus DP70 microscope by means of a 20x/0.75 NA objective and DPManager software v.2.1.1 (Olympus).

A

B

C

D

E

Figure 2. Synergistic antiproliferative effect due to modulation of MYC and Ikaros by CPI203/Len/Dex. (A) JJN-3, RPMI-8226 and ARP-1 were treated with increasing doses of CPI203 and/or Len/Dex and the drug cytostatic effect was analyzed by MTT proliferation assay after 48 hours (h). (B) Relative proliferation of 7 myeloma cell lines after 48 h of treatment with CPI203 (0.1 mM), Len/Dex (5 mM/100 nM) and the 3-drug combination. Data are shown as meanÂąStandard Error of Mean (SEM). One-way ANOVA test was performed; P=0.0003 was considered statistically significant. *P<0.05, **P<0.01,***P<0.001. (C) Heatmap of the leading edge of Ikaros-related gene sets identified as negatively enriched in the combo by GSEA using an increasing profile analysis. Threshold FDR=0.087 and NES=-1.65. (D) IKZF1 and IKZF3 mRNA expression in all myeloma cell lines tested after 6-h treatment with CPI203 (0.1 mM) and/or Len/Dex (5 mM/100 nM). Results are referred to the untreated control and GUSB was used as endogenous control. (E) Changes in the expression of the selected factors (Ikaros, MYC and Aiolos) after 24 h of CPI203 and/or Len/Dex treatment were confirmed by Western blot in three cell lines. Data are shown as meanÂąSEM. t-test was performed comparing CPI203 and Len/Dex as single agents with combo. *P<0.05, **P<0.01, ***P<0.001.

haematologica | 2017; 102(10)

1779

T. Diaz et al.

A Figure 3. Synergistic antitumor effect of the 3-drug combination (CPI203/Len/Dex) in primary myeloma cells. (A) Relative proliferation of primary bone marrow cells from 9 relapsed multiple myeloma (MM) patients, after 48 hours (h) of treatment with CPI203 (0.1 mM), Len/Dex (5 mM/100 nM) and the 3-drug combination. Results are referred to the untreated control and each point represents one patient. Data are shown as meanÂąStandard Error of Mean (SEM). *P<0.05, **P<0.01, ***P<0.001. (B) Basal levels of mRNA for IKZF1, IRF4 and MYC (using GUSB as endogenous control) in all the patients included in the study related to response to the drug combination treatment. Combo: combinational therapy.

B

Statistical analysis All statistical analyses were performed using GraphPad software 5.0 (GraphPad Software Inc., San Diego, CA, USA). The comparisons between all analyzed groups were evaluated with a Kruskal-Wallis test and the comparisons between two groups were analyzed with Student t-test or non-parametric MannWhitney test. P<0.05 was considered statistically significant; data are represented as meanÂąStandard Error of Mean (SEM) of 3 independent experiments (*P<0.05, **P<0.01, ***P<0.001).

Results Anti-myeloma activity of CPI203 is independent of cell sensitivity to lenalidomide and involves MYC and IKZF1 downregulation CPI203 has recently been shown to exert a significant antitumor activity in the low micromolar range of concentration in MM cell lines, irrespective of the primary response to the proteasome inhibitor bortezomib or to the alkylating agent melphalan.18 To investigate the activity of the compound with regards to the response to a lenalidomide-based therapy, a panel of 5 lenalidomide-responsive and 2 lenalidomide-resistant cell lines was exposed for 48 hours (h) to CPI203 doses ranging from 0.05 to 1 mM and cell viability was measured by MTT assay. The compound exerted a dose-dependent inhibition of proliferation in all the MM cell lines tested, with the optimal reduction in cell proliferation achieved at the 0.5 mM dose (median response: 65.4%, range: 40-81%) (Figure 1A), while in the most sensitive cell lines (MM.1R, JJN-3 and KMM-1) the GI50 value decreased to below 100 nM (Table 1). Of interest, cell response to the compound was independent of primary response to lenalidomide, as resistant cells showed a similar response to the sensitive ones (median response: 54.1%, range: 43-66%). In agreement with previous reports, the compound failed to evoke apoptotic cell 1780

death in MM cells since its activity mainly related to a significant blockade of the cell cycle at the G1 phase (mean increase of apoptotic cells: 15.1%) in the three representative cell lines analyzed: JJN-3 (32%), ARP-1 (42%) and RPMI-8226 (9%) (Figure 1B). This effect was accompanied by a decrease in MYC protein levels in all the cell lines, although a strict correlation could not be observed with the efficacy of the compound (Figure 1C), thus arguing in favor of a role for additional mechanism(s). To better characterize the main factors involved in MM response to CPI203, we then performed gene expression profiling (GEP) analysis with the three cell lines used previously, either untreated or treated for 6 h with 0.1 mM CPI203.

We performed GSEA using well-defined and previously described gene signatures.20 As expected, there was an enrichment of genes up-regulated by MYC and genes down-regulated by the transcription factor BLIMP1 in the control cells when compared to CPI203-treated samples (Online Supplementary Table S1 and Online Supplementary Figure S1). It is noteworthy that among the different proliferation-associated gene sets analyzed there was a simultaneous, marked upregulation of Ikaros-repressed genes and downregulation of Ikaros-induced genes in cells exposed to the BET inhibitor (Figure 1D and Online Supplementary Table S1). As Ikaros has been shown to be a crucial regulator of the G1 to S transition of the cell cycle,22 we checked for the presence of G1 regulatory factors among the top 10 genes regulated by CPI203 and included in the two Ikaros gene sets considered; we identified GADD45B, a well-known negative regulator of cell cycle progression.23 Consistently, we observed a concomitant downregulation of Ikaros (mean reduction: 70.4%; range: 55.1-82.2%) and increase in GADD45B protein levels (mean increment: 45.2%; range: 34.5-58.9%) after 24 h of treatment with 0.1 mM CPI203 in the three representative cell lines (Figure 1E). As expected, CPI203 induced a decrease in MYC and IKZF1 mRNA levels (mean decrease: haematologica | 2017; 102(10)

Combinational effect of CPI203 with Len/Dex in myeloma

40-50%) (Figure 1F); this transcriptional repression could explain the upregulation of GADD45B.

CPI203 synergistically increases lenalidomide-induced blockade of MYC and Ikaros signaling in vitro Reduction in Ikaros protein levels has been well documented and related to cereblon-dependent proteasomal degradation of the transcription factor in MM cells after treatment with lenalidomide.7,9,24 To determine if transcriptional repression of the corresponding gene by means of CPI203 could offer an improvement in lenalidomide antitumor activity, we treated our panel of 7 MM cell lines with standard doses of the Len-Dex regimen, in the presence or the absence of the BET inhibitor at the two optimal doses previously described (0.1 and 0.5 mM). We had previously tested all three drugs individually, and the combination of lenalidomide and/or dexamethasone with CPI203 in four MM cell lines (MM1.S, KMM.1, U266 and RPMI-8226). In this preliminary study, we observed that neither lenalidomide nor dexamethasone as single agent had a combinational effect with the BRD4 inhibitor (Online Supplementary Figure S2). In contrast, cell treatment with the Len/Dex combination, which corresponds to the regimen proposed to patients with MM in clinical practice, allowed a remarkable improvement of CPI203 activity with both concentrations of the drug in all seven cell lines tested (Figure 2A and Online Supplementary Figure S3).

In order to better evaluate the co-operation between the two drugs, we calculated the Combination index (CI) in each cell line, based on the Chou-Talalay method. The best combinational activity was achieved when combining the 0.1 ÂľM dose of CPI203 with the treatment by lenalidomide (5 mM) and dexamethasone (0.1 mM), obtaining CI values ranging from 0.082 to 0.514 (mean: 0.280) (Table 1). As CI values between 0.3 and 0.7 indicate a synergistic effect, and values between 0.1 and 0.3 suggest a strong synergism between the two drugs of interest, our results suggest a high synergistic effect of the 3-drug combination. These doses were those used in all the validation experiments. At these doses, single agent CPI203 induced a 42.7% reduction (range: 13-74%; P<0.05), which achieved up to 76.1% reduction (range: 53-96%; P<0.001) when combined with Len/Dex treatment (Figure 2B). Although neither CPI203 nor Len/Dex individually caused a significant increase in cell apoptosis, the combination of these drugs resulted in a mean relative increase in the number of apoptotic cells of 37.9%. In order to validate the main gene signatures involved in this effect, we ran a new GSEA analysis using an increasing profile mode, comparing control cells with CPI203-treated and CPI203/Len/Dex-treated samples, in the same conditions as described previously. As shown in Online Supplementary Table S2, and as exemplified in Figure 2C and Online Supplementary Figure S4, gene sets related to MYC function as transcriptional regulator or with plasma cell differentiation (i.e. BLIMP-1 and IRF4-dependent gene sets) were significantly more affected by the CPI203/Len/Dex combination than by CPI203 alone. Of special interest, the group of genes positively regulated by Ikaros was also disrupted by the combination treatment (Figure 2C). Western blot analysis confirmed the increasing reduction in Ikaros and MYC protein levels with the sequential addition of the different drugs, being both factors dramatically decreased in the CPI203/Len/Dex drug combination treated cells (Figure 2E). As expected, while cells exposed to single haematologica | 2017; 102(10)

agent CPI203 or CPI203/Len/Dex combination harbored a 40-50% reduction in IKZF1 mRNA levels, respectively, Len/Dex treatment alone did not significantly affect Ikaros transcript levels, confirming a post-transcriptional regulation of Ikaros expression (Figure 2D). As previously described, lenalidomide also induces the degradation of Aiolos via interaction with CRBN. Thus we analyzed both mRNA expression and protein levels of this factor upon myeloma cell exposure to the drug. The modulation of IKZF3 mRNA after CPI203 and/or Len/Dex treatment was very similar to that seen with IKZF1 (Figure 2D).

However, we observed that CPI203, as single agent, had no effect on Aiolos protein expression, and downregulation of this protein in cells receiving the 3-drug combination was mainly due to lenalidomide activity (Figure 2E).

BET bromodomain inhibition increases Len/Dex efficacy in MM primary cultures To validate the above results in MM primary samples ex vivo, bone marrow aspirates from 9 symptomatic MM patients with high contents in tumor cells (mean percentage CD38+ cells: 60%) were cultured for 48 h upon a monolayer of the mesenchymal stromal cell line stromaNKtert25 plus IL-6,26 and in the presence or the absence of CPI203 (0.1 mM) and/or lenalidomide (5 mM)/dexamethasone (0.1 mM) combination. In the absence of drug, a subset of primary CD38+ cells underwent cell cycle entry and proliferation as evaluated by cell labeling with a CellTracker dye (mean: 15%). When referred to control cells, CPI203- and Len/Dex-exposed samples showed a 33% and 30% reduction in stromamediated cell proliferation, respectively, while this effect was increased up to 53% in the case of the drug combination (*P<0.05, **P<0.01, ***P<0.001) (Figure 3A). Of special interest, in this set of primary samples there was a significant correlation between IKZF1 mRNA levels and cell response to the CPI203/Len/Dex combination, as those cases with lower levels of Ikaros showed a higher response to this treatment (P=0.04) (Figure 3B). Such a correlation was not found when considering IRF4 or MYC expression (Figure 3B); thus highlighting the crucial role of Ikaros towards BETi/Len/Dex combinational activity in primary MM cells.

The CPI203/Len/Dex combination inhibits MM tumor growth in vivo To further characterize the co-operative role of CPI203 and Len/Dex in vivo, SCID mice inoculated with RPMI8226 cells were randomly assigned into three treatment groups (CPI203 2.5 mg/kg BID, lenalidomide 50 mg/kg daily plus dexamethasone 1 mg/kg twice weekly and combination) and the vehicle-treated groups. While CPI203 and Len/Dex achieved a 62% and 61% reduction in tumor volume (P=0.031 and P=0.023, respectively), when compared to the vehicle group, the combination of BETi and Len/Dex resulted in complete arrest of tumor growth in mice receiving the CPI203/Len/Dex combination (P=0.0012) (Figure 4A). Accordingly, tumor glucose uptake was reduced to 47% and 45% in animals treated with CPI203 and Len/Dex, respectively, while the combination therapy resulted in a 64% reduction (Figure 4B).

Immunohistochemical analysis of the corresponding tumors confirmed an additional decrease in the mitotic index as shown by phospho-histone H3 staining, together with the almost complete disappearance of MYC-, IRF41781

T. Diaz et al.

and Ikaros-positive cells, and a remarkable accumulation of GADD45B-expressing cells in the group receiving the drug combination (Figure 4C). These results confirmed our in vitro data, showing that the combination of the BET inhibitor CPI203 with Len/Dex augments the antitumor properties of each single agent, and results in the abrogation of Ikaros and MYC signaling and consequent blockade of tumor growth.

Discussion Constitutive activation of MYC signaling is detected in more than 60% of patient-derived myeloma cells27 and can be involved in the pathogenesis of MM through different mechanisms related to the progression from early stages, such as MGUS, to symptomatic disease. One of the most common somatic genomic aberrations in early-

A

B

C

Figure 4. CPI203 plus Len/Dex synergistically inhibits tumor growth in multiple myeloma (MM) mouse model. (A) Mice were inoculated with 1.2x107 RPMI-8226 cells and treatment began at day (d)7 post-cell inoculation. Evolution of tumor burden during the treatment; volumes were recorded every 3-4 days by external calipers. (B) Intratumoral glucose uptake images obtained with an Odyssey infra-red scanner and their corresponding fluorescence quantifications from representative mice at the day of sacrifice. Data are shown as meanÂąStandard Error of Mean (SEM). *P<0.05, **P<0.01, ***P=0.001. (C) Tumor size (mm) and immunohistochemical staining of hematoxylin and eosin (H&E), p-histone H3, GADD45B, Ikaros, MYC and IRF4 in consecutive tissue sections from tumors after two weeks of indicated treatment.

1782

haematologica | 2017; 102(10)

Combinational effect of CPI203 with Len/Dex in myeloma

and late-stage MM is rearrangement or translocation of MYC.28 Moreover, oncogenic super enhancers can recruit the BET family and consequently augment the aberrant MYC expression.29 Given that BET bromodomain inhibition has previously been shown to disrupt MYC signaling among other pathways in different hematologic cancers, we sought to determine if the BETi might represent a new therapeutic option for MM. In this study, we have demonstrated that the BRD4 selective inhibitor CPI203 could be used either as a single agent (considering its remarkable activity as monotherapy in vitro and in vivo) or in combination with standard therapies, as shown previously with the proteasome inhibitor bortezomib.18 Interestingly, as observed in mantle cell lymphoma,20 the synergistic activity of CPI203 with bortezomib in MM also overcame resistance to this proteasome inhibitor. Here, we followed a similar design by including two cell lines with reduced sensitivity to lenalidomide and demonstrated by in vitro and in vivo approaches the antitumoral activity of this molecule, as well as its capacity to enhance MM response to the immunomodulatory drug, thus highlighting possible therapeutic implications. Moreover, our results are in agreement with a previous publication in primary effusion lymphoma which demonstrated a synergistic effect on cytotoxicity between IMIDs and BDR4 inhibitors.30 While, despite the advances in the management of the disease, MM remains incurable, strategies based on the combination of IMiDs with other agents have improved the prognosis of these patients. For example, the VTD (bortezomib, thalidomide and dexamethasone) combination is a highly effective induction regimen prior to autologous stem cell transplantation (ASCT) to treat patients with standard- and high-risk MM, although 15% of patients fail to respond.6 Moreover, the duration of responses is limited and nearly all patients relapse and require salvage chemotherapy. In this sense, rescue therapy with Len/Dex is effective in increasing the response rate, the time to progression and overall survival in patients with relapsed or refractory MM,31-33 being an established treatment option for this group of patients.

Len/Dex also constitutes the back-bone of combination therapy with newer agents, such as proteasome inhibitors or monoclonal antibodies.34,35 According to our in vivo results, the addition of CPI203 to Len/Dex allowed for an almost complete and prolonged inhibition of tumor growth, where probably dexamethasone plays a crucial role (either direct or indirectly) as seen when used in patients. In this sense, this 3-drug strategy may improve responses compared to the effects of combining new-generation IMiDs with dexamethasone.6,36-38 Thus, following the current phase I clinical trials testing the BET bromodomain inhibitor in different hematologic malignancies, including patients with previously treated MM (clinicaltrials.gov identifier: 02157636), it would be interesting to design phase I/II clinical trials including this triple drug combination in relapsed/refractory (R/R) MM patients.

Mechanistically, lenalidomide is known to bind cereblon, with the subsequent activation of the E3-ubiquitin ligase activity that results in the degradation of key transcription factors like Ikaros. Moreover, lenalidomide indirectly inhibits IRF4 expression, mainly through downregulation of Ikaros and Aiolos.9 CPI203 has been reported to cause significant decreases in MYC expression, which may be sufficient to reduce proliferation and aggresomehaematologica | 2017; 102(10)

mediated survival.18 Using a GEP approach, we have identified a new role for BETi in the regulation of Ikaros-regulated factors at both the transcript and protein levels, in addition to the established role of BETi in the MYC/IRF4 signaling axis in MM. Among this group of genes, the expression of the negative regulator of cell cycle progression, GADD45B, tightly correlated with the G1 cell cycle blockade observed in MM cell lines upon CPI203 exposure. In agreement with these data, the combination of CPI203 with Len/Dex therapy induced a synergistic effect on proliferation in all the MM cell lines and a co-operative effect on primary cases analyzed, reaching between 50% and 80% global antiproliferative activity, in close correlation with a decrease in Ikaros protein in the case of the cell lines, or in basal Ikaros transcript levels in the cases of MM primary cultures. The identification of Ikaros-dependent signaling as a constant parameter involved in CPI203/Len/Dex response may be of special interest, as the identification of potential predictive response biomarkers may allow an individualized selection of patients to receive these specific treatments. In this regard, an association has been described between cereblon expression and response to lenalidomide and dexamethasone in patients with MM.39,40 Specifically, response and survival in patients with MM treated with lenalidomide improved when protein levels of cereblon were higher. In our study, higher basal levels of Ikaros in patients correlated with the poorest in vitro responses to the drug combination, in accordance with the concept that overexpression of Ikaros in MM cells could induce resistance to lenalidomide.8,9 Nevertheless, further studies in larger series of patients evaluating Ikaros expression as a potential marker of response to BETi-based regimens would be required to confirm these first observations.

In summary, following our initial aim to explore the combinational activity of therapy targeting MYC based on CPI203-mediated bromodomain inhibition with a lenalidomide-based regimen in malignant plasma cells, we demonstrate here that CPI203 is as efficient in vitro and in vivo as the approved Len/Dex therapy at the concentrations currently used in clinical settings to treat patients with R/R MM. More interestingly, we show a constant and rationally-based co-operation between CPI203 and Len/Dex therapy in both in vitro and in vivo models of MM. The combination of BET bromodomain inhibitors with the Len/Dex therapy is a logistically feasible approach, and could be considered as an option to improve the response in R/R patients with MM, even in cases with suboptimal prior response to IMIDs. Acknowledgments The authors would like to thank Constellation Pharmaceuticals for kindly providing CPI203.

Funding This work was supported in part by Instituto de Salud Carlos III and Fondo Europeo de Desarrollo Regional (FEDER) â&#x20AC;&#x153;Una manera de hacer Europaâ&#x20AC;? (RD12/0036/0046 to JB, PI12/01847 and PI15/00102 to G.R., and PI16/0423 to CFL), Generalitat de Catalunya (2014SGR-552 to JB), Josep Carreras Leukaemia Research Institute (Celgene grant: CEL029 to JB and CFL) and an IDIBAPS starting grant (II040060 to CFL). We thank the Hematopathology Unit from Hospital ClĂ­nic de Barcelona for their assistance during the work.

1783

T. Diaz et al.

References 1. Kyle RA, Rajkumar SV. Multiple myeloma.

Blood. 2008;111(6):2962-2972.

2. Palumbo A, Anderson K. Multiple myeloma. N Engl J Med. 2011;364(11):1046-1060.

3. Blade J, Rosinol L, Cibeira MT. Are all myelomas preceded by MGUS? Blood.

2009;113(22):5370.

4. Ludwig H, Sonneveld P, Davies F, et al.

European perspective on multiple myeloma treatment strategies in 2014. Oncologist.

2014;19(8):829-844.

5. Moreau P, Avet-Loiseau H, Harousseau JL, Attal M. Current trends in autologous stem-cell transplantation for myeloma in the era of novel therapies. J Clin Oncol.

2011;29(14):1898-1906.

6. Rosinol L, Oriol A, Teruel AI, et al.

Superiority of bortezomib, thalidomide, and dexamethasone (VTD) as induction pretransplantation therapy in multiple myeloma: a randomized phase 3 PETHEMA/GEM study. Blood. 2012;120(8):15891596.

7. Fink EC, Ebert BL. The novel mechanism of lenalidomide activity. Blood. 2015; 126(21):2366-2369.

8. Kronke J, Udeshi ND, Narla A, et al.

Lenalidomide causes selective degradation of IKZF1 and IKZF3 in multiple myeloma cells. Science. 2014;343(6168):301-305.

9. Lu G, Middleton RE, Sun H, et al. The myeloma drug lenalidomide promotes the cereblon-dependent destruction of Ikaros proteins. Science. 2014;343(6168):305-309.

10. Ludwig H, Bolejack V, Crowley J, et al.

Survival and years of life lost in different age cohorts of patients with multiple myeloma. J Clin Oncol. 2010;28(9):15991605.

11. Devaiah BN, Lewis BA, Cherman N, et al.

BRD4 is an atypical kinase that phosphorylates serine2 of the RNA polymerase II carboxy-terminal domain. Proc Natl Acad Sci USA. 2012;109(18):6927-6932.

12. Dey A, Chitsaz F, Abbasi A, Misteli T, Ozato K. The double bromodomain protein Brd4 binds to acetylated chromatin during interphase and mitosis. Proc Natl Acad Sci USA. 2003;100(15):8758-8763.

13. Rodriguez RM, Huidobro C, Urdinguio RG, et al. Aberrant epigenetic regulation of bromodomain BRD4 in human colon cancer. J Mol Med (Berl). 2012;90(5):587-595.

14. Segura MF, Fontanals-Cirera B, GazielSovran A, et al. BRD4 sustains melanoma proliferation and represents a new target for epigenetic therapy. Cancer Res. 2013; 73(20):6264-6276.

15. Delmore JE, Issa GC, Lemieux ME, et al.

BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell. 2011; 146(6):904-917.

16. Filippakopoulos P, Qi J, Picaud S, et al.

1784

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

Selective inhibition of BET bromodomains.

Nature. 2010;468(7327):1067-1073.

Mertz JA, Conery AR, Bryant BM, et al.

Targeting MYC dependence in cancer by inhibiting BET bromodomains. Proc Natl Acad Sci USA. 2011;108(40):16669-16674.

Siegel MB, Liu SQ, Davare MA, et al. Small molecule inhibitor screen identifies synergistic activity of the bromodomain inhibitor CPI203 and bortezomib in drug resistant myeloma. Oncotarget. 2015; 6(22):18921-18932.

King B, Trimarchi T, Reavie L, et al. The ubiquitin ligase FBXW7 modulates leukemia-initiating cell activity by regulating MYC stability. Cell. 2013;153(7):15521566.

Moros A, Rodriguez V, Saborit-Villarroya I, et al. Synergistic antitumor activity of lenalidomide with the BET bromodomain inhibitor CPI203 in bortezomib-resistant mantle cell lymphoma. Leukemia. 2014; 28(10):2049-2059.

Iacobucci I, Iraci N, Messina M, et al.

IKAROS deletions dictate a unique gene expression signature in patients with adult B-cell acute lymphoblastic leukemia. PLoS One. 2012;7(7):e40934.

Gomez-del Arco P, Maki K, Georgopoulos K. Phosphorylation controls Ikaros's ability to negatively regulate the G(1)-S transition.

Mol Cell Biol. 2004;24(7):2797-2807.

Zhang W, Bae I, Krishnaraju K, et al. CR6: A third member in the MyD118 and Gadd45 gene family which functions in negative growth control. Oncogene. 1999; 18(35):4899-4907.

Kronke J, Hurst SN, Ebert BL. Lenalidomide induces degradation of IKZF1 and IKZF3.

Oncoimmunology. 2014;3(7):e941742.

Kawano Y, Kobune M, Yamaguchi M, et al.

Ex vivo expansion of human umbilical cord hematopoietic progenitor cells using a coculture system with human telomerase catalytic subunit (hTERT)-transfected human stromal cells. Blood. 2003; 101(2):532-540.

Zhang Y, Moschetta M, Huynh D, et al.

Pyk2 promotes tumor progression in multiple myeloma. Blood. 2014;124(17):26752686.

Chng WJ, Huang GF, Chung TH, et al.

Clinical and biological implications of MYC activation: a common difference between MGUS and newly diagnosed multiple myeloma. Leukemia. 2011;25(6):10261035.

Shou Y, Martelli ML, Gabrea A, et al.

Diverse karyotypic abnormalities of the cmyc locus associated with c-myc dysregulation and tumor progression in multiple myeloma. Proc Natl Acad Sci USA. 2000; 97(1):228-233.

Chapuy B, McKeown MR, Lin CY, et al.

Discovery and characterization of super-

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

enhancer-associated dependencies in diffuse large B cell lymphoma. Cancer Cell.

2013;24(6):777-790.

Gopalakrishnan R, Matta H, Tolani B, Triche T Jr, Chaudhary PM. Immunomodulatory drugs target IKZF1IRF4-MYC axis in primary effusion lymphoma in a cereblon-dependent manner and display synergistic cytotoxicity with BRD4 inhibitors. Oncogene. 2016; 35(14):1797-1810.

Dimopoulos M, Spencer A, Attal M, et al.

Lenalidomide plus dexamethasone for relapsed or refractory multiple myeloma. N Engl J Med. 2007;357(21):2123-2132.

Wang M, Dimopoulos MA, Chen C, et al.

Lenalidomide plus dexamethasone is more effective than dexamethasone alone in patients with relapsed or refractory multiple myeloma regardless of prior thalidomide exposure. Blood. 2008;112(12):4445-4451.

Weber DM, Chen C, Niesvizky R, et al.

Lenalidomide plus dexamethasone for relapsed multiple myeloma in North America. N Engl J Med. 2007;357(21):21332142.

Ocio EM, Richardson PG, Rajkumar SV, et al. New drugs and novel mechanisms of action in multiple myeloma in 2013: a report from the International Myeloma Working Group (IMWG). Leukemia. 2014; 28(3):525-542.

Sonneveld P, Broijl A. Treatment of relapsed and refractory multiple myeloma.

Haematologica. 2016;101(4):396-406.

Kumar S, Flinn I, Richardson PG, et al.

Randomized, multicenter, phase 2 study (EVOLUTION) of combinations of bortezomib, dexamethasone, cyclophosphamide, and lenalidomide in previously untreated multiple myeloma. Blood. 2012; 119(19):4375-4382.

Richardson PG, Weller E, Lonial S, et al.

Lenalidomide, bortezomib, and dexamethasone combination therapy in patients with newly diagnosed multiple myeloma.

Blood. 2010;116(5):679-686.

Stewart AK, Rajkumar SV, Dimopoulos MA, et al. Carfilzomib, lenalidomide, and dexamethasone for relapsed multiple myeloma. N Engl J Med. 2015;372(2):142152.

Gandhi AK, Mendy D, Waldman M, et al.

Measuring cereblon as a biomarker of response or resistance to lenalidomide and pomalidomide requires use of standardized reagents and understanding of gene complexity. Br J Haematol. 2014;164(2):233244.

Heintel D, Rocci A, Ludwig H, et al. High expression of cereblon (CRBN) is associated with improved clinical response in patients with multiple myeloma treated with lenalidomide and dexamethasone. Br J Haematol. 2013;161(5):695-700.

haematologica | 2017; 102(10)

ARTICLE

Cell Therapy & Immunotherapy

Notch2 blockade enhances hematopoietic stem cell mobilization and homing

Weihuan Wang,1 Shuiliang Yu,1 Jay Myers,2 Yiwei Wang,1 William W. Xin,3 Marwah Albakri,1 Alison W. Xin,4 Ming Li,5 Alex Y. Huang,1,2 Wei Xin,1 Christian W. Siebel,6 Hillard M. Lazarus7 and Lan Zhou1

Department of Pathology, Case Western Reserve University, Cleveland, OH; Department of Pediatrics, Case Western Reserve University, Cleveland, OH; 3School of Arts and Sciences, University of Pennsylvania, Philadelphia, PA; 4Hathaway Brown School, Shaker Heights, OH; 5Biostatistics and Bioinformatics Core Facility, Case Comprehensive Cancer Center, School of Medicine, Case Western Reserve University, Cleveland, OH; 6Department of Molecular Biology Oncology, Genentech Inc., South San Francisco, CA and 7Department of Medicine, Case Western Reserve University, Cleveland, OH, USA

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

1 2

Haematologica 2017 Volume 102(10):1785-1795

ABSTRACT

D

espite use of newer approaches, some patients being considered for autologous hematopoietic cell transplantation (HCT) may only mobilize limited numbers of hematopoietic progenitor cells (HPCs) into blood, precluding use of the procedure, or being placed at increased risk of complications due to slow hematopoietic reconstitution. Developing more efficacious HPC mobilization regimens and strategies may enhance the mobilization process and improve patient outcome. Although Notch signaling is not essential for homeostasis of adult hematopoietic stem cells (HSCs), Notch-ligand adhesive interaction maintains HSC quiescence and niche retention.

Using Notch receptor blocking antibodies, we report that Notch2 blockade, but not Notch1 blockade, sensitizes hematopoietic stem cells and progenitors (HSPCs) to mobilization stimuli and leads to enhanced egress from marrow to the periphery. Notch2 blockade leads to transient myeloid progenitor expansion without affecting HSC homeostasis and self-renewal. We show that transient Notch2 blockade or Notch2-loss in mice lacking Notch2 receptor lead to decreased CXCR4 expression by HSC but increased cell cycling with CXCR4 transcription being directly regulated by the Notch transcriptional protein RBPJ. In addition, we found that Notch2-blocked or Notch2-deficient marrow HSPCs show an increased homing to the marrow, while mobilized Notch2-blocked, but not Notch2-deficient stem cells and progenitors, displayed a competitive repopulating advantage and enhanced hematopoietic reconstitution. These findings suggest that blocking Notch2 combined with the current clinical regimen may further enhance HPC mobilization and improve engraftment during HCT.

Correspondence: lxz47@case.edu or lan.zhou@case.edu

Received: March 13, 2017.

Accepted: July 13, 2017.

Pre-published: July 20, 2017. doi:10.3324/haematol.2017.168674 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1785 Š2017 Ferrata Storti Foundation

Introduction Hematopoietic cell transplantation (HCT) is the only curative option for various neoplastic and a few non-neoplastic diseases.1 The vast majority of clinical autologous HCT procedures utilize hematopoietic progenitor cells (HPCs) mobilized into the blood. For a variety of reasons, some patients may not mobilize adequate numbers of HPCs and thus are not candidates for the autologous HCT procedure. In addition, in some subjects, less than an optimal number of HPCs may be obtained, resulting in slower hematopoietic reconstitution and increased risk of complications during the transplant.2-4 In recent years, the use of CXCR4 antagonizing molecules/peptides (i.e. AMD3100 or plerixafor) has enhanced HPC mobilization and overcome some of these limitations.5 Inadequate mobilization, however, still remains a problem for many patients and the development of more efficacious strategies may enhance patient outcome.6 haematologica | 2017; 102(10)

Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1785

W. Wang et al.

The signaling molecule Notch is important for stem cell self-renewal and fate determination in many tissues, including the hematopoietic system. An important feature of Notch is its adhesive nature which was first described by cell aggregation assays in Drosophila studies.7,8 There are 4 Notch receptors (Notch1-4), and 2 families of Notch ligand: Jagged (JAG1-2), and delta-like (DLL1-4) ligand.

Notch2 is the major isotype expressed on hematopoietic stem cells (HSC) and non-lymphoid progenitor cells.9-12 However, the precise role and the physiological significance of Notch receptors, either as adhesion and/or signaling molecules, in HSC homeostasis and functional support are still not completely understood.

Notch signaling transactivation is consequent to a functional engagement of the Notch receptor with the Notch ligand. We previously reported that hematopoietic stem cell and progenitors (HSPCs) with faulty Notch-ligand interaction due to the loss of O-fucose modification of Notch display increased cell cycling and decreased adhesion to marrow osteoblastic lineage cells.11 These HSPCs exhibit enhanced egress from the marrow. However, the significance and the mechanism of Notch downstream signaling in the maintenance of HSC quiescence are not clear. Here we report that prior treatment with Notch2 blocking antibody sensitizes HSPC to the mobilizing stimuli of G-CSF and AMD3100 with a 3-4-fold increase in mobilization without affecting the overall bone marrow HSC homeostasis and self-renewal. Moreover, we demonstrate that Notch signaling directly regulates CXCR4 expression, and hence transient Notch2 blockade decreases CXCR4 concentration and increases cell cycling.

Consistent with these findings, transient Notch2 blockade leads to greater HSPC homing to the marrow and a competitive repopulating advantage of the progenitors with enhanced recovery of hematopoietic elements.

Methods Mice The Institutional Animal Care and Use Committee of Case Western Reserve University approved all aspects of the animal research described in this study. C57Bl/6 (Ly5.2) and B6.SJL-Ptrca Pep3b/BoyJ (B6.BoyJ:Ly5.1) mice were maintained in the lab. VavCre/Notch2F/F mice were generated by crossing Vav-Cre mice (008610; Jackson Laboratory, Bar Harbor, ME, USA) with Notch2F/F mice (010525; Jackson Laboratory, Bar Harbor, ME, USA).

In vivo Notch receptor blockade Humanized anti-Notch1 (anti-NRR1, Genentech), anti-Notch2 (anti-NRR2, Genentech) or control antibody (anti-ragweed, Genentech, South San Francisco, CA, USA) have been described previously.11 Antibodies were injected i.p. either at 15 mg/mL for anti-NRR1 and anti-ragweed, or at 25 mg/mL for anti-NRR2 as a single dose or twice weekly three days apart for a total of 4 doses.

HSPC mobilization assays Hematopoietic stem cell and progenitor mobilization was performed as described.13 Briefly, mice were injected subcutaneously with 2.5 mg G-CSF (Amgen, Thousand Oaks, CA, USA), twice daily for two days, followed by subcutaneous injection of 5 mg/kg AMD3100 (Sigma-Aldrich, St. Louis, MO, USA). Blood (250 mL) and hematopoietic tissues were collected 1 hour (h) later for the determination of circulating, splenic and marrow HSPC frequencies.

1786

Bone marrow analysis, transplantation, qRT-PCR, cell cycle analysis, chromatin immunoprecipitation, luciferase reporter analysis, and multi-photon intravital imaging See details in the Online Supplementary Appendix.

Statistical analysis Data are presented as meanÂąStandard Deviation (S.D.) unless otherwise stated. Statistical significance was assessed by Student t-test, Pearson Ď&#x2021;2 test, and Wilcoxon rank sum test.

Results Transient Notch2 signaling blockade promotes HSPC egress in response to mobilizing reagents To examine the effects of Notch signaling blockade in HSPC egress, we applied Notch blocking antibody or isotype control antibody to wild-type (WT) mice. The Notch receptor-specific antibody does not interfere with receptor-ligand interaction but blocks the cleavage of Notch receptor and hence the downstream signaling activation.

Because a single dose of Notch1 or Notch2 blocking antibody did not increase HSPC circulation in the periphery (data not shown), we applied a single dose of Notch blocking antibody followed by treatment with mobilizing reagents, using a commonly adopted regimen in mouse studies [4 doses of granulocyte-colony stimulating factor (G-CSF) or one dose of AMD3100, either alone or combined].13 We found that a single dose of anti-Notch2 followed by G-CSF or AMD3100 resulted in a 56% and 111% increase in white blood cell (WBC) counts compared with either reagent alone. Anti-Notch2 plus G-CSF increased circulating LSK and LK cells by 63% and 2.4-fold more, respectively, than G-CSF alone, while anti-Notch2 plus AMD3100 had a milder effect. When anti-Notch2 was followed by combined G-CSF and AMD3100 stimulation, it further increased WBC count and mobilized more LSKs and LKs (Figure 1A-C). We did not observe any significant change in circulating HSPCs in mice receiving anti-Notch1 (data not shown).

Previous studies showed that up to 4 doses of Notch2 blocking antibody were required to achieve the complete on-target biological effects of Notch2 inhibition.14 We tested this observation by applying 4 doses of antibodies to mice (Online Supplementary Figure S1A). We observed a 68% increase in WBC counts in mice receiving 4 doses of anti-Notch2 alone. We also observed a moderate increase in circulating LK cells in mice receiving either anti-Notch1 or anti-Notch2 (Online Supplementary Figure S1B and C).

Spleen-residing LSK and LK cells, however, increased after Notch2 blockade but not after Notch1 blockade (Online Supplementary Figure S1D).

We then compared a single dose versus 4 doses of Notch2 antibody combined with G-CSF or/and AMD3100 (Online Supplementary Figure S1E). Four doses of anti-Notch2 combined with G-CSF or AMD3100 alone resulted in an 87% and 120% increase in WBC counts, increased circulating LKs by 2.0- and 2.8-fold more, and increased LSKs by 61% and 68% more, respectively, than either reagent alone (Figure 1D-F). When 4 doses of antiNotch2 were combined with both reagents, WBC counts further increased to 46.6x109/L, and LSKs and LKs increased by 2.0-fold and 1.7-fold more in the periphery compared to controls that were stimulated with both haematologica | 2017; 102(10)

Notch2 blockade enhances HSPC egress

A

B

C

D

E

F

Figure 1. Notch2 blocking antibody sensitizes hematopoietic stem cell and progenitors (HSPC) egress in response to granulocyte-colony stimulating factor (G-CSF) or/and AMD3100. Mice were given 4 doses of G-CSF in two days, or a single dose of AMD3100, or 4 doses of G-CSF followed by AMD3100 the next day, two days after a single dose (A-C) or 4 doses of Notch2-blocking or control antibody (D-F) (see details of treatment scheme in Online Supplementary Figure S1). Twenty-four hours (h) after the last dose of G-CSF, or 1 h after AMD3100, peripheral blood (PB) was analyzed for white blood cell (WBC) counts (A and D) and the presence of LSK (B and E) and LK (C and F) cells by FACS in PB (n=4-7/group). Results are pooled from 3 independent experiments and presented as meanÂąStandard Deviation (S.D.). Student t-test *P<0.05, **P<0.01.

reagents but without anti-Notch2 (Figure 1D-F). In comparison, mice receiving Notch1 antibody had no significant increase in HSPCs in the circulation or in the spleen (Online Supplementary Figure S1F-H). Both splenic LSK and LK frequencies were also increased by combining antiNotch2 with either G-CSF or AMD3100, and further increased by combining both reagents (Online Supplementary Figure S1I). We concluded from these observations that: 1) blocking Notch2 but not Notch1 induces enhanced HSPC egress; 2) although Notch2 blockade synergizes with either G-CSF or AMD3100 in promoting HSPC egress, a stronger stimulating effect on WBC counts and mobilized LSKs is achieved when anti-Notch2 is combined with both reagents; while 3) a single dose antiNotch2 in the combined regimen showed a similar effect on increasing WBC and LK numbers as the 4-dose regimen, it induced a lower increment in circulating LSK than the 4-dose regimen.

Notch2 deficiency causes increased HSPC egress To confirm that the observed HSPC egress after Notch2 blockade is indeed caused by the loss of Notch2 signaling, haematologica | 2017; 102(10)

we analyzed mice with Notch2 deficiency in the hematopoietic system using the Vav-Cre/Notch2F/F mice that had efficient deletion of Notch2 expression on HSPCs (Online Supplementary Figure S2). We found that circulating LSK and LK cells increased from 208/mL and 591/mL in control mice to 526/mL and 927/mL in Notch2-deficient mice (Figure 2A), respectively. Spleen-residing LSKs and LKs also increased by 2.4- and 2.0-fold more compared to the control mice (Figure 2B). Moreover, the increased HSPC egress persisted after Notch2-deficient bone marrow cells were transferred into lethally irradiated WT mice (Figure 2C and D). The greater HSPC egress was accompanied by a 46% increase in WBC counts (Figure 2E) and 30% increase in circulating granulocytes, but a 28% relative reduction in circulating T cells at 12 weeks after transplantation (Figure 2F). In addition, there were 1.3-fold and 78% increases in LSKs and LKs in the spleen (Figure 2G and H). In comparison, consistent with other reports, the bone marrow HSPC homeostasis of the de novo Vav-Cre/Notch2F/F mice or of the recipients receiving Notch2-deficient cells remained largely unaffected (Figure 2I and J).12,15 These findings are consistent with the 1787

W. Wang et al.

A

C

B

E

G

H

I

D

F

J

Figure 2. Notch2 deficiency causes increased hematopoietic stem cell and progenitors (HSPC) egress. (A and B) HSPCs present in the periphery (A) and in the spleen (B) of 8-week old Vav-Cre/Notch2F/F mice and control mice (Vav-Cre/Notch2+/+) were determined by FACS. Representative FACS profile (gated on Lin–c-kit+ cells) and bar graphs of total numbers of LSK cells and LK cells in the blood (A) and in the spleen (B) pooled from 3 similar experiments in which Vav-Cre/Notch2F/F and control mice (n=5) were examined. (C-F) Total numbers of LSK cells (C), LK cells (D), white blood cell counts (E), and frequencies of T cells (CD4/CD/8), B cells (B220) and granulocytes (Gr-1) (F) present in the peripheral blood (PB) of recipient mice 12 weeks after receiving bone marrow (BM) transplantation from control (n=6) or VavCre/Notch2F/F mice (n=6). (G-J) Spleen-residing LSK (G) and LK (H) frequencies, as well as BM HSC subpopulations, including LT-HSC (Flt3–CD34–LSK), ST-HSC (Flt3–CD34+LSK) and MPP (Flt3+CD34+LSK) (I), and LK subsets including CMP (Lin–c-kit+Sca–1-CD34+FcγRII/III–), MEP (Lin–c-kit+Sca-1-CD34-FcγRII/III–) and GMP (Lin–c-kit+Sca-1-CD34+FcγRII/III+) (J) were determined by FACS in the same group of transplanted mice. Results are pooled from 2 experiments and presented as mean±Standard Deviation (S.D.). Student t-test *P<0.05, **P<0.01.

increased circulating and spleen-residing HSPCs in mice receiving Notch2 blocking antibody, and suggest that Notch2 signaling loss is responsible for increased HSPC egress.

Notch2 blockade mildly alters bone marrow HSPC homeostasis but does not affect HSC self-renewal It is important to determine if Notch2 blockade would adversely affect HSPC homeostasis and HSC self-renewal and differentiation. We therefore analyzed marrow HSPC populations after 4 doses of Notch2 antibody treatment.

The numbers of marrow total LSK, LK, and common lymphoid progenitor (CLP) cells were decreased modestly in Notch2 antibody-treated mice compared to control-treated mice (Online Supplementary Figure S3A-C). A more detailed analysis of LSK and LK subpopulations revealed a mild reduction of multi-potential progenitors (MPPs), common myeloid progenitors (CMPs), and megakaryocyte-erythroid progenitors (MEPs) in Notch2 antibodytreated mice compared to control mice (Online Supplementary Figure S3D and E). While it is possible that Notch2 blockade caused a transient reduction of the marrow HSPC generation,12 these alterations also are likely reflective of HSPC redistribution by Notch2 blockade, 1788

considering that HSPC numbers were increased in the periphery and in the spleen (Figure 1) and there is no change in HSPC apoptosis (Online Supplementary Figure S3F). On the other hand, we did not observe any significant alterations in any of the HSPC subsets in mice that received Notch2 antibody combined with G-CSF and AMD3100 (Online Supplementary Figure S3G-I). These data suggest that mild HSPC alterations caused by Notch2 blockade can be compensated by the proliferative stimuli through G-CSF and/or AMD3100. We also noted that mice receiving Notch1 antibody had a 32% reduction in CLP cells (Online Supplementary Figure S3C), a finding consistent with other reports that Notch1 is involved in the CLP development in the marrow.9,16 We then assessed the effects of Notch blockade on marrow HSC differentiation and self-renewal by transplanting Notch1- or Notch2-blocked bone marrow cells into lethally irradiated WT mice. On day 10, we observed a transient high WBC count and a higher hemoglobin concentration which continued until ten weeks after transplantation, and higher platelets at four and eight weeks in mice receiving Notch2-blocked marrow cells than in mice receiving control antibody-treated cells (Figure 3A and B) or Notch1-blocked donor cells (Online Supplementary Figure haematologica | 2017; 102(10)

Notch2 blockade enhances HSPC egress

A

B

C

D

E

Figure 3. Faster hematopoietic recovery from Notch2-blocked marrow progenitors after transplantation. (A and B) Platelet counts, white blood cell (WBC) counts, and hemoglobin levels on days 7, 10, 16 and 21 (A) (n=4-5/group), and at 4, 8, 10, and 13 weeks (B) after transplantation were determined (n=6/group, pooled from 2 experiments). (C-E) Bone marrow frequencies of LSK subsets (C), CMP/MEP/GMP cells (D), and CLP cells (E) were determined in the marrow of mice three months after receiving transplantation (n=6-7/group, pooled from 2 experiments). Results are presented as meanÂąStandard Deviation (S.D.). Student t-test *P<0.05, **P<0.01.

S4). We did not find any significant alteration in the numbers of LSK cells, stem cell subpopulations, or CLP in mice receiving Notch2-blocked marrow cells (Figure 3C-E).

However, the MEPs and CMPs derived from Notch2blocked donors, but not from Notch1-blocked donors, were expanded by approximately 92% and 75%, respectively, compared to those derived from control antibodytreated donors (Figure 3D).

We then took marrow cells from the primary recipients and performed secondary transplantation (Online Supplementary Figure S5). We did not observe a significant difference in the frequency of stem cells or progenitors from Notch2-blocked (Online Supplementary Figure S5A and B) or Notch2-deficient marrow in the secondary transplant recipients than from controls (Online Supplementary Figure S5C and D). As expected, Notch1 blockade had no effect on the HSPC homeostasis after primary or secondary haematologica | 2017; 102(10)

transplantation either (Figure 3C-E and Online Supplementary Figure S5B). These findings suggest that transient Notch2 blockade induces a short-term expansion of myeloid progenitors during stress hematopoiesis, such as bone marrow transplantation, resulting in faster reconstitution of platelet and hemoglobin, whereas the more primitive bone marrow stem cells and lymphoid progenitors are unaffected by Notch2 blockade.

Notch2 blockade induces enhanced homing and reconstitution of the stem cells and myeloid progenitors from mobilized HSPCs To determine the reconstitution potential of mobilized and Notch2-blocked HSPC, we used a competitive transplantation assay in which circulating HSPCs (Ly5.2) mobilized by Notch2 blockade (combined with G-CSF and AMD3100) compete with congenic bone marrow cells 1789

W. Wang et al.

A

B

C

D

E

Figure 4. Notch2 blockade induces enhanced reconstitution of the stem cells and myeloid progenitors from mobilized hematopoietic stem cell and progenitors (HSPC). (A) Scheme of competitive reconstitution by mobilized circulating HSPCs in which mononuclear cells from 200 mL blood collected from control antibody or Notch2 antibody-treated mice (Ly5.2) were mixed with 0.4x106 competitor marrow cells (Ly5.1), and transplanted into lethally irradiated recipient mice (Ly5.1). (B) The percentage of donor Ly5.2+ cell chimerism in the peripheral blood (PB) mononuclear cells of recipient mice (n=5-7/group) at various time points after transplantation. (C-E) The percentage of peripheral B cells, T cells and granulocytes (C), bone marrow megakaryocyte-erythroid progenitors (MEPs) and granulocyte-macrophage progenitors (GMPs) (D), and bone marrow total LSK and LT-hematopoietic stem cells (LT-HSCs) (E) derived from Ly5.2+ mobilized HSPCs in recipient mice at various time points after transplantation. Results are presented as meanÂąStandard Deviation (S.D.). Student t-test *P<0.05, **P<0.01. wk: weeks.

(Ly5.1) for engraftment in lethally irradiated recipients (Ly5.1) (Figure 4A). Evaluation of donor reconstitution revealed a trend toward a higher chimerism at three weeks and significantly higher chimerism sustained by HSPCs mobilized by Notch2 blockade at eight and 12 weeks after transplantation (Figure 4B). Granulocyte numbers derived from mobilized and Notch2-blocked HSPCs were significantly higher than those derived from control cells by ten and 13 weeks after transplantation, while lymphocytes contributed by Notch2-blocked HSPCs also showed a trend of increase (Figure 4C). Accordingly, bone marrow GMPs and MEPs derived from Notch2-blocked HSPCs increased compared to those derived from control donors (Figure 4D). Similarly, proportions of LSKs and long-term HSCs (LT-HSCs) derived from Notch2-blocked HSPCs were much higher than those derived from control donors (Figure 4E). These findings suggest a competitive advantage of Notch2-blocked stem cells and myeloid progenitors over controls after transplantation.

The enhanced chimerism from Notch2-blocked mobilized HSPC suggests that Notch2 blockade may enhance HSPC homing and/or engraftment. Because the number of mobilized HSPC was not sufficient to allow a direct imag1790

ing analysis for homing, we assessed the homing and niche locations of adoptively transferred bone marrow HSPCs from Notch2-deficient or from Notch2 antibodytreated mice. We observed that 1.6-fold more of Notch2deficient (Figure 5A) and 1.4-fold more Notch2-blocked progenitors (Online Supplementary Figure S6A) homed to the bone marrow compared to their corresponding controls (P<0.001 by 2-sample Pearson Ď&#x2021;2 test). Notch2-deficient (Figure 5A and C) or Notch2-blocked progenitors (Online Supplementary Figure S6A and C) showed similar spatial locations to the blood vessels as their controls. In comparison, Notch2-deficient progenitors were positioned more distal from the endosteum than control progenitors, with a median distance of 12.6 and 6.7 Âľm, respectively (P<0.001) (Figure 5B). However, there was no great difference between Notch2-blocked progenitors and the controls regarding their mean distance to the endosteum progenitors (Online Supplementary Figure S6B).

Notch2 signaling blockade increases HSPC cell cycling and down-regulates CXCR4 expression To understand the mechanism underlying the enhanced sensitization to mobilizing reagents and an expansion of haematologica | 2017; 102(10)

Notch2 blockade enhances HSPC egress

A

B

C

Figure 5. Notch2 loss enhances hematopoietic stem cell and progenitors (HSPC) homing and leads to altered localizations relative to the endosteum. Isolated bone marrow LK cells (0.2x106) from Vav-Cre/Notch2F/F mice and control mice were labeled with CFSE and transferred into lethally-irradiated wild-type (WT) mice. Twentyfour hours later, 2-photon imaging was performed to locate CFSE+ cells in the calvarium. The endosteum is highlighted by the blue second harmonic signal, while the vessel was labeled by TRITC-Dextran dye. (A) The shortest 3D distances between the LK cells and the blood vessel or the endosteum (mm) were compared for control and Notch2-deficient cells. Wilcoxon rank sum test was performed. Data shown were pooled from 3 mice (3 experiments) in each group. Cell numbers counted in the entire calvarium of each recipient were 30, 63, and 57 (total n=150) derived from Vav-Cre/Notch2F/F mice, and 9, 26, and 22 (total n=57) derived from control (ctrl) mice, in experiments 1, 2 and 3, respectively. (B) Representative 2D images show the locations of control versus Notch2-deficient LK cells relative to the blood vessel. Bar size=100 mm.

myeloid progenitors from Notch2-blocked donors during transplantation, we assessed the cell-cycling status of the marrow HSPCs by the proliferation marker Ki67 in conjunction with the DNA-specific dye 7-AAD as a measure of quiescence. Compared to the HSPCs from mice receiving control antibody, Notch2-blocked LSKs but not HSCs had decreased quiescent cells in G0 fraction but increased proliferative cells in G1 fraction (Figure 6A). A similar reduction of G0 and an increase of G1 LSK cells were observed in Notch2-deficient LSK cells (data not shown).

These findings suggest that Notch2 helps maintain HSPC quiescence. To confirm that the observed reduction in HSPC quiescence following Notch2 blockade is a result of direct effect on the HSPC cells, we applied the antibody in an in vitro co-culture system where isolated LSK cells were co-cultured with Notch ligand DLL4 or JAG1-expressing OP9 cells. In this co-culture system, Notch receptor interaction with Notch ligand in vitro promotes HSPC adhesion to ligand-expressing osteoblasts and inhibits HSPC cycling, an effect that could be blocked by ligand-neutralizing antibody.11 Here we found that treating cells with anti-Notch2 caused a similar reduction in quiescent LSKs haematologica | 2017; 102(10)

but an increase in cycling LSKs in G1 and S/G2/M fractions (Figure 6B).

Because the cell surface chemokine receptor CXCR4 is essential for the colonization of bone marrow by HSPC as well as for the maintenance of stem cell quiescence,17-19 we investigated the expression of CXCR4 on LSK cells from Notch2-blocked bone marrow or Notch2-deficient mice.

Cell surface CXCR4 expressed by Notch2-blocked LSKs decreased by approximately 50% when compared to the controls (Figure 7A). Similarly, CXCR4 expression by Notch2-deficient LSKs was also decreased by approximately 50% when compared to control LSKs (Figure 7B).

These observations raised the possibility that Notch2 signaling directly regulates CXCR4 expression. Indeed, analysis of CXCR4 promoter identified several RBPJ/CSL binding sequences (TGGGAA) (Figure 7C). CHIP analysis revealed that RBPJ binds strongly to motif -6.1kb and weakly to motif -4.1kb but not to motif -1kb (Figure 7D).

In addition, co-transfection of the CXCR4 promoter construct harboring the two RBPJ binding motifs together with NOTCH2 siRNA or RBPJ siRNA resulted in an 1791

W. Wang et al.

A

B

Figure 6. Notch2 signaling blockade increases hematopoietic stem cell and progenitor (HSPC) cell cycling. (A) Cell-cycling status of the marrow LSK cells and CD150+CD48-LSKs hematopoietic stem cells (HSCs) was determined by the proliferation marker Ki67 in conjunction with the DNA-specific dye 7-AAD in mice receiving 4 doses of control or Notch2 blocking antibody. One representative FACS profile of LSK cells, and the relative proportion of cells in G0, G1 and S-G2/M phase of the cell cycle in LSK and HSC (bottom). Results are pooled from 2 experiments, and are presented as averageÂąStandard Deviation (S.D.) (n=5/each group). (B) 1.5x105 LK cells were co-cultured with confluent OP9-DLL4 or OP9-JAG1 cells in the presence of control (ctrl) or Notch2 blocking antibody (400 ng/mL) for four days before cell-cycling analysis on gated LK cells. One representative FACS profile of 3 similar experiments. Student t-test *P<0.05.

approximately 40% and approximately 20% decrease in CXCR4 reporter activity, respectively, compared to the control siRNA treatment (Figure 7E). The reporter activity decreased by 32% and 75% when the promoter construct was transfected with either motif 6.1 (Del6.1) or both motifs 4.1 and 6.1 were deleted (Del4.1/6.1) in the luciferase reporter assay. Furthermore, when cells were transfected with the Del6.1 or the Del4.1/6.1 construct, the reporter activity showed even greater reduction after Notch2 was down-regulated by siRNA (Figure 7F). In comparison, Notch1 or Notch2 overexpression resulted in increased CXCR4 reporter activity; this increased activity was also dependent on the two essential RBPJ binding motifs (Figure 7G). Finally, when bone marrow LK cells were co-cultured with OP9-DLL4 cells, blocking Notch2, but not Notch1, there was a reduction in LK cell surface expression of CXCR4 (Figure 7H). Together, these observations suggest that CXCR4 expression is directly regulated by Notch signaling, and that CXCR4 on marrow HSPC is down-regulated by Notch2 blockade.

Discussion In this study, we demonstrated that blocking Notch2 combined with the current clinical regimen further enhances HSPC mobilization and homing without affecting HSC homeostasis. We found that Notch2 blockade leads to increased HSPC cell cycling and down-regulated CXCR4 expression. We showed that CXCR4 is directly regulated by the Notch transcriptional protein RBPJ. In addition, we showed that Notch2 blockade leads to a competitive repopulating advantage of mobilized HSPC after transplantation. These results suggest that Notch2 may serve as a new target for promoting HSPC mobilization and HSPC engraftment during transplantation.

Notch is a signaling molecule important for stem cell self-renewal and fate determination in many tissues, 1792

including the hematopoietic system. Despite in vitro evidence that activation of Notch stimulates HSC self-renewal,20-23 in vivo studies do not support the concept that Notch has an essential role in adult HSC steady state homeostasis.15,24 Nevertheless, Notch2 was found to be responsible for the rate of generation of repopulating stem cells during stress hematopoiesis in the early phase of hematopoietic recovery.12 Consistently, we found that Notch2 blockade alone is associated with a transient reduction in the bone marrow progenitors and HSCs, likely caused by the decreased immediate replenishment of the HSPCs in the absence of Notch2. However, since blocking Notch2 decreases HSC quiescence and niche retention, Notch2 blockade leads to increased HSPC cell cycling and egress from the marrow. In comparison, blocking Notch1 neither affects HSC quiescence, nor causes HSPC exit from the marrow. Therefore, Notch2 but not Notch1 is important for HSC quiescence maintenance and proper niche localization. There is no significant effect on HSPC homeostasis when Notch2 blockade is employed in conjunction with G-CSF and AMD3100, and no apparent negative effect on short-term or long-term HSC reconstitution when these bone marrow cells are used in the transplantation procedures.

Long-term hematopoiesis is maintained by a small pool of HSCs that ensure balanced proliferation, differentiation, and quiescence. One of the major mechanisms that retains HSPCs in their bone marrow niches and directs their migration and homing from blood to bone marrow involves interaction of the CXCR4 receptor with Îą-chemokine stromal-derived factor 1 (SDF-1). CXCR4 promotes HSC quiescence and blocking CXCR4/SDF-1 signaling hampers HSC retention.18,19,25 The importance of the CXCR4/SDF-1 as a retention mechanism for HSPC is underscored by the extensive efforts to develop mobilization reagents by targeting this axis, e.g. by antagonizing CXCR4,13,26 by downregulation of SDF-1 expression or suppression of SDF-1-producing cells,27,28 or by proteolytic haematologica | 2017; 102(10)

Notch2 blockade enhances HSPC egress

A

B

D

E

C H F

G

Figure 7. Notch2 signaling regulates hematopoietic stem cell and progenitor (HSPC) CXCR4 expression. (A and B) Cell surface expression of CXCR4 in mean fluorescence intensity (MFI) was assessed by FACS analysis of bone marrow LSK cells from mice receiving 4 doses of control (ctrl) or Notch1/2 blocking antibody (n=5/each group) (A), or from Vav-Cre/Notch2F/F mice and control mice (Vav-Cre/Notch2+/+) (n=4/each group) (B). (C) CXCR4 promoter region has 3 potential RBP-J binding sites. (D) CHIP assay of wild-type LK bone marrow cells. Immunoprecipitation was conducted with control antibody, or anti-RBPJ followed by PCR of the CXCR4 promoter. Results shown are mean±Standard Deviation (S.D.) of 3 biological replicates. Student t-test *P<0.05, ***P<0.001. (E) CXCR4-Luc report construct with 2.0 kb CXCR4 promoter sequence containing motif 4.1 and motif 6.1 was transfected into 293T cells expressing RBPJ (RBPJ KD), Notch2 (N2 KD), or control siRNA (ctrl KD). (F) The dependence of the CXCR4-Luc reporter activity on motif 4.1 and 6.1 was assessed by transfecting the CXCR4 reporter construct with single motif 6.1 deleted (Del6.1) or both motif 4.1 and 6.1 deleted (Ddel4.1/6.1) into 293T cells that expressed control siRNA or Notch2 siRNA. (G) CXCR4-Luc reporter activity was determined by transfecting the wild-type CXCR4-Luc reporter construct (WT) or the construct with both 4.1- and 6.1-motif deleted (DD) into 293T cells expressing ICN1-expression plasmid (ICN1 OE), ICN2-expression plasmid (ICN2 OE), or control plasmid (empty vector; EV). (H) Bone marrow WT LK cells were isolated and cocultured with OP9-DLL4 cells for 96 hours in the presence of Notch1, Notch2, or control antibody (400 ng/mL). CXCR4 expression was assessed by FACS. (E-H) Mean±S.D of 3 biological replicates. Student t-test *P<0.05, **P<0.01, ***P<0.001.

cleavage of CXCR4 and SDF-1.29,30 Other potential agents and novel strategies to promote HPC mobilization have been tested.31 These include, but are not limited to, S1P agonists,32 VCAM/VLA-4 inhibitors,33 parathyroid hormone,34 proteosome inhibitors,35 Groβ,36,37 CD26,38 and heparan sulphate.39 Here, we report that blocking Notch2 combined with G-CSF and AMD3100 induces enhanced HSPC mobilization, and that Notch signaling directly regulates CXCR4 expression. Because we did not find any significant alterations in the expression of other major cell surface adhesion molecules caused by the lack of Notch2 expression (Online Supplementary Figure S7), presumably this down-regulated CXCR4 is responsible, at least partially, for the increased cell cycling and egress of HSPCs in Notch2-deficient mice or after Notch2 blockade. Other possible mechanisms that should not be excluded may involve a direct adhesive effect by Notch2-ligand interaction with stromal niche cells independent of CXCR4 in promoting HSC niche retention, as Notch ligand neutralizing antibodies are still able to induce HSPC egress even in mice deficient in RBPJ-mediated canonical Notch signaling.11 Interestingly, we find that Notch2-blocked or Notch2deficient HSPCs display enhanced homing to the bone haematologica | 2017; 102(10)

marrow. This is accompanied with increased repopulating ability in peripheral HSPCs mobilized by Notch2 blockade, enhanced hematopoietic recovery, and expansion of the GMPs and MEPs derived from Notch2-blocked cells.

These properties are likely due to the increased proliferative capacity and rapid differentiation associated with Notch2 deficiency itself12 or with downregulation of CXCR4 by Notch2-blockade, as a similar observation was found in CXCR4 haplo-insufficient HSPCs.40 Alternatively, a transient downregulation of CXCR4 by Notch2 antibody during HSPC egress is followed by a compensatory increase in CXCR4 that mediates the increased homing after HSPCs leave the bone marrow.

However, the insufficient numbers of mobilized HSCs mean a precise assessment of CXCR4 expression cannot be made. In addition, this argument cannot be applied to Notch2 deficient mice. We do not exclude the possibility that the enhanced homing of Notch2-blocked/deficient HSPCs is mediated by other molecules that are directly or indirectly regulated by Notch2. Consistent with other reports,12 we found that, unlike Notch2-blocked HSPCs, Notch2-deficient HSPCs showed similar or a mild reduction in chimerism in the competitive setting when compared to the control HSPCs (Online Supplementary Figure 1793

W. Wang et al.

S8), and displayed a trend of reduction in frequency in the secondary transplant recipients (Online Supplementary Figure S5E and F). This indicates that HSC function related to stress hematopoiesis is somewhat impaired by Notch2 deficiency despite increased homing. This paradoxical finding may be explained by a distal spatial location of Notch2-deficient HSPC relative to the endosteum; however, the underlying cause needs further investigation.

Endothelial cells41-43 and perivascular stromal cells19,44 are important components of HSC niches. A direct contact of HSCs with endothelial cells through Notch-ligand interactions has been shown to prevent HSC exhaustion.42 On the other hand, we showed that HSPCs with faulty Notch-ligand interaction display a similar distal spatial location relative to the endosteum.11 Whether this reflects a disrupted interaction between HSPCs and the Notch ligand-expressing supporting cells in the quiescent niche or in the vicinity of endosteum, or is caused by altered cell metabolism, or some other still unidentified mechanism, needs further investigation. Importantly, transient Notch2 blockade does not affect HSC self-renewal, hence it is unlikely that transplant of Notch2-blocked HSPCs will adversely affect HSC long-term reconstitution in recipients.

The more effective HSPC mobilization and the enhanced engraftment related to Notch2 blockade may be in general useful as a strategy to achieve satisfactory engraftment and improve patient outcome in HCT. Nearly all autologous HCT procedures are performed using mobilized HPCs. Successful outcome is dependent on infusing an adequate number of functionally active HPCs. Until recently, a mobilization strategy of G-CSF, either alone or in combination with chemotherapy, failed to mobilize an ‘optimal’ CD34 cell dose in up to 40% of patients.45 AMD3100, in combination with G-CSF, increases total

References 1. Copelan EA. Hematopoietic stem-cell transplantation. N Engl J Med. 2006;354 (17):1813-1826.

2. Moog R. Management strategies for poor peripheral blood stem cell mobilization.

Transfus Apher Sci. 2008;38(3):229-236.

3. Perseghin P, Terruzzi E, Dassi M, et al.

Management of poor peripheral blood stem cell mobilization: incidence, predictive factors, alternative strategies and outcome. A retrospective analysis on 2177 patients from three major Italian institutions. Transfus Apher Sci. 2009;41(1):33-37.

4. Jantunen E, Kvalheim G. Mobilization strategies in hard-to-mobilize patients with lymphoid malignancies. Eur J Haematol.

2010;85(6):463-471.

5. Moreb JS, Salmasinia D, Hsu J, et al. LongTerm Outcome after Autologous Stem Cell Transplantation with Adequate Peripheral Blood Stem Cell Mobilization Using Plerixafor and G-CSF in Poor Mobilizer Lymphoma and Myeloma Patients. Adv Hematol. 2011;2011:517561.

6. Stiff PJ, Micallef I, Nademanee AP, et al.

Transplanted CD34(+) cell dose is associated with long-term platelet count recovery following autologous peripheral blood stem cell transplant in patients with nonHodgkin lymphoma or multiple myeloma.

1794

7.

8.

9.

10.

11.

12.

13.

CD34+ cells mobilized, and is often used for HPC mobilization in myeloma and non-Hodgkin lymphoma patients.46 Unfortunately, this approach does not always lead to adequate HPC collection in patients who have previously been exposed to extensive cytotoxic therapy; furthermore, in some instances, multiple HCT procedures may be contemplated that require greater numbers of HPCs.5,47,48 Since findings from this study were performed using a ‘standard’ animal model, an optimal regimen needs to be better defined in order to achieve the maximum mobilization in humans. Nevertheless, our findings from animal studies suggest that Notch2 is a potential target to be considered for the mobilizing regimen in conjunction with AMD3100 and G-CSF for patients thought to be potentially at risk of not being able to mobilize sufficient numbers of CD34 cells. This strategy, as the data suggest, could improve collection efficacy and reduce procedure-related costs and the risks associated with large volume or multiple collections.49,50 Potentially, this approach could accelerate hematopoietic recovery, considering the enhanced homing and engraftment of the HSPCs and increased granulocyte recovery which could potentially reduce the risk of infection, and ultimately improve both short- and long-term patient outcome.

Acknowledgments We thank Dr. Hal Broxmeyer for valuable suggestions about our experimental design.

Funding This study was supported in part by research funding from NIH HL103827 and American Cancer Society LIB-125064 to LZ, and by the Department of Pathology Case Western Reserve University faculty startup fund to WX and LZ.

Biol Blood Marrow Transplant. 2011;17 (8):1146-1153.

Fehon RG, Kooh PJ, Rebay I, et al.

Molecular interactions between the protein products of the neurogenic loci Notch and Delta, two EGF-homologous genes in Drosophila. Cell. 1990;61(3):523-534.

Rebay I, Fleming RJ, Fehon RG, et al.

Specific EGF repeats of Notch mediate interactions with Delta and Serrate: implications for Notch as a multifunctional receptor. Cell. 1991;67(4):687-699.

Oh P, Lobry C, Gao J, et al. In vivo mapping of notch pathway activity in normal and stress hematopoiesis. Cell Stem Cell. 2013; 13(2):190-204.

Yao D, Huang Y, Huang X, et al. Protein Ofucosyltransferase 1 (Pofut1) regulates lymphoid and myeloid homeostasis through modulation of Notch receptor ligand interactions. Blood. 2011;117(21):5652-5662.

Wang W, Yu S, Zimmerman G, et al. Notch Receptor-Ligand Engagement Maintains Hematopoietic Stem Cell Quiescence and Niche Retention. Stem Cells. 2015;33 (7):2280-2293.

Varnum-Finney B, Halasz LM, Sun M, et al.

Notch2 governs the rate of generation of mouse long- and short-term repopulating stem cells. J Clin Invest. 2011;121(3):12071216.

Broxmeyer HE, Orschell CM, Clapp DW, et

14.

15.

16.

17.

18.

19.

al. Rapid mobilization of murine and human hematopoietic stem and progenitor cells with AMD3100, a CXCR4 antagonist.

J Exp Med. 2005;201(8):1307-1318.

Wu Y, Cain-Hom C, Choy L, et al.

Therapeutic antibody targeting of individual Notch receptors. Nature. 2010; 464(7291):1052-1057.

Maillard I, Koch U, Dumortier A, et al.

Canonical notch signaling is dispensable for the maintenance of adult hematopoietic stem cells. Cell Stem Cell. 2008;2(4):356-366.

Yu VW, Saez B, Cook C, et al. Specific bone cells produce DLL4 to generate thymusseeding progenitors from bone marrow. J Exp Med. 2015;212(5):759-774.

Broxmeyer HE, Kohli L, Kim CH, et al.

Stromal cell-derived factor-1/CXCL12 directly enhances survival/antiapoptosis of myeloid progenitor cells through CXCR4 and G(alpha)i proteins and enhances engraftment of competitive, repopulating stem cells. J Leukoc Biol. 2003;73(5):630638.

Nie Y, Han YC, Zou YR. CXCR4 is required for the quiescence of primitive hematopoietic cells. J Exp Med. 2008;205(4):777-783.

Sugiyama T, Kohara H, Noda M, et al.

Maintenance of the hematopoietic stem cell pool by CXCL12-CXCR4 chemokine signaling in bone marrow stromal cell niches. Immunity. 2006;25(6):977-988.

haematologica | 2017; 102(10)

Notch2 blockade enhances HSPC egress

20. Duncan AW, Rattis FM, DiMascio LN, et al.

Integration of Notch and Wnt signaling in hematopoietic stem cell maintenance. Nat Immunol. 2005;6(3):314-322.

21. Karanu FN, Murdoch B, Gallacher L, et al.

The notch ligand jagged-1 represents a novel growth factor of human hematopoietic stem cells. J Exp Med. 2000; 192(9):1365-1372.

22. Stier S, Cheng T, Dombkowski D, et al.

Notch1 activation increases hematopoietic stem cell self-renewal in vivo and favors lymphoid over myeloid lineage outcome.

Blood. 2002;99(7):2369-2378.

23. Varnum-Finney B, Xu L, Brashem-Stein C, et al. Pluripotent, cytokine-dependent, hematopoietic stem cells are immortalized by constitutive Notch1 signaling. Nat Med. 2000;6(11):1278-1281.

24. Mancini SJ, Mantei N, Dumortier A, et al.

Jagged1-dependent Notch signaling is dispensable for hematopoietic stem cell selfrenewal and differentiation. Blood. 2005; 105(6):2340-2342.

25. Tzeng YS, Li H, Kang YL, et al. Loss of Cxcl12/Sdf-1 in adult mice decreases the quiescent state of hematopoietic stem/progenitor cells and alters the pattern of hematopoietic regeneration after myelosuppression. Blood. 2011;117(2):429-439.

26. Abraham M, Biyder K, Begin M, et al.

Enhanced unique pattern of hematopoietic cell mobilization induced by the CXCR4 antagonist 4F-benzoyl-TN14003. Stem Cells. 2007;25(9):2158-2166.

27. Christopher MJ, Liu F, Hilton MJ, et al.

Suppression of CXCL12 production by bone marrow osteoblasts is a common and critical pathway for cytokine-induced mobilization.

Blood. 2009;114(7):1331-1339.

28. Winkler IG, Barbier V, Wadley R, et al.

Positioning of bone marrow hematopoietic and stromal cells relative to blood flow in vivo: serially reconstituting hematopoietic stem cells reside in distinct nonperfused niches. Blood. 2010;116(23):375-385.

29. Greenbaum AM, Link DC. Mechanisms of G-CSF-mediated hematopoietic stem and progenitor mobilization. Leukemia. 2011; 25(2):211-217.

30. Levesque JP, Hendy J, Takamatsu Y, et al.

Disruption of the CXCR4/CXCL12 chemotactic interaction during hematopoietic stem cell mobilization induced by GCSF or

haematologica | 2017; 102(10)

31.

32.

33.

34.

35.

36.

37.

38.

39.

cyclophosphamide. J Clin Invest. 2003; 111(2):187-196.

Hopman RK, DiPersio JF. Advances in stem cell mobilization. Blood Rev. 2014;28(1):3140.

Ratajczak MZ, Lee H, Wysoczynski M, et al. Novel insight into stem cell mobilization-plasma sphingosine-1-phosphate is a major chemoattractant that directs the egress of hematopoietic stem progenitor cells from the bone marrow and its level in peripheral blood increases during mobilization due to activation of complement cascade/membrane attack complex.

Leukemia. 2010;24(5):976-985.

Papayannopoulou T, Nakamoto B. Peripheralization of hemopoietic progenitors in primates treated with anti-VLA4 integrin. Proc Natl Acad Sci USA. 1993; 90(20):9374-93748.

Brunner S, Zaruba MM, Huber B, et al.

Parathyroid hormone effectively induces mobilization of progenitor cells without depletion of bone marrow. Exp Hematol.

2008;36(9):1157-1166.

Niesvizky R, Mark TM, Ward M, et al.

Overcoming the response plateau in multiple myeloma: a novel bortezomib-based strategy for secondary induction and highyield CD34+ stem cell mobilization. Clin Cancer Res. 2013;19(6):1534-1546.

King AG, Horowitz D, Dillon SB, et al.

Rapid mobilization of murine hematopoietic stem cells with enhanced engraftment properties and evaluation of hematopoietic progenitor cell mobilization in rhesus monkeys by a single injection of SB-251353, a specific truncated form of the human CXC chemokine GRObeta. Blood. 2001; 97(6):1534-1542.

Fukuda S, Bian H, King AG, et al. The chemokine GRObeta mobilizes early hematopoietic stem cells characterized by enhanced homing and engraftment. Blood.

2007;110(3):860-869.

Broxmeyer HE, Hangoc G, Cooper S, et al.

AMD3100 and CD26 modulate mobilization, engraftment, and survival of hematopoietic stem and progenitor cells mediated by the SDF-1/CXCL12-CXCR4 axis. Ann N Y Acad Sci. 2007;1106:1-19.

Saez B, Ferraro F, Yusuf RZ, et al. Inhibiting stromal cell heparan sulfate synthesis improves stem cell mobilization and

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

enables engraftment without cytotoxic conditioning. Blood. 2014;124(19):29372947.

McDermott DH, Gao JL, Liu Q, et al.

Chromothriptic cure of WHIM syndrome.

Cell. 2015;160(4):686-699.

Kiel MJ, Yilmaz OH, Iwashita T, et al.

SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells.

Cell. 2005;121(7):1109-1121.

Butler JM, Nolan DJ, Vertes EL, et al.

Endothelial cells are essential for the selfrenewal and repopulation of Notch-dependent hematopoietic stem cells. Cell Stem Cell. 2010;6(3):251-264.

Ding L, Saunders TL, Enikolopov G, et al.

Endothelial and perivascular cells maintain haematopoietic stem cells. Nature. 2012; 481(7382):457-462.

Sacchetti B, Funari A, Michienzi S, et al.

Self-renewing osteoprogenitors in bone marrow sinusoids can organize a hematopoietic microenvironment. Cell.

2007;131(2):324-336.

Pusic I, Jiang SY, Landua S, et al. Impact of mobilization and remobilization strategies on achieving sufficient stem cell yields for autologous transplantation. Biol Blood Marrow Transplant. 2008;14(9):1045-1056.

Mohty M, Hubel K, Kroger N, et al.

Autologous haematopoietic stem cell mobilisation in multiple myeloma and lymphoma patients: a position statement from the European Group for Blood and Marrow Transplantation. Bone Marrow Transplant.

2014;49(7):865-872.

Moskowitz CH, Glassman JR, Wuest D, et al. Factors affecting mobilization of peripheral blood progenitor cells in patients with lymphoma. Clin Cancer Res. 1998; 4(2):311-316.

Keating GM. Plerixafor: a review of its use in stem-cell mobilization in patients with lymphoma or multiple myeloma. Drugs.

2011;71(12):1623-1647.

Michon B, Moghrabi A, Winikoff R, et al.

Complications of apheresis in children.

Transfusion. 2007;47(10):1837-1842.

Bojanic I, Dubravcic K, Batinic D, et al.

Large volume leukapheresis: Efficacy and safety of processing patient's total blood volume six times. Transfus Apher Sci. 2011; 44(2):139-147.

1795

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Complications in Hematology

Ferrata Storti Foundation

Haematologica 2017 Volume 102(10):1796-1805

Characterization of atrial fibrillation adverse events reported in ibrutinib randomized controlled registration trials

Jennifer R. Brown,1 Javid Moslehi,2 Susan O’Brien,3 Paolo Ghia,4 Peter Hillmen,5 Florence Cymbalista,6 Tait D. Shanafelt,7 Graeme Fraser,8 Simon Rule,9 Thomas J. Kipps,10 Steven Coutre,11 Marie-Sarah Dilhuydy,12 Paula Cramer,13 Alessandra Tedeschi,14 Ulrich Jaeger,15 Martin Dreyling,16 John C. Byrd,17 Angela Howes,18 Michael Todd,19 Jessica Vermeulen,20 Danelle F. James,21 Fong Clow,21 Lori Styles,21 Rudy Valentino,21 Mark Wildgust,19 Michelle Mahler19 and Jan A. Burger22

Dana-Farber Cancer Institute, Boston, MA, USA; 2Division of Cardiovascular Medicine and Cardio-Oncology Program Vanderbilt School of Medicine, Nashville, TN, USA; 3Chao Family Comprehensive Cancer Center, University of California, Irvine, Orange, CA, USA; 4 Università Vita-Salute San Raffaele and IRCCS Istituto Scientifico San Raffaele, Milano, Italy; 5CA Leeds Teaching Hospitals, St. James Institute of Oncology, Leeds, UK; 6Hôpital Avicenne, AP-HP, UMR Paris13/INSERM U978, Bobigny, France; 7Mayo Clinic, Rochester, MN, USA; 8Juravinski Cancer Centre, McMaster University, Hamilton, ON, Canada; 9 Department of Haematology, Plymouth University Medical School, Plymouth, UK; 10 Moores UCSD Cancer Center, San Diego, CA, USA; 11Stanford University School of Medicine and Stanford Cancer Institute, Stanford, CA, USA; 12Hôpital Haut-Lévêque, Bordeaux, Pessac, France; 13Department I of Internal Medicine and German CLL Study Group, University of Cologne, Germany; 14Azienda Ospedaliera Niguarda Cà Granda, Milano, Italy; 15Medical University of Vienna, Austria; 16Department of Medicine III, Klinikum der Ludwig-Maximilians-Universität München, Campus Grosshadern, Germany; 17 Ohio State University Comprehensive Cancer Center, Columbus, OH, USA; 18Janssen Research & Development, High Wycombe, UK; 19Janssen Research & Development, LLC, Raritan, NJ, USA; 20Janssen Research & Development, LLC, Leiden, the Netherlands; 21 Pharmacyclics, Sunnyvale, CA, USA and 22Leukemia Department, University of Texas MD Anderson Cancer Center, Houston, TX, USA 1

ABSTRACT

Correspondence: jbrown2@partners.org

Received: April 17, 2017.

Accepted: July 18, 2017.

Pre-published: July 27, 2017. doi:10.3324/haematol.2017.171041 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/10/1796 ©2017 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode.

Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1796

T

he first-in-class Bruton’s tyrosine kinase inhibitor ibrutinib has proven clinical benefit in B-cell malignancies; however, atrial fibrillation (AF) has been reported in 6-16% of ibrutinib patients. We pooled data from 1505 chronic lymphocytic leukemia and mantle cell lymphoma patients enrolled in four large, randomized, controlled studies to characterize AF with ibrutinib and its management. AF incidence was 6.5% [95% Confidence Interval (CI): 4.8, 8.5] for ibrutinib at 16.6months versus 1.6% (95%CI: 0.8, 2.8) for comparator and 10.4% (95%CI: 8.4, 12.9) at the 36-month follow up; estimated cumulative incidence: 13.8% (95%CI: 11.2, 16.8). Ibrutinib treatment, prior history of AF and age 65 years or over were independent risk factors for AF. Multiple AF events were more common with ibrutinib (44.9%; comparator, 16.7%) among patients with AF. Most (85.7%) patients with AF did not discontinue ibrutinib, and more than half received common anticoagulant/antiplatelet medications on study. Low-grade bleeds were more frequent with ibrutinib, but serious bleeds were uncommon (ibrutinib, 2.9%; comparator, 2.0%). Although the AF rate among older non-trial patients with comorbidities is likely underestimated by this dataset, these results suggest that AF among clinical trial patients is generally manageable without ibrutinib discontinuation (clinicaltrials.gov identifier: 01578707, 01722487, 01611090, 01646021).

Introduction Clinical trials of ibrutinib have demonstrated consistent benefits and improvements in overall survival (OS) and progression-free survival (PFS) among patients with chronic lymphocytic leukemia (CLL)/small lymphocytic lymphoma (SLL) and relapsed or refractory mantle cell lymphoma (MCL), including those with high-risk haematologica | 2017; 102(10)

Pooled AF analysis in ibrutinib studies

disease features.1-4 Ibrutinib therapy is generally well tolerated, but has been associated with atrial fibrillation (AF), with an overall clinical trial incidence of 6-16%.1-6 In a phase II study that enrolled 86 patients with CLL/SLL, incidence of AF reached 16% in longer-term follow up.5 In a meta-analysis,7 the pooled rate of AF was 3.3 [95% Confidence Interval (CI): 2.5, 4.1] per 100 person-years in ibrutinib-treated patients versus 0.84 (95%CI: 0.32, 1.6) per 100 person-years in non-ibrutinib-treated patients.

However, risk factors, natural history, and management strategies of AF associated with ibrutinib treatment are largely unknown. Because continuous treatment is required to maintain benefit from single-agent ibrutinib therapy, understanding a patientâ&#x20AC;&#x2122;s natural history and optimizing AF management should improve the safe use of ibrutinib in B-cell malignancies. Management of AF typically relies on rate and/or rhythm control, depending on underlying structural cardiovascular disease (CVD).8-10 Systemic thromboembolic events (specifically stroke) are the most frequent major complication of AF, along with other cardiovascular (CV) complications and increased mortality.11,12 Anticoagulation, commonly with vitamin K antagonists, reduces the risk of stroke by approximately two-thirds, while increasing bleeding risks.

Thus, risk calculators (i.e. CHA2DS2-VASC) have been developed to weigh benefits against risks of anticoagulation in patients without an underlying malignancy.

Most patients with CLL/SLL and non-Hodgkin lymphomas are diagnosed at 65 years or over and have multiple medical comorbidities.5,13,14 It has been reported that 6% of patients aged 65 years or over and diagnosed with CLL/SLL had AF at baseline (higher than the 1.0-1.8% of an age-matched general population), and 6% more developed AF over a 5-year treatment period,13-15 suggesting that patients with CLL/SLL may have a higher risk of developing AF than the normal population.

Data regarding management of AF in ibrutinib-treated patients are limited, and the association between ibrutinib therapy and increased rates of bleeding needs to be considered in the context of AF management in these patients. To further characterize ibrutinib-associated AF and describe management thereof, we report here a pooled analysis of all cases of AF across four randomized controlled trials (RCTs).

Methods Study populations from initial data reports of the four RCTs [PCYC-1112 (RESONATE, clinicaltrials.gov identifier: 01578707), PCYC-1115 (RESONATE-2, clinicaltrials.gov identifier: 01722487), CLL3001 (HELIOS, clinicaltrials.gov identifier: 01611090), and MCL3001 (RAY, clinicaltrials.gov identifier: 01646021)] were pooled, including patients randomized to receive ibrutinib [alone or with bendamustine plus rituximab (BR)] and patients receiving comparator therapy (ofatumumab, chlorambucil, placebo plus BR, or temsirolimus). The studies were approved by the institutional review board or independent ethics committee at each institution.

Data from the initial study reports were used for the detailed pooled analysis. Limited data on incidence of AF were available in the extended follow-up period and were included for patients randomized to ibrutinib. Patients were subject to similar eligibility criteria; specifically, patients requiring vitamin K antagonists, such as warfarin, or strong CYP3A4/5 inhibitors were excluded, although other anticoagulants and antiplatelet agents were permithaematologica | 2017; 102(10)

ted. Patients were censored at crossover. Patients were excluded if they had uncontrolled or clinically significant CVD, including uncontrolled arrhythmia or class III or IV congestive heart failure or a history of myocardial infarction, unstable angina, or acute coronary syndrome within six months prior to randomization.

Complete study methodologies are detailed elsewhere (Online Supplementary Appendix 1).1-4 The incidences of AF and atrial flutter events were referred to collectively as AF. All treatment-emergent AF events (defined as events occurring after first dose of study drug until 30 days after last dose) are reported. CV events captured using Standardised Medical Dictionary for Regulatory Activities (MedDRA) Queries (SMQ) were grouped into five CVD categories: arrhythmia, congestive heart failure, ischemic heart disease, hypertension, and ischemic CNS vascular conditions (Online Supplementary Appendix 2). Bleeding events listed were captured using SMQ. CHA2DS2-VASc scores estimating risk of stroke in patients with AF were evaluated using patientsâ&#x20AC;&#x2122; characteristics at baseline. Cox regression models were used to perform univariate and multivariate analyses of risk factors for developing AF. These analyses evaluated age increase, gender (male), AF risk increase (prior history of AF/flutter), increase in body mass index, Rai stage and prior history of AF/abnormal heart rhythm, coronary artery disease, diabetes, hyperlipidemia, hypertension, and valvular heart disease.13 The univariate model included each single factor plus treatment group. Cumulative incidence of AF was estimated in a Cox regression model accounting for deaths and disease progression without prior AF as competing risk events.5,16 The AF risk score for each CLL patient with no prior history of AF was computed using the Shanafelt predictive model.13

Results Patientsâ&#x20AC;&#x2122; characteristics and incidence of AF In total, 1505 patients were included, with 756 randomized to ibrutinib (alone or with BR) and 749 to comparator (Table 1). One hundred thirty-nine patients had previously treated MCL and the remainder had newly diagnosed or previously treated CLL. At the time of the initial study reports, the median follow up in the pooled analysis was 16.6 months; median duration of exposure was 13.3 months for the ibrutinib group and 5.8 months for the comparator (Online Supplementary Table S1). With a median follow up of 16.6 months, 6.5% (95%CI: 4.8, 8.5) of patients receiving ibrutinib and 1.6% (95%CI: 0.8, 2.8) receiving the comparator [relative risk 4.1 (95%CI: 2.2, 7.5)] reported AF while on treatment (Figure 1). Most AF events developed de novo in patients without a history of AF. The incidence of AF was 7.0% (95%CI: 5.1, 9.3) in CLL patients and 4.3% (95%CI: 1.6, 9.2) in MCL patients treated with ibrutinib. Patients treated with ibrutinib combination therapy (HELIOS study) had a 7.7% (95%CI: 4.9, 11.4) incidence of AF, compared with 5.8% (95%CI: 3.8, 8.3) in ibrutinib monotherapy patients. The exposureadjusted incidence rates of AF per 100 patient-months were 0.503 for the ibrutinib group and 0.199 for the comparator. The estimated cumulative incidence of AF was higher in patients treated with ibrutinib versus comparators [7.4% (95%CI: 5.6, 9.6) vs. 1.9% (95%CI:1.0, 3.4)] (Figure 2A and B). Median age of patients developing AF was 71 years for both groups, which is older than the overall median age of 67 years. History of prior AF/abnormal heart rhythm was more common in patients who had AF on study (ibrutinib, 26.5%; comparator, 25.0%) than in 1797

J.R. Brown et al.

Table 1. Baseline demographic and clinical characteristics of patients in the pooled analysis.

All patients (n=756) Median age, years (range) <65, n (%) 65-75, n (%) >75, n (%) Male, n (%) Ethnicity (White), n (%) BMI, n (%)a â&#x2030;¤18 >18-24.9 25-29.9 â&#x2030;Ľ30 Anticoagulant at baseline, n (%) Antiplatelet at baseline, n (%) Prior history for patients, n (%) AF/abnormal heart rhythm Coronary artery disease Diabetes Hyperlipidemia Hypertension Infection

Ibrutinib Patients with AF (n=49)

All patients (n=749)

Comparator Patients with AF (n=12)

67.0 (30-89) 274 (36.2) 324 (42.9) 158 (20.9) 508 (67.2) 670 (88.6)

71.0 (59-84) 9 (18.4) 26 (53.1) 14 (28.6) 33 (67.3) 47 (95.9)

67.0 (34-90) 288 (38.5) 331 (44.2) 130 (17.4) 506 (67.6) 686 (91.6)

71.0 (58-88) 1 (8.3) 8 (66.7) 3 (25.0) 9 (75.0) 12 (100)

6 (0.8) 284 (37.6) 276 (36.5) 166 (22.0) 41 (5.4) 131 (17.3)

0 17 (34.7) 21 (42.9) 9 (18.4) 4 (8.2) 17 (34.7)

6 (0.8) 256 (34.2) 288 (38.5) 166 (22.2) 41 (5.5) 148 (19.8)

0 2 (16.7) 3 (25.0) 7 (58.3) 2 (16.7) 2 (16.7)

88 (11.6) 41 (5.4) 98 (13.0) 49 (6.5) 328 (43.4) 289 (38.2)

13 (26.5) 3 (6.1) 7 (14.3) 8 (16.3) 31 (63.3) 26 (53.1)

80 (10.7) 36 (4.8) 122 (16.3) 48 (6.4) 327 (43.7) 300 (40.1)

3 (25.0) 0 5 (41.7) 1 (8.3) 10 (83.3) 5 (41.7)

AF: atrial fibrillation; n: number; BMI: body mass index. aTwenty-four patients with ibrutinib and 33 with placebo had missing BMI at baseline.

patients who did not (ibrutinib, 10.6%; comparator, 10.4%). Patients with a history of hypertension were more likely to develop AF than those without [31 of 328 (9.5%) vs. 18 of 428 (4.2%)] in the ibrutinib group. The majority of patients with prior hypertension did not develop clinically evident AF on ibrutinib (ibrutinib, 90.5%; comparator, 96.9%) during the observation period. In patients without a history of hypertension, 38 developed de novo hypertension; only one patient developed de novo hypertension and AF. Longer-term follow up in patients randomized to ibrutinib provided an additional 8467 patient-months for analysis. During this period, 29 additional patients experienced AF. Newly reported cases of AF occurred at a continuous low rate over time. With extended follow up, 78 ibrutinibtreated patients [10.4% (95%CI: 8.4, 12.9)] experienced AF. Estimated cumulative incidence rate of AF at 36 months was 13.8% (95%CI: 11.2, 16.8) (Figure 2C). After adjusting for competing risks of progressive disease and death, estimated cumulative incidence rate of AF was 11.2% (95%CI: 9.0, 13.8) (Figure 2D).

Clinical features of treatment-emergent AF In the first six months, 5.3% of ibrutinib patients developed AF with a continued low rate over time. The median time to onset of AF was 2.8 months (range 0.3-17.5) for the ibrutinib group and 2.0 months (range 0.6-18.9) for the comparator, with a median follow up of 16.6 months. In 2 patients in the ibrutinib group and 4 in the comparator, an AF event occurred after the patient had permanently dis1798

continued study drug (within 30 days) for other reasons.

Overall, median duration of AF episodes was three days for both groups; however, the range varied widely. The mean (SD) duration of AF episodes was 12.6 (29.5) days for the ibrutinib group and 5.1 (5.5) days for comparator.

The majority of patients experiencing AF had only one episode [27 of 49 (55.1%) for ibrutinib; 10 of 12 (83.3%) for comparator] (Online Supplementary Table S2); 22 patients (44.9%) in the ibrutinib group had multiple episodes and 2 patients (16.7%) in the comparator had two episodes (Online Supplementary Tables S2 and S3).

Among patients who had two or more AF episodes in the ibrutinib group, the median time between events was 1.1 months.

Common Toxicity Criteria grade 1 or 2 AF occurred in 27 (3.6%) patients in the ibrutinib group and 8 (1.1%) patients in the comparator group, accounting for more than half of the AF events that occurred in either group (Online Supplementary Appendix 3). AF events leading to hospitalization (including grade 3 and 4 events) were reported as serious adverse events (SAEs) in 23 (3.7%) patients receiving ibrutinib and 6 (1.0%) receiving comparator. Among these SAEs, 17 patients in the ibrutinib group and 3 patients in the comparator group reported grade 3 events. Only one grade 4 event was reported, which was in the ibrutinib group. No deaths were attributed to AF in either group.

With extended follow up, the median time to onset of AF in patients randomized to ibrutinib was 5.7 (range 0.340.2) months. Of the 78 patients with AF, almost twohaematologica | 2017; 102(10)

Pooled AF analysis in ibrutinib studies

Figure 1. Onset of first atrial fibrillation event by treatment.

thirds [49 (62.8%)] had only one episode of AF and more than half [43 (55.1%)] had AF events of grade 2 or lower (Online Supplementary Table S4).

Predictors of AF in trial patients Univariate analyses identified prior history of AF, ibrutinib therapy, age over 65 years, hypertension, and hyperlipidemia as significant risk factors for developing AF. Multivariate analyses showed prior history of AF, ibrutinib therapy, and age over 65 years as independent predictors of AF (Figure 3). The influence of prior coronary artery disease, valvular heart disease, and diabetes were also evaluated and not identified as significant risk factors for developing AF while on ibrutinib.

In CLL patients without a history of AF who were treated with ibrutinib, the incidence and risk of de novo AF increased with Shanafelt risk score category (Table 2 and Online Supplementary Figure S1). Estimated 5-year de novo AF rates were 0.4% in category 0-1, 2.8% in category 2-3, 7.6% in category 4, and 17.9% in category â&#x2030;Ľ5.

Management of study therapy concurrent with AF events Twenty-four of 49 patients with AF in the ibrutinib (49.0%) and 4 of 12 in the comparator (33.3%) group were managed without any interruption or modification of study drug. Patients who required dose modification or interruption were slightly older; all other baseline demographic and clinical characteristics were similar between the two groups (Online Supplementary Table S5).

No patient in either group had a dose reduction attributed to AF; however, a similar proportion in each group had dose interruptions due to AF [16 of 49 (32.7%) for ibrutinib; 4 of 12 (33.3%) for comparator]. The median duration of interruption was 11 days for ibrutinib and 17 days for comparator. With the caveat of small numbers, there was no statistical difference between the 18-month PFS rate in 6 patients with AF who had ibrutinib dose interruption for seven days or more versus those with dose interruption for fewer than seven days [66.7% (95%CI: haematologica | 2017; 102(10)

19.5, 90.4) vs. 71.4% (95%CI: 54.0, 83.2)]. Seven of 49 (14.3%) patients in the ibrutinib group and no patients in the comparator group discontinued study treatment due to AF. Approximately one-half of patients with multiple AF events had dose interruptions (Online Supplementary Figure S2), and 5 of 22 (22.7%) discontinued. Plots of AF events, dose interruptions, and concomitant therapy for individual patients with AF are found in the Online Supplementary Figure S2. Of ibrutinib patients with AF and extended follow up, approximately half [41 of 78 (52.6%)] were managed without dose reduction or interruption of study treatment (Online Supplementary Table S4).

Medical management of AF Atrial fibrillation was primarily managed with treatment commonly used for rate and rhythm control, with the most frequently used agents digoxin [11 of 49 (22.4%)], bisoprolol [10 of 49 (20.4%)], and amiodarone [8 of 49 (16.3%)] in the ibrutinib group, and amiodarone [4 of 12 (33.3%)] and diltiazem [3 of 12 (25.0%)] in the comparator group (Online Supplementary Table S6A). In the ibrutinib group, 2 patients were managed using cardioversion and one had a pacemaker inserted due to concomitant bradycardia. Patients did not have serial electrocardiographic monitoring to evaluate return to normal sinus rhythm; however, 36 of 49 patients receiving ibrutinib (73.5%) and 9 of 12 receiving comparator (75.0%) had their AF events reported as recovered or resolved.

More than one-third [17 of 49 (34.7%)] of ibrutinib patients who had an AF event on study were taking antiplatelet medications at study entry, and 4 (8.2%) were taking an anticoagulant. In the comparator group, 2 patients each were on antiplatelet and anticoagulant medications (Table 1). Many patients who experienced AF received concomitant anticoagulant and/or antiplatelet medication; the most commonly used antiplatelet agent was aspirin, and the most commonly used anticoagulant was low-molecular-weight heparin (Online Supplementary Table S6A and B). Among ibrutinib-treated patients who received anticoagulant/antiplatelet medications at any 1799

J.R. Brown et al.

Table 2. Incidence of de novo atrial fibrillation (AF) by Shanafelt risk score category13 in chronic lymphocytic leukemia patients with no history of AF treated with ibrutinib.

Risk score category

N. patients (%) n=588

N. patients with AF (%)

HR (95% CI)

Estimated 5-year AF ratea % (95%CI)

Estimated 5-year AF rate % (95%CI)13,b,c

0-1

282 (48.0)

10 (3.5)

Ref

2-3

209 (35.5)

16 (7.7)

4

71 (12.1)

6 (8.5)

â&#x2030;Ľ5

26 (4.4)

4 (15.4)

2.294 (1.016-5.183) 2.997 (1.089-8.250) 3.289 (0.868-12.467)

0.40 (0.1-2.8) 2.8 (1.2-6.5) 7.6 (2.9-19.2) 17.9 (6.8-42.1)

1.8 (0.8-2.8) 4.1 (2.5-5.6) 8.0 (4.8-11.0) 17.2 (10.3-23.5)

HR: Hazards Ratio; CI: Confidence Interval; N, n: number; Ref: reference. aEstimated by Cox regression model for time to acquired AF from diagnosis. Median follow up on study for the patients was 15.4 months. Median follow-up time from diagnosis was 89.1 months. bMedian follow up was 7.3 years. cRisk score was the sum of the risk values across each factor independently associated with AF, older age (65-74: 2, â&#x2030;Ľ75+: 3), male gender (1), valvular heart disease (2), and hypertension (1).

Table 3. Rates of bleeding in ibrutinib and comparator patient groups in terms of the presence of atrial fibrillation (AF) and use of antiplatelet and/or anticoagulant agents.

N. (%) Any bleeding event Grade 1/2 events Grade 3/4 events Antiplatelet/anticoagulant use in patients with bleeding event of any gradea Anticoagulant medication Antiplatelet medication Anticoagulant and antiplatelet medications given in combination No anticoagulant or antiplatelet medication

Patients with AF (n=49)

Ibrutinib (n=756) Patients All without AF patients (n=707) (n=756)

25/49 (51.0) 24/49 (49.0) 1/49 (2.0) n=25

268/707 293/756 (37.9) (38.8) 244/707 268/756 (34.5) (35.4) 21/707 (3.0) 22/756 (2.9) n=268 n=293

Patients with AF (n=12)

Comparator (n=749) Patients without AF (n=737)

All patients (n=749)

2/12 (16.7) 1/12 (8.3) 1/12 (8.3) n=2

127/737 (17.2) 113/737 (15.3) 14/737 (1.9) n=127

129/749 (17.2) 114/749 (15.2) 15/749 (2.0) n=129

18/127 (14.2) 19/127 (15.0) 3/127 (2.4)

18/129 (14.0) 19/129 (14.7) 3/129 (2.3)

93/127 (73.2)

95/129 (73.6)

13/25 (52.0) 6/25 (24.0) 4/25 (16.0)

27/268 (10.1) 37/268 (13.8) 3/268 (1.1)

40/293 (13.7) 43/293 (14.7) 7/293 (2.4)

0

10/25 (40.0)

210/268 (78.4)

220/293 (75.1)

2/2 (100)

0 0

N, n: number. aBleeding events are counted if they are between seven days prior and one day post-administration of anticoagulant/antiplatelet.

time on study, median treatment duration was longest for aspirin and novel oral anticoagulants (Online Supplementary Table S6A and B and Online Supplementary Figure S2).

Patients with AF tended to have higher CHA2DS2-VASc scores regardless of study treatment, with 35 of 49 ibrutinib patients and 10 of 12 comparator patients having scores â&#x2030;Ľ2 (Online Supplementary Table S7). All 35 ibrutinib patients received anticoagulation: 23 (65.7%) received an anticoagulant alone and 12 (34.3%) received both anticoagulant and antiplatelet medications. All 10 patients in the comparator group received an anticoagulant and 20.0% of those also received an antiplatelet agent.

1800

Bleeding events with the use of anticoagulant and antiplatelet medications Overall, 38.8% (293 of 756) of patients in the ibrutinib group and 17.2% (129 of 749) of patients in the comparator experienced a bleeding event; most in each group were grade 1 or 2 (91.5% and 88.3%, respectively). Bleeding occurred irrespective of whether patients experienced AF (Table 3). Of the patients with AF, 25 of 49 (51.0%) in the ibrutinib group had bleeding events; timing of the bleeding events relative to onset of AF was not evaluated.

Bleeding events were grade 1 or 2 in severity in all patients in the ibrutinib group (Online Supplementary Figure S3). In haematologica | 2017; 102(10)

Pooled AF analysis in ibrutinib studies

A

B

C

D

Figure 2. Cumulative incidence (95% CI) of atrial fibrillation with ibrutinib. (A) unadjusted for competing risks (death and progressive disease) and (B) adjusted. With extended follow up: unadjusted (C) and adjusted (D).

the ibrutinib group, 10 of 25 (40.0%) patients had not received anticoagulant/antiplatelet medication. Nine (36.0%) patients were receiving a single anticoagulant/antiplatelet medication at the time of the bleeding event (aspirin, n=1; novel oral anticoagulants, n=5; low-molecular-weight heparin, n=3), and 6 (24%) were receiving at least two medications (2 of whom received three). All 35 patients with CHA2D22-VASc scores â&#x2030;Ľ2 who developed AF on ibrutinib received anticoagulant/antiplatelet medication.

During the study, 12 (34.3%) of these patients had a grade 1 or 2 bleeding event, 7 of whom were on a single anticoagulant/antiplatelet medication and 5 of whom were on two or more medications.

Three bleeding events resulted in death, all in the ibrutinib group (ruptured abdominal aortic aneurism, subdural hematoma, post-procedural hemorrhage); one patient was on aspirin at the time of the fatal bleed, the other 2 patients had not received concomitant anticoagulants/antiplatelets, while none of the patients had AF while on treatment (Online Supplementary Table S8). haematologica | 2017; 102(10)

Clinical sequelae of AF Cardiovascular disease clinical sequelae were captured using MedDRA SMQ and grouped into five CVD categories: arrhythmia, congestive heart failure, ischemic heart disease, hypertension, and ischemic CNS vascular conditions (Online Supplementary Appendixn 2). Among patients who had a single AF episode, these CVD clinical sequelae were seen with similar frequency in both groups: 5 of 27 (18.5%) of patients in the ibrutinib group compared with 2 of 10 (20.0%) patients in the comparator (Online Supplementary Table S9). Thirteen patients on ibrutinib who had multiple AF episodes, and both patients on comparator, developed clinical CVD sequelae; 9 (69.2%) and 2 (100%) patients, respectively, had a history of one of these conditions. One comparator-treated patient with an AF event had an ischemic CNS vascular condition within the observation time; no ibrutinib-treated patients had an ischemic CNS vascular condition.

Given that clinical complications of AF can occur in the absence of clinically symptomatic AF, incidences of cardiovascular events (as defined above) were also evaluated in the full cohort. Hypertension was the only group 1801

J.R. Brown et al.

Figure 3. Significant factors for development of atrial fibrillation using univariate and multivariate Cox regression. HR: Hazards Ratio; CI: Confidence Interval.

term that occurred on study at a significantly higher rate in the ibrutinib group compared with the comparator (Table 4). Furthermore, CLL patients experiencing AF on ibrutinib had similar PFS duration as patients who did not (Figure 4).

Discussion Ibrutinib has shown a highly favorable benefit-risk ratio for patients with CLL/SLL and relapsed or refractory MCL, albeit with certain side effects including AF. To date, the risk factors, natural history, therapeutic management, and outcomes of ibrutinib-related AF have not been well characterized. In this pooled analysis of four RCTs, with a median follow up of 16.6 months, the incidence of AF in patients treated with ibrutinib was 6.5% (95%CI: 4.8, 8.5%). The incidence of AF was 10.4% (95%CI: 8.4, 12.9) with additional follow up, which is relatively consistent with prior clinical studies and independent reports.5,13,17-20 The incidence of AF was highest in the first six months, and then continued at a low rate. Multivariate analysis showed that use of ibrutinib, prior history of AF, and age over 65 years were associated with a higher risk of AF. Older patients, in general, have a higher propensity for CVD including AF, so it is not surprising that the patients developing AF in this pooled analysis were older than the overall study population. In addition, although history of AF was a predictor of AF in this pooled analysis, we noted that 85.2% of patients with a history of AF did not have a recurrence while being treated with ibrutinib at a median follow up of 16.6 months. However, the small number of patients with AF and the exclusion of patients with significant cardiac disease from the clinical trials may limit the 1802

interpretation of the findings in other patient populations.

Given the rate of AF is highest in the first six months of ibrutinib therapy and may be higher in older, non-trial populations, patients should be carefully monitored for signs of AF, particularly if they are older or have a preexisting history.

In a separate analysis investigating only those patients on ibrutinib without a prior history of AF, we investigated the utility of the Shanafelt risk score in estimating the likelihood of developing AF among our CLL study population on ibrutinib, using a similar methodology as the original report.13 In the Shanafelt predictive model, older age, male gender, valvular heart disease, and hypertension were identified as risk factors independently associated with de novo AF at the time of CLL diagnosis.13 We observed a similar increased risk of de novo AF among patients with higher score categories, which may be useful for counseling patients. The estimated 5-year AF rates were similar to those observed in the original report.13 Current recommendations for management of AF in patients on ibrutinib indicate that therapy should be interrupted for any new onset or worsening grade â&#x2030;Ľ3 nonhematologic toxicity, including AF. Once symptoms have resolved to grade 1 or baseline, treatment may be reinitiated at the starting dose. For patients who develop AF, physicians should follow the appropriate guidelines for clinical management of AF. In this pooled analysis, the approach to AF management was heterogeneous and included practices in accordance with the ibrutinib protocol recommendations and prescribing information as well as CV standard of care for AF.21,22 Approximately half of patients developing AF on ibrutinib were able to be managed without dose interruption or modificiation. In a recent report by Thompson et al., patients with CLL who haematologica | 2017; 102(10)

Pooled AF analysis in ibrutinib studies

Figure 4. Progression-free survival in patients with and without atrial fibrillation (AF).

Table 4. Cardiovascular (CV) events occurring while on therapy listed by MedDRA SMQ grouping.a

Higher level term, n (%)

Ibrutinib (n=756)

Comparator (n=749)

P

Hypertension Congestive heart failure Ischemic cardiac disease Ischemic CNS vascular conditions Arrhythmia

71 (9.4) 151 (20.0) 17 (2.2) 10 (1.3) 55 (7.3)

26 (3.5) 163 (21.8) 17 (2.3) 6 (0.8) 46 (6.1)

<0.0001 0.7118 0.4802 0.3044 0.1148

SMQ: Standardised MedDRA Queries; n: number; CNS: central nervous system. aCV events captured using MedDRA SMQ were grouped into five cardiovascular disease categories (Online Supplementary Appendix 2).

had ibrutinib interrupted at the the onset of AF had an inferior PFS compared with that seen in patients who continued ibrutinib or had dose reductions.23 In this analysis, interrupting ibrutinib therapy for seven days or more in the context of AF did not appear to significantly impact 18-month PFS; given the limited sample size, however, these findings should be interpreted with caution. The majority of the patients did not discontinue ibrutinib due to AF so limited data were available on patients receiving an alternative therapy due to AF in our study. Among 7 patients who discontinued ibrutinib, one patient received subsequent chemoimunotherapy with BR. However, there are a number of agents in the CLL landscape that provide alternative options for CLL patients who may need to discontinue ibrutinib treatment due to AF, including novel agents like idelalisib and venetoclax. A few reports have been published on patient outcomes after ibrutinib discontinuation due to AEs, and suggest that outcomes are better than those seen in the setting of disease progression.24 The approach to medical management of AF in patients receiving ibrutinib should take into account the potential risk of pharmacokinetic interactions with commonly used anticoagulant/antiplatelet medications. US prescribing information for ibrutinib currently recommends a dose reduction to 140 mg daily for co-administration with moderate CYP3A/4 inhibitors.21,22 At the time the studies haematologica | 2017; 102(10)

in this pooled analysis were conducted, however, dose interruption or modification was at the physicianâ&#x20AC;&#x2122;s discretion, and many patients continued on full-dose ibrutinib while receiving moderate CYP3A/4 inhibitors to manage AF, specifically amiodarone and diltiazem. It was beyond the scope of the current analysis to determine whether increased toxicity resulted from this approach; however, current practice and prudence would dictate avoidance of CYP3A/4 inhibitors if possible or dose reduction of ibrutinib consistent with the prescribing information.21,22 In vitro studies suggest that ibrutinib induces platelet aggregation defects due to the inhibition of Brutonâ&#x20AC;&#x2122;s tyrosine kinase and TEC in the glycoprotein VI collagen-activated pathway,25,26 and therefore concomitant use of anticoagulants or aspirin with ibrutinib could enhance bleeding risk. In the current analysis, serious bleeding events occurred in 2.9% of patients treated with ibrutinib overall and in 2.0% of patients with AF. Due to the heterogeneous approach to AF management, it was difficult to characterize the impact of specific medication combinations on the risk of bleeding; however, the very low incidence of grade â&#x2030;Ľ3 bleeding events, even among patients receiving more than one anticoagulant/antiplatelet agent, is reassuring. These results, however, are in contrast to a recent report of real-world experience with 56 patients who developed AF on ibrutinib, in whom a 14% rate of major bleeding was seen.6 These results suggest that more 1803

J.R. Brown et al.

data are needed on the specific risk factors for major bleeds in ibrutinib-treated patients, and until those are available, additional caution and monitoring are warranted in clinical practice, particularly for older patients with comorbidities who are likely to have a higher bleeding risk.6 Notably, in this analysis, no thromboembolic events were observed among ibrutinib-treated patients developing AF. It will be of interest for future studies to determine whether ibrutinib use itself has sufficient antiplatelet activity to confer some of the benefits of aspirin or other antiplatelet agents.

In this pooled analysis of relatively young and healthy clinical trial subjects, we did not find an increased rate of congestive heart failure, ischemic cardiac disease, stroke, or other arrhythmias among patients with AF in the ibrutinib group relative to the comparator. This finding should be interpreted with caution given the small sample size and limited follow up; however, it is notable given that patients with AF generally had higher rates of comorbidities like hypertension and hyperlipidemia at study entry.

When we evaluated the study population as a whole (regardless of the presence of symptomatic AF), we noted an increase in the incidence of de novo and recurrent or ongoing hypertension in patients treated with ibrutinib.

Patients with a history of hypertension were more likely to develop AF; however, in patients without a history of hypertension, 38 of 428 patients developed de novo hypertension but only one patient developed de novo hypertension and AF, suggesting that at least so far, these events are not highly correlated. Ibrutinib therapy has been associated with hypertension in clinical trials, making it difficult to discern if the signal we saw was a direct treatment effect or clinical sequelae related to subclinical AF in a subset of patients.

This analysis has inherent limitations, particularly in its post hoc assessment of completed studies, which focused primarily on oncological rather than CV outcomes. Patient numbers in certain subgroups were low, despite having access to data from four large RCTs, and the inclusion of patients with MCL who received a different dose of ibrutinib and have different disease biology, as well as the inclusion of patients treated with both ibrutinib alone or in combination with BR, may impact on the interpretation of the findings. The method of capturing AEs and concomitant medications limited our ability to evaluate dose intensity, sequencing of anticoagulation in patients with

References 1. Dreyling M, Jurczak W, Jerkeman M, et al.

Ibrutinib versus temsirolimus in patients with relapsed or refractory mantle-cell lymphoma: an international, randomised, open-label, phase 3 study. Lancet.

2016;387(10020):770-778.

2. Burger JA, Tedeschi A, Barr PM, et al.

Ibrutinib as initial therapy for patients with chronic lymphocytic leukemia. N Engl J Med. 2015;373(25):2425-2437.

3. Chanan-Khan A, Cramer P, Demirkan F, et al. Ibrutinib combined with bendamustine and rituximab compared with placebo, bendamustine, and rituximab for previous-

1804

AF, and the temporal relationship between AF and bleeding events. Furthermore, due to study exclusion criteria related to certain serious comorbidities, patients on the clinical trials were undoubtedly healthier than most treated in general practice. Given these limitations, this study may underestimate the incidence of AF among older patients treated outside a clinical trial setting with ibrutinib. A recently published retrospective study of CLL patients treated at several cancer centers found that AF persisted in 62% of 56 ibrutinib-treated patients despite AF-directed therapy.6 Three episodes of cardiac failure, one stroke, and major bleeding events in 14% of patients were observed in that study.6 Algorithm-based guidelines have been proposed to manage ibrutinib-associated AF but have not yet been validated.27 Additional larger datasets, perhaps population-based, will be required to determine representative rates of AF with ibrutinib in different patient groups, to better characterize the incidence of AF-related complications, and to evaluate the value of proposed guidelines outside of a clinical trial setting.

Prudence dictates that clinicians consider the benefitrisk profile of ibrutinib therapy in patients with a history of AF or other predisposing risk factors. Results of this pooled analysis of more than 1500 patients in four RCTs suggest that, with appropriate vigilance and monitoring, the majority of patients with known risk factors for AF can be safely treated with ibrutinib. Alternative treatment options are available for those who discontinued ibrutinib due to AF. However, most patients who develop AF on treatment will not require treatment discontinuation and many can be managed safely with commonly used anticoagulant/antiplatelet medications. Prospective clinical studies focusing on detailed evaluation of the cardiac effects of ibrutinib are warranted to further elucidate the potential mechanisms of AF.28,29 Acknowledgments The authors thank the patients, families, caregivers, research nurses, study co-ordinators and support staff who contributed to all of the studies.

Funding This analysis was sponsored by Janssen Research & Development, LLC and Pharmacyclics, LLC. Medical writing and editorial assistance was provided by PAREXEL International and was funded by Janssen Global Services, LLC.

ly treated chronic lymphocytic leukaemia or small lymphocytic lymphoma (HELIOS): a randomised, double-blind, phase 3 study.

Lancet Oncol. 2016;17(2):200-211.

4. Byrd JC, Brown JR, O'Brien S, et al.

Ibrutinib versus ofatumumab in previously treated chronic lymphoid leukemia. N Engl J Med. 2014;371(3):213-223.

5. Farooqui M, Valdez J, Soto S, Bray A, Tian X, Wiestner A. Atrial fibrillation in CLL/SLL patients on ibrutinib. Presented at: American Society of Hematology (ASH) 57th Annual Meeting; December 5-8, 2015; Orlando, FL. 6. Thompson PA, Levy V, Tam CS, et al. Atrial fibrillation in CLL patients treated with ibrutinib. An international retrospective

study. Br J Haematol. 2016;175(3):462-466.

7. Leong DP, Caron F, Hillis C, et al. The risk of atrial fibrillation with ibrutinib use: a systematic review and meta-analysis.

Blood. 2016;128(1):138-140.

8. Camm AJ, Kirchhof P, Lip GY, et al.

Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Eur Heart J. 2010;31(19):2369-2429.

9. January CT, Wann LS, Alpert JS, et al. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on practice

haematologica | 2017; 102(10)

Pooled AF analysis in ibrutinib studies

10.

11.

12.

13.

14.

15.

guidelines and the Heart Rhythm Society.

Circulation. 2014;130(23):2071-2104.

Lip GY, Nieuwlaat R, Pisters R, Lane DA, Crijns HJ. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: the euro heart survey on atrial fibrillation. Chest.

2010;137(2):263-272.

Andersson T, Magnuson A, Bryngelsson IL, et al. All-cause mortality in 272,186 patients hospitalized with incident atrial fibrillation 1995-2008: a Swedish nationwide long-term case-control study. Eur Heart J. 2013;34(14):1061-1067.

Leong DP, Eikelboom JW, Healey JS, Connolly SJ. Atrial fibrillation is associated with increased mortality: causation or association? Eur Heart J. 2013;34(14):1027-1030.

Shanafelt TD, Parikh SA, Noseworthy PA, et al. Atrial fibrillation in patients with chronic lymphocytic leukemia (CLL). Leuk Lymphoma. 2017;58(7):1630-1639.

Barrientos JC, Meyer N, Song X, Rai KR. Characterization of atrial fibrillation and bleeding risk factors in patients with chronic lymphocytic leukemia (CLL): a population-based retrospective cohort study of administrative medical claims data in the United States (US). Presented at: American Society of Hematology (ASH) 57th Annual Meeting; December 5-8, 2015; Orlando, FL. Benjamin EJ, Wolf PA, D'Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the

haematologica | 2017; 102(10)

16.

17.

18.

19.

20.

21.

22.

23.

Framingham Heart Study. Circulation.

1998;98(10):946-952.

Fine JP, Gray RJ. Proportional hazards model for the subdistribution of a competing risk. J Am Statistical Assoc.

1999;94:496-509.

Byrd JC, Furman RR, Coutre SE, et al.

Three-year follow-up of treatment-naive and previously treated patients with CLL and SLL receiving single-agent ibrutinib.

Blood. 2015;125(16):2497-2506.

O'Brien S, Jones JA, Coutre SE, et al.

Ibrutinib for patients with relapsed or refractory chronic lymphocytic leukaemia with 17p deletion (RESONATE-17): a phase 2, open-label, multicentre study. Lancet Oncol. 2016;17(10):1409-1418.

Wang ML, Blum KA, Martin P, et al. Longterm follow-up of MCL patients treated with single-agent ibrutinib: updated safety and efficacy results. Blood. 2015; 126(6):739-745.

Wang ML, Lee H, Chuang H, et al. Ibrutinib in combination with rituximab in relapsed or refractory mantle cell lymphoma: a single-centre, open-label, phase 2 trial. Lancet Oncol. 2016;17(1):48-56.

IMBRUVICA (ibrutinib) [US prescribing information]. Horsham, PA: Janssen Biotech, Inc.; 2016.

IMBRUVICA (ibrutinib) [summary of product characteristics]. Beerse, Belgium: Janssen Pharmaceutica NV; 2016.

Thompson PA, Levy V, Tam CS, et al. The impact of atrial fibrillation on subsequent

24.

25.

26.

27.

28.

29.

survival of patients receiving ibrutinib as treatment of chronic lymphocytic leukemia (CLL): an international study. Blood.

2015;126(23)3301.

Mato AR, Nabhan C, Barr PM, et al.

Outcomes of CLL patients treated with sequential kinase inhibitor therapy: a real world experience.

Blood.

2016; 128(18):2199-2205.

Kamel S, Horton L, Ysebaert L, et al.

Ibrutinib inhibits collagen-mediated but not ADP-mediated platelet aggregation.

Leukemia. 2015;29(4):783-787.

Levade M, David E, Garcia C, et al.

Ibrutinib treatment affects collagen and von Willebrand factor-dependent platelet functions. Blood. 2014;124(26):3991-3995.

Vrontikis A, Carey J, Gilreath JA, Halwani A, Stephens DM, Sweetenham JW. Proposed algorithm for managing ibrutinibrelated atrial fibrillation. Oncology (Williston Park). 2016;30(11):pii:219802.

McMullen JR, Boey EJ, Ooi JY, Seymour JF, Keating MJ, Tam CS. Ibrutinib increases the risk of atrial fibrillation, potentially through inhibition of cardiac PI3K-Akt signaling.

Blood. 2014;124(25):3829-3830.

Yang T, Moslehi JJ, Roden DM. Proarrhythmic effects of ibrutinib, a clinically approved inhibitor of Bruton's tyrosine kinase (BTK) used in cancer therapy [abstract]. Circulation. 2015;132(Suppl 3):A14587.

1805

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

The origin of a name that reflects Europe’s cultural roots.

Ancient Greek

aÂma [haima] = blood a·matow [haimatos] = of blood lÒgow [logos]= reasoning

Scientific Latin

haematologicus (adjective) = related to blood

Scientific Latin

haematologica (adjective, plural and neuter, used as a noun) = hematological subjects

Modern English

The oldest hematology journal, publishing the newest research results.

2016 JCR impact factor = 7.702

Haematologica, as the journal of the European Hematology Association (EHA), aims not only to serve the scientific community, but also to promote European cultural identify.