Page 1


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 Maggio 2017.


haematologica calendar of events

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

6th Training Course for Paediatricians & Paediatric Nurses on HSCT in Children and Adolescents The European Group for Blood and Marrow Transplantation (EBMT) Chairs: P Bader, M Ifversen June 8-10, 2017 Copenhagen, Denmark

27 Regional Congress of the ISBT The International Society of Blood Transfusion (ISBT) Chairs: J Georgsen, MB Hansen, E van der Schoot, E Wood June 17-21, 2017 Copenhagen, Denmark th

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

2nd EBMT International Transplant Course The European Group for Blood and Marrow Transplantation (EBMT) Chairs: M Mohty, J Kuball, R Duarte September 8-10, 2017 Barcelona, Spain

13th Educational Course of the Lymphoma Working Party on "Treatment of Malignant Lymphoma: State-of-the-Art and Role of Stem Cell Transplantation" The European Group for Blood and Marrow Transplantation (EBMT) Chairs: S Montoto, A Sureda, M Trneny September 21-22, 2017 Prague, Czech Republic

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

EHA Tutorial on Biology and Management of Myeloid Malignancies October 20-22, 2017 Yerevan, Armenia

Russian Onco-Hematology Society's Conference on Malignant Lymphoma - Joint Symposium October 25-26, 2017 Moscow, Russian Federation

Turkish Society of Hematology - EHA Joint Symposium November 1 - 4, 2017 Antalya, Turkey

Argentinian Society of Hematology - EHA Joint Education Day November 17 - 18, 2017 Mar del Plata, Argentina

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

Calendar of Events updated on May 4, 2017


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

Table of Contents Volume 102, Issue 6: June 2017 Cover Figure Hematology (image generated by www.somersault1824.com).

Editorials 967

Myelodysplasia in younger adults: outlier or unique molecular entity? David A. Sallman and Eric Padron

968

Immunoglobulin genes in chronic lymphocytic leukemia: key to understanding the disease and improving risk stratification Lesley-Ann Sutton et al.

972

Unmet needs in the scientific approach to older patients with lymphoma Dominique Bron et al.

Articles Red Cell Biology & its Disorders

976

Family cord blood banking for sickle cell disease: a twenty-year experience in two dedicated public cord blood banks Hanadi Rafii et al.

984

Histone deacetylase 6 regulates cytokinesis and erythrocyte enucleation through deacetylation of formin protein mDia2 Xuehui Li et al.

995

Cdk6 contributes to cytoskeletal stability in erythroid cells Iris Z. Uras et al.

Platelet Biology & its Disorders

1006

Macrothrombocytopenia and dense granule deficiency associated with FLI1 variants: ultrastructural and pathogenic features Paul Saultier et al.

Bone Marrow Failure

1017

An abnormal bone marrow microenvironment contributes to hematopoietic dysfunction in Fanconi anemia Yuan Zhou et al.

Myelodysplastic Syndromes

1028

Molecular features of early onset adult myelodysplastic syndrome Cassandra M. Hirsch et al.

Myeloproliferative Disorders

1035

The clinical and molecular diversity of mast cell leukemia with or without associated hematologic neoplasm Mohamad Jawhar et al.

Acute Myeloid Leukemia

1044

Higher HOPX expression is associated with distinct clinical and biological features and predicts poor prognosis in de novo acute myeloid leukemia Chien-Chin Lin et al.

Haematologica 2017; vol. 102 no. 6 - June 2017 http://www.haematologica.org/


haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation 1054

Targeted therapy for a subset of acute myeloid leukemias that lack expression of aldehyde dehydrogenase 1A1 Maura Gasparetto et al.

1066

Impact of ABO incompatibility on patients’ outcome after haploidentical hematopoietic stem cell transplantation for acute myeloid leukemia - a report from the Acute Leukemia Working Party of the EBMT Jonathan Canaani et al.

Acute Lymphoblastic Leukemia

1075

Targeting the 5T4 oncofetal glycoprotein with an antibody drug conjugate (A1mcMMAF) improves survival in patient-derived xenograft models of acute lymphoblastic leukemia Owen J. McGinn et al.

Chronic lymphocytic Leukemia

1085

Distinct molecular genetics of chronic lymphocytic leukemia in Taiwan: clinical and pathogenetic implications Shang-Ju Wu et al.

Non-Hodgkin Lymphoma

1091

Relevance of ID3-TCF3-CCND3 pathway mutations in pediatric aggressive B-cell lymphoma treated according to the non-Hodgkin lymphoma-Berlin-Frankfurt-MĂźnster protocols Marius Rohde et al.

Plasma Cell Disorders

1099

Prognostic impact of circulating plasma cells in patients with multiple myeloma: implications for plasma cell leukemia definition Miquel Granell et al.

1105

Monitoring multiple myeloma by next-generation sequencing of V(D)J rearrangements from circulating myeloma cells and cell-free myeloma DNA Anna Oberle et al.

Cell Therapy & Immunotherapy

1112

A risk factor analysis of outcomes after unrelated cord blood transplantation for children with Wiskott-Aldrich syndrome Zhanna Shekhovtsova e al.

Complications in Hematology

1120

Human rhinovirus detection in the lower respiratory tract of hematopoietic cell transplant recipients: association with mortality Sachiko Seo et al.

Letters to the Editor Letters are available online only at www.haematologica.org/content/102/6.toc

e216

Mutational analysis in serial marrow samples during azacitidine treatment in patients with post-transplant relapse of acute myeloid leukemia or myelodysplastic syndromes Janghee Woo et al. http://www.haematologica.org/content/102/6/e216

e219

No correlation of intensity of phlebotomy regimen with risk of thrombosis in polycythemia vera: evidence from European Collaboration on Low-Dose Aspirin in Polycythemia Vera and Cytoreductive Therapy in Polycythemia Vera clinical trials Tiziano Barbui et al. http://www.haematologica.org/content/102/6/e219

Haematologica 2017; vol. 102 no. 6 - June 2017 http://www.haematologica.org/


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

e222

Molecular landscape of acute promyelocytic leukemia at diagnosis and relapse Annette Fasan et al. http://www.haematologica.org/content/102/6/e222

e225

Acute lymphoblastic leukemia with aleukemic prodrome: preleukemic dynamics and possible mechanisms of immunosurveillance Olga Zimmermannova et al. http://www.haematologica.org/content/102/6/e225

e229

Survival of patients with lymphoplasmacytic lymphoma and solitary plasmacytoma in Germany and the United States of America in the early 21st century Janick Weberpals et al. http://www.haematologica.org/content/102/6/e229

e233

Demystification of enigma on antigen-presenting cell features of human basophils: data from secondary lymphoid organs Emmanuel Stephen-Victor et al. http://www.haematologica.org/content/102/6/e233

Case Report Case Reports are available online only at www.haematologica.org/content/102/6.toc

e238

Ruxolitinib, a potent JAK1/JAK2 inhibitor, induces temporary reductions in the allelic burden of concurrent CSF3R mutations in chronic neutrophilic leukemia Arief S. Gunawan et al. http://www.haematologica.org/content/102/6/e238

Haematologica 2017; vol. 102 no. 6 - June 2017 http://www.haematologica.org/


EDITORIALS Myelodysplasia in younger adults: outlier or unique molecular entity? David A. Sallman and Eric Padron Malignant Hematology Department, H. Lee Moffitt Cancer Center, Tampa, FL, USA E-mail: eric.padron@moffitt.org

M

doi:10.3324/haematol.2017.165993

yelodysplastic syndrome (MDS) is a disease of the elderly with a median age at diagnosis of 71 years and a sharp increase in incidence reported after the sixth decade of life.1 This age-associated increase in incidence also applies to pre-MDS and pre-leukemic states such as clonal hematopoiesis of indeterminate potential, clonal cytopenias of uncertain significance, and idiopathic cytopenias of uncertain significance, with 10% of individuals over 65 years of age harboring these conditions while they are extremely rare in individuals less than 40 years of age.2-4 Despite this, MDS can sporadically affect younger adults and occurs rarely in the pediatric population. Although evidence suggests that pediatric MDS, compared to its adult counterpart, is associated with a distinct pathophysiology hallmarked by an increased incidence of hereditary syndromes and disparate mutational landscape,5,6 whether MDS in younger adults is a unique molecular entity or simply a continuum of “classic adult disease� is unknown. Comprehensive molecular annotation of MDS and secondary acute myeloid leukemia has identified somatic variants in the vast majority of patients, with these having a significant impact on diagnosis, prognosis and treatment selection.7-9 Using this evidence as a benchmark, Hirsch and colleagues present data on 634 patients with MDS, MDS/myeloproliferative neoplasms (including chronic myelomonocytic leukemia) or secondary acute myeloid leukemia and investigate the landscape of mutations based on age at presentation.10 The authors utilized whole exome sequencing or a targeted next-generation sequencing panel of 60 genes representing the most common myeloid neoplasm-associated mutations. As expected, the age distribution of the cohort was unimodal with early onset MDS defined as that occurring in patients less than 50 years of age (10% of cohort). Classified according to World Health Organization criteria, patients with early onset MDS more frequently had refractory anemia with excess blasts. Targeted next-generation sequencing and/or whole exome sequencing both demonstrated an increasing number of mutations, independently of World Health Organization classification, associated with age by both linear correlation and average median number of mutations. Interestingly, the increased mutational complexity was also independent of cytogenetic complexity, which was not significantly different based on age. The evaluation of differences in gene mutation frequencies between younger and older adults with MDS showed that the older patients had a higher incidence of mutations in genes associated with the spliceosome and epigenetic regulator families, data which correlate strongly with those in the literature on aging clonal hematopoiesis.7-9 Specifically, TET2 and SRSF2 mutations were significantly more common in older patients, a finding corroborating recent investigations using ultra-deep sequencing in a cohort of 4,000 patients in which SRSF2 mutations were observed exclusively in patients greater than 70 years of age.11 Together, these data and those presented by Hirsch et al. suggest that spliceosome haematologica | 2017; 102(6)

mutations appear to be particularly associated with an aging hematopoietic environment. In a similar analysis of acute myeloid leukemia in the elderly (>65 years of age, n=100), Silva and colleagues also showed spliceosome (SRSF2 in 23% of patients) and epigenetic modifiers (TET2 and ASXL1 in 24% and 21% of patients, respectively) to be the most commonly mutated genes in their cohort which was significantly increased compared to younger patients from the TCGA database.12,13 However, molecular profiling did not demonstrate any specific mutations or family of mutations as being present uniquely in patients with early onset MDS. Rather, gene mutation rates of the known drivers described above appear to increase linearly with age, suggesting a continuum rather than a unique molecular disease. These data differ from those of patients with juvenile myelomonocytic leukemia, for example, in whom RAS family mutations are found in the vast majority while being present in only 30% of patients with chronic myelomonocytic leukemia.14-16 In fact in this study, there was a trend for an increase in RAS family mutations in patients greater than 50 years of age. As expected, familial mutations were more common in the early onset cohort, which correlates with the known earlier age at diagnosis of familial MDS/acute myeloid leukemia.17 This was particularly exemplified by the fact that the majority of asymptomatic carriers had clonal hematopoiesis by the age of 50 (>80% with RUNX1 germline mutations).17 In addition to performing comprehensive assessment of mutational frequency based on age, Hirsch and colleagues additionally described the clonal architecture in selected patients. Although there were differences regarding dominant clonal mutations based on age (i.e. RUNX1, SF3B1, TP53 in early onset patients and TET2, SF3B1, and STAG2 in older patients), these differences were not universally associated with age but rather a gradual representation of a biological continuum. Unlike other hematologic neoplasms that occur with bimodal distributions such as aplastic anemia, cases of early onset MDS do not constitute a distinct, molecularly defined subgroup.18 Instead, they represent a continuum of MDS in older patients with molecular distinction largely attributable to mutations associated with clonal hematopoiesis of aging. As age-related clonal hematopoiesis leads to hematologic malignancy in only a minority of cases, these data support a growing body of literature describing serial acquisition of genomic events over time with consequent progression to overt disease. Although it is plausible that early onset MDS may simply be a consequence of age-related outliers with characteristic pre-leukemic states described above, future studies are required to understand the inciting molecular events in this population. For example, do patients with early onset MDS have co-morbidities that contribute to accelerated aging or inflammatory states conducive to the pathological events that lead to MDS? Or do they harbor occult germline variants with low penetrance that have yet to be described 967


Editorials

thereby provoking somatic variants characteristic of classic MDS? Although these important questions require further investigation, the work by Hirsch et al. suggests a clear distinction between MDS and other hematologic diseases with bimodal distributions. For example in aplastic anemia, it has become increasingly clear that younger patients have distinct pathophysiologies and therapeutic vulnerabilities with important clinical implications.19 However, the analysis by Hirsch et al. indicates that the molecular underpinnings of adult MDS, regardless of age, are likely more similar than they are different and that our clinical management, including enrollment in interventional studies, should reflect this.

References 1. Ma X, Does M, Raza A, Mayne ST. Myelodysplastic syndromes: incidence and survival in the United States. Cancer. 2007;109(8):1536-1542. 2. Xie M, Lu C, Wang J, et al. Age-related mutations associated with clonal hematopoietic expansion and malignancies. Nat Med. 2014;20(12):14721478. 3. Jaiswal S, Fontanillas P, Flannick J, et al. Age-related clonal hematopoiesis associated with adverse outcomes. N Engl J Med. 2014;371(26):24882498. 4. Genovese G, Kahler AK, Handsaker RE, et al. Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequence. N Engl J Med. 2014;371(26):2477-2487. 5. Niemeyer CM, Kratz CP. Paediatric myelodysplastic syndromes and juvenile myelomonocytic leukaemia: molecular classification and treatment options. Br J Haematol. 2008;140(6):610-624.

6. Li W, Morrone K, Kambhampati S, Will B, Steidl U, Verma A. Thrombocytopenia in MDS: epidemiology, mechanisms, clinical consequences and novel therapeutic strategies. Leukemia. 2016;30(3):536-544. 7. Bejar R, Stevenson K, Abdel-Wahab O, et al. Clinical effect of point mutations in myelodysplastic syndromes. N Engl J Med. 2011;364(26):2496-2506. 8. Papaemmanuil E, Gerstung M, Malcovati L, et al. Clinical and biological implications of driver mutations in myelodysplastic syndromes. Blood. 2013;122(22):3616-3627; quiz 99. 9. Papaemmanuil E, Gerstung M, Bullinger L, et al. Genomic classification and prognosis in acute myeloid leukemia. N Engl J Med. 2016;374(23): 2209-21. 10. Hirsch CM, Przychodzen BP, Radivoyevitch T, Patel B, Thota S, Clemente MJ, et al. Molecular features of early onset adult myelodysplastic syndrome. Haematologica. 2017;102(6):1028-1034. 11. McKerrell T, Park N, Moreno T, et al. Leukemia-associated somatic mutations drive distinct patterns of age-related clonal hemopoiesis. Cell Rep. 2015;10(8):1239-1245. 12. Silva P, Neumann M, Vosberg S, et al. Acute Myeloid leukemia in the elderly is characterized by a distinct genetic landscape. Blood. 2015;126(23):804. 13. 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. 14. Loh ML, Sakai DS, Flotho C, et al. Mutations in CBL occur frequently in juvenile myelomonocytic leukemia. Blood. 2009;114(9):1859-1863. 15. Chan RJ, Cooper T, Kratz CP, Weiss B, Loh ML. Juvenile myelomonocytic leukemia: a report from the 2nd International JMML Symposium. Leuk Res. 2009;33(3):355-362. 16. Emanuel PD. Juvenile myelomonocytic leukemia and chronic myelomonocytic leukemia. Leukemia. 2008;22(7):1335-1342. 17. 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. 18. Montané E, Ibáñez L, Vidal X, et al. Epidemiology of aplastic anemia: a prospective multicenter study. Haematologica. 2008;93(4):518-523. 19. Scheinberg P, Young NS. How I treat acquired aplastic anemia. Blood. 2012;120(6):1185-1196.

Immunoglobulin genes in chronic lymphocytic leukemia: key to understanding the disease and improving risk stratification Lesley-Ann Sutton,1,2 Anastasia Hadzidimitriou,3 Panagiotis Baliakas,1 Andreas Agathangelidis,4 Anton W. Langerak,5 Stephan Stilgenbauer,6 Sarka Pospisilova,7 Zadie Davis,8 Francesco Forconi,9 Frederic Davi,10 Paolo Ghia,4 Richard Rosenquist1,2 and Kostas Stamatopoulos,1,3 on behalf of the European Research Initiative on CLL (ERIC) 1

Department of Immunology, Genetics and Pathology, Science for Life Laboratory, Uppsala University, Sweden; 2Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden; 3Institute of Applied Biosciences, Center for Research and Technology Hellas, Thessaloniki, Greece; 4Division of experimental Oncology, IRCCS Istituto Scientifico San Raffaele e Università Vita-Salute San Raffaele, Milan, Italy; 5Department of Immunology, Erasmus MC, University Medical Center Rotterdam, The Netherlands; 6Department of Internal Medicine III, Ulm University, Germany; 7Central European Institute of Technology, Masaryk University and University Hospital Brno, Czech Republic; 8Department of Haematology, Royal Bournemouth Hospital, UK; 9Haematology Department and Cancer Sciences Unit, Cancer Research UK and NIHR Experimental Cancer Medicine Centres, University of Southampton, UK and 10Hematology Department and University Pierre et Marie Curie, Hopital Pitie-Salpetriere, Paris, France E-mail: ghia.paolo@hsr.it

W

doi:10.3324/haematol.2017.165605

hile triggering through the B-cell receptor (BcR) facilitates B-cell development and maintenance, it also carries intertwined risks for the emergence of lymphoid malignancies, since malignant B cells can exploit BcR signaling pathways in order to initiate and fuel clonal expansion. Indeed, substantial research into chronic lymphocytic leukemia (CLL), largely based on immunogenetic data, supports the notion that the clonotypic BcR immunoglobulin (IG) engages in the recognition 968

of and selection by putative (auto)antigen.1 This highlights the critical role of the BcR IG in the pathophysiology of CLL and implies that disease development is functionally driven and dynamic, rather than being a simple stochastic process. From a clinical perspective, the remarkable therapeutic efficacy of novel drugs such as ibrutinib and idelalisib which target effectors of the BcR signaling pathway (BTK and PI3Kδ, respectively), further vouch for this idea, and herald a major paradigm shift which may ultimately haematologica | 2017; 102(6)


Editorials

lead to changes in the natural history of the disease.2 The IG molecule is an essential component of the multimeric BcR complex and forms a unique genetic identity that is the perfect contender for a clonal marker since it is present from the birth of every B cell onwards, and thus also includes CLL tumor cells, as they derive from activated B cells. Moreover, in contrast to other markers, most notably genomic aberrations, the clonal BcR IG remains stable and unchanged as the disease evolves.3 From the inception of immunogenetics analyses in CLL (Figure 1), reports began to emerge indicating pronounced skewing in IG gene usage and differences from the repertoire of normal B cells, alluding to (super)antigen selection.4 Soon thereafter it was realized that a significant fraction of patients with CLL, approximately 50%, carried somatic hypermutations (SHM) within their BcR IG.4 Additionally, a varying imprint of SHM was seen in clonal BcR IG utilizing different IG heavy variable (IGHV) genes, pointing to functional selection. A further twist in the CLL immunogenetics story was provided by the discovery that the SHM status of the rearranged IGHV gene segregates CLL cases into two broad categories with markedly differ-

ent outcomes. Cases with no or a limited SHM burden (germline identity (GI) ≥ 98%) constituted “unmutated CLL” (U-CLL), wherein patients generally follow an aggressive disease course with short time-to-first-treatment (TTFT), poor response to chemoimmunotherapy and inferior overall survival, thus starkly contrasting cases with GI < 98% (“mutated CLL”, M-CLL) wherein patients usually have a more indolent form of the disease.5,6 Within the clinical arena, IGHV mutation burden allowed us to predict the clinical course of the disease based on the number of SHMs within the expressed IG genes. In more recent years, the strongest molecular evidence for antigen selection in CLL emerged from the finding that unrelated patients can carry identical or almost identical BcR IGs, a phenomenon that cannot be attributed to chance alone and is now aptly termed “stereotypy”.4,7 The aforementioned stratification of patients based on the SHM status of the clonotypic BcR IG has proved to be one of the most robust prognosticators in CLL,8 superseding the clinical impact of other prognostic markers that may fluctuate or change over time. This division reflects

Figure 1. Historical timeline of immunogenetic studies in chronic lymphocytic leukemia (CLL). IG: immunoglobulin; BcR: B cell receptor; SHM: somatic hypermutation; PFS: progression-free survival; M-CLL: mutated CLL; U-CLL: unmutated CLL; FCR: fludarabine, cyclophosphamide and rituximab; OR: overall response.

haematologica | 2017; 102(6)

969


Editorials

A

B

Figure 2. Kaplan-Meier curves for overall survival (OS) and time-to-first-treatment (TTFT) in CLL patients carrying del(17p). M-CLL harboring del(17p) exhibit significantly longer OS (A) and TTFT (B) compared to U-CLL carrying the same genetic defect. Cases included in this analysis are part of a multi-institutional cohort from our collaborative consortium comprising 8563 CLL patients. TTFT analysis was performed in early stage (Binet A) patients. M-CLL: mutated CLL; U-CLL: unmutated CLL.

fundamental biologic differences alluding to a different ontogeny for the two mutational groups and, as a consequence, the BcR IG holds much promise and may be central for developing a biologically-oriented prognostication scheme for CLL.8-10 However, it should be kept in mind that within both M-CLL and U-CLL, a sizeable proportion of cases exhibit clinicobiological behavior that deviates from that associated with its mutational categorization. This highlights the fact that the renowned heterogeneity of CLL persists even after categorization based on the level of SHMs within the IG molecule.9-11 A classical example is offered by stereotyped CLL subset #2, defined by the expression of a distinctive BcR IG utilizing IGHV321/IGLV3-21, which has emerged as a prototype of aggressive disease independently of the SHM load.9,10,12 Indeed, evidence suggests that the immunogenetic subclassification of CLL based on BcR IG stereotypy has clinical potential beyond subset #2, with individual subsets differing significantly in terms of demographics, clinical presentation and the presence or absence of prognostically relevant mutations or cytogenetic aberrations.9 Furthermore, it appears that the particular clinical and cellular background of a genetic lesion, shaped by distinctive signaling through a particular clonotypic BcR IG, may mediate the prognostic or predictive value of recurrent genetic lesions. This idea is exemplified by the finding that within M-CLL, patients harboring trisomy 12 have a TTFT similar to that of patients carrying TP53 aberrations, whereas, in contrast, trisomy 12 is associated with a favorable outcome within U-CLL.13 These findings may explain why trisomy 12 emerges as an intermediate-risk aberration in prognostic indices when the SHM status of the CLL IG is not taken into consideration.14 A similar finding is evidenced when analyzing patients carrying del(17p), with M-CLL cases exhibiting a significantly longer overall survival and TTFT compared to U-CLL patients carrying the same genetic defect (Figure 2). Also worth mentioning is the finding of an asymmetric distribution of certain gene mutations amongst patients bearing distinct immuno970

genetic features e.g., MYD88 mutations are exclusively found within M-CLL while the vast majority of NOTCH1 mutations are detected within U-CLL.15,16 Taken collectively, and bearing in mind that the clinical impact of several biological features is strongly influenced by the SHM status, it is increasingly apparent that definitive conclusions about the clinical implications of any given biomarker should be drawn only after also taking into consideration the SHM status of the rearranged BcR IG. This holds even for well-established prognosticators such as TP53 aberrations. Additional support for the pivotal role of immunogenetic analysis in CLL was recently provided by studies demonstrating that IGHV gene SHM status is a strong marker for predicting the response to chemoimmunotherapy, in particular the fludarabine, cyclophosphamide and rituximab (FCR) regimen which is the gold standard treatment for medically fit CLL patients lacking TP53 defects.17 More specifically, M-CLL cases treated with FCR in the context of clinical trials or general practice were independently reported to achieve prolonged responses, often with no detectable minimal residual disease, thus differing significantly from U-CLL cases.18-20 Interestingly, upon treatment with newer therapeutic agents such as ibrutinib and idelalisib, CLL patients appear to benefit equally and experience similar overall responses irrespective of the IGHV mutational status;21-23 however, the follow-up time is still limited thus precluding definitive conclusions. That said, differences have been noted between the two mutational groups regarding certain clinical parameters following the administration of novel drugs, for example, the initial ibrutinib-induced rise in lymphocyte count and also the duration of lymphocytosis is reported to be greater in M-CLL than that found in U-CLL patients.21 While further investigations into the patterns of lymphocytosis and their association with a clinical response are warranted, these observations may hold value for follow-up assessment. Altogether, the aforementioned examples strongly indicate that determining the SHM status of the IG molecule haematologica | 2017; 102(6)


Editorials

is imperative not only for general assessment of the disease course in CLL, but also for guiding treatment decisions; put simply, IG gene analysis should no longer be viewed only as a prognostic test but also as a predictive test for the use of certain therapies. This idea of following an IG-centric model in order to better stratify CLL patients will likely continue to gain value in the near future due to the emergence of novel treatments and the growing concept of precision therapy, and will have a direct impact on the clinical management of patients with CLL. When broaching the topic of immunogenetic analysis in CLL, what is irrefutable is that the accurate reporting of results obtained from such analyses is paramount, and rigorous standards and meticulous attention to detail are critically important. ERIC, the European Research Initiative on CLL, has been at the forefront of setting standards for immunogenetic research in CLL through pioneering the adoption of good practices by: (i) arranging dedicated educational workshops for the international community; (ii) formulating recommendations for determining the SHM status of IG genes in CLL aimed at harmonizing IG gene sequence analysis in CLL in order to ensure that results are reliable and comparable among different laboratories;24,25 (iii) establishing the IG Network, which promotes and advances immunogenetic analysis across the medical community; and (iv) launching a certification system with external quality control, an asset for accreditation of laboratories performing IG analysis. In conclusion, immunogenetic analysis has proved essential for understanding CLL pathophysiology. We argue that it is equally essential for predicting responses to therapies in this most unpredictable and clinically heterogeneous disease. Indeed, IG-centric risk stratification appears more appealing and relevant today now that signaling inhibition has emerged as a powerful, nonchemotherapeutic approach towards eventually curing CLL, a still incurable disease. Acknowledgements This work was supported in part by the Swedish Cancer Society, the Swedish Research Council, the Lion’s Cancer Research Foundation, the Marcus Borgström Foundation and the Selander’s Foundation, Uppsala; H2020 “AEGLE, An analytics framework for integrated and personalized healthcare services in Europe” by the EU; “MEDGENET, Medical Genomics and Epigenomics Network” (H2020, No.692298) by the EU; research grant AZV-MZ-CR 15-30015A-4/2015, Bloodwise, UK (grants 16003 and 14037), GCH-CLL: funded by the General Secretariat for Research and Technology (GSRT) of Greece and the Italian Ministry of Health (MoH) and IMI2 “HARMONY” funded by the EU. AWL receives unrestricted research grants from Roche-Genentech.

3. 4. 5. 6. 7.

8. 9.

10. 11.

12.

13.

14. 15. 16. 17.

18. 19. 20.

21.

22. 23.

References 1. Vardi A, Agathangelidis A, Sutton LA, Ghia P, Rosenquist R, Stamatopoulos K. Immunogenetic studies of chronic lymphocytic leukemia: revelations and speculations about ontogeny and clinical evolution. Cancer Res. 2014;74(16):4211-4216. 2. Wiestner A. BCR pathway inhibition as therapy for chronic lympho-

haematologica | 2017; 102(6)

24. 25.

cytic leukemia and lymphoplasmacytic lymphoma. Hematology Am Soc Hematol Educ Program. 2014;2014(1):125-134. Sutton LA, Rosenquist R. The complex interplay between cell-intrinsic and cell-extrinsic factors driving the evolution of chronic lymphocytic leukemia. Semin Cancer Biol. 2015;34:22-35. Fais F, Ghiotto F, Hashimoto S, et al. Chronic lymphocytic leukemia B cells express restricted sets of mutated and unmutated antigen receptors. J Clin Invest. 1998;102(8):1515-1525. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38 expression as novel prognostic indicators in chronic lymphocytic leukemia. Blood. 1999;94(6):1840-1847. Hamblin TJ, Davis Z, Gardiner A, Oscier DG, Stevenson FK. Unmutated Ig V(H) genes are associated with a more aggressive form of chronic lymphocytic leukemia. Blood. 1999;94(6):1848-1854. Agathangelidis A, Darzentas N, Hadzidimitriou A, et al. Stereotyped Bcell receptors in one-third of chronic lymphocytic leukemia: a molecular classification with implications for targeted therapies. Blood. 2012;119(19):4467-4475. International CLLIPIwg. An international prognostic index for patients with chronic lymphocytic leukaemia (CLL-IPI): a meta-analysis of individual patient data. Lancet Oncol. 2016;17(6):779-790. Baliakas P, Hadzidimitriou A, Sutton L, et al. Clinical effect of stereotyped B-cell receptor immunoglobulins in chronic lymphocytic leukaemia: a retrospective multicentre study. Lancet Haematol. 2014;1(2):74-84. Baliakas P, Agathangelidis A, Hadzidimitriou A, et al. Not all IGHV3-21 chronic lymphocytic leukemias are equal: prognostic considerations. Blood. 2015;125(5):856-859. Jeromin S, Haferlach C, Dicker F, Alpermann T, Haferlach T, Kern W. Differences in prognosis of stereotyped IGHV3-21 chronic lymphocytic leukaemia according to additional molecular and cytogenetic aberrations. Leukemia. 2016;30(11):2251-2253. Tobin G, Thunberg U, Johnson A, et al. Chronic lymphocytic leukemias utilizing the VH3-21 gene display highly restricted Vlambda2-14 gene use and homologous CDR3s: implicating recognition of a common antigen epitope. Blood. 2003;101(12):4952-4957. Baliakas P, Moysiadis T, Hadzidimitriou A, et al. Tailored approaches for refined prognostication in chronic lymphocytic leukemia patients with mutated versus unmutated immunoglobulin receptors. Blood. 2016;128(22):3199. Dohner H, Stilgenbauer S, Benner A, et al. Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med. 2000;343(26):1910-1916. Baliakas P, Hadzidimitriou A, Sutton LA, et al. Recurrent mutations refine prognosis in chronic lymphocytic leukemia. Leukemia. 2015;29(2):329-336. Sutton LA, Young E, Baliakas P, et al. Different spectra of recurrent gene mutations in subsets of chronic lymphocytic leukemia harboring stereotyped B-cell receptors. Haematologica. 2016;101(8):959-967. Hallek M, Fischer K, Fingerle-Rowson G, et al. Addition of rituximab to fludarabine and cyclophosphamide in patients with chronic lymphocytic leukaemia: a randomised, open-label, phase 3 trial. Lancet. 2010;376(9747):1164-1174. Rossi D, Terzi-di-Bergamo L, De Paoli L, et al. Molecular prediction of durable remission after first-line fludarabine-cyclophosphamide-rituximab in chronic lymphocytic leukemia. Blood. 2015;126(16):1921-1924. Fischer K, Bahlo J, Fink AM, et al. Long-term remissions after FCR chemoimmunotherapy in previously untreated patients with CLL: updated results of the CLL8 trial. Blood. 2016;127(2):208-215. Thompson PA, Tam CS, O'Brien SM, et al. Fludarabine, cyclophosphamide, and rituximab treatment achieves long-term disease-free survival in IGHV-mutated chronic lymphocytic leukemia. Blood. 2016;127(3):303-309. 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. 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. 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. Ghia P, Stamatopoulos K, Belessi C, et al. ERIC recommendations on IGHV gene mutational status analysis in chronic lymphocytic leukemia. Leukemia. 2007;21(1):1-3. Langerak AW, Davi F, Ghia P, et al. Immunoglobulin sequence analysis and prognostication in CLL: guidelines from the ERIC review board for reliable interpretation of problematic cases. Leukemia. 2011;25(6):979984.

971


Editorials

Unmet needs in the scientific approach to older patients with lymphoma Dominique Bron,1 Igor Aurer,2 Marc P. E. André,3 Christophe Bonnet,4 Dolores Caballero,5 Claire Falandry,6 Eva Kimby,7 Pierre Soubeyran,8 Emanuele Zucca,9 Andre Bosly10 and Bertrand Coiffier;11 on behalf of the European Lymphoma Institute (ELI) Group and the EHA SWG “Aging and Hematology” 1

Department of Hematology, Institut Jules Bordet (ULB), Brussels, Belgium; 2Hematological Malignancies Unit, University Hospital Centre Zagreb, Croatia; 3Department of Hematology, CHU Dinant Godinne (CHU UCL Namur), Belgium; 4Department of Hematology, CHU Liège, Belgium; 5Department of Hematology, Complejo Asistencial Universitario de Salamanca, Spain; 6Unité de Gériatrie, Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Université Lyon 1, France; 7Department of Hematology, Karolinska Institute Huddinge University Hospital Stockholm, Sweden; 8Institut Bergonié, Bordeaux, France; 9Lymphoma Unit Division of Research – IOSI / Oncology Institute of Southern Switzerland, Bellinzona, Switzerland; 10CHU Dinant Godinne (CHU UCL Namur), Belgium and 11Department of Hematology, Hospices Civils de Lyon, France E-mail: dbron@ulb.ac.be

doi:10.3324/haematol.2017.167619

L

ymphomas in older patients require special attention because these patients have potentially curative diffuse large B-cell lymphomas (DLBCLs) but may have other diseases that could alter their ability to tolerate treatment. The incidence of lymphomas in older patients has increased in recent years. As detailed in Table 1, most of the subtypes of lymphomas, with the exception of Burkitt lymphoma, lymphocyte predominant nodular Hodgkin lymphoma and classical Hodgkin lymphoma, increase after the age of 50, with a median age of onset of 67 years. Most frequent entities in this age group are DLBCLs, followed by marginal zone lymphomas and follicular lymphomas. Because of the predominance of DLBCLs, this review will focus mainly on this subtype. For optimal management of older patients with lymphoma, three types of issues have to be taken into account. These are: 1) patient-related; 2) disease-related; 3) treatment-related.

Patient-related issues Aging is a natural process leading to the progressive loss of physiological functions, increased inflammatory status, decreased naïve immune T-cell population, increased mutational events and epigenetic modifications resulting in an impaired health status. With the aging of the population, the number of older patients with lymphoma will continue to increase, namely, very old patients over 80 years of age. This population is characterized by its heterogeneity in terms of comorbidities, life expectancy, physical fitness and socio-economic situation. It is clear today that chronological age alone is meaningless, as it is even in very fit patients; marrow reserve and renal function are decreased, and neurological tolerance to toxic drugs is severely impaired. A common characteristic of older patients is the presence of comorbidities that necessitate polymedication, increasing the risk of interaction between these drugs and those used to treat lymphoma. In very old patients, we also have to be mindful of geriatric syndromes (falls, cognitive disturbances, incontinence, dementia and loss of autonomy) that are correlated with a shorter life expectancy. This population should be protected from treatmentrelated toxicity and be allowed to benefit from optimal supportive care to preserve their quality of life. The socioeconomic situation of these patients is also a concern, as it can significantly impair the way the patient deals with unexpected adverse events, hospital visits and the high cost of new medications. 972

The median age of patients with non-Hodgkin lymphoma (NHL) is 67 years and in many international prognostic scoring systems, such as the International Prognostic Index (IPI) or the Follicular Lymphoma International Prognostic Index 2 (FLIPI 1-2), age over 60 remains an adverse risk factor correlated with lower response to chemotherapy and poorer survival.1,2 However, according to the publications from the Groupe d'Etude des Lymphomes de l'Adulte (GELA), the benefit of R-CHOP compared to CHOP is significant in all categories of older patients, including very old selected patients aged between 75-80 years.3 Since life-expectancy for an 80 year-old patient can vary between 3 and 11 years according to his health status, it is of major importance to identify patients without irreversible comorbidities who have a significant life-expectancy and who would benefit most from the optimal treatment regimen. We know from the exhaustive multivariate analysis from Hamaker that there are 3 major predictive factors for mortality in the treatment of older patients with malignant hemopathies4 (Table 2). These factors are: 1) the physical capacity of the patient, such as performance status or the “up and go” evaluation; followed by 2) nutritional deficiencies; and finally 3) the comorbidity index. Therefore, functional, physiological, psycho-cognitive and socio-economic evaluations should be part of the multidimensional geriatric assessment of patients over 70 years of age, taking into account the patient-related, and not only the disease-related, comorbidities. While physicians are quite good at evaluating functional status, comorbidities or nutritional status, the psycho-cognitive function is usually poorly investigated and the socioeconomic status rarely evaluated. The value of a multidimensional geriatric assessment was prospectively investigated in 2000 successive Belgian patients.5 This study confirmed that multidimensional geriatric assessments gave prognostic information in terms of overall survival (OS), morbidity and loss of quality of life, but more importantly, they detected multiple problems that influenced the choice of a better targeted treatment. In a population of clinically fit lymphoma patients sent by their physician to receive chemotherapy treatment, we observed that cognitive problems were completely underestimated: 30% of the patients had an abnormal Mini Mental State Evaluation (MMSE) test and 51% had an abnormal Montreal Cognitive Assessment (MoCA) test (which is more sensitive in its evaluation of executive functions), which could reach 70% in patients over 80 haematologica | 2017; 102(6)


Editorials

years. In this clinically “fit” population, admitted for the treatment of hematologic malignancies, we observed a significant deleterious impact on the 2-year overall survival rate when the MoCA test was below 26 or the MMSE test was below 23/27 (P<0.01).6 Therefore, to better evaluate the fitness of older patients, close collaboration between geriatricians and hematologists is highly recommended. However, a prerequisite should be the harmonization of the terminology. Indeed, many trials conducted in patients aged over 60

use the term “elderly patients”, while geriatricians consider “elderly patients” as those having severe comorbidities, loss of autonomy and requiring a lot of attention. At the same time, a patient onco-hematologists would call “vulnerable” would be considered “frail” by geriatricians, leading to confusion in the interpretation of “frailty” in the population receiving chemotherapy and potentially excluding them from clinical trials. An optimal, easy-to-use screening tool is essential to determine patient “fitness”. The G8 screening tool has been validated in solid tumors and malignant

Table 1. Non-Hodgkin lymphoma histological subtypes according to age.

Table 2. Prognostic factors for mortality. Hematologic malignancies in the elderly.4

ADL: activities of daily living; IADL: instrumental activities of daily living; AML: acute myeloid leukemia; MDS: myelodysplastic syndromes; RAEB: refractory anemia with excess blasts; DLBCL: diffuse large B-cell lymphoma.

haematologica | 2017; 102(6)

973


Editorials

hemopathies.7,8 However, in “clinically fit” patients, the weight of nutritional and psychological status is too important. Indeed, many DLBCL patients experience weight loss, and often feel tired and in poor health. The score of their G8 questionnaire one week after steroids shows a complete improvement across all categories by the reversibility of these lymphoma-related comorbidities. The strict application of the G8 screening, completed by a Comprehensive Geriatric Assessment (CGA), could lead to under-treatment of patients, undermining their chances to be cured.

Disease-related problems The paradigm in diffuse large B-cell lymphomas is that we are facing a curable disease with adequate immunochemotherapy in a population for whom this adequate treatment can lead to life-threatening side effects. We also know that the OS of lymphoma patients is correlated with achievement of complete remission (CR) and that a minor reduction in the dosage or dose intensity compromises their chances to be cured. In addition, older NHL patients present with more advanced stages at diagnosis, a more unfavorable biological profile (non-germinal center vs. germinal center), more genetic mutations (MYC positive), and an increased incidence of anemia, which is an adverse prognostic factor whatever the cause. However, the disease remains the major cause of death. The GELA study and the German study have both demonstrated that RCHOP21 is superior to CHOP14 and to CHOP21 in terms of OS and toxicity.3,9,10 The benefit of rituximab was observed in all age categories (60-70, 70-75, 75-80 years of age) but this trial was limited to “fit” patients under 80. In a prospective phase II trial, Peyrade and colleagues investigated the R miniCHOP regimen in a population of DLBCL patients aged 80 or over.11 This trial confirmed the feasibility and efficacy (49% OS at 4 years) of this reduced R-CHOP in very old patients, as reported also in retrospective studies.12 However, this population was also selected on the basis of comorbidities and an interesting retrospective study reported by Marchesi et al. showed that “frail” patients [aged 85 years or over, activities of daily living (ADL) < 6, 1 geriatric syndrome or > 3 comorbidities] did not benefit from any chemotherapy.13 Indeed, there have been no randomized clinical trials (RCT) in frail DLBCL patients nor in other subtypes of lymphomas (HL, Burkitt, T lymphoma, etc.). A recent review outlines the reasons for this.14

Treatment-related problems The treatment of older patients with NHL, and specifically the R-CHOP regimen, is associated with short-term toxicity: primarily hematologic toxicities with the increased risk of febrile neutropenia, cardiac toxicity and neurotoxicity. This regimen also results in long-term toxicity with secondary myelodysplasia or acute myeloblastic leukemia (MDS or AML), functional and cognitive decline, and cardiomyotoxicity, leading to the loss of autonomy. If the toxic death rate is important in all clinical trials, in older patients this rate is exacerbated. The prognostic factors associated with poor survival are discussed by Hamaker and are usually patient-related. Screening tests to predict this toxicity, mostly used in oncology, are the Chemotherapy Risk 974

Assessment Scale for High-Age Patients (CRASH) test and the Vulnerable Elders Survey (VES-13).15,16 However, they have not yet been validated in hematologic patients and are thus rarely used. Soubeyran reviewed predictive factors for unacceptable toxicities such as early toxic death, loss of autonomy and unplanned hospitalization (P Soubeyran, unpublished data, 2017) identified in 3 recent trials. The significant predictive factors were: Mini Nutritional Assessment (MNA) < 24, “up and go” test > 20 sec, the Lawton Instrumental Activities of Daily Living (IADL) < 8, and Geriatric Depression Scale-15 (GDS15) > 5. The major side effect of chemotherapy in older patients is neutropenia (>60% over 80 years) with a mortality ranging from 9% to 23%. Marrow reserve is reduced in patients aged over 60 but their response to granulocytecolony stimulating factor (G-CSF) is similar to that of younger patients. Although it has been known for more than ten years that the addition of G-CSF as primary prophylaxis reduces mortality in the treatment of NHL, in fact, only 36% of older patients receive G-CSF as primary prophylaxis; 20% receive a G-CSF as secondary prophylaxis and up to 30% receive no growth factor at all.17 Another event causing early death is tumor lysis syndrome during the first cycle. This risk can be significantly reduced by a pre-phase treatment using steroids for one week with or without rituximab,18 as confirmed at the Lymphoma Study Association (LYSA) meeting (F Peyrade, unpublished data, 2017). Other concerns after treatment with R-CHOP are cardiovascular problems and late heart failure, diabetes and high sensitivity to neurotoxic drugs such as vincristine. One should be very careful when monitoring heart rate, glycemia and searching for peripheral neuropathies. Older patients are also more sensitive to secondary tumors such as lung cancers and MDS. Pharmacokinetic data are dramatically lacking in very old patients and trials such as the GELA trial, with reduced doses of anthracycline leading to a similar CR rate and OS rate, suggest that older patients may have an increased half-life of drugs or drug metabolites. Table 3. Unmet needs in the multidisciplinary approach of lymphoma patients. Patient-related - Harmonization in language (older vs. elderly, frail vs. vulnerable) - Simple tool to evaluate patient fitness that allows optimal curative treatment - Better distinction between patient-related and disease-related comorbidities (multidisciplinary approach) Disease-related - RCT adapted to “real life” - Informed consent of the patient after cognitive evaluation - Patient-centered RCT with “OS with quality of life” as a primary end point! Treatment-related - Primary prophylaxis with G-CSF - Pre-phase treatment before aggressive chemotherapy - Optimal survey of neuropathies - Involvement of family caregivers RCT: randomized clinical trials; OS: overall survival; G-CSF: granulocyte-colony stimulating factor.

haematologica | 2017; 102(6)


Editorials

Another group published similar data in DLBCL patients above 80 years where less than 85% anthracycline relative dose intensity achieved a better overall survival.19 Finally, as pointed out by Hamaker, most of the ongoing trials for elderly patients with hematologic malignancies are not addressing the right end points.20 Indeed, the outcomes of primary importance for this population, such as quality of life, health care utilization and loss of functional capacity, were only measured in 8%, 4% and 0.7% of the trials! As a scientific community, we must support patientfocused cancer care in RCT to further elucidate all these unresolved issues in order to significantly improve our knowledge of the optimal treatment of older patients.

Conclusion The treatment paradigm in aggressive NHL is, on the one hand, an effective conservative treatment that preserves quality of life and controls the disease, and on the other, an intensive potentially curative treatment with more toxicities. A multidisciplinary approach using harmonized language is mandatory (Table 3). It is important to determine the risk-benefit ratio for each patient. We are entering an era in which the patientphysician relationship has evolved from paternalism to a face-to-face dialogue. The patient is involved in the decision-making process and expresses his wishes regarding his quality of life. However, this patient involvement is conditioned by his ability to understand the risk-benefit ratio of the treatment, and to read and sign an informed consent. Today, the management of lymphoma in older patients is a multistep approach starting with the prognostic evaluation of the lymphoma and the potential severe adverse events induced by the treatment. A second step evaluates the physical, physiological, cognitive and socio-economic status of the patient, raising the question of life expectancy with or without the tumor. Finally, and probably more importantly, the patient should express his expectations in terms of quality of life.

References 1. [No authors listed] A predictive model for aggressive non-Hodgkin's lymphoma. The International Non-Hodgkin's Lymphoma Prognostic Factors Project. N Engl J Med. 1993;329(14):987-994. 2. Solal-Céligny P, Roy P, Colombat P, et al. Follicular lymphoma international prognostic index. Blood. 2004;104(5):1258-1265. 3. Coiffier B, Thieblemont C, Van Den Neste E, et al. Long-term outcome of patients in the LNH-98.5 trial, the first randomized study comparing rituximab-CHOP to standard CHOP chemotherapy in DLBCL patients: a study by the Groupe d'Etudes des Lymphomes de

haematologica | 2017; 102(6)

l'Adulte. Blood. 2010;116(12):2040-2045. 4. Hamaker ME, Prins MC, Stauder R. The relevance of a geriatric assessment for elderly patients with a haematological malignancy--a systematic review. Leuk Res. 2014;38(3):275-283. 5. Kenis C, Heeren P, Bron D, et al. Multicenter implementation of geriatric assessment in Belgian patients with cancer: a survey on treating physicians' general experiences and expectations. J Geriatr Oncol. 2014;5(4):431-438. 6. Dubruille S, Libert Y, Maerevoet M, et al. Prognostic value of neuropsychological and biological factors in clinically fit older patients with hematological malignancies admitted to receive chemotherapy. Geriatr Oncol. 2015;6(5):362-369. 7. Petit-Monéger A, Rainfray M, Soubeyran P, Bellera CA, MathoulinPélissier S. Detection of frailty in elderly cancer patients: Improvement of the G8 screening test. J Geriatr Oncol. 2016;7(2):99-107. 8. Hamaker ME, Mitrovic M, Stauder R. The G8 screening tool detects relevant geriatric impairments and predicts survival in elderly patients with a haematological malignancy. Ann Hematol. 2014;93(6):1031-1040. 9. Coiffier B, Lepage E, Briere J, et al. CHOP chemotherapy plus rituximab compared with CHOP alone in elderly patients with diffuse large-B-cell lymphoma. N Engl J Med. 2002;346(4):235-242. 10. Pfreundschuh M, Ho AD, Cavallin-Stahl E, et al. MabThera International Trial (MInT) Group. Prognostic significance of maximum tumour (bulk) diameter in young patients with good-prognosis diffuse large-B-cell lymphoma treated with CHOP-like chemotherapy with or without rituximab: an exploratory analysis of the MabThera International Trial Group (MInT) study. Lancet Oncol. 2008;9(5):435-444. 11. Peyrade F, Gastaud L, Ré D, Pacquelet-Cheli S, Thyss A. Treatment decisions for elderly patients with haematological malignancies: a dilemma. Lancet Oncol. 2012;13(8):e344-352. 12. Ricciuti G, Finolezzi E, Luciani S, et al. Combination of rituximab and nonpegylated liposomal doxorubicin (R-NPLD) as front-line therapy for aggressive non-Hodgkin lymphoma (NHL) in patients 80 years of age or older: a single-center retrospective study. Hematol Oncol. 2017 Feb 3. [Epub ahead of print] 13. Marchesi F, Cenfra N, Altomare L, et al. A retrospective study on 73 elderly patients (≥75years) with aggressive B-cell non Hodgkin lymphoma: clinical significance of treatment intensity and comprehensive geriatric assessment. J Geriatr Oncol. 2013;4(3):242-248. 14. Molina-Garrido MJ, Guillen-Ponce C. Where Are We Headed With Research in Frail Elderly Patients With Cancer? J Clin Oncol. 2016;34(33):4049-4050. 15. Extermann M, Boler I, Reich RR, et al. Predicting the risk of chemotherapy toxicity in older patients: the Chemotherapy Risk Assessment Scale for High-Age Patients (CRASH) score. Cancer. 2012;118(13):3377-3386. 16. Luciani A, Biganzoli L, Colloca G, et al. Estimating the risk of chemotherapy toxicity in older patients with cancer: The role of the Vulnerable Elders Survey-13 (VES-13). J Geriatr Oncol. 2015;6(4): 272-279. 17. Choi MR, Solid CA, Chia VM, et al. Granulocyte colony-stimulating factor (G-CSF) patterns of use in cancer patients receiving myelosuppressive chemotherapy. Support Care Cancer. 2014;22(6):1619-1628. 18. Pfreundschuh M. How I treat elderly patients with diffuse large Bcell lymphoma. Blood. 2010;116(24):5103-5110. 19. Carson KR, Riedell P, Lynch R, et al. Comparative effectiveness of anthracycline-containing chemotherapy in United States veterans age 80 and older with diffuse large B-cell lymphoma. J Geriatr Oncol. 2015;6(3):211-218. 20. Hamaker ME, Mitrovic M, Stauder R. The G8 screening tool detects relevant geriatric impairments and predicts survival in elderly patients with a haematological malignancy. Ann Hematol. 2014;93(6):1031-1040.

975


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Red Cell Biology & its Disorders

Ferrata Storti Foundation

Haematologica 2017 Volume 102(6):976-983

Family cord blood banking for sickle cell disease: a twenty-year experience in two dedicated public cord blood banks

Hanadi Rafii,1,2 Françoise Bernaudin,3 Helene Rouard,4 Valérie Vanneaux,5,6 Annalisa Ruggeri,1,2 Marina Cavazzana,7,8,9 Valerie Gauthereau,10 Aurélie Stanislas,7,8 Malika Benkerrou,11 Mariane De Montalembert,12 Christele Ferry,13 Robert Girot,14 Cecile Arnaud,3 Annie Kamdem,3 Joelle Gour,15 Claudine Touboul,15 Audrey Cras,5,6 Mathieu Kuentz,16 Claire Rieux,17 Fernanda Volt,1,2 Barbara Cappelli,2 Karina T. Maio,1,2 Annalisa Paviglianiti,1,2 Chantal Kenzey,1,2 Jerome Larghero5,6 and Eliane Gluckman1,2

Eurocord, Paris-Diderot University EA 3518, Saint-Louis Hospital, Assistance PubliqueHôpitaux de Paris, France; 2Monacord, International Observatory for Sickle Cell Disease, Centre Scientifique de Monaco, Monaco; 3Department of Pediatrics, Referral Center for Sickle Cell Disease, Centre Hospitalier Intercommunal, Paris XII University, Créteil, France; 4Cell Therapy Facility, EFS Ile de France, Créteil, France; 5Cell Therapy Facility, Saint-Louis Hospital, Assistance Publique-Hôpitaux de Paris, France; 6 Biotherapy Clinical Investigation Center, Paris-Diderot University, Sorbonne Paris Cité, INSERM, F-75010, France; 7Biotherapy Department, Necker Children’s Hospital, Assistance Publique-Hôpitaux de Paris, France; 8Biotherapy Clinical Investigation Center, Groupe Hospitalier Universitaire Ouest, Assistance Publique-Hôpitaux de Paris, INSERM, France; 9Paris Descartes–Sorbonne Paris Cité University, Imagine Institute, France; 10 Fédération Parisienne Pour le Dépistage et la Prévention des Handicaps de l'Enfant (FPDPHE), Necker Children’s Hospital, Assistance Publique-Hôpitaux de Paris, France; 11 Department of Pediatrics, Referral Center for Sickle Cell Disease, Robert Debré Hospital, Assistance Publique-Hôpitaux de Paris, France; 12Department of Pediatrics, Necker Children’s Hospital, Assistance Publique-Hôpitaux de Paris, France; 13 Department of Stem Cell Transplantation, Saint-Louis Hospital, Assistance PubliqueHôpitaux de Paris, France; 14Department of Hemato-Biology, Tenon Hospital, Assistance Publique-Hôpitaux de Paris, France; 15Department of Gynecology, Centre Hospitalier Intercommunal, Créteil, France; 16Department of Hematology, Groupe Hospitalier Universitaire Henri-Mondor, Créteil, France; 17Unité d'Hémovigilance, Groupe Hospitalier Universitaire Henri-Mondor, Créteil, France 1

Correspondence: hanadi.rafii-elayoubi@aphp.fr

Received: December 24, 2016. Accepted: March 10, 2017. Pre-published: March 16, 2017. doi:10.3324/haematol.2016.163055 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/976 ©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.

976

ABSTRACT

E

fforts to implement family cord blood banking have been developed in the past decades for siblings requiring stem cell transplantation for conditions such as sickle cell disease. However, public banks are faced with challenging decisions about the units to be stored, discarded, or used for other endeavors. We report here 20 years of experience in family cord blood banking for sickle cell disease in two dedicated public banks. Participants were pregnant women who had a previous child diagnosed with homozygous sickle cell disease. Participation was voluntary and free of charge. All mothers underwent mandatory serological screening. Cord blood units were collected in different hospitals, but processed and stored in two public banks. A total of 338 units were stored for 302 families. Median recipient age was six years (11 months15 years). Median collected volume and total nucleated cell count were 91 mL (range 23-230) and 8.6x108 (range 0.7-75x108), respectively. Microbial contamination was observed in 3.5% (n=12), positive hepatitis B serology in 25% (n=84), and homozygous sickle cell disease in 11% (n=37) of the collections. Forty-four units were HLA-identical to the intended recipient, and 28 units were released for transplantation either alone (n=23) or in combination with the bone marrow from the same donor (n=5), reflecting a utilization rate of 8%. Engraftment rate was 96% with 100% survival. Family cord blood banking yields good quality units for sibling transplantation. More comprehensive banking based on close collaboration among banks, clinical and transplant teams is recommended to optimize the use of these units. haematologica | 2017; 102(6)


Sibling cord blood banking for sickle cell disease

Introduction Sickle cell disease (SCD) is the most common inherited hemoglobin disorder. Particularly frequent in people of African descent, the disease is associated with numerous complications and early mortality.1,2 Major progress has been made in the management of SCD and this has improved patients' survival without curing the disease;3 this includes implementation of antenatal counseling and neonatal screening programs,3 screening and prevention of neurological complications,4 prevention of pneumococcal infections,5,6 and the use of hydroxycarbamide5,7 and chronic transfusion therapy.8 However, hematopoietic stem cell transplantation (HSCT) remains, to date, the only known curative therapy for SCD,9-12 offering a cure rate exceeding 90%.9-17 Walters et al., in a recent review of HSCT in SCD children after HLA-identical sibling HSCT reported overall survival and event-free survival rates of 95% and 92%, respectively.17 Similar results were reported in retrospective registrybased surveys from the Center for International Blood and Marrow Transplant Research18 and the European Society for Blood and Marrow Transplantation9,10,13 with overall survival rates of over 91%. Despite a cure rate exceeding 90%, regardless of the stem cell source, HSCT is still under-used for patients with SCD.9,10 Limiting factors include the availability of suitable donors and, for the majority of patients for whom a matched sibling donor has been identified, the reluctance of families and physicians to consider transplantation because of possible transplant-related toxicity. Umbilical cord blood transplantation (CBT) from a related family member has proved to be an effective alternative for patients with SCD, resulting in survival rates similar or superior to adult donor transplant9,10,13,19,20 and lower probability of acute and chronic graft-versus-host diseases (GvHD).9,10 More recently, a promising, but still experimental, approach based on gene and cellular therapy methods has been proposed aiming to correct the sickle gene defect in the patient’s own stem cells.21,22 With the emergence of the gene-therapy biotechnology to treat genetic disorders22 and the improved outcomes of CBT, many directed public or private cord blood (CB) banking programs have been established for affected siblings who would benefit from a related HSCT. Consequently, cord blood banks (CBB) are now confronted with major challenges regarding storage and disposal of the units. We report here our 20-year experience in two public family-directed cord blood banks for SCD.

Methods Participants were pregnant women who had a child homozygous for SCD. Recruitment was carried out through referral addressed to the CBB by the clinician treating the affected child. All mothers, including those who refused prenatal diagnosis, were offered sibling CB banking. Participation was voluntary. Informed consent was obtained from the mother before delivery, in accordance with local ethical requirements. Collections were organized by two public banks, free of charge for the families. Both banks had a wide experience in unrelated CB banking and were affiliated to the French network of CBB accredited by the “Agence Nationale de Biomedecine”. haematologica | 2017; 102(6)

Table 1. Characteristics of collected cord blood units. Total CBUs collected Median volume collected (mL) Median TNC count (x108) Median CD34+ cell count (x106) Median CFU-GM cell count (x105) Median cryopreservation duration (years)

N

Range

338 91 8.6 2.5 3.4 7

23 - 230 0.7 - 75 0.1 - 61 0.01 - 63 1 - 20

CBUs: cord blood units; TNC: total nucleated cells; CFU-GM: colony forming unit granulocyte macrophages.

Table 2. HLA typing and utilization of cord blood units.

Cord blood units HLA not performed Non-HLA identical HLA identical CBT CBT + BMT (same donor) BMT same donor (CBU not used) Not transplanted No HSCT indication Patient declined HSCT CBU has ( S/ S) hemoglobin

N

(%)

193 101 44 23 5 1 15 9 4 2

(57%) (30%) (13%) (7%) (1%) (<1%) (5%)

CBU: cord blood units; CBT: cord blood transplant; BMT: bone marrow transplant; HSCT: hematopoietic stem cell transplant.

The CBUs were reserved for family use and were shipped to the transplant center once the decision to proceed with transplantation was made. All mothers underwent a panel of mandatory serology testing prior to banking, which included hepatitis B and C viruses, HIV 1-2, HTLV I-II and syphilis. The CBU were collected in the health care institution selected by the mother for child delivery. Twenty-seven institutions were involved and were contacted by the CBB. Informed consents were collected by midwives or designated medical personnel locally before delivery. Detailed instructions and training to local staff were provided by the CBB team. CBU collection kits, including a collection pack and materials necessary for harvesting, were sent to the contact team in the delivery hospital along with the standard regulatory forms. All CBUs were transported to the designated CBB to be processed within 24 hours of collection. The CBBs policy was to process and store all the collected CBUs, independent of the unit volume, cell counts and HLA compatibility. Cell counts, cell viability, sterility tests, progenitor cell quantification and functional assays were performed for all CBUs. Due to related cost and absence of immediate patient indication for transplantation, histocompatibility testing was not conducted routinely, unless requested by the referring clinician. There was no volume reduction before cryopreservation. The CBUs were cryopreserved using a controlled-rate freezer, then transferred to the vapor phase of a liquid nitrogen storage tank and maintained at less than -150°C. The CBUs and aliquots collected from mothers having positive infectious markers or awaiting results were stored in quarantine tanks. Abnormal results that could affect donor suitability were considered individually at the time a CBU was requested for transplantation. Hemoglobin genotyping was not performed on CBU samples. Information about donor SCD status was available through the national neonatal screening database for genetic diseases. CBUs 977


H. Rafii et al.

collected from infants found to have SS genotype, were not discarded unless a written request was formulated by the referring physician. Once the decision to proceed with transplantation was made, confirmatory HLA typing was conducted. CBUs were thawed and shipped to the transplant unit for infusion. Total nucleated cells and CD34+ cell counts, viability and sterility tests were performed after thawing. Data were collected from CBB databases, Eurocord registry and patient records. Neutrophil engraftment was defined as the first of three consecutive days with neutrophils 0.5x109/L or more, and platelet engraftment as the first of three consecutive days with platelets 20x109/L or more with no platelet transfusions for at least seven days before reconstitution.

Results From 1995 to 2014, a total of 338 CBUs were collected with a sustained increase in the number of referrals for family CB banking for SCD over the years. These units were stored for 302 families, including 32 mothers recruited more than once. All collections were for an existing homozygous sibling. Fifteen families (5%) had more than one affected child. Figure 1 details the distribution of the CBUs collected in each CBB.

Characteristics of banked CBUs Median values for specific quality parameters of the collected CBUs are shown in Table 1. Sixty-one percent (n=207) of the collections exceeded 80 mL (Figure 2) and 40% (n=135) exceeded 10.0x108 total nucleated cell count (TNC). Sixty-five percent (n=218) of the units had a minimal TNC of 5x108 (based on 20 kg recipient, a target dose of ≥2.5x107 nucleated cells/kg), considered suitable for pediatric transplantation according to the US FDA23 criteria for banking CBUs and 44% of the units (n=148) exceeded the minimal pre-freezing thresholds for cell counts and volume adopted by many unrelated CBBs24-27 (i.e. volume >40 mL and TNC >9x108). A microbial contamination rate of 3.5% was observed in our cohort as 12 CBUs failed sterility testing; these units were stored because antimicrobial sensitivity results did not preclude their use for transplantation with the provision of appropriate antibiotics. Eighty-four CBUs (25%) had positive hepatitis B serological markers: 76% (n=64) of these were due to positive anti-HBs and/or anti-HBc antibodies with negative HBs antigen. Two units tested positive for hepatitis C (HCV) antibody, but HCV RNA was not detected by PCR, despite maternal active HCV infection. All CBUs were negative for HIV. HLA-typing was carried out for 145 donor-recipient pairs (43% of the collected CBU) upon request of the referring physicians (Table 2). Forty-four units had a full HLA antigen match with the intended recipient: these represented 30% of the typed units and 13% of all collected units. Eleven percent (n=37) of the collections had homozygous SCD and 45% (n=152) a carrier disease status (Figure 3). Of the 145 HLA typed CBU, 8% (n=11) were homozygous for SCD, and 26% (n=38) were both HLA identical to the recipient and had normal hemoglobin or sickle cell carrier status, thus considered as “potential graft source” for the intended sibling. 978

Table 3. Transplant characteristics. CBUs used for transplant, n CBT with available data, n Median recipient age at HSCT, years (range) Graft type, n (%) CBU CB + BM (same donor) HLA compatibility, n (%) 10/10 Conditioning regimen, n (%) Cy + Bu ATG pre-transplant, n (%) GVHD prophylaxis, n (%) CSA CSA + MTX G-CSF post-transplant, n (%)

28 25 6.8

(3.2 - 13.6)

23 5

(82%) (18%)

25

(100%)

25 25

(100%) (100%)

24 1 11

(80%) (20%) (44%)

CBU: cord blood units; CBT: cord blood transplants; HSCT: hematopoietic stem cell transplant; BM: bone marrow; CB: cord blood; Cy: cyclophosphamide; Bu: busulfan; ATG: anti-thymocyte globulin; GvHD: graft-versus-host disease; CSA: cyclosporine A; MTX: methotrexate; G-CSF: granulocyte colony stimulating factor.

To date, 28 units were released for transplantation either alone (CBT, n=23) or in combination with a bone marrow (BM) graft from the same donor (CBT+BMT, n=5), reflecting a utilization rate of 8% for the whole cohort over 20 years, and 19% transplant rate for the typed units. Another 13 HLA-identical CBUs with adequate cell dose to perform transplantation remained in storage for future use, as their intended recipient was not an immediate indication for transplantation (n=9) or decided not to undergo the procedure (n=4) because of transplant-related risks. Three HLA identical CBUs were deemed unsuitable for transplantation due to their biological features: 2 units were homozygous for HbS and one unit had low cell counts (2.7 x107 TNC/kg and 0.2 x106 CD34+ cell/kg) (Table 2). Of further interest, 41% (n=137) of the collected CBUs were referred from a single hematologic center.9,11 Histocompatibility testing was requested and performed for 87% (n=119) of these CBUs and 30% (n=35) were HLA identical to the affected sibling. To date, 25 of these collections were used for CBT. This represents an 18% (25 of 137) utilization rate for collections referred from this center, 21% (25 of 119) transplant rate for typed units, and a 71% (25 of 35) transplant rate when a matched sibling is available. These are noteworthy rates and reflect a 2.3-fold increase in the utilization rate when the collections are requested by clinicians willing to use CBUs to transplant the affected child once transplantation is indicated. When compared to other referrals that were not affiliated to a transplant center (n=201), only 29 CBUs (14%) were typed for HLA, of which 3 units were used for transplantation; this reflects a 1.5% (3 of 201) utilization rate for the units referred from these centers, 10% (3 of 29) of the typed units, and 33% (3 of 9) of the HLA-identical units.

Transplant characteristics Over a period of 20 years, 28 units were released for transplantation to 28 patients. Characteristics of transplanted patients are shown in Table 3. The median age of recipients at transplantation was 6.8 (3.2-13.6) years. All patients for whom data were available (n=25) received a haematologica | 2017; 102(6)


Sibling cord blood banking for sickle cell disease

Figure 1. Family cord blood banking program. Flow diagram. SCD: sickle cell disease; CBU: cord blood units.

conditioning regimen including busulfan, cyclophosphamide and anti-thymocyte globulin (ATG). All CBUs were HLA identical to the intended recipient, and 54% of the donors were sickle cell carriers. The CBUs were transplanted after a median storage time of 2.2 years (range 0.48.1) (Table 4). The median cell dose collected was 4.0x107/kg (0.5-13.1 x107) TNC and 1.6 x105/kg (0.2-14.8 x105) CD34+ cells, and the median infused cell dose was 3.1x107/kg (0.2- 7.6x107) TNC and 1.4 x105/kg (0.2-11.8x105) CD34+ cells, with a median recipient body weight of 22 kg (14-56) for the 28 patients transplanted (Table 4). Five CBUs with a collected TNC dose less than 2x107/kg (0.5-1.7x107/kg) were combined with the BM collected from the same donor (Table 5). After adding the BM, the median infused TNC and CD34+ cell doses were 18.5x107/kg (7.3-26.7x107) and 5.4x106/kg (2.3-10 x106), respectively, of which a median dose of 1x107/kg TNC (0.2-1.2x106) and 0.03x106/kg CD34+ (0.02-0.04 x106) were provided by the CBUs. The median time to engraftment was 28 days (range 1760) for neutrophils, and 54 days (range 12-91) for platelets (n=18). All but one patient engrafted for neutrophils before day 60, with a median time of 33 days after single CBT and 26 days after BMT+CBT. Chimerism at day 100 for those patients with available data (n=22) was 100% donor for 9 patients (41%) and 60-80% donor for 12 patients (55%). At last follow up, donor chimerism exceeded 80% for all patients with available data. Notably, one patient transplanted in 1996 with an adehaematologica | 2017; 102(6)

quate cell dose (4x107 TNC/kg) had primary graft failure; this patient also failed to engraft after a second transplant using the BM collected from the same sibling donor. At last follow up, the patient was alive with autologous reconstitution and remained anemic but free of vasoocclusive events. This is due to sustained increase in fetal hemoglobin production, as previously reported after autologous reconstitution in patients affected by SCD11,28 and β-hemoglobin disorders.29 Five patients who received single CBT experienced grade I-II acute GvHD that required steroids in 4 cases. One recipient of combined CB+BM transplant developed limited liver chronic GvHD, which was the only chronic GvHD event reported for the entire cohort. All 28 patients are alive and free of SCD symptoms to date (except the sole case of persistent anemia after primary graft failure), with a median follow up of 4.3 years (4 months-14 years).

Discussion The aim of this study was to investigate whether sibling CBU collections had sufficient cellular quantity and quality to support allogeneic CBT of the intended recipient and to assess the utilization pattern of these collections. Family CB banking differs in many important aspects from public CBU donations. Besides the cost related to this stem cell source, when less expensive alternatives 979


H. Rafii et al.

using the sibling BM are possible, there are major challenges related to cellular characteristics of the collections, such as the long-term storage and low utilization rate noted by many teams13,30-32 and confirmed in our cohort. The cellular and biological characteristics of our banked units were comparable to those reported by other directed CBB programs13,26,33 with a utilization rate reaching 8% of the 338 CBUs collected over 20 years, and 19% of the typed units. Similar low utilization rates of banked sibling CBUs for hematologic disorders have been reported by many authors.30-33 Both of our CBBs had previously published their experience13 in banking sibling CBUs for malignant and nonmalignant disorders, with more than 400 directed CBUs collected in one bank and 111 units (80% for SCD affected siblings) in the second, a utilization rate for HLA identical sibling transplants of 3% and 9%, respectively, and 5-year overall survival of 83% and 100%, respectively, in the subgroup of patients with hemoglobin disorders.11,13 Most of

the patients in our cohort were included in these studies. Adequate volume and TNC dose are critical factors in achieving a successful CB transplantation; therefore, the use of procedures that might increase collection volume or decrease cell loss during processing is especially important in family CB banking. While 65% of our CBUs had cellular characteristics (minimal collected TNC count of 5x108) adequate to support allogeneic CBT in the pediatric setting,26,34 only 13% (n=45) of our collections provided the TNC dose and volume recommended for unrelated CBU banking (initial volume >80 mL and TNC >18x108). However, the low cellular content should not be a limitation for the use of these CBU,31,33 particularly when there is an opportunity to increase the cell dose by combining CB with BM collected from the same sibling donor, an approach that proved to be effective with excellent overall survival and low incidence of GvHD, as noted by many authors.13,30,31,35 Indeed, all but one of the patients transplanted in our

Table 4. Characteristics of transplanted cord blood units. CBUs transplanted, n (%) Median volume, mL (range) Median recipient weight, kg Collected cell dose, n (range) Median NC x107/kg Median CD34+ x105/kg Median CFU x104/kg Infused cell dose, n (range) Median NC x107/kg Median CD34+ x105/kg Storage duration, years (range)

CBUs infused alone

CBUs infused with BM*

Total CBUs

23 (82%) 109 (49 - 230) 22 (14 - 45)

5 (18%) 55 (49 - 65) 33 (14 - 56)

28 (100%) 99 (49 - 230) 22 (14 - 56)

5.2 (2.9 - 13.1) 1.8 (0.6 - 14.8) 1.6 (0.4 - 4.9)

1.1 (0.5 - 1.7) 0.3 (0.2 - 0.4) 0.3 (0.1 - 0.9)

4.0 (0.5 - 13.1) 1.6 (0.2 - 14.8) 1.1 (0.1 - 4.9)

3.7 (1.5 - 7.6) 2.1 (0.5 - 11.8) 1.7 (0.4 - 8.1)

1 (0.2 - 1.5) 0.3 (0.2 - 0.4) 3.1 (2.6 - 4.0)

3.1 (0.2 - 7.6) 1.4 (0.2 - 11.8) 2.2 (0.4 - 8.1)

CBU: cord blood units; NC: nucleated cells; CFU: colony forming units; BM: bone marrow. *Characteristics of BM infused are reported in Table 5.

Figure 2. Volume distribution (%) of collected cord blood units.

980

haematologica | 2017; 102(6)


Sibling cord blood banking for sickle cell disease

cohort engrafted within less than 60 days. Adding the same donor’s BM to the CBUs with low cell counts probably improved the engraftment rate of our patients. The 3.5% bacterial contamination rate observed in our cohort was comparable to those reported in other studies;33,34 bacterial contaminations could be addressed at the time of transplantation with prophylactic antibiotics if necessary. Homozygous SCD was detected in 11% of the collections, a lower rate than might be expected in autosomal recessive genetic disorders. This might be explained by the quite frequent antenatal diagnosis testing performed in this high-risk population. Many important questions require further investigation, including the cost related to long-term storage of this graft source and its low utilization rate.

Ideally, the CBUs reported in our cohort should have been collected from families at risk of transmitting homozygous SCD, rather than those with an affected child. Storage-related cost remains a major limitation to such an ambitious program, when faced with the low utilization rate of these units. In our cohort, 43% of the units were typed for HLA upon the request of the referring physicians, and 30% of the typed units were HLA-identical to the intended recipient; an expected proportion that reflects the 25% probability of finding a matched sibling donor. To limit the unnecessary long-term storage and its associated costs, HLA testing and identification of SCD status for all collections could be implemented, so a better-informed decision could be made about keeping or disposing of those units with a low usage rate, such as non-HLA-identical units or those with HbS homozygous status. However, the deci-

Figure 3. Hemoglobin genotype of collected cord blood units. AA: normal hemoglobin genotype; AS: heterozygous βS; SS: homozyous βS/βS.

Table 5. Characteristics of the combined cord blood + bone marrow grafts. Patient weight at HSCT Patient age at HSCT, years Donor age at BM harvest, years CBU characteristics Collected volume (mL) Infused TNC (x107/kg) Infused CD34+ (x106/kg) Viability (%) BM characteristics Collected volume (mL) Infused TNC (x107/kg) Infused CD34+ (x106/kg) Viability (%) Total TNC infused (x107 kg) Total CD34 infused (x106/ kg)

Patient 1

Patient 2

Patient 3

Patient 4

Patient 5

56 12 4

26 10 3.5

14 4 3.1

33 13.6 2.9

35 12.5 2.6

65 0.6 0.03 90

52 1 0.02 73

49 1.2 0.03 83

57 1.1 0.04 76

55 0.2 0.02 76

405 26 10 96 26.6 10.03

310 6.3 2.27 94 7.3 2.29

264 25.5 5.4 94 26.7 5.43

330 17.4 2.44 97 18.5 2.48

300 13.6 5.89 96 13.8 5.91

HSCT: hematopoietic stem cell transplant; BM: bone marrow; CB: cord blood; CBU: cord blood unit; TNC: total nucleated cell.

haematologica | 2017; 102(6)

981


H. Rafii et al.

sion-making process must be managed with caution; nonHLA-identical CBUs could be kept to transplant homozygous identical siblings resulting from future pregnancies. Moreover, alternative experimental treatments might emerge in the future to support haploidentical-related allogeneic CBT36 or autologous CBT in the setting of gene therapy.37-42 Finally, enhancing the utilization rate of the banked unit will also achieve a significant reduction in storage needs and related costs. The transplantation experience reported here with a single center being involved in most of the transplants, strongly suggests that close collaboration with clinical and transplant teams willing to support CBT of the affected sibling is likely to increase the proportion of transplants using directed CBUs. An increase in the awareness of the excellent outcomes of CBT in this setting is also crucial to overcome the reluctance of physicians and families to consider transplant. However, despite the increase in number of homozygous cases of SCD in the developed countries and the excellent transplant outcomes, referrals for transplantation remain limited, as reported by our group in a recent review of sibling transplants.10

References 1. Modell B, Darlison M. Global epidemiology of haemoglobin disorders and derived service indicators. Bull World Health Organ. 2008;86(6):480-487. 2. Weatherall DJ. Hemoglobinopathies worldwide: present and future. Curr Mol Med. 2008;8(7):592-599. 3. Evidence-Based Management of Sickle Cell Disease: Expert Panel Report - National Heart, Lung, and Blood Institute; 2014. https://www.nhlbi.nih.gov/sites/www.nhl bi.nih.gov/files/sickle-cell-diseasereport.pdf. Last accessed 14 Dec 2016. 4. Adams RJ, McKie VC, Hsu L, et al. Prevention of a first stroke by transfusions in children with sickle cell anemia and abnormal results on transcranial Doppler ultrasonography. N Engl J Med. 1998; 339(1):5-11. 5. Quinn CT, Rogers ZR, McCavit TL, Buchanan GR. Improved survival of children and adolescents with sickle cell disease. Blood. 2010;115(17):3447-3452. 6. Gaston MH, Verter JI, Woods G, et al. Prophylaxis with oral penicillin in children with sickle cell anemia. A randomized trial. N Engl J Med. 1986;314(25):1593-1599. 7. Wang WC, Ware RE, Miller ST, et al. Hydroxycarbamide in very young children with sickle-cell anaemia: a multicentre, randomised, controlled trial (BABY HUG). Lancet. 2011;377(9778):1663-1672. 8. Yawn BP, Buchanan GR, Afenyi-Annan AN, et al. Management of sickle cell disease: summary of the 2014 evidence-based report by expert panel members. JAMA. 2014;312(10):1033-1048. 9. Gluckman E. Allogeneic transplantation strategies including haploidentical transplantation in sickle cell disease. Am Soc Hematol Educ Program Hematol. 2013; 2013:370-376. 10. Gluckman E, Cappelli B, Bernaudin F, et al. Sickle cell disease: an international survey

982

11.

12.

13.

14.

15.

16.

17.

18.

19.

The current study demonstrates that, despite the challenges associated with public directed CBB programs, sibling CBU represents a useful stem cell resource for families with specific medical indications, such as a child with sickle cell anemia or other genetic conditions that could benefit from a related transplant. These families should be encouraged to bank the CBUs from siblings. Family banking may yield good quality HLA-identical CBUs, with post-transplant outcomes and survival similar or superior to other stem cell sources. The potential scope of medically indicated sibling cord blood banking is considerable, including hemoglobin disorders, as well as genetic and metabolic diseases. For this reason, banks should adopt a more comprehensive approach for the identification and collection of CBUs for family-directed uses, and work closely with the clinical and transplant teams to optimize the use of these units. Acknowledgments The authors thank all participating families, transplant centers and delivery teams for the valuable contributions to the study and the organization â&#x20AC;&#x153;Cordon de vieâ&#x20AC;&#x153; in Monaco, for continuous support.

of results of HLA-identical sibling hematopoietic stem cell transplantation. Blood. 2017;129(11):1548-1556. Bernaudin F, Socie G, Kuentz M, et al. Long-term results of related myeloablative stem-cell transplantation to cure sickle cell disease. Blood. 2007;110(7):2749-2756. Hsieh MM, Fitzhugh CD, Tisdale JF. Allogeneic hematopoietic stem cell transplantation for sickle cell disease: the time is now. Blood. 2011;118(5):1197-1207. Gluckman E, Ruggeri A, Rocha V, et al. Family-directed umbilical cord blood banking. Haematologica. 2011;96(11):17001707. Arnold SD, Bhatia M, Horan J, Krishnamurti L. Haematopoietic stem cell transplantation for sickle cell disease - current practice and new approaches. Br J Haematol. 2016;174(4):515-525. Angelucci E, Matthes-Martin S, Baronciani D, et al. Hematopoietic stem cell transplantation in thalassemia major and sickle cell disease: indications and management recommendations from an international expert panel. Haematologica. 2014; 99(5):811-820. Hsieh MM, Fitzhugh CD, Weitzel RP, et al. Nonmyeloablative HLA-matched sibling allogeneic hematopoietic stem cell transplantation for severe sickle cell phenotype. JAMA. 2014;312(1):48-56. Walters MC, De Castro LM, Sullivan KM, et al. Indications and Results of HLAIdentical Sibling Hematopoietic Cell Transplantation for Sickle Cell Disease. Biol Blood Marrow Transplant. 2016;22(2):207211. Panepinto JA, Walters MC, Carreras J, et al. Matched-related donor transplantation for sickle cell disease: report from the Center for International Blood and Transplant Research. Br J Haematol. 2007;137(5):479485. Locatelli F, Rocha V, Reed W, et al. Related umbilical cord blood transplantation in

20.

21.

22.

23.

24.

25.

26.

patients with thalassemia and sickle cell disease. Blood. 2003;101(6):2137-2143. Locatelli F, Kabbara N, Ruggeri A, et al. Outcome of patients with hemoglobinopathies given either cord blood or bone marrow transplantation from an HLA-identical sibling. Blood. 2013;122(6):1072-1078. Kumar SR, Markusic DM, Biswas M, High KA, Herzog RW. Clinical development of gene therapy: results and lessons from recent successes. Mol Ther Methods Clin Dev. 2016;3:16034. Hoban MD, Cost GJ, Mendel MC, et al. Correction of the sickle cell disease mutation in human hematopoietic stem/progenitor cells. Blood. 2015;125(17):2597-2604. Cellular & Gene Therapy Guidances Guidance for Industry and FDA Staff: IND Applications for Minimally Manipulated, Unrelated Allogeneic Placental/Umbilical Cord Blood Intended for Hematopoietic and Immunologic Reconstitution in Patients with Disorders Affecting the Hematopoietic System. http:// www.fda.gov/biologicsbloodvaccines/guidancecomplianceregulatoryinformation/guidances/cellularandgenetherapy/ucm38821.htm. Last accessed 11 Dec 2016. Rocha V, Gluckman E, Eurocord-Netcord registry and European Blood and Marrow Transplant group. Improving outcomes of cord blood transplantation: HLA matching, cell dose and other graft- and transplantation-related factors. Br J Haematol. 2009; 147(2):262-274. Saccardi R, Tucunduva L, Ruggeri A, et al. Impact of cord blood banking technologies on clinical outcome: a Eurocord/Cord Blood Committee (CTIWP), European Society for Blood and Marrow Transplantation and NetCord retrospective analysis. Transfusion. 2016;56(8):20212029. Querol S, Mufti GJ, Marsh SGE, et al. Cord blood stem cells for hematopoietic stem

haematologica | 2017; 102(6)


Sibling cord blood banking for sickle cell disease

27.

28.

29.

30.

31.

cell transplantation in the UK: how big should the bank be? Haematologica. 2009; 94(4):536-541. Kurtzberg J, Cairo MS, Fraser JK, et al. Results of the cord blood transplantation (COBLT) study unrelated donor banking program. Transfusion. 2005;45(6):842-855. Ferster A, Corazza F, Vertongen F, et al. Transplanted sickle-cell disease patients with autologous bone marrow recovery after graft failure develop increased levels of fetal haemoglobin which corrects disease severity. Br J Haematol. 1995; 90(4):804-808. Paciaroni K, Gallucci C, De Angelis G, et al. Sustained and full fetal hemoglobin production after failure of bone marrow transplant in a patient homozygous for beta 0-thalassemia: a clinical remission despite genetic disease and transplant rejection. Am J Hematol. 2009;84(6):372-373. Goussetis E, Peristeri I, Kitra V, et al. Low usage rate of banked sibling cord blood units in hematopoietic stem cell transplantation for children with hematological malignancies: implications for directed cord blood banking policies. Blood Cells Mol Dis. 2011;46(2):177-181. Smythe J, Armitage S, McDonald D, et al. Directed sibling cord blood banking for

haematologica | 2017; 102(6)

32.

33.

34.

35.

36.

37.

transplantation: the 10-year experience in the national blood service in England. Stem Cells. 2007;25(8):2087-2093. Reed W, Smith R, Dekovic F, et al. Comprehensive banking of sibling donor cord blood for children with malignant and nonmalignant disease. Blood. 2003; 101(1):351-357. Screnci M, Murgi E, Valle V, et al. Sibling cord blood donor program for hematopoietic cell transplantation: the 20-year experience in the Rome Cord Blood Bank. Blood Cells Mol Dis. 2016;57:71-73. Allan D, Petraszko T, Elmoazzen H, Smith S. A review of factors influencing the banking of collected umbilical cord blood units. Stem Cells. 2013;2013:463031. Tucunduva L, Volt F, Cunha R, et al. Combined cord blood and bone marrow transplantation from the same human leucocyte antigen-identical sibling donor for children with malignant and non-malignant diseases. Br J Haematol. 2015;169(1):103-110. BolaĂąos-Meade J, Fuchs EJ, Luznik L, et al. HLA-haploidentical bone marrow transplantation with post transplant cyclophosphamide expands the donor pool for patients with sickle cell disease. Blood. 2012;120(22):4285-4291. Levasseur DN, Ryan TM, Pawlik KM,

38.

39.

40.

41.

42.

Townes TM. Correction of a mouse model of sickle cell disease: lentiviral/antisickling beta-globin gene transduction of unmobilized, purified hematopoietic stem cells. Blood. 2003;102(13):4312-4319. Chou Y-C, Chen R-L, Lai Z-S, et al. Pharmacological Induction of Human Fetal Globin Gene in Hydroxyurea-Resistant Primary Adult Erythroid Cells. Mol Cell Biol. 2015;35(14):2541-2553. Xu J, Peng C, Sankaran VG, et al. Correction of sickle cell disease in adult mice by interference with fetal hemoglobin silencing. Science. 2011;334(6058):993-996. Cavazzana M, Ribeil JA, Payen E, et al. Outcomes of Gene Therapy for Severe Sickle Disease and Beta-Thalassemia Major Via Transplantation of Autologous Hematopoietic Stem Cells Transduced Ex Vivo with a Lentiviral Beta AT87Q-Globin Vector. Blood. 2015;126(23):202. Malik P. Gene therapy for hemoglobinopathies: tremendous successes and remaining caveats. Mol Ther. 2016; 24(4):668-670. Kanter J, Walters MC, Hsieh MM, et al. Interim Results from a Phase 1/2 Clinical Study of Lentiglobin Gene Therapy for Severe Sickle Cell Disease. Blood. 2016; 128(22):1176-1176.

983


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Red Cell Biology & its Disorders

Ferrata Storti Foundation

Histone deacetylase 6 regulates cytokinesis and erythrocyte enucleation through deacetylation of formin protein mDia2

Xuehui Li,1# Yang Mei,2# Bowen Yan,1# Eric Vitriol,1 Suming Huang,3,4 Peng Ji2 and Yi Qiu1

Department of Anatomy and Cell Biology, College of Medicine, University of Florida, Gainesville, FL, USA; 2Department of Pathology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA; 3Department of Biochemistry and Molecular Biology, College of Medicine, University of Florida, Gainesville, FL, USA and 4Macau Institute for Applied Research in Medicine and Health, State Key Laboratory of Quality Research in Chinese Medicine, Macau University of Science and Technology, Avenida Wai Long, Taipa, Macau 1

Haematologica 2017 Volume 102(6):984-994

#These authors contributed equally to the manuscript.

ABSTRACT

T Correspondence: qiuy@ufl.edu or peng-ji@fsm.northwestern.edu Received: December 1, 2016. Accepted: February 27, 2017. Pre-published: March 2, 2017. doi:10.3324/haematol.2016.161513 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/984 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.

984

he formin protein mDia2 plays a critical role in a number of cellular processes through its ability to promote nucleation and elongation of actin filaments. In erythroblasts, this includes control of cytokinesis and enucleation by regulating contractile actin ring formation. Here we report a novel mechanism of how mDia2 is regulated: through acetylation and deacetylation at lysine 970 in the formin homology 2 domain. Ectopic expression of an acetyl-mimic mDia2 mutant in mouse erythroblasts is sufficient to abolish contractile actin ring formation at the cleavage furrow and subsequent erythrocyte cytokinesis and enucleation. We also identified that class II histone deacetylase 6 deacetylates and subsequently activates mDia2. Knockdown or inhibition of histone deacetylase 6 impairs contractile actin ring formation, and expression of a non-acetyl-mimic mDia2 mutant restores the contractile actin ring and rescues the impairment of enucleation. In addition to revealing a new step in mDia2 regulation, this study may unveil a novel regulatory mechanism of formin-mediated actin assembly, since the K970 acetylation site is conserved among Dia proteins Introduction Formin proteins play important roles in many different cellular processes such as cell adhesion, cell migration, cytokinesis, phagocytosis, endosomal trafficking, synaptic growth and the maintenance of cell polarity by directing actin nucleation and polymerization.1,2 The Diaphanous-related formins are a subfamily of formins that are effectors of Rho-family GTPases.1,3 The mammalian Diaphanous-related formin (mDia2) is important for cell motility, cell polarity, vesical trafficking and cytokinesis.1,3-6 All formin proteins contain the formin homology 2 (FH2) domain, which directly mediates actin assembly.3,7,8 Two FH2 domains form a doughnutshaped dimer that nucleates new, unbranched actin filaments and moves along the actin filaments to promote the addition of further actin monomers.2,9-11 The adjacent formin homology 1 domain accelerates actin assembly by recruitment of profilinbound G-actin.2,11,12 The process of erythropoiesis starts with the multipotent hematopoietic stem cell. The erythroid progenitor cells derived from these multipotent cells undergo terminal erythroid differentiation through a series of maturation stages to produce enucleated reticulocytes which subsequently mature into red blood cells. At the late stage of differentiation, the nuclear chromosomes become highly condensed and Factin bundles accumulate between the extruding nucleus and nascent reticulocyte. Resembling the process in cytokinesis, non-muscle myosin and F-actin form a contractile actin ring (CAR) at the cleavage furrow between the incipient reticulocyte and pyrenocyte, resulting in subsequent enucleation.13,14 Mice deficient in mDia2 haematologica | 2017; 102(6)


HDAC6 regulates mDia2 activity

survive until E11.5 and exhibit severe anemia with multinucleated erythroblasts. The mDia2-deficient erythroid cells fail to complete cytokinesis and have decreased accumulation of F-actin in the cleavage furrow during late differentiation from proerythroblasts.15 Conditional knockout of mDia2 at adult stage shows ineffective erythropoiesis with bi-and multi-nucleated erythroblasts,16 demonstrating the critical roles of mDia2 in erythropoiesis. Lysine acetylation is a widely occurring post-translational protein modification involved in various functions in transcriptional regulation and cellular processes in eukaryotic cells.17 The enzymes responsible for acetyl group addition to or removal from the ε-amino groups of target proteins are known as histone acetyltransferases and histone deacetylases (HDAC), respectively. HDAC6, a class IIb member of HDAC, has been shown to play essential roles in regulating cell migration, protein trafficking and accumulation of misfolded proteins into the aggressome.17,18 HDAC6 is a primary tubulin deacetylase and regulates tubulin-mediated cell motility, migration and cytokinesis.19-21 HDAC6 also regulates cell motility and endocytosis through regulating actin remodeling.22,23 The role of HDAC6 in hematopoiesis, especially erythropoiesis, is completely unknown. In this study, we found that knockdown of HDAC6 or inhibition of HDAC6 activity prevents cytokinesis and enucleation in cultured mouse fetal erythroblasts through disruption of CAR formation. We further discovered that HDAC6 regulates these processes through deacetylation of mDia2. The overexpression of non-acetyl-mimic mDia2 rescues the HDAC6 knockdown defect in enucleation. Our study therefore unveils a novel regulatory mechanism of formin protein by which mDia2 mediates actin organization and subsequently controls terminal erythroblast maturation.

Bone marrow transplantation Bone marrow cell infection and transplantation have been described elsewhere16 and were approved by the Northwestern University IACUC (2015-IS00001416). Briefly, c-Kit-positive stem/progenitor cells from mDia2 knockout mice were purified by mouse CD117 (c-KIT)-positive selection kit (STEMCELL Tech.). The cells were cultured in serum-free StemSpan SFEM medium and spin-infected with concentrated virus supernatants supplied with 8 μg/mL of polybrene at 2500 rpm for 90 min. The infected cells were recovered for 24 h at 37°C and then 2×106 cells were injected retro-orbitally into lethally irradiated CD45.1-positive recipient mice (gamma-irradiation, 1000 rad). One month after transplantation, reconstitution was achieved and recipient mice were sacrificed for the flow cytometric bone marrow enucleation assay staining with Ter119 and Hoechst 33342.

Immunocytochemistry and confocal microscopy Immunocytochemistry is described in the Online Supplementary Methods. Staining was examined with a Leica TCS-SP5 confocal imaging system (Leica Microsystems Inc.). To quantify polarized florescent signals, a line, with a thickness of 20 pixels, was drawn across the center of the cell towards the polarization direction using Fiji software. The polarization of the target signal was defined as a more than 2-fold change of florescence intensity between the maximal peak values from the side of the polarized cytoplasm and the nuclear side. At least five random pictures were taken from one slide, and more than 30 cells were measured for each experiment.

Flow cytometric analysis

Methods

For the in vitro analysis of differentiation, the mouse fetal erythroid progenitors were cultured in vitro for 24 h with erythropoietin and then the culture was continued for an additional 24 h without erythropoietin. Detached cells from culture were subjected to FACS analysis using antibodies against Ter119 PE and CD71 FITC (BD Pharmingen). For the in vitro analysis of enucleation, cells were stained with Hoechst 33342 (Sigma-Aldrich) and Ter119 PE. The FACS analysis was carried out using a BD LSR II (BD Biosciences).

Retroviral production and infection

Statistics

The generation of retroviral particles and viral infection have been described previously.24 The infection rate was determined at around 60% by infection of green-fluorescent protein (GFP)-containing virus. For cells co-infected with MSCV–GFP-mDia2 shRNA and MICD4 mDia2, positive cells were selected by GFP and CD4 surface marker. For cells co-infected with pSuper HDAC6 shRNA and MICD4 mDia2, positive cells were selected by CD4 surface marker.

Statistical significance was determined by a Student t test. P values <0.05 are considered statistically significant.

Purification and culture of mouse fetal liver cells All mouse procedures were conducted in accordance with the National Institutes of Health guidelines for the care and use of laboratory animals for research purposes and the collection of mouse fetal liver cells was approved by the University of Florida Institutional Animal Care and Use Committee (IACUC) (#201309309; #201609309). Mouse fetal liver cells were obtained from E13.5 C57BL/6 embryos. The purification and in vitro culture of mouse fetal liver erythroblast precursors (Ter119-negative cells) were performed as previously described.25 Purified cells were seeded in fibronectin-coated wells (BD Pharmingen) and differentiation induced in Iscove modified Dulbecco medium (IMDM) containing 1 U/mL erythropoietin (Amgen). The medium was changed to erythropoietin-free IMDM 24 h after induction. haematologica | 2017; 102(6)

Results Inhibition of HDAC6 impairs cytokinesis and erythrocyte enucleation During erythropoiesis, erythroid progenitors undergo a series of differentiation stages with cell division, nuclear condensation and eventually enucleation, a type of cytokinesis which requires actin remodeling.26-28 It has been shown that the class IIb HDAC6 plays a role in cytokinesis and endocytosis.20,23 We, therefore, investigated whether HDAC6 is also involved in erythrocyte maturation. We first examined the expression and localization of HDAC6 during erythroid differentiation using Ter119negative erythroid progenitor cells purified from E13.5 mouse fetal livers. The progenitor cells were cultured and induced to differentiate into erythroblasts in vitro by addition of erythropoietin.25 HDAC6 was evenly distributed in cytoplasm during the early stage of erythropoiesis. As the nucleus became polarized during the later stage of erythropoiesis, HDAC6 gradually accumulated at the boundary of the cytoplasm and nucleus (Figure 1A). During enu985


X. Li et al.

A

B

C

D

E

Figure 1. Inhibition of HDAC6 impairs cytokinesis and erythrocyte enucleation. (A) Confocal microscopy analysis of HDAC6 cellular distribution during Ter119-negative mouse fetal progenitor cell differentiation. The Ter119-negative erythroid progenitor cells were harvested from E13.5 mouse fetal livers, and cultured in vitro with erythropoietin for 24 h, and then the culture continued for an additional 24 h without erythropoietin. The cells were fixed at 0, 6 h, 12 h, and 48 h during the culture and immunostained with anti-HDAC6 conjugated with Alexa Fluor 488 and DAPI. Scale bar is 5 μm. The dashed line indicates the cell boundary. (B) Flow cytometric analysis of cultured Ter119-negative mouse fetal progenitors. DMSO, TubA (5 μM) or TubA (10 μM) was added at the time of erythropoietin induction and the cells were treated for 48 h. Cells were stained with Ter119 and CD71, and analyzed by FACS. The percentages of cells in an undifferentiated state (R1+R2) and differentiated state (R3+R4) were analyzed. (C) Flow cytometric analysis of induced Ter119-negative mouse fetal progenitors. DMSO, TubA (5 μM) or TubA (10 μM) was added at the time of erythropoietin induction and the cells were treated for 48 h. Cells were stained with Ter119 and Hoechst 33342, and analyzed by FACS. The percentage of the gated cells (Ter119-positive and Hoechst-negative cells), which represent the pool of enucleated reticulocytes, was analyzed. (D and E) DMSO, TubA (5 μM) or TubA (10 μM) was added 36 h after erythropoietin induction and the cells were treated for an additional 12 h. Cells were stained with (D) Ter119 and Hoechst 33342, or (E) Ter119 and CD71, and analyzed by FACS. The error bars represent mean +SD (n=3), *P<0.05 compared to DMSO treatment.

986

haematologica | 2017; 102(6)


HDAC6 regulates mDia2 activity

B

A

C

D

F

E

G

Figure 2. Inhibition of HDAC6 blocks contractile actin ring formation. (A) Immunostaining of HDAC6 and F-actin in cultured Ter119-negative mouse fetal progenitors. i-iv indicate progressive stages during enucleation. (B) Immunostaining of HDAC6 and F-actin in differentiating Ter119-negative mouse fetal progenitors treated with DMSO or TubA. (C) Quantification of cells with polarized F-actin [as shown in stages ii to iv in (A)] in cells treated with DMSO or TubA. (D) Immunostaining of HDAC6 and F-actin in Ter119-negative mouse fetal progenitors infected with retrovirus harboring pSuper vector or HDAC6 shRNA. (E) Quantification of polarized F-actin in cells with HDAC6 knockdown. (F) Immunostaining of HDAC6 and F-actin in cultured Ter119-negative mouse fetal progenitors treated with TubA or HDAC6 shRNA. (G) Quantification of enucleated cells in induced mouse fetal liver progenitor cells with HDAC6 shRNA or treated with TubA. Scale bar is 5 ÎźM. The error bars represent mean +SD (n=3), *P<0.05 compared to DMSO.

haematologica | 2017; 102(6)

987


X. Li et al.

A

C

B

D Figure 3. HDAC6 is required for recruitment of mDia2 at the contractile actin ring. (A) Immunostaining of mDia2 and Factin in cultured erythropoietininduced Ter119-negative mouse fetal progenitors treated with DMSO or TubA. (B) Quantification of cells with polarized mDia2. (C) Immunostaining of mDia2 and Factin in cultured uninduced Ter119-negative mouse fetal progenitors treated with DMSO or TubA. (D) Quantification of multinuclear cells as described in (C). Scale bar is 5 Îźm. The error bars represent mean +SD (n=3), *P<0.05 compared to DMSO.

cleation, HDAC6 localized at the sites of CAR formation between the incipient reticulocytes and pyrenocytes (Figure 1A). The total protein and gene expression level of HDAC6 did not change significantly during erythropoesis (Online Supplementary Figure S1A, S1B). The dynamic localization of HDAC6 to the cytokinetic furrow suggests that it may play an important role there. To determine whether HDAC6 is involved in erythroid cell maturation, the mouse erythroid progenitor cells were induced to differentiate by addition of erythropoietin and treated with tubastatin A (TubA), a HDAC6-specific inhibitor.29 The expression level of HDAC6 was not affected by TubA treatment (Online Supplementary Figure S1A-C); however the deacetylase activity of HDAC6 was inhibited (Online Supplementary Figure S1B). FACS analysis was utilized to study the erythrocyte differentiation. Cell populations were gated and quantified based on expression levels of CD71 and Ter119.25 The population of differentiated erythrocytes, which were positive for both CD71 and Ter119 (R3 + R4), was profoundly reduced by TubA (Figure 1B). The enucleation efficiency was also analyzed by FACS using co-staining with Hoechst for DNA and Ter119. The population of enucleated cells, which were positive for Ter119 and negative for Hoechst, was also reduced by TubA treatment (Figure 1C). These results demonstrate that inhibition of HDAC6 affects erythroid progenitor cell differentiation and enucleation. An alternate explanation 988

for these results could be that the loss of enucleated cells is due to a reduction of differentiated cells. However, when the erythroid progenitor cells were induced to differentiate by erythropoietin for 36 h prior to treatment with TubA for 12 h, there was a significant decrease in the enucleated cell population (Figure 1D) without the R3 + R4 population being affected significantly (Figure 1E). It is known that the class III HDAC SIRT2 and HDAC6 have an overlapping function as both enzymes can deacetylate tubulin.30 We, therefore, tested whether SIRT2 also affects enucleation. However, treatment with class III HDAC inhibitors, sirtinol or nicotinamide, did not affect enucleation (Online Supplementary Figure S1D,E). These results reveal a functional significance of HDAC6 in differentiation and enucleation during the late stage of terminal erythropoiesis.

Inhibition of HDAC6 blocks recruitment of mDia2 at the contractile actin ring Remodeling of actin filaments is critical to erythroblast enucleation.27,28 The formation of a CAR at the boundary between the incipient reticulocyte and pyrenocyte (condensed nucleus) is a phenomenon distinct to enucleation. At the late erythroblast stage, HDAC6 gradually moves toward the boundary of the nucleus and cytoplasm (Figure 1A), which led us to further examine HDAC6 function on the formation of the CAR. During the early stages of eryhaematologica | 2017; 102(6)


HDAC6 regulates mDia2 activity

A

C

D

B

thropoiesis, actin filaments are distributed at the plasma membrane and are partially co-localized with HDAC6 (Figure 2A). During nuclear condensation, actin filaments gradually accumulate at the cleavage furrow to form the CAR (Figure 2A). In cells treated with TubA, F-actin remains at the cell periphery and CAR formation is blocked, even though there is no significant defect in nuclear polarization (Figure 2B,C, Online Supplementary Figure S2A,B). To further confirm the role of HDAC6 in enucleation, HDAC6 was knocked down with retrovirus encoding shRNA. Knockdown of HDAC6 also significantly blocked F-actin polarization to the CAR (Figure 2D,E). Interestingly, a few erythroblasts with inactivated HDAC6 or HDAC6 knockdown eventually enucleated (Figure 2F,G). However, even in these cells, CAR formation was disrupted. Taken together, these results suggest that HDAC6 is required for the organization of actin filaments into the CAR. An earlier study indicated that formin protein mDia2 is required for the formation of the CAR and subsequent enucleation.6 We therefore examined whether inhibition of HDAC6 affects mDia2 function. In normal cultured mouse fetal erythroblasts, mDia2 gradually polarizes to one side of the plasma membrane and partially co-localizes with actin filaments where the CAR is formed during the enucleation process (Figure 3A). Treatment with TubA significantly impaired the polarization of mDia2 (Figure 3A,B). Since mDia2 is also important for cytokinesis,4,15 we then tested whether HDAC6 is also important for mDia2mediated cytokinesis. Control dividing cells form the CAR at the cleavage furrow (Figure 3C, left panel). mDia2 is polarized to the front edge of the CAR (Figure 3C, left haematologica | 2017; 102(6)

Figure 4. mDia2 interacts with HDAC6 during erythropoiesis. Confocal microscopy analysis of HDAC6 and mDia2 co-localization in cultured Ter119-negative mouse fetal progenitors treated with (A) DMSO or (B) TubA. The florescent intensity was measured using Volocity imaging software. Scale bar is 5 Îźm. (C) 293T cells were transfected with GFP-Flagtagged mDia2 and Flag-tagged HDAC6 or HDAC6 mutant as indicated. The cell lysate was immunoprecipitated with GFP antibody. Associated proteins were detected by western blotting with antibodies as indicated. (D) Endogenous HDAC6 in MEL cells treated with TubA and immunoprecipitated with HDAC6 antibody. Associated mDia2 was detected by western blotting using anti-mDia2 antibody.

panel). TubA-treated cells failed to form the cleavage furrow or CAR and lacked mDia2 localization at the cleavage site. Some TubA-treated cells had two or more nuclei, indicating that nuclear division had occurred (Figure 3C, top right panel; Figure 3D; Online Supplementary Figure S3A). These cells grew at a slower rate (Online Supplementary Figure S3B,C), supporting the notion that cytokinesis is blocked. A few treated cells did manage to go through complete cytokinesis without the formation of an actin ring (Figure 3C, bottom right panel), suggesting an alternative mechanism for cytokinesis without the CAR. Other cell types, such as NIH3T3 fibroblasts, also require HDAC6 activity for cytokinesis (Online Supplementary Figure S3D). Our study therefore demonstrates that HDAC6 is important for mDia2-mediated cytokinesis and enucleation.

mDia2 is a novel target of HDAC6 To further investigate how HDAC6 affects mDia2mediated enucleation, we first examined whether HDAC6 interacts with mDia2 in erythroid cells. mDia2 co-localized with HDAC6 during both early and late stages of erythropoiesis in mouse fetal erythroid progenitors, suggesting that these two proteins may have an interaction (Figure 4A). The co-localization was not affected by TubA treatment (Figure 4B). The interaction between HDAC6 and mDia2 was confirmed by immunoprecipitation analysis. GFP-Flag-mDia2 was co-transfected into 293T cells with either wild-type Flag-HDAC6 or catalytically inactive Flag-HDAC6 mutant. HDAC6 was co-precipitated with GFP mDia2 (Figure 4C). Consistent with the imaging results described above, both wild-type 989


X. Li et al. A

D

B

E

C

Figure 5. mDia2 is acetylated in vivo and HDAC6 is responsible for deacetylation of mDia2. (A) Total acetylated protein was immunoprecipitated by anti-acetyl-K antibody from the cell extracts treated with or without TSA in Ter119-negative mouse fetal progenitors. mDia2 protein was detected by specific anti-mDia2 antibody using western blotting. (B) Schematic representation of mDia2 protein structure. K970 is the acetylation site in the FH2 domain of mDia2. (C) 293T cells were transfected with Flag-tagged mDia2 WT (F-WT), mDia2 K970R (F-KR), or mDia2 K970Q (F-KQ). The cell extracts were incubated with anti-Flag antibody and acetylated mDia2 and total Flag-tagged mDia2 were detected by anti-acetyl-lysine (anti-acetyl-K) and anti-Flag antibodies. (D) Total acetylated protein was immunoprecipitated by anti-acetyl-K antibody from the cell extracts treated with or without inhibitors in MEL cells as indicated. Acetylated mDia2 protein was detected by anti-mDia2 antibody. The numbers indicate the relative density of the bands. (E) Total acetylated protein was immunoprecipitated by anti-acetyl-K antibody from the cell extracts of stable HDAC6 knockdown (KD) or scramble control MEL cells. mDia2 protein was detected by anti-mDia2 antibody. The numbers indicate the relative density of the bands. Each experiment was repeated three times.

HDAC6 and catalytically inactive HDAC6 were found in the mDia2 complex (Figure 4C), indicating that deacetylase activity is not required for the interaction. We further confirmed this interaction with endogenous proteins in cells of the murine erythroleukemia (MEL) line. The endogenous HDAC6 protein interacted with mDia2 in both TubA-treated and control cell lysates (Figure 4D). These results indicate that mDia2 forms a complex with HDAC6 and that the inhibition of deacetylase activity does not interfere with the integrity of the complex. As HDAC6 is a deacetylase for a number of cellular proteins, we hypothesized that mDia2 is a novel target of HDAC6. First, we examined whether mDia2 can be acetylated in vivo. In mouse fetal erythroid progenitors, mDia2 was acetylated at a low level and the acetylation was dramatically increased after treatment with trichostatin A (TSA), a broad-spectrum HDAC inhibitor (Figure 5A). This shows that mDia2 may be dynamically acetylated and deacetylated during erythropoiesis and inhibition of HDAC activity enhances mDia2 acetylation in vivo. Next, we wanted to identify the acetylation site on mDia2. Since TubA treatment blocks the recruitment of mDia2 and F-actin at the CAR, it is conceivable that the acetylation site is localized at the FH2 domain, the domain that mediates actin binding and polymerization.2 Through a protein alignment study with mDia proteins among model organisms, we found that lysine 970 of the FH2 domain of mDia2 is highly conserved (Figure 5B, 990

Online Supplementary Figure S4). To determine whether K970 is a potential acetylation site, we generated Flagtagged K970R and K970Q mDia2 mutants to mimic unacetylated and acetylated forms, respectively, and expressed them in 293T cells. The wild-type (WT) mDia2 in 293T cells was acetylated. Mutations on lysine 970 abolished the acetylation (Figure 5C). Taken together, these results suggest that mDia2 is an acetylated protein and lysine 970 in the FH2 domain of mDia2 is a major acetylation site in vivo. To examine whether HDAC6 is responsible for mDia2 deacetylation, we treated MEL cells with various HDAC inhibitors, TubA, valproic acid (a class I and IIa inhibitor), or sirtinol (a class III HDAC inhibitor). Total acetylated lysine proteins were immunoprecipitated with an acetyllysine antibody from cell lysates and the presence of mDia2 protein was detected. mDia2 acetylation was dramatically increased in TubA-treated cells, but not in the other inhibitor-treated cells (Figure 5D). Furthermore, mDia2 acetylation was significantly increased with HDAC6 knockdown (Figure 5E), indicating that HDAC6 is indeed a major deacetylase for mDia2 in vivo.

HDAC6 regulates enucleation through deacetylation of mDia2 In order to determine whether deacetylation of mDia2 is required for CAR formation, we simultaneously knocked down mDia2 with shRNA (Online Supplementary haematologica | 2017; 102(6)


HDAC6 regulates mDia2 activity

B

A

C

D

F

E

G

H

I

Figure 6. Acetylation of mDia2 impairs enucleation. (A) Ter119-negative mouse fetal progenitors were infected with retrovirus carrying GFP or mDia2 shRNA (KD), and shRNA resistant mDia2 WT, mDia2 K970R (KR), mDia2 K970Q (KQ), or control vector as indicated. Cells were fixed and subjected to immunostaining at 48 h after erythropoietin induction. Scale bar is 5 μm. (B) Quantification of cells with polarized F-actin in cells described in (A). The error bars represent mean +SD (n=3), *P<0.05 compared to the GFP/vector control. (C) Immunostaining of mDia2 in cultured Ter119-negative mouse fetal progenitors infected with mDia2 shRNA (KD) and shRNA resistant mDia2 WT, or mDia2 K970Q (KQ). Representative images are shown. Scale bar is 5 μm. (D) Quantification of cells with polarized mDia2 in cells described in (C). The error bars represent mean +SD (n=3), *P<0.05 compared to mDia2 KD/WT. (E) Flow cytometric analysis of cultured Ter119-negative mouse fetal progenitors described in (A). Cells were harvested 48 h after erythropoietin induction and GFP and CD4-positive cells, which harbor mDia2 shRNA and mDia2 KD, were collected and subjected to FACS analysis. The percentage of enucleated reticulocytes from a representative experiment is indicated. (F) Quantification of enucleation in cultured Ter119-negative mouse fetal progenitors described in (E) at 48 h. The enucleation rate was normalized to the GFP/vector control. The error bars represent mean +SD (n=3), *P<0.05 compared to the vector control. (G to I) c-Kit-positive stem/progenitor cells from control or mDia2 conditional knockout (cKO) mice were infected with empty retroviral vector or retrovirus encoding mDia2 WT and KQ mutant as indicated. The cells were then transplanted into lethally irradiated CD45.1-positive recipient mice. One month after bone marrow transplantation, mice were sacrificed for analysis of (G) circulating red blood cell count; (H) hemoglobin level and (I) flow cytometry analysis of enucleation in bone marrow cells by Ter119/Hoechst 33342 staining. Results are representative of three experiments. Data were shown as mean ± SEM. *P<0.05; **P<0.01 and ***P<0.001.

haematologica | 2017; 102(6)

991


X. Li et al. A

B

C

Figure 7. mDia2 rescues enucleation in HDAC6 knockdown mouse fetal progenitors. (A) Immunostaining of actin in cultured Ter119-negative mouse fetal progenitors infected with HDAC6 shRNA (KD) and mDia2 K970R or control viruses. Representative images are shown. Scale bar is 5 Îźm. (B) Quantification of cells with polarized F-actin in cells described in (A). (C) Schematic representation of the HDAC6mDia2 pathway in cytokinesis and enucleation. Deacetylation of mDia2 by HDAC6 promotes formation of the CAR and subsequent cytokinesis and enucleation. Inhibition of HDAC6 could cause the accumulation of acetylated mDia2, which impairs CAR formation and subsequent cytokinesis and enucleation processes.

Figure S5A,B) and re-expressed shRNA resistant mDia2 WT, non-acetyl-mimicking mDia2 K970R, or acetyl-mimicking mDia2 K970Q (Online Supplementary Figure S5C). The WT and K970R mutant, but not the K970Q mutant, rescued the defective F-actin polarization phenotype in mDia2 knockdown cells (Figure 6A,B). The K970Q mutant also failed to rescue the mDia2 polarization phenotype (Figure 6C,D). Importantly, WT or K970R rescued the enucleation in mDia2 knockdown cells (Figure 6E,F). As expected, the mDia2 K970Q mutant was not able to rescue the defects of enucleation in mDia2 knockdown cells (Figure 6E,F). To test whether this is also the case in vivo, bone marrow cells from mDia2 knockout mice were infected with retrovirus encoding WT mDia2 or K970Q mutant. Infected bone marrow cells were transplanted into lethally irradiated recipient mice and bone marrow was collected 1 month later. The mice with mDia2 knockout had low red blood counts and low hemoglobin levels. This phenotype was not rescued in mice transplanted with K970Q bone marrow, while it was partially rescued in those transfected with WT mDia2 (Figure 6G,H). Importantly, the number of enucleated red cells in bone marrow was significantly reduced in K970Q-rescued mice 992

(Figure 6I). The more severe effect of K970Q in bone marrow than in the in vitro assay may be due to the fact that K970Q affects both cytokinesis and enucleation in bone marrow. We, thus, demonstrated that deacetylation of mDia2 is required for terminal erythropoiesis in vitro and in vivo. Since we identified that HDAC6 is the deacetylase for mDia2, and mDia2 deacetylation is essential for CAR formation, we hypothesized that HDAC6 mediates erythroblast enucleation through deacetylation of mDia2. If this is the case, then overexpression of an unacetylated mDia2 mimic mutant should rescue CAR and enucleation defects in HDAC6 knockdown cells. To test this, we overexpressed mDia2 K970R or control vector in HDAC6 knockdown Ter119-negative erythroid progenitors (Online Supplementary Figure S6). The re-expression of mDia2 K970R mutant rescued F-actin polarization at the cleavage furrow in the HDAC6 knockdown cells (Figure 7A,B). These results provide direct evidence that mDia2 acts downstream of the HDAC6-regulated pathway in erythroblast enucleation and deacetylation of mDia2 by HDAC6 is required for CAR formation and proper enucleation during erythropoiesis (Figure 7C). haematologica | 2017; 102(6)


HDAC6 regulates mDia2 activity

Discussion HDAC6 regulates various cellular processes, such as cell migration, vesicle trafficking, endocytosis and cytokinesis.18,19,23,31-33 Our study here demonstrates that HDAC6 can regulate actin nucleation through deacetylation of the formin protein mDia2, which reveals a RhoGTPase-independent mechanism that regulates mDia2 activity. Other reports show that HDAC6 plays an important role in regulating actin-mediated endocytosis and osteoclast maturation.20,25,34 It is yet to be determined whether other HDAC6-dependent processes, such as endocytosis or cell motility, also involve deacetylation of mDia2 or other formin proteins. It has been well demonstrated that HDAC6 interacts with and deacetylates cytoplasmic α-tubulin and subsequently affects microtubule organization.19,25,33,35 Although there is no direct evidence showing that tubulin acetylation affects cytokinesis, it has been shown that inhibition of Arl3 results in increased tubulin acetylation and failure of cytokinesis.36 In addition, mDia2 stabilizes microtubules by enhancing HDAC6 tubulin deacetylase activity.34,37 It is, therefore, possible that tubulin deacetylation by HDAC6 may also be involved in cytokinesis. The F-actin binding protein cortactin is an acetylated protein and is important for CAR formation.22,38 Acetylated cortactin dissociates from actin filaments and HDAC6 deacetylates cortactin.22 It is, therefore, conceivable that HDAC6-mediated cortactin deacetylation may also play an important role in cytokinesis. Thus, HDAC6 may have multiple roles in regulating cytokinesis and enucleation through modulating acetylation levels of various proteins that are involved in the regulation of actin and tubulin networks. However, tubulin and cortactin deacetylation may not play an essential role in cytokinesis or enucleation as non-acetyl-mimicking mDia2 can rescue HDAC6 knockdown defects. Enucleation is considered as an asymmetric cytokinesis event.13 In mammalian cells, formation of the CAR is required for successful cytokinesis. However, in our study, we found that although in low percentage, erythrocytes without functional HDAC6 can divide or enucleate without apparent accumulation of mDia2 or F-actin bundles at the cleavage farrow (Figures 2F and 3C). This is interesting but consistent with a recent finding in vivo using a mDia2 whole body knockout mouse model. These mice die as embryo on day 13.5 with severe defects in primitive erythropoiesis. Similar to what we observed here, although enucleation is significantly influenced, some primitive erythroblasts managed to enucleate.15 We suspect that an alternative pathway may be utilized when the CAR is absent. Indeed, yeast and Dictyostelium cells are able to divide without actin or myosin II, two key components of the CAR.3941 Furthermore, it has been shown that an actin-binding protein, cortexillin, may rescue cytokinesis in the absence of the CAR.42 It remains unclear what mechanism is involved in erythrocyte enucleation in the absence of the CAR in mammalian cells. It is worth mentioning that HDAC6 knockout mice are viable with no developmental defect.23,43 Although the

haematologica | 2017; 102(6)

mice have hyperacetylated tubulin and elevated acetylated HSP90, they do not have significant cellular defects, as shown in transient knockdown cell lines,18,43 suggesting that germline deletion of HDAC6 may result in a compensation of HDAC6 function. Consistent with this notion, except no response to HDAC6 inhibitor, the Ter119-negative fetal progenitor cells isolated from knockout mice differentiate and enucleate normally (data not shown). These cells can form a normal CAR and mDia2 recruitment is also normal (data not shown). Since HDAC6 is a unique deacetylase for cellular functions, it remains to be determined which protein or deacetylase plays a role in compensating HDAC6 function. Formin proteins play critical roles in cell morphology, motility, polarity, and division, which are essential for cell growth and differentiation.2 Formin protein activities are regulated through interactions with various binding proteins. Post-translational modifications are also important in modulating formin activity. There is one report describing that phosphorylation of mDia2 on the diaphanous autoregulatory domain (DAD) enhances its activity by reducing the interaction with the diaphanous inhibitory domain (DID).44 In this study, we uncovered a novel posttranslational modification for formin proteins at their FH2 domain. We found that mDia2 is acetylated and that the acetylation abolishes mDia2-mediated CAR formation. We further discovered that the acetylation is reversible and HDAC6-mediated deacetylation of mDia2 restores CAR formation. Interestingly, the acetylation site of mDia2, K970, is located in the FH2 domain, the key domain that forms dimers and promotes actin filament nucleation and actin polymerization through an interaction with the barbed end of actin filaments. The FH2 domain is highly conserved throughout formin proteins (Online Supplementary Figure S4).2 Importantly, the acetylation site on FH2 is sequence- and structurally-conserved within model organisms, indicating that acetylation may have a conserved function (Online Supplementary Figure S4).11,45 The crystal structure shows that FH2 domains form a stable, but flexible dimeric “donut” structure.11,12 Interestingly, the crystal structure of the FH2 domain with actin suggests that the acetylation site may not be important for direct actin binding and FH2 dimerization.12 The function of FH2 domain acetylation on actin nucleation does, therefore, require further investigation. In addition, the acetylation site is also conserved in mDia1 and the acetylation may, therefore, have a more general effect on actin nucleation beyond cytokinesis and erythrocyte enucleation processes. Acknowledgments We thank Dr. Narumiya (Kyoto University, Kyoto, Japan) for the mDia2 plasmid. We are very grateful to Tao Yang, Yuanjing Liu for their preliminary study and technical assistance, and to Tracy Read for technical assistance. This work was supported by the National Institutes of Health (grants R01 HL095674 to YQ; DK102718 to PJ and DK110108 to SH). The work was also supported by the Department of Defense grant CA140119 to PJ.

993


X. Li et al.

References 1. Bogdan S, Schultz J, Grosshans J. Formin' cellular structures: physiological roles of Diaphanous (Dia) in actin dynamics. Commun Integr Biol. 2013;6(6):e27634. 2. Wallar BJ, Alberts AS. The formins: active scaffolds that remodel the cytoskeleton. Trends Cell Biol. 2003;13(8):435-446. 3. Watanabe N, Madaule P, Reid T, et al. p140mDia, a mammalian homolog of Drosophila diaphanous, is a target protein for Rho small GTPase and is a ligand for profilin. EMBO J. 1997;16(11):3044-3056. 4. Watanabe S, Ando Y, Yasuda S, et al. mDia2 induces the actin scaffold for the contractile ring and stabilizes its position during cytokinesis in NIH 3T3 cells. Mol Biol Cell. 2008;19(5):2328-2338. 5. Wallar BJ, Deward AD, Resau JH, Alberts AS. RhoB and the mammalian Diaphanous-related formin mDia2 in endosome trafficking. Exp Cell Res. 2007;313(3):560-571. 6. Ji P, Jayapal SR, Lodish HF. Enucleation of cultured mouse fetal erythroblasts requires Rac GTPases and mDia2. Nat Cell Biol. 2008;10(3):314-321. 7. Romero S, Le Clainche C, Didry D, Egile C, Pantaloni D, Carlier MF. Formin is a processive motor that requires profilin to accelerate actin assembly and associated ATP hydrolysis. Cell. 2004;119(3):419-429. 8. Kovar DR, Harris ES, Mahaffy R, Higgs HN, Pollard TD. Control of the assembly of ATPand ADP-actin by formins and profilin. Cell. 2006;124(2):423-435. 9. Higgs HN. Formin proteins: a domain-based approach. Trends Biochem Sci. 2005;30(6): 342-353. 10. Paul AS, Pollard TD. Review of the mechanism of processive actin filament elongation by formins. Cell Motil Cytoskeleton. 2009;66(8):606-617. 11. Shimada A, Nyitrai M, Vetter IR, et al. The core FH2 domain of diaphanous-related formins is an elongated actin binding protein that inhibits polymerization. Mol Cell. 2004;13(4):511-522. 12. Thompson ME, Heimsath EG, Gauvin TJ, Higgs HN, Kull FJ. FMNL3 FH2-actin structure gives insight into formin-mediated actin nucleation and elongation. Nat Struct Mol Biol. 2013;20(1):111-118. 13. Ji P, Murata-Hori M, Lodish HF. Formation of mammalian erythrocytes: chromatin condensation and enucleation. Trends Cell Biol. 2011;21(7):409-415. 14. An X, Mohandas N. Erythroblastic islands, terminal erythroid differentiation and reticulocyte maturation. Int J Hematol. 2011;93(2): 139-143. 15. Watanabe S, De Zan T, Ishizaki T, et al. Loss of a Rho-regulated actin nucleator, mDia2, impairs cytokinesis during mouse fetal erythropoiesis. Cell Rep. 2013;5(4):926-932. 16. Mei Y, Zhao B, Yang J, et al. Ineffective erythropoiesis caused by binucleated late-stage erythroblasts in mDia2 hematopoietic specif-

994

17. 18.

19. 20.

21.

22.

23.

24.

25.

26. 27.

28.

29.

30.

31.

ic knockout mice. Haematologica. 2016;101 (1):e1-5. Yang XJ, Seto E. Lysine acetylation: codified crosstalk with other posttranslational modifications. Mol Cell. 2008;31(4):449-461. Valenzuela-Fernandez A, Cabrero JR, Serrador JM, Sanchez-Madrid F. HDAC6: a key regulator of cytoskeleton, cell migration and cell-cell interactions. Trends Cell Biol. 2008;18(6):291-297. Hubbert C, Guardiola A, Shao R, et al. HDAC6 is a microtubule-associated deacetylase. Nature. 2002;417(6887):455-458. Wickstrom SA, Masoumi KC, Khochbin S, Fassler R, Massoumi R. CYLD negatively regulates cell-cycle progression by inactivating HDAC6 and increasing the levels of acetylated tubulin. EMBO J. 2010;29(1):131-144. Zhang Y, Li N, Caron C, et al. HDAC-6 interacts with and deacetylates tubulin and microtubules in vivo. EMBO J. 2003;22(5):11681179. Zhang X, Yuan Z, Zhang Y, et al. HDAC6 modulates cell motility by altering the acetylation level of cortactin. Mol Cell. 2007;27(2):197-213. Gao YS, Hubbert CC, Lu J, Lee YS, Lee JY, Yao TP. Histone deacetylase 6 regulates growth factor-induced actin remodeling and endocytosis. Mol Cell Biol. 2007;27(24): 8637-8647. Liu Y, Peng L, Seto E, Huang S, Qiu Y. Modulation of histone deacetylase 6 (HDAC6) nuclear import and tubulin deacetylase activity through acetylation. J Biol Chem. 2012;287(34):29168-29174. Zhang J, Socolovsky M, Gross AW, Lodish HF. Role of Ras signaling in erythroid differentiation of mouse fetal liver cells: functional analysis by a flow cytometry-based novel culture system. Blood. 2003;102(12):39383946. Mohandas N, Gallagher PG. Red cell membrane: past, present, and future. Blood. 2008;112(10):3939-3948. Chasis JA, Prenant M, Leung A, Mohandas N. Membrane assembly and remodeling during reticulocyte maturation. Blood. 1989;74(3): 1112-1120. Koury ST, Koury MJ, Bondurant MC. Cytoskeletal distribution and function during the maturation and enucleation of mammalian erythroblasts. J Cell Biol. 1989;109(6 Pt 1):3005-3013. Butler KV, Kalin J, Brochier C, Vistoli G, Langley B, Kozikowski AP. Rational design and simple chemistry yield a superior, neuroprotective HDAC6 inhibitor, tubastatin A. J Am Chem Soc. 2010;132(31):10842-10846. North BJ, Marshall BL, Borra MT, Denu JM, Verdin E. The human Sir2 ortholog, SIRT2, is an NAD+-dependent tubulin deacetylase. Mol Cell. 2003;11(2):437-444. Gao YS, Hubbert CC, Yao TP. The microtubule-associated histone deacetylase 6 (HDAC6) regulates epidermal growth factor receptor (EGFR) endocytic trafficking and degradation. J Biol Chem. 2010;285(15):

11219-11226. 32. Serrador JM, Cabrero JR, Sancho D, Mittelbrunn M, Urzainqui A, SanchezMadrid F. HDAC6 deacetylase activity links the tubulin cytoskeleton with immune synapse organization. Immunity. 2004;20(4):417-428. 33. Matsuyama A, Shimazu T, Sumida Y, et al. In vivo destabilization of dynamic microtubules by HDAC6-mediated deacetylation. Embo J. 2002;21(24):6820-6831. 34. Destaing O, Saltel F, Gilquin B, et al. A novel Rho-mDia2-HDAC6 pathway controls podosome patterning through microtubule acetylation in osteoclasts. J Cell Sci. 2005;118(Pt 13):2901-2911. 35. Haggarty SJ, Koeller KM, Wong JC, Grozinger CM, Schreiber SL. Domain-selective small-molecule inhibitor of histone deacetylase 6 (HDAC6)-mediated tubulin deacetylation. Proc Natl Acad Sci USA. 2003;100(8):4389-4394. 36. Zhou C, Cunningham L, Marcus AI, Li Y, Kahn RA. Arl2 and Arl3 regulate different microtubule-dependent processes. Mol Biol Cell. 2006;17(5):2476-2487. 37. Bartolini F, Moseley JB, Schmoranzer J, Cassimeris L, Goode BL, Gundersen GG. The formin mDia2 stabilizes microtubules independently of its actin nucleation activity. J Cell Biol. 2008;181(3):523-536. 38. Chen S, Tang DD. c-Abl tyrosine kinase regulates cytokinesis of human airway smooth muscle cells. Am J Respir Cell Mol Biol. 2014;50(6):1076-1083. 39. Bi E, Maddox P, Lew DJ, et al. Involvement of an actomyosin contractile ring in Saccharomyces cerevisiae cytokinesis. J Cell Biol. 1998;142(5):1301-1312. 40. Naqvi NI, Eng K, Gould KL, Balasubramanian MK. Evidence for F-actin-dependent and independent mechanisms involved in assembly and stability of the medial actomyosin ring in fission yeast. EMBO J. 1999;18(4):854862. 41. Weber I, Niewohner J, Faix J. Cytoskeletal protein mutations and cell motility in Dictyostelium. Biochem Soc Symp. 1999;65:245-265. 42. Stock A, Steinmetz MO, Janmey PA, et al. Domain analysis of cortexillin I: actinbundling, PIP(2)-binding and the rescue of cytokinesis. EMBO J. 1999;18(19):5274-5284. 43. Zhang Y, Kwon S, Yamaguchi T, et al. Mice lacking histone deacetylase 6 have hyperacetylated tubulin but are viable and develop normally. Mol Cell Biol. 2008;28(5):16881701. 44. Staus DP, Taylor JM, Mack CP. Enhancement of mDia2 activity by Rho-kinase-dependent phosphorylation of the diaphanous autoregulatory domain. Biochem J. 2011;439(1):5765. 45. Otomo T, Tomchick DR, Otomo C, Panchal SC, Machius M, Rosen MK. Structural basis of actin filament nucleation and processive capping by a formin homology 2 domain. Nature. 2005;433(7025):488-494.

haematologica | 2017; 102(6)


ARTICLE

Red Cell Biology & its Disorders

Cdk6 contributes to cytoskeletal stability in erythroid cells

Iris Z. Uras,1* Ruth M. Scheicher,1* Karoline Kollmann,1 Martin Glösmann,2 Michaela Prchal-Murphy,1 Anca S. Tigan,1 Daniela A. Fux1, Sandro Altamura,3,4 Joana Neves,3,4 Martina U. Muckenthaler,3,4 Keiryn L. Bennett,5 Stefan Kubicek,5 Philip W. Hinds,6 Marieke von Lindern7 and Veronika Sexl1

Institute of Pharmacology and Toxicology, University of Veterinary Medicine, Vienna, Austria; 2VetCORE, University of Veterinary Medicine, Vienna, Austria; 3Department of Pediatric Hematology, Oncology, and Immunology, University of Heidelberg, Germany; 4 Molecular Medicine Partnership Unit, University of Heidelberg, Germany; 5CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences, Vienna, Austria; 6Department of Developmental, Molecular and Chemical Biology, Tufts University School of Medicine, and Tufts Cancer Center, Boston, MA, USA and 7 Department of Hematopoiesis, Sanquin Research, Amsterdam, the Netherlands

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

1

*These authors contributed equally to this work

Haematologica 2017 Volume 102(6):995-1005

ABSTRACT

M

ice lacking Cdk6 kinase activity suffer from mild anemia accompanied by elevated numbers of Ter119+ cells in the bone marrow. The animals show hardly any alterations in erythroid development, indicating that Cdk6 is not required for proliferation and maturation of erythroid cells. There is also no difference in stress erythropoiesis following hemolysis in vivo. However, Cdk6-/- erythrocytes have a shortened lifespan and are more sensitive to mechanical stress in vitro, suggesting differences in cytoskeletal architecture. Erythroblasts contain both Cdk4 and Cdk6, while mature erythrocytes apparently lack Cdk4 and their Cdk6 is partly associated with the cytoskeleton. We used mass spectrometry to show that Cdk6 interacts with a number of proteins involved in cytoskeleton organization. Cdk6-/- erythroblasts show impaired F-actin formation and lower levels of gelsolin, which interacts with Cdk6. We also found that Cdk6 regulates the transcription of a panel of genes involved in actin (de-)polymerization. Cdk6-deficient cells are sensitive to drugs that interfere with the cytoskeleton, suggesting that our findings are relevant to the treatment of patients with anemia – and may be relevant to cancer patients treated with the new generation of CDK6 inhibitors.

Correspondence: veronika.sexl@vetmeduni.ac.at

Received: November 11, 2016. Accepted: February 22, 2017. Pre-published: March 2, 2017. doi:10.3324/haematol.2016.159947

Introduction Cyclins and cyclin-dependent kinases (Cdk) are critical regulators of the cell cycle. Cyclin/Cdk complexes trigger cell-cycle progression with D-type cyclins and Cdk responsible for controlling the early G1 phase of the cell cycle. In the early G1 phase, cyclin-D-Cdk4/6 complexes phosphorylate and inactivate the retinoblastoma protein to activate E2F-dependent transcription.1 Cdk4 and Cdk6 are 71% identical at the amino acid level, are ubiquitously expressed and bind all three D-type cyclins.2 Only the combined deletion of Cdk4 and Cdk6 induces late embryonic lethality due to defects in hematopoiesis,3,4 while deletion of either gene alone is not fatal. Accordingly, Cdk4 and Cdk6 are considered to have redundant functions in regulating cell-cycle progression. Nevertheless, the lack of either of the kinases leads to specific defects in particular types of cells. Cdk4-/- mice have defective postnatal pancreatic beta cells and pituitary cells,5,6 while the lack of Cdk6 leads to disorders in the hematopoietic compartment such as altered thymocyte development and lower numbers of red blood cells and cells of other hematopoietic lineages.4,7,8 Recent evidence indicates that Cdk6 is a key player in the differentiation of a variety of cell types, a function not shared by Cdk4. A decline in the level of Cdk6 is required for terminal differentiation of haematologica | 2017; 102(6)

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/995 ©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.

995


I.Z. Uras et al.

murine erythroid leukemia cells9 and Cdk6 expression in astrocytes has been associated with the expression of progenitor cell markers.10 Osteoblast differentiation has been described to be inhibited by overexpression of Cdk6 but not of Cdk4.11 A related study showed that Cdk6 protein levels are downregulated by Rankl-induced osteoclast differentiation.12 Moreover, Cdk6 inhibits the transcriptional activation of Runx1 and thus blocks myeloid differentiation.13 Cdk6 kinase activity is also required for thymocyte development, in which it functions by modulating the expression of Notch target genes.8 Cdk6 but not Cdk4 regulates transcription of key players in lymphoma formation and myeloid leukemia in a kinase-dependent and/or kinase-independent manner.14–16 Cdk6 has also been ascribed a unique role in actin dynamics in astrocytes:10,17 overexpression of Cdk6 in primary mouse astrocytes results in altered gene expression and modified cytoskeletal architecture, as well as in loss of stress fibers and enhanced motility. Erythrocytes transport oxygen through the body. They have a unique biconcave shape that allows a certain flexibility when travelling through narrow capillary vessels.18 Changes of form are made possible by the delicate interplay between the fluid cell membrane and the structure of the membrane cytoskeleton.18 Mutations in genes encoding actin-binding proteins such as adducin, dematin, tropomyosin, tropomodulin and protein 4.1 are known to cause altered actin stability and increased erythrocyte fragility leading to anemia. The correct polymerization and organization of actin is vital for the strength and flexibility of erythrocytes.18 Besides, actin-remodeling proteins are required for regulating actin filament length. Among those, gelsolin, a Ca++-dependent protein, severs preassembled actin filaments and is hence crucial in actin remodeling.19 We show that the anemic phenotype of mice lacking Cdk6 kinase activity is not attributable to a reduced production of erythrocytes but is the consequence of a reduced erythroid life span. Our findings reveal a novel role for Cdk6 as a stabilizer of the erythrocyte cytoskeleton via the deregulation of genes involved in cytoskeleton stability.20–22

Methods

antibodies: Ter119-PE (eBioscience), CD71 Biotin (eBioscience), Streptavidin eflour 780 (eBioscience), CD49e PE (integrin α5), and CD49d FITC (integrin α4).

In vivo biotin labeling The life span of red blood cells in vivo was measured by injecting 3mg EZ-Link-sulfo-NHS-biotin (Pierce, Rockford, IL, USA) into mice. Decay of the labeled red blood cells was measured by FACS analysis.23

Osmotic fragility Fresh blood from age-matched mice was washed twice in phosphate-buffered saline. Osmotic fragility was measured by mixing 25 μL blood with 2.5 mL saline solutions (with salt concentrations between 0 and 0.9% NaCl). After gentle mixing the suspension was incubated for 15 min at room temperature and centrifuged at 500 x g for 10 min. The osmotic fragility curve was obtained by plotting the measured absorbance of hemoglobin released into the supernatants at 540 nm for each solution against NaCl concentrations. Duplicates for each NaCl concentration point were read.

Measurement of membrane mechanical stability Fresh murine blood was taken, diluted in 10 volumes of phosphate-buffered saline and subjected to a constant shear stress of 29G, 0.33 x 12 mm. The suspension was centrifuged at 500 x g for 10 min. The absorbance of hemoglobin supernatant was measured at 540 nm. Fresh human erythrocytes were diluted in 10 volumes of phosphate-buffered saline and incubated with palbociclib (manufactured by Pfizer) at indicated concentrations overnight at 37°C, in 5% CO2 and 95% humidity. Dimethyl sulfoxide (DMSO) was used as the vehicle control. The degree of resistance of red blood cells to mechanical stress was measured as above.

Study approval Venous blood was drawn from healthy volunteers after informed consent. Animal experiments were performed in accordance with protocols approved by Austrian law and the Animal Welfare Committee at the University of Veterinary Medicine, Vienna.

Statistical analysis Statistical analysis was carried out using a two-tailed unpaired Student t test and one-way or two-way analysis of variance (ANOVA) as appropriate. Data are presented as mean values ± standard error of the mean (SEM) and were processed using GraphPad software.

Mouse strains Mice were maintained under pathogen-free conditions at the Institute of Pharmacology and Toxicology, University of Veterinary Medicine, Vienna (Austria). C57BL/6 mice are referred to as Cdk6+/+. Cdk6-/- 4 and Cdk6K43M/K43M 8 mice were generated on the C57BL/6 background. Mice aged 6-8 weeks were used unless otherwise indicated. Blood parameters were analyzed with a VetABC blood counter.

Flow cytometry Bone marrow, spleen and fetal liver cells were stained for erythroid subsets using the antibodies listed below. Samples were analyzed by a FACSCantoII flow cytometer using FACSDiva software (Becton-Dickinson). Erythrocyte development in the bone marrow and spleen was analyzed as follows: Ter119medCD71high (proerythroblasts; R1), Ter119highCD71high (basophilic erythroblasts; R2), Ter119highCD71med (late basophilic and polychromatophilic erythroblasts; R3) and Ter119highCD71low (orthochromatophilic erythroblasts; R4). Cells were stained with the fluorescently labeled 996

Results Cdk6-/- mice have anemia accompanied by an increased number of Ter119+ cells in the bone marrow Cdk6-/- mice are viable and fertile and display minor defects in hematopoiesis with anemia and reduced cellularity in thymus and spleen.4 The reduction in erythrocyte numbers is paralleled by an increase in mean cell volume and mean corpuscular hemoglobin.4 We confirmed these observations in C57BL/6 Cdk6-/- mice under our housing conditions (Figure 1A,B and Online Supplementary Figure S1A). Anemia may be caused by enhanced hemolysis or altered iron metabolism. We analyzed levels of plasma iron, plasma hepcidin (a key iron regulator) and erythropoietin (a major mitogen for erythroid progenitors). No haematologica | 2017; 102(6)


Cdk6 regulates the erythrocyte cytoskeleton

significant changes were detected although there was a tendency towards higher plasma iron and erythropoietin levels in Cdk6-/- mice (Online Supplementary Figure S1B-D). Likewise, the ability to form myeloid colonies was not altered among genotypes (Online Supplementary Figure S1E). Analysis of the maturation of erythroid progenitors

A

B

C

D

E

F

in the bone marrow and spleen by FACS revealed minor differences: a profiling developed by Socolovsky et al.,24 showed slightly enhanced numbers of cells in the R2 (basophilic erythroblasts) and R4 (orthochromatophilic erythroblasts) fractions in the bone marrow but not in the spleen (Figure 1C,D). The increased amount of erythro-

Figure 1. Cdk6-/- mice are anemic with compensatory upregulation of erythrocytes in the bone marrow. (A) Cdk6-/- mice have a reduced red blood cell count (n of Cdk6+/+ mice = 7; n of Cdk6-/- mice = 11). A two-tailed unpaired Student t test was used for the statistical analysis. Error bars indicate ± SEM (***P<0,001). (B) One representative Giemsa staining of Cdk6-/- and Cdk6+/+ blood is depicted (scale bar 20 μm; microscope: Zeiss AxioImager Z2; lens: Zeiss plan-apochromat 63x/1.4 oil lens; acquisition software: ZEN 2012). (C-D) FACS staining of erythroid development in the bone marrow (BM) and spleen of Cdk6-/- and Cdk6+/+ mice determined by the surface marker CD71/Ter119: Ter119medCD71high (proerythroblasts), Ter119highCD71high (basophilic erythroblasts), Ter119highCD71med (late basophilic and polychromatophilic erythroblasts) and Ter119highCD71low (orthochromatophilic erythroblasts) in regions R1-R4, respectively. The distribution of the respective populations is summarized in the lower panels. Statistical analysis was performed with a two-tailed unpaired Student t test in bone marrow (n=17 per genotype) and spleen (n≥7 per genotype); n.s.: not significant; *P<0.05; **P<0.01; ***P<0.001). (E) Cdk6-/- mice have increased total numbers of Ter119+ erythrocytes in the bone marrow (BM) (n=6 per genotype). A two-tailed unpaired Student t test was used for the statistical comparison. Error bars indicate ±SEM (***P<0.001). (F) Reticulocyte numbers in the peripheral blood (PB) were analyzed by FACS using thiazole orange (n=5 per genotype). Reticulocyte counts were increased in Cdk6-/- mice, balancing the reduced red blood cell count. Statistical analysis was carried out using a two-tailed unpaired Student t test. Error bars indicate ± SEM (*P<0.05).

haematologica | 2017; 102(6)

997


I.Z. Uras et al.

cyte precursors became most evident from the highly significant increase in total Ter119+ erythroid cells in the bone marrow of Cdk6-/- animals (Figure 1E). No differences were detected in the spleen (Online Supplementary Figure S2A) or during erythrocyte maturation in fetal liver erythropoiesis (at day E13) between Cdk6-/- and wildtype controls (Online Supplementary Figure S2B). One explanation for the increased proportion of Ter119+ cells in the bone marrow accompanying anemia in the peripheral blood is a reduced ability of mature red blood cells to exit the bone marrow. To enter the periphery, red cells downregulate surface integrins during erythropoiesis. FACS staining revealed no differences in integrin patterns: levels of downregulation of integrin Îą4 (Online Supplementary Figure S2C) and integrin Îą5 (Online Supplementary Figure S2D) were comparable irrespective of the genotype. The unaltered ability of Cdk6-/- erythrocytes to migrate efficiently from the bone marrow to the periphery was also visible from the elevated numbers of reticulocytes (which contain residual amounts of RNA that can be stained with thiazole orange) in the blood. Cdk6-/- mice had significantly more reticulocytes in the peripheral blood than had Cdk6+/+ controls (Figure 1F) reflecting the enhanced production of red blood cells. In summary, a reduced erythroid development in the bone marrow does not account for anemia in Cdk6-/- mice.

erythroid cells in the wildtype bone marrow increased from ~50% to ~60%, it remained consistently high at 60% in Cdk6-/- bone marrow (Figure 2E). This difference was not seen in the spleen: numbers of Ter119+ cells increased to the same extent in Cdk6-/- and Cdk6+/+ mice upon phenylhydrazine treatment (Figure 2F).

Cdk6-/- erythrocytes have a decreased life span in vivo We next analyzed the possibility that the anemia in Cdk6-/- mice is linked to a shortened lifespan of the cells. Peripheral erythrocytes were labeled in vivo with EZ-Link sulfo-NHS-biotin and blood samples were taken every 10 days. In line with published data,28,29 we found that under these conditions wildtype erythrocytes had an average halflife of 19 days, while that of erythrocytes from Cdk6-/- mice was significantly shorter (14 days; Figure 3A). The shortened half-life may be indicative of a lower mechanical stability of the cells as the long-term survival of mature erythrocytes in the periphery demands mechanical strength and flexibility. Freeze-thaw cycles also pose a challenge to the mechanical stability of cells. Notably, we found that erythroblasts from Cdk6-/- fetal livers did not survive a freeze-thaw cycle (Figure 3B). Changes in membrane stability may also be associated with reduced tolerance to osmotic stress which was tested using various concentrations of NaCl. However, osmotic sensitivity was not altered between Cdk6+/+ and Cdk6-/- erythrocytes (Figure 3C).

Cdk6-/- mice respond normally to erythropoietic stress The bone marrow reacts immediately to conditions of high need by increasing the production of erythrocytes through a process referred to as stress erythropoiesis. We reasoned that subtle changes in erythropoiesis may become more pronounced under conditions of stress. Two independent experimental settings were used to investigate whether Cdk6-/- mice respond to hypoxia by increasing erythroid production. First, we mimicked stress erythropoiesis in vitro: fetal liver erythroblasts (day E13) can be cultured and stimulated to proliferate and differentiate by use of distinct cytokines.25 Cdk6+/+ and Cdk6-/- erythroblasts displayed superimposable growth curves (Figure 2A) and similar morphology (Figure 2B). A proliferation assay using CellTrace Violet showed no differences among genotypes (Online Supplementary Figure S3A). Differentiating erythrocytes divide approximately four times before they mature to fully differentiated enucleated red cells.25 The kinetics of differentiation of Cdk6-/- and Cdk6+/+ erythroblasts were comparable, as shown by growth curves (Figure 2C), the Violet Cell proliferation assay (Online Supplementary Figure S3B) and benzidine staining (Online Supplementary Figure S3C). The in vitro findings were echoed by in vivo studies: mice were exposed to phenylhydrazine which induces hemolytic anemia and forces the subsequent rapid expansion of the erythroid lineage in the bone marrow and spleen.26,27 Cdk6-/- and Cdk6+/+ mice showed a comparable recovery response when analyzed 10 days after receipt of a single dose of phenylhydrazine. Red blood cell count (Figure 2D) and hematocrit (Online Supplementary Figure S3D) were restored to normal levels. Massive increases in numbers of all premature stages (R1-R2) of the erythroid lineage were detectable in the bone marrow (Online Supplementary Figure S3E) and the spleen (Online Supplementary Figure S3F) 3 days after phenylhydrazine induction irrespective of the genotype. We only observed one subtle difference: whereas the proportion of Ter119+ 998

Cdk6 is tethered to the cell membrane in erythrocytes We next tested whether the absence of Cdk6 is associated with changes of the cytoskeleton. The major cytoskeletal components (spectrin, ankyrin, band 3 and protein 4.1/4.2) were present at comparable levels irrespective of the genotype (Online Supplementary Figure S4). However, phalloidin staining of fixed erythroblasts revealed a pronounced reduction in filamentous-actin (F-actin) polymerization upon loss of Cdk6 (Figure 4A and Online Supplementary Figure S5A,B) although the level of total betaactin (B-actin, non-polymerized actin) in erythroblasts was unaffected (Figure 4B). In mature erythrocytes differences in phalloidin staining were still observed, albeit to a lesser extent: in wildtype erythrocyte cytoskeleton F-actin structures were symmetrical and had a pronounced ring shape whereas the actin oligomers in Cdk6-/- erythrocytes were distributed diffusely (Online Supplementary Figure S5C). In line, Cdk6-deficient lymphoid cells had a decreased filamentous to globular actin ratio suggesting the interplay between Cdk6 and actin structures is not restricted to erythroid cells (Online Supplementary Figure S5D). Proliferating erythroblasts express both cell-cycle kinases Cdk4 and Cdk6 (Figure 4B). Mature erythrocytes still contain Cdk6 protein (Figure 4C) whereas Cdk4 is degraded during cell maturation. Cdk4 was only present at low levels in Cdk6-/- erythrocytes (Figure 4C). Erythrocyte membranes (ghosts) contain mainly cytoskeletal and membrane proteins but also some peripheral and membrane-associated proteins. It is therefore of particular interest that we detected significant amounts of Cdk6 but not Cdk4 in ghosts (Figure 4D and Online Supplementary Figure S5E). These data indicate that Cdk6 is maintained in mature erythrocytes; a proportion is associated with the membrane and may play an active role in regulating cytoskeletal proteins. These findings are in line with a model that Cdk6 is tethered to the cytoskeleton where it may be protected from degradation and have a membrane stabilizing role.17 haematologica | 2017; 102(6)


Cdk6 regulates the erythrocyte cytoskeleton

Cdk6 exerts kinase-dependent effects on erythrocyte stability Cdk6 may affect cytoskeleton and membrane stability in a kinase-dependent and/or kinase-independent manner. Mice harboring a kinase dead mutant of Cdk6 (Cdk6K43M/K43M)8 recapitulate the phenotype of Cdk6-/- ani-

mals: we found decreased red blood cell counts accompanied by an increase in mean cell volume (Figure 5A and Online Supplementary Figure S6A). No gross morphological alterations were detected (Figure 5B). Reminiscent of Cdk6-/- mice we found alterations in the stages of erythroid development in the bone marrow and an increased total

A

B

C

D

E

F

Figure 2. Stress erythropoiesis is unaffected upon loss of Cdk6. (A) Proliferation of in vitro cultured Cdk6-/- and Cdk6+/+ erythroblasts is comparable (n=3 per genotype). Error bars indicate ± SEM. (B) Immunohistochemical staining of erythroblasts using neutral benzidine and histological dyes. (C) Differentiation was induced by a change in growth factors added to the medium. Erythroblasts divide a few times before enucleation and contraction starts. Cell numbers are shown. Error bars indicate ± SEM (n=3 per genotype). (D-F) Cdk6-/- and Cdk6+/+ animals were treated twice with phenylhydrazine (PHZ) to induce hemolytic anemia. (D) Cdk6-/- and Cdk6+/+ mice are able to restore red blood cell counts to normal levels to a comparable extent on day 10 after PHZ challenge. Error bars indicate ± SEM (n≥4 per genotype). Two-way ANOVA was used for the statistical comparison. (E) The total amount of Ter119+ erythrocytes in the bone marrow (BM) is already enriched in untreated Cdk6deficient bone marrow with no further increase upon phenylhydrazine (PHZ) treatment. A two-tailed unpaired Student t test was used for the statistical comparison. Error bars indicate ± SEM (n≥4 per genotype; **P<0.01; n.s.: not significant). (F) Phenylhydrazine (PHZ) treatment results in elevated total amount of Ter119+ erythrocytes in spleen independent of the genotype. Statistical analysis was performed with a two-tailed unpaired Student t test. Error bars indicate ± SEM (n≥4 per genotype; ****P<0.0001 n.s.: not significant).

haematologica | 2017; 102(6)

999


I.Z. Uras et al.

number of Ter119+ erythroid cells (Figure 5C,D). Analysis of erythroid progenitors in the spleen and numbers of reticulocytes in the peripheral blood were not significantly altered (Online Supplementary Figure S6B,C). However, as seen in Cdk6-deficiency, Cdk6K43M/K43M erythroid cells showed a decreased filamentous to globular actin ratio (Figure 5E). These data suggest a kinase-dependent function of Cdk6 in cytoskeletal integrity.

Bcr/Ablp185+-transformed lymphoid cells (Figure 6B). Lymphoid cell expression of gelsolin, a known regulator of actin filament (de-)polymerization,19 was also decreased in Cdk6K43M/K43M mice (Online Supplementary Figure S7B). Intrigued by the observation that Cdk6 regulates the expression of genes involved in cytoskeleton organization in all our model systems we investigated which components of the cytoskeleton or cell membrane are bound to Cdk6. We immunoprecipitated Cdk6 from wildtype peripheral blood followed by mass spectrometry and complemented the analysis by including Cdk6 immunoprecipitation from wildtype lymphoid cells. The detection of Cdk6 protein served as a positive control. Mass spectrometry verified a set of potential Cdk6 interactors (Online Supplementary Figure S8). In mature erythrocytes we confirmed the interaction with gelsolin and also detected a set of proteins that have been associated with anemia and/or erythroid cytoskeletal organization/ dynamics (Figure 6C and Online Supplementary Figure S8). Tubulin was an attractive candidate in lymphoid cells (Online Supplementary Figure S8).

Cdk6 transcriptionally regulates genes involved in cytoskeleton organization in a kinase-dependent manner To investigate which cytoskeleton components are altered upon loss of Cdk6 kinase activity, we performed quantitative polymerase chain reaction studies of a set of genes that had been implicated in cytoskeleton remodeling in the literature using erythroid progenitors with intact nuclei [Ter119highCD71low (orthochromatophilic erythroblasts; R4)] (Figure 6A and Online Supplementary Figure S7A). In the absence of Cdk6 kinase activity we found significant changes in the mRNA levels of four genes: Gelsolin, Tubulin alpha-8, Baiap2 and Pip5k1b were reduced to ~50% (Figure 6A). Cdk6 directly regulates the transcription of Gelsolin and Baiap2 as we were able to further show enhanced binding of Cdk6 to the respective promoters by chromatin immunoprecipitation assays in the nontransformed mouse progenitor HPC7 cell line and in

A

Cdk6-/- cells display an increased sensitivity to shear stress and drugs interfering with cytoskeleton organization On this basis, we also reasoned that Cdk6-/- cells should be more susceptible to any pharmacological disturbance of the cytoskeleton. We perturbed either actin metabolism

B

C

Figure 3. Erythrocytes in Cdk6-/- animals have a decreased life span. (A) Erythrocytes were biotinylated in vivo to determine the erythroid life span in the periphery. Decay of the biotinylation was analyzed every 10 days. Cdk6-/- erythrocytes have a half-life decreased by ~5 days. The statistical comparison was performed using a two-tailed unpaired Student t test. Error bars indicate Âą SEM (n=7 per group; **P<0.01; ***P<0.001; ****P<0.0001). (B) Increased susceptibility of Cdk6-/- erythroblasts to stress was observed. Wildtype erythroblasts derived from fetal liver grew normally after a freeze-thaw cycle but Cdk6-/- cells failed to grow (n=3 per genotype). A two-tailed unpaired Student t test was used for the statistical analysis. Error bars indicate Âą SEM (*P<0.05). (C) Osmotic fragility was analyzed by incubating fresh blood with different concentrations of NaCl. Specific hemolysis was measured at 540 nm. No differences were observed between genotypes (n=4 per group).

1000

haematologica | 2017; 102(6)


Cdk6 regulates the erythrocyte cytoskeleton

by the addition of jasplakinolide,19 which stabilizes actin filaments, or microtubule metabolism by the addition of two benzimidazoles (albendazole and thiabendazole),30 which block tubulin polymerization. The treatments showed that our prediction was respected: Cdk6-/- cells were significantly more sensitive (Figure 6D). Combined treatment with CDK6 kinase inhibitor and microtubule poison mimicked the absence of Cdk6 in wildtype cells: the Bliss additivity model revealed an in vitro synergistic cytotoxicity between palbociclib and albendazole (Online Supplementary Figure S9A). Fragility of erythrocytes can be tested by exposing cells to shear stress: Cdk6-/- and Cdk6K43M/K43M erythrocytes reacted with an enhanced lysis (Figure 6E). Similar observations although not reaching the level of statistical significance were made upon CDK4/6 kinase inhibitor treatment in

human erythroid cells (Online Supplementary Figure S9B). Collectively, these data indicate a global role for Cdk6 as a regulator of membrane and cytoskeleton stability which manifests itself as anemia in the red cell compartment.

Discussion Anemia is a frequent disorder with a variety of possible causes. It most frequently stems from reduced erythrocyte formation due to a lack of iron, vitamin B12 or folic acid31– 33 but may also be associated with increased or premature destruction of erythrocytes in the periphery.34 Anemia of this latter type is generally caused by alterations in the cytoskeleton and the membrane and a large number of erythrocyte disorders are associated with mutations in

A

B

C

Figure 4. Altered cytoskeletal composition in Cdk6-/- erythroid cells. (A) Phalloidin immunofluorescence staining revealed decreased amounts of filamentous (F-) actin in Cdk6-/- erythroblasts (scale bar 50 μm; microscope: Zeiss LSM 510 Meta; Axiovert 200M; lens: Zeiss plan-apochromat 63x/1.4 oil lens; fluorochromes: DAPI, Alexa 546 (coupled phalloidin); acquisition software: ZEN 2009). The right panel depicts the summary of three experiments. Statistical analysis was conducted by a two-tailed unpaired Student t test. Error bars indicate ± SEM (*P<0.5). (B) Proliferating erythroblasts exhibited normal expression levels of Cdk4 and Cdk6 and no differences in beta-actin (in its non-polymerized form) as quantified by western blots. Three biological replicates per genotype are shown. Anti-Hsc70 antibody was used as loading control. A lymphoid cell line served as positive control. (C) Mature erythrocytes showed elevated levels of Cdk6. Only a weak Cdk4 signal was detectable in Cdk6-/- erythrocytes by immunoblotting. Actin levels were similar. Three biological replicates per genotype are depicted. Whole bone marrow and a lymphoid cell line were used as positive controls. Anti-Hsc70 antibody served as loading control. (D) In erythrocyte membranes (ghosts) Cdk6 but not Cdk4 protein was detectable. Actin levels remained comparable. Anti-Hsc70 antibody was used as loading control. Three biological replicates per genotype are shown. Whole bone marrow and a lymphoid cell line served as positive controls.

D

haematologica | 2017; 102(6)

1001


I.Z. Uras et al.

A

B

C

D

E

Figure 5. Cdk6K43M/K43M mice show mild anemia with compensatory upregulation of erythrocytes in the bone marrow. (A) Cdk6K43M/K43M mice have a reduced red blood cell count (n of Cdk6+/+ mice=19; n of Cdk6K43M/K43M mice=17). Statistical analysis was carried out using a two-tailed unpaired Student t test. Error bars indicate ± SEM (****P<0.0001). (B) One representative Giemsa staining of Cdk6K43M/K43M and Cdk6+/+ blood is depicted (scale bar 60 μm). (C) FACS staining of the erythroid development in the bone marrow (BM) was determined by the surface marker CD71/Ter119 as described in Figure 1C,D (n≥16 per genotype). A two-tailed unpaired Student t test was used for the statistical analysis. Error bars indicate ± SEM (n.s.: not significant; *P<0.05; **P<0.01; ****P<0.0001). R1: Ter119medCD71high (proerythroblasts); R2: Ter119highCD71high (basophilic erythroblasts); R3: Ter119highCD71med (late basophilic and polychromatophilic erythroblasts) and R4: Ter119highCD71low (orthochromatophilic erythroblasts). (D) Cdk6K43M/K43M mice had increased total numbers of Ter119+ erythrocytes in the bone marrow (BM) (n of Cdk6+/+=36; n of Cdk6K43M/K43M=21). Statistical comparison was conducted by a two-tailed unpaired Student t test. Error bars indicate ± SEM (****P<0.0001). (E) One representative blot shows the amount of F-actin content versus G-actin content in mature red blood cells of indicated genotypes. The right panel depicts the summary of three experiments. A two-tailed unpaired Student t test was used for the statistical analysis. Error bars indicate ± SEM (*P<0.05; n.s.: not significant). F: filamentous F-actin; G: free globular G-actin.

1002

haematologica | 2017; 102(6)


Cdk6 regulates the erythrocyte cytoskeleton

genes required for membrane and cytoskeleton stability.18 Such mutations render erythrocytes susceptible to mechanical stress and lead to a premature degradation of red cells. We now define a deficiency in Cdk6 kinase activity as the cause of a novel form of anemia characterized by enhanced mechanical instability of erythrocytes. Mice lacking Cdk6 or its kinase activity (Cdk6K43M/K43M) suffer from a mild form of anemia with enhanced numbers of Ter119+ cells in the bone marrow and increased numbers of reticulocytes in the peripheral blood. This observation is in line with enhanced erythropoiesis, which may be viewed as the organism’s attempt to compensate

A

B

C

D

E

haematologica | 2017; 102(6)

for the enhanced loss of erythrocytes. We did not detect any reduction of erythroblast proliferation or differentiation in in vitro systems that mimic stress erythropoiesis or when mice were treated with phenylhydrazine. Cdk6 does not have a critical role in cell proliferation, since its function can be performed in its absence by Cdk4. Differentiation along the erythroid lineage is largely unaltered by the elimination of Cdk6. This finding is in contrast to a report that Cdk6 is involved in the differentiation of the murine erythroleukemia cell line, MEL. The apparent discrepancy probably relates to the different experimental systems used. Transformed cell lines harbor multi-

Figure 6. Cdk6 regulates transcription of genes involved in cytoskeleton organization in a kinasedependent manner. (A) Gene expression was analyzed by quantitative reverse transcriptase polymerase chain reaction in region R4 (orthochromatophilic erythroblasts) of indicated bone marrow. Relative expression levels were normalized to Rplp0 mRNA. Statistical analysis was carried out using a two-tailed unpaired Student t test. Error bars indicate ± SEM (n≥6 per group; **P<0.01; ***P<0.001; ****P<0.0001). (B) Chromatin immunoprecipitation experiments were performed in the mouse HPC7 Cdk6+/+ progenitor cell line and Bcr/Ablp185+ Cdk6+/+ lymphoid cells. Protein-DNA complexes were immunoprecipitated using home-made sera against Cdk6 and analyzed by quantitative polymerase chain reaction for the presence on the indicated promoter regions. p16INK4a and Egr1 promoter regions served as positive controls. Bar graphs depict fold enrichment over a negative region downstream of CD19 as described in the Methods section. (C) Lysates from mature erythrocytes were subjected to Cdk6 immunoprecipitation (IP) and blotted with an anti-gelsolin antibody. The asterisk indicates an unspecific cross-reacting band. Beta-actin served as loading control. sn: supernatant. (D) Viability measurements upon treatment with an actin-specific agent Jasplakinolide (4 nM) and microtubule depolymerizing agents albendazole (1 μM) and thiabendazole (10 μM) for 72 h were conducted using the CellTiterGlo Viability Assay. The analysis was carried out in triplicate. A two-tailed unpaired Student t test was used for the statistical comparison. Error bars indicate ± SEM (***P<0.001; ****P<0.0001). (E) The absorbance of hemoglobin supernatant from indicated mature erythrocytes at 540 nm after a constant shear stress is depicted. Statistical analysis was carried out using a two-tailed unpaired Student t test. Error bars indicate ± SEM (n=12 per group; ****P<0.0001; n.s.: not significant).

1003


I.Z. Uras et al.

ple mutations that may interfere with normal or regular differentiation and so may not mimic the situation in untransformed cells. For example, loss of CDK6 in MLLtransformed myeloid acute myelogenous leukemia cells induces differentiation35 while the myeloid compartment in Cdk6-deficient mice is normal and contains cells at all stages of maturation.4 Signaling and transcriptional control are rewired in transformed cells so that molecules that are not required in non-transformed cells may become important for differentiation or proliferation. Erythroblast protein expression changes during differentiation. Certain proteins are degraded or removed with the nucleus at enucleation or by vesiculation during reticulocyte maturation.36–38 Most proteins are degraded during development. Levels of Cdk4 decrease dramatically and the protein is not detectable in mature erythrocytes, although Cdk6 is still present. Together with the observation that Cdk6- but not Cdk4-deficient mice5,6 suffer from anemia, this provides strong evidence that Cdk6 has a function in murine erythrocytes that cannot be performed by Cdk4. Interestingly, in the absence of Cdk6, mature erythrocytes do contain detectable levels of Cdk4. This paradoxical finding can be explained by postulating that in the absence of Cdk6, Cdk4 can interact with cytoskeletal structures usually occupied by Cdk6, thereby being at least partially protected from degradation, although the interaction may not have any functional consequence. Previous studies indicated that expression of cell-cycle components correlate with variations in size and cell divisions during erythropoiesis.39 A direct role has been assigned to cyclin D3: its levels orchestrate the number of cell divisions during terminal erythropoiesis, thereby controlling the number and size of erythrocyte progeny.40 In contrast, proliferation and differentiation of erythrocytes are not impaired in mice lacking Cdk6 and mature erythrocytes contain Cdk6. We thus speculated that the anemic phenotype might be linked to alterations in the cytoskeleton. Cdk6 has been reported to associate with the cytoskeleton in astrocytes, where forced overexpression of Cdk6 induces changes to the cytoskeletal organization.10,17 Staining of erythroid cells for F-actin revealed a dependence on the presence of Cdk6, with Cdk6-deficiency associated with markedly lower levels of F-actin. A number of clinical conditions have been linked to impaired F-actin formation18,41,42 and these may be associated with a shortened half-life of erythrocytes in the peripheral blood. We found an altered F:G-actin composition in transformed lymphoid cells, indicating that the role of Cdk6 in actin remodeling is not restricted to erythroid cells. A possible mechanism is suggested by the observation that Cdk6 interacts in mature erythrocytes with gelsolin, a known regulator of actin remodeling.19 The cytoskeleton of erythrocytes faces particular challenges. Erythrocytes must pass through narrow capillaries, placing special requirements on the stability and flexibility of the cytoskeleton.18 Our results point to a further unique feature of the erythrocyte cytoskeleton, namely its dependence on Cdk6 for structural integrity and flexibility. Cdk6 is tethered to the erythrocyte cytoskeleton and this finding is consistent with the localization of the protein when it is overexpressed in astrocytes.17 We and others have shown that Cdk6 directly regulates transcription in both kinase-dependent and kinase-inde-

1004

pendent manners, interacting with a variety of transcription factors including Stat and AP-1 as well as with nuclear factor-κB.14–16,43 We now report that Cdk6 promotes the transcription of genes involved in cytoskeletal organization. The function depends on Cdk6 kinase activity: upon loss of Cdk6 kinase activity in erythroid progenitors (R4) there is a significant decrease in mRNA levels of Tubulin alpha-8 implicated in microtubule assembly and Baiap2, Gelsolin and Pip5k1b involved in actin dynamics. BAIAP2 participates in F-actin rearrangements when activated by small GTPases.21 Loss of gelsolin in mice has been shown to change the balance between polymerized and depolymerized actin in red blood cells.19 PIP5K1B has a function in the dynamics of the actin cytoskeleton44 and PIP5 kinases synthesize phosphatidylinositol-4,5-bisphosphate [PI(4,5)P2], which binds to gelsolin and thereby promote actin polymerization.45–47 Loss of Cdk6 kinase activity thus causes impaired actin remodeling, which is likely to result in increased fragility of mature erythrocytes.48 Interestingly, mass spectrometry analysis of both erythroid and lymphoid cells confirmed that Cdk6 interacts with a number of proteins involved in cytoskeletal organization, suggesting a global role of Cdk6 in cytoskeletal integrity. Two sets of experiments show that our results have functional significance. First, loss of Cdk6 causes mechanical instability of red blood cells in the shear test, which mimics the entry of erythrocytes into the narrow capillaries of the body. Secondly, Cdk6-deficient cells are more sensitive to actin and microtubule inhibitors. Our work defines Cdk6 as a unique member of the Cdk family with a potential dual function for the cytoskeleton. In erythroid cells, Cdk6 affects cytoskeletal stability by transcriptionally regulating a set of genes that control cytoskeletal organization. This function depends on Cdk6 kinase activity. Furthermore, Cdk6 has a direct structural role in erythrocytes. It is tethered to the cytoskeleton, where it may phosphorylate unknown proteins, be a stabilizing anchoring factor or simply be protected from degradation. Further research is needed to unravel the mechanisms behind this function and its consequences. The phenotype of Cdk6-/- mice is recapitulated in Cdk6K43M/K43M animals. This finding may have consequences for therapies that target CDK6 kinase activity over a longer time. Inhibitors of the CDK4/6 kinases are being tested for use in the treatment of many forms of cancer.49 Our findings provide a possible explanation for the observation that patients receiving inhibitors of CDK4/6 kinases are prone to reduced erythroid stability which contributes to anemia.50 Acknowledgments We thank Philipp Jodl and Sabine Fajmann for their excellent technical help. We are deeply indebted to Graham Tebb for numerous scientific discussions and editing the manuscript. We thank Anja Redlinghofer for technical assistance, Botond Cseh for support and Michael Freissmuth, Richard Moriggl and Helmut Dolznig for critical scientific input. This work was made possible by the Austrian Science Fund (FWF) Grant SFB F47 (to SK and VS) and FWF Grant P 24297-B23 (to VS), by an ERC-advanced grant (to VS) and by EMBO ASTF 103.00-2011a (to RMS).

haematologica | 2017; 102(6)


Cdk6 regulates the erythrocyte cytoskeleton

References 1. Sherr CJ, Roberts JM. Living with or without cyclins and cyclin-dependent kinases. Genes Dev. 2004;18:2699–2711. 2. Meyerson M, Enders GH, Wu CL, et al. A family of human cdc2-related protein kinases. EMBO J. 1992;11(8):2909–2917. 3. Kozar K, Sicinski P. Cell cycle progression without cyclin D-CDK4 and cyclin D-CDK6 complexes. Cell Cycle. 2005;4(3):388–391. 4. Malumbres M, Sotillo R, Santamaría D, et al. Mammalian cells cycle without the D-type cyclin-dependent kinases Cdk4 and Cdk6. Cell. 2004;118(4):493–504. 5. Rane SG, Dubus P, Mettus R V, et al. Loss of Cdk4 expression causes insulin-deficient diabetes and Cdk4 activation results in betaislet cell hyperplasia. Nat Genet. 1999;22(1):44–52. 6. Tsutsui T, Hesabi B, Moons DS, et al. Targeted disruption of CDK4 delays cell cycle entry with enhanced p27(Kip1) activity. Mol Cell Biol. 1999;19(10):7011–7019. 7. Hu MG, Deshpande A, Enos M, et al. A requirement for cyclin-dependent kinase 6 in thymocyte development and tumorigenesis. Cancer Res. 2009;69(3):810–818. 8. Hu MG, Deshpande A, Schlichting N, et al. CDK6 kinase activity is required for thymocyte development. Blood. 2011;117(23): 6120–6131. 9. Matushansky I, Radparvar F, Skoultchi AI. Manipulating the onset of cell cycle withdrawal in differentiated erythroid cells with cyclin-dependent kinases and inhibitors. Blood. 2000;96(8):2755–2764. 10. Ericson KK, Krull D, Slomiany P, Grossel MJ. Expression of cyclin-dependent kinase 6, but not cyclin-dependent kinase 4, alters morphology of cultured mouse astrocytes. Mol. Cancer Res. 2003;1(9):654–664. 11. Ogasawara T, Kawaguchi H, Jinno S, et al. Bone morphogenetic protein 2-induced osteoblast differentiation requires Smadmediated down-regulation of Cdk6. Mol Cell Biol. 2004;24(15):6560–6568. 12. Ogasawara T, Katagiri M, Yamamoto A, et al. Osteoclast differentiation by RANKL requires NF-kappaB-mediated downregulation of cyclin-dependent kinase 6 (Cdk6). J Bone Miner Res. 2004;19(7):1128–1136. 13. Fujimoto T, Anderson K, Jacobsen SEW, Nishikawa S-I, Nerlov C. Cdk6 blocks myeloid differentiation by interfering with Runx1 DNA binding and Runx1C/EBPalpha interaction. EMBO J. 2007;26(9):2361–2370. 14. Kollmann K, Heller G, Schneckenleithner C, et al. A kinase-independent function of CDK6 links the cell cycle to tumor angiogenesis. Cancer Cell. 2013;24(2):167–181. 15. Scheicher R, Hoelbl-Kovacic A, Bellutti F, et al. CDK6 as a key regulator of hematopoietic and leukemic stem cell activation. Blood. 2015;125(1):90–102. 16. Uras IZ, Walter GJ, Scheicher R, et al. Palbociclib treatment of FLT3-ITD+ AML cells uncovers a kinase-dependent transcriptional regulation of FLT3 and PIM1 by CDK6. Blood. 2016;127(23):2890–2902. 17. Slomiany P, Baker T, Elliott ER, Grossel MJ. Changes in motility, gene expression and

haematologica | 2017; 102(6)

18. 19. 20.

21.

22.

23. 24.

25.

26.

27.

28.

29.

30. 31. 32.

33.

34.

actin dynamics: Cdk6-induced cytoskeletal changes associated with differentiation in mouse astrocytes. J Cell Biochem. 2006;99(2):635–646. Mohandas N, Gallagher PG. Red cell membrane: past, present, and future. Blood. 2008;112(10):3939–3948. Cantù C, Bosè F, Bianchi P, et al. Defective erythroid maturation in gelsolin mutant mice. Haematologica. 2012;97(7):980–988. Tolias KF, Hartwig JH, Ishihara H, et al. Type Ialpha phosphatidylinositol-4-phosphate 5kinase mediates Rac-dependent actin assembly. Curr Biol. 2000;10(3):153–156. Yamagishi A, Masuda M, Ohki T, Onishi H, Mochizuki N. A novel actin bundling/filopodium-forming domain conserved in insulin receptor tyrosine kinase substrate p53 and missing in metastasis protein. J Biol Chem. 2004;279(15): 14929– 14936. Govind S, Kozma R, Monfries C, Lim L, Ahmed S. Cdc42Hs facilitates cytoskeletal reorganization and neurite outgrowth by localizing the 58-kD insulin receptor substrate to filamentous actin. J Cell Biol. 2001;152(3):579–594. de Jong K. Short survival of phosphatidylserine-exposing red blood cells in murine sickle cell anemia. Blood. 2001;98(5):1577–1584. Socolovsky M, Nam H, Fleming MD, et al. Ineffective erythropoiesis in Stat5a(-/-)5b(-/) mice due to decreased survival of early erythroblasts. Blood. 2001;98(12):3261–3273. von Lindern M, Deiner EM, Dolznig H, et al. Leukemic transformation of normal murine erythroid progenitors: v- and c-ErbB act through signaling pathways activated by the EpoR and c-Kit in stress erythropoiesis. Oncogene. 2001;20(28):3651–3664. Dornfest BS, Naughton BA, Johnson R, Gordon AS. Hepatic production of erythropoietin in a phenylhydrazine-induced compensated hemolytic state in the rat. J Lab Clin Med. 1983;102(2):274–285. Dornfest BS, Lapin DM, Adu S, Naughton BA. Dexamethasone suppresses the generation of phenylhydrazine-induced anemia in the rat. Proc Soc Exp Biol Med. 1992;199(4):491–500. Neumann CA, Krause DS, Carman CV, et al. Essential role for the peroxiredoxin Prdx1 in erythrocyte antioxidant defence and tumour suppression. Nature. 2003;424(6948): 561– 565. Hoffmann-Fezer G, Mysliwietz J, Mörtlbauer W, et al. Biotin labeling as an alternative nonradioactive approach to determination of red cell survival. Ann Hematol. 1993;67(2):81–87. Lacey E. Mode of action of benzimidazoles. Parasitol Today. 1990;6(4):112–115. Stabler SP. Clinical practice. Vitamin B12 deficiency. N Engl J Med. 2013;368(2): 149– 160. Abbaspour N, Hurrell R, Kelishadi R. Review on iron and its importance for human health. J Res Med Sci. 2014;19(2): 164–174. Brabin BJ, Premji Z, Verhoeff F. An analysis of anemia and child mortality. J Nutr. 2001;131(2S-2):636S–645S; discussion 646S– 648S. Alaarg A, Schiffelers RM, van Solinge WW,

35.

36.

37.

38.

39.

40.

41. 42. 43.

44.

45.

46. 47.

48.

49. 50.

van Wijk R. Red blood cell vesiculation in hereditary hemolytic anemia. Front Physiol. 2013;4:365. Placke T, Faber K, Nonami A, et al. Requirement for CDK6 in MLL-rearranged acute myeloid leukemia. Blood. 2014;124 (1):13–23. Pasini EM, Kirkegaard M, Mortensen P, et al. In-depth analysis of the membrane and cytosolic proteome of red blood cells. Blood. 2006;108(3):791–801. Goodman SR, Daescu O, Kakhniashvili DG, Zivanic M. The proteomics and interactomics of human erythrocytes. Exp Biol Med (Maywood). 2013;238(5):509–518. Goodman SR, Kurdia A, Ammann L, Kakhniashvili D, Daescu O. The human red blood cell proteome and interactome. Exp Biol Med (Maywood). 2007;232(11): 1391– 1408. Dolznig H, Bartunek P, Nasmyth K, Müllner EW, Beug H. Terminal differentiation of normal chicken erythroid progenitors: shortening of G1 correlates with loss of Dcyclin/cdk4 expression and altered cell size control. Cell Growth Differ. 1995;6(11):1341–1352. Sankaran VG, Ludwig LS, Sicinska E, et al. Cyclin D3 coordinates the cell cycle during differentiation to regulate erythrocyte size and number. Genes Dev. 2012;26(18): 2075– 2087. Fowler VM. Regulation of actin filament length in erythrocytes and striated muscle. Curr Opin Cell Biol. 1996;8(1):86–96. Iolascon A, Perrotta S, Stewart GW. Red blood cell membrane defects. Rev Clin Exp Hematol. 2003;7(1):22–56. Handschick K, Beuerlein K, Jurida L, et al. Cyclin-dependent kinase 6 is a chromatinbound cofactor for NF-κB-dependent gene expression. Mol Cell. 2014;53(2):193–208. van den Bout I, Divecha N. PIP5K-driven PtdIns(4,5)P2 synthesis: regulation and cellular functions. J Cell Sci. 2009;122(Pt 21):3837–3850. Tolias KF, Hartwig JH, Ishihara H, et al. Type Iα phosphatidylinositol-4-phosphate 5kinase mediates Rac-dependent actin assembly. Curr. Biol. 2000;10(3):153–156. Janmey PA, Stossel TP. Modulation of gelsolin function by phosphatidylinositol 4,5bisphosphate. Nature. 325(6102):362–364. Janmey PA, Iida K, Yin HL, Stossel TP. Polyphosphoinositide micelles and polyphosphoinositide-containing vesicles dissociate endogenous gelsolin-actin complexes and promote actin assembly from the fast-growing end of actin filaments blocked by gelsolin. J Biol Chem. 1987;262(25): 12228–12236. Kalfa TA, Pushkaran S, Mohandas N, et al. Rac GTPases regulate the morphology and deformability of the erythrocyte cytoskeleton. Blood. 2006;108(12):3637–3645. Johnson N, Shapiro GI. Cyclin-dependent kinase 4/6 inhibition in cancer therapy. Cell Cycle. 2012;11(21):3913. Walker AJ, Wedam S, Amiri-Kordestani L, et al. FDA approval of palbociclib in combination with fulvestrant for the treatment of hormone receptor-positive, HER2-negative metastatic breast cancer. Clin Cancer Res. 2016;22(20):4968–4972.

1005


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Platelet Biology & its Disorders

Ferrata Storti Foundation

Haematologica 2017 Volume 102(6):1006-1016

Macrothrombocytopenia and dense granule deficiency associated with FLI1 variants: ultrastructural and pathogenic features

Paul Saultier,1 Léa Vidal,1 Matthias Canault,1 Denis Bernot,1 Céline Falaise,2 Catherine Pouymayou,2 Jean-Claude Bordet,3 Noémie Saut,1,2 Agathe Rostan,1,2 Véronique Baccini,1,2 Franck Peiretti,1 Marie Favier,1 Pauline Lucca,4,5,6 Jean-François Deleuze,7 Robert Olaso,7 Anne Boland,7 Pierre Emmanuel Morange,1,2 Christian Gachet,8,9,10,11 Fabrice Malergue,12 Sixtine Fauré,1 Anita Eckly,8,9,10,11 David-Alexandre Trégouët,4,5,6 Marjorie Poggi1* and Marie-Christine Alessi1,2*

Aix Marseille Univ, INSERM, INRA, NORT, Marseille; 2APHM, CHU Timone, French Reference Center on Inherited Platelet Disorders, Marseille; 3Unité d’Hémostase Biologique, Bron; 4ICAN Institute for Cardiometabolism and Nutrition, Paris; 5Inserm, UMR_S 1166, Team Genomics and Pathophysiology of Cardiovascular Diseases, Paris; 6 Sorbonne Universités, Université Pierre et Marie Curie (UPMC Univ Paris 06), UMR_S 1166; 7Centre National de Génotypage, Institut de Génomique, CEA, Evry; 8UMR_S949 INSERM, Strasbourg; 9Etablissement Français du Sang (EFS)-Alsace, Strasbourg; 10 Fédération de Médecine Translationnelle de Strasbourg (FMTS); 11Université de Strasbourg and 12Beckman Coulter Immunotech, Life Sciences Global Assay and Applications Development, Marseille, France 1

*MP and M-CA contributed equally to this work

ABSTRACT

C

Correspondence: paul.saultier@gmail.com

Received: September 6, 2016. Accepted: February 24, 2017. Pre-published: March 2, 2017. doi:10.3324/haematol.2016.153577 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/10006 ©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.

1006

ongenital macrothrombocytopenia is a family of rare diseases, of which a significant fraction remains to be genetically characterized. To analyze cases of unexplained thrombocytopenia, 27 individuals from a patient cohort of the Bleeding and Thrombosis Exploration Center of the University Hospital of Marseille were recruited for a highthroughput gene sequencing study. This strategy led to the identification of two novel FLI1 variants (c.1010G>A and c.1033A>G) responsible for macrothrombocytopenia. The FLI1 variant carriers’ platelets exhibited a defect in aggregation induced by low-dose adenosine diphosphate (ADP), collagen and thrombin receptor-activating peptide (TRAP), a defect in adenosine triphosphate (ATP) secretion, a reduced mepacrine uptake and release and a reduced CD63 expression upon TRAP stimulation. Precise ultrastructural analysis of platelet content was performed using transmission electron microscopy and focused ion beam scanning electron microscopy. Remarkably, dense granules were nearly absent in the carriers’ platelets, presumably due to a biogenesis defect. Additionally, 2529% of the platelets displayed giant α-granules, while a smaller proportion displayed vacuoles (7-9%) and autophagosome-like structures (03%). In vitro study of megakaryocytes derived from circulating CD34+ cells of the carriers revealed a maturation defect and reduced proplatelet formation potential. The study of the FLI1 variants revealed a significant reduction in protein nuclear accumulation and transcriptional activity properties. Intraplatelet flow cytometry efficiently detected the biomarker MYH10 in FLI1 variant carriers. Overall, this study provides new insights into the phenotype, pathophysiology and diagnosis of FLI1 variant-associated thrombocytopenia. Introduction Congenital reduced platelet count or impaired platelet function lead to a family of diseases, which are increasingly recognized as a significant cause of bleeding in children and adults.1 Congenital macrothrombocytopenias (CMTPs) constitute a large subgroup of inherited thrombocytopenia characterized by enlarged platelets. haematologica | 2017; 102(6)


FLI1-associated thrombocytopenia

Several genetic variants have been associated with CMTP, of which the most frequent alter genes that encode platelet surface proteins and megakaryocyte (MK) cytoskeletal proteins.2 However, mounting evidence has shown that several forms of CMTP are caused by variants in genes encoding hematopoietic transcription factors, which lead to altered downstream expression of genes that regulate platelet formation and function.3-10 In 2013, Stockley et al. used nextgeneration sequencing (NGS) to analyze 13 unrelated patients suspected of having an inherited qualitative platelet defect in the UK Genotyping and Phenotyping of Platelets (UK-GAPP) study.7 The group reported two substitutions and a 4-bp frameshift deletion in the transcription factor gene FLI1. Two of the index cases also had mild thrombocytopenia, and one of them exhibited enlarged platelets. These findings highlight the role of FLI1 in megakaryopoiesis and platelet function. To analyze cases of unexplained thrombocytopenia, 27 individuals from the patient cohort of the Bleeding and Thrombosis Exploration Center of the University Hospital of Marseille were recruited for a high-throughput gene sequencing investigation. In the study herein, we report the discovery of two novel FLI1 dominant variants linked to CMTP in two unrelated pedigrees. We investigated the pathogenesis of FLI1-associated thrombocytopenia by analyzing subcellular FLI1 localization and MK differentiation from hematopoietic progenitors derived from patients. We characterized the genotype-phenotype relationship and described the intraplatelet flow cytometry-based quantification of MYH10, a biomarker of FLI1 and RUNX1 alterations. Focused ion beam scanning electron microscopy (FIB-SEM) is an imaging approach based on the serial sectioning of entire platelets, which allows a 3D representation of the granules with respect to their shape, content and number. Using this method, we showed a severe dense granule deficiency in platelets of the FLI1 variant carriers.

Methods Methods concerning high-throughput gene sequencing, structural model of FLI1-DNA interactions, platelet phenotyping,

platelet-rich plasma (PRP) serotonin level, ATP secretion, mepacrine uptake and release, flow cytometric quantification of MYH10 expression (whole blood-based assay), western blot assay, site-directed mutagenesis, luciferase reporter assay, epifluorescence microscopy and in vitro MK differentiation and proplatelet formation are described in the Online Supplementary Material.

Recruitment of patients and genetic analysis strategy Patients were recruited from the cohort of the Bleeding and Thrombosis Exploration Center (University Hospital of Marseille) after informed written consent was obtained, in accordance with protocols approved by the local Institutional Review Board and the Declaration of Helsinki. Two high-throughput sequencing methods were used to identify candidate variants: (A) Wholeexome sequencing performed at the Centre National de Génotypage (Evry, France), or (B) sequencing of a 308 gene panel (selected genes associated with platelet diseases; available on request).

Flow-cytometric quantification of MYH10 expression (PRP-based assay) PRP was fixed in 1% paraformaldehyde phosphate buffered saline (PBS). Next, the fixed PRP was centrifuged at 1000 G for five min. The platelet pellet was suspended in permeabilization buffer (0.5% Triton X-100, 2mM Ethylenediaminetetraacetic acid, 0.5% bovine serum albumin (BSA) PBS) and labeled with rabbit antiMYH10 antibody (Cell Signaling; #3404) for one hour. After a washing step, the platelets were incubated with goat anti-rabbit Alexa-488-labeled secondary antibody (Abcam; ab150077) for 30 min in permeabilization buffer. After another washing step, the data were acquired using Navios Cytometer (Beckman Coulter) and analyzed with FlowJo software (Tree Star, Inc.).

Electron microscopy For ultrastructural analysis, platelets in citrated PRP were diluted and fixed in PBS pH 7.2 containing 1.25% glutaraldehyde for one hour. After centrifugation and two PBS washings, they were postfixed in 150 mM cacodylate-hydrochloric acid (HCl) buffer containing 1% osmium tetroxide pH 7.4 for 30 min at 4°C. After dehydration in a graded alcohol, embedding in EPON was performed by polymerization at 60°C for 72 hours. Ultrathin sections (~70 nm thick) were mounted on 200 mesh copper grids, contrasted with uranyl acetate and lead citrate and examined using a JEOL

Table 1. Platelet phenotyping in patients carrying the FLI1 variants.

Patient ID

FLI1 variant

Age‡ Platelet MPV¶ Serum (years) counts (fl) PAI1 (x109/l) levels (ng/ml)

Reference ranges§

F1-II2 F1-III1 F2-II4

152-402 7.1-9.6 92-283

c.1010 G>A (p.R337Q) c.1010 G>A (p.R337Q) c.1033 A>G (p.K345E)

PRP serotonin levels (μg/109plt) 0.30-1.2

46

154

10.7

430

0.26

19

131

13.0

443

NA

52

140

11.1

273

0.22

Platelet aggregation maximal intensity (%) ADP Coll* AA* 2.5 μM:75-95 78-90 5.0 μM:78-91 NA 2.5 μM:31 5.0 μM:81 2.5 μM:58 5.0 μM:76

79-94

NA

NA

83

85

81

83

Platelet glycoprotein MFI† (a.u.) αIIbβ3 GPIbα CD63 P-selectin Base: 20-47 TRAP: 28-60 Base: 21 TRAP: 30 Base: 34 TRAP: 42 Base: 22 TRAP: 30

Base: 1.3-5.1 TRAP: 0.7-2.2 Base: 1.2 TRAP: 0.9 Base: NA TRAP: NA Base: 2.9 TRAP: 2.9

Base: 0.4-1.3 TRAP: 1.3-4.2 Base: 0.6 TRAP: 0.8 Base: 0.7 TRAP: 1 Base: 0.9 TRAP: 1.3

Base: 0.2-0.8 TRAP: 1.9-7.4 Base: 0.3 TRAP: 1.7 Base: NA TRAP: NA Base: 0.6 TRAP: 5

Age at last evaluation. ¶MPV: mean platelet volume (measured by optical method using ADVIA 120, Siemens). §The reference ranges were defined as the [minimum-maximum] interval of values obtained from healthy individuals in our laboratory. *The collagen and arachidonic acid aggregation assays were performed using agonist concentrations of 3.3 μg/mL and 0.5 mg/mL, respectively. †Platelet glycoprotein expression was measured via flow cytometry at basal conditions and after stimulation using TRAP-14 (50 μM). PAI1: plasminogen activator inhibitor-1; PRP: platelet-rich plasma; plt: platelet; ADP: adenosine diphosphate; Coll: collagen; AA: arachidonic acid; a.u.: arbitrary unit; NA: not available; TRAP: thrombin receptor-activating peptide; MFI: mean fluorescence intensity. ‡

haematologica | 2017; 102(6)

1007


P. Saultier et al. A

B

D

C

Figure 1. Identification of two novel FLI1 variants. (A) Pedigrees for the affected families. Squares denote males and circles denote females. Black filled symbols represent family members carrying heterozygous c.1010 G>A and c.1033 A>G FLI1 variants. Dotted line symbols represent non-tested members. Arrows indicate the probands. (B) Schematic diagram of the FLI1 protein. The functional N-terminal Pointed domain (PNT), and C-terminal ETS DNA-binding domain (ETS) are depicted. The positions of the alterations in FLI1 are indicated in red (alterations reported in this study) or black (previously reported alterations).7,8 (C) Sequence alignment of the FLI1 protein (variant NM_002017.3). The variants reported in this study are indicated in red (top). Alignments of various members of the ETS-domain transcription factor family (middle) and different species (bottom) are provided. (D) Diagram of the simulated interactions between FLI1 and double-stranded DNA (dsDNA). Left: Interaction of native FLI1 with dsDNA. The FLI1 dsDNA-interacting domain is represented as a green ribbon. The dsDNA is represented as a stick model (carbons: pink; nitrogen: blue; oxygen: red; phosphate: orange). The interactions of R337 with Guanine 5 atoms N7 and O6 as well as K345 with Adenine 15 atom OP2 (side chain carbons: yellow; nitrogen: blue; oxygen: red) are indicated in red. Right: model of FLI1 structure with mutated residues p.R337Q and p.K345E. An expanded view of the interaction between altered residues and DNA is shown at the bottom.

JEM1400 transmission electron microscope equipped with a Gatan Orius 600 camera and Digital Micrograph software (Lyon Bio Image, Centre d'Imagerie Quantitative de Lyon Est, France). Morphometric measurements were done using ImageJ software (National Institutes of Health, USA). Preparations of platelet whole mounts were obtained after brief contact of formvar coated grids on small drops of PRP. Then, grids were rinsed rapidly with distilled water, dried with the edge of a filter paper, waved in the air and examined under the transmission electron microscope. Dense granules were identified and counted as previously described.11 FIB-SEM samples were processed as for transmission electron microscopy (TEM), and examined using a Helios NanoLab microscope (FEI). The 3D models were computed using Amira software.

Statistical analyses Statistical significance was determined via a 2-tailed MannWhitney test unless otherwise specified. P<0.05 was considered statistically significant. Quantitative variables were presented as mean ± standard error of the mean. Analyses were performed using GraphPad Prism software.

Results Identification of two novel FLI1 variants Two high-throughput sequencing methods were used to identify candidate variants in 27 patients with suspected congenital thrombocytopenia. DNA samples from 12 individuals (three members of four unrelated families) with unexplained autosomal dominant macrothrombocytope1008

nia were analyzed via whole-exome sequencing. In one family (Family F1), this analysis revealed a heterozygous single nucleotide change in FLI1. This missense variant, c.1010G>A, encodes an arginine to glutamine change (p.R337Q) in the highly conserved E twenty-six (ETS) DNA-binding domain of FLI1 (Figure 1A,B). The presence of the variant was confirmed using capillary sequencing. There was a segregation between the FLI1 variant and a history of thrombocytopenia in all cases for which DNA samples were available. In the three other families, no candidate variant was detected. Next, 15 additional individuals from 14 unrelated families were screened using a 308 gene panel NGS strategy. This screening led to the genetic diagnosis of four patients with macrothrombocytopenia. Aside from the identification of MYH9, ACTN1 and GP1BA variants in three unrelated individuals, one patient carried c.1033A>G heterozygous FLI1 variant, yielding p.K345E substitution (Family F2). In this family, the clinical and genetic status of the proband’s father (F2-I1) was unknown and the proband’s mother (F2-I2) displayed a normal platelet count, but the FLI1 gene was not sequenced. This second FLI1 variant also affected the ETS domain (Figure 1A,B). Sequence alignment of FLI1 orthologs and paralogs demonstrated that both variants affected amino acid residues in a highly conserved sequence (Figure 1C). As these variants were located in the ETS domain, a simulation of the protein-DNA interactions was performed. Regarding the p.R337Q variant, glutamine 337 retained the ability to interact with DNA; however, the haematologica | 2017; 102(6)


FLI1-associated thrombocytopenia

A

C

putative hydrogen (H)-bonds should be weaker than those established with the positively charged arginine guanidinium moiety. In the p.K345E variant, glutamic acid 345 is negatively charged and likely has a repelling effect on DNA binding (Figure 1D).

Description of the families The affected members of the first family (F1-II2 and F1III1) were both males, aged 46 and 19, respectively. They had not experienced significant bleeding (International Society on Thrombosis and Haemostasis Bleeding Assessment Tool (ISTH BAT) score: 1 for both), whereas the female proband in the second family (F2-II4), who was 52 years of age, had displayed excessive bleeding (ISTH BAT score: 16) with predominantly gynecological, obstetrical and oral cavity hemorrhaging. The three affected members did not display any extra hematologic abnormality. The first-line laboratory tests are reported in Table 1. All affected members presented with mild thrombocytopenia (mean value: 142 ± 7 x 109/l). The mean platelet volume (MPV) was elevated (11.6 ± 1.2 fl; reference range 7.1 - 9.6 fl). In these families, coagulation, fibrinolysis and von Willebrand factor levels were normal (data not shown).

Platelet receptor expression and platelet function evaluation The CD63 expression did not increase significantly upon TRAP stimulation (Table 1). The basal expression of αIIbβ3 was in the normal range among FLI1 variant carriers (Table 1). The platelet light transmission aggregometry (Figure 2A and Table 1) revealed a significant aggregation defect upon stimulation with different agonists, when used at a low concentration (2 and 2.5 µM ADP, 2 μg/ml collagen, 10 μM TRAP-6). The use of higher concentrahaematologica | 2017; 102(6)

B Figure 2. Platelet function analysis. (A) Light transmission aggregometry upon ADP (2 μM), collagen (2 and 10 μg/ml) and TRAP-6 (10 and 50 μM) stimulation in a FLI1 variant carrier (F1-III1) and in a control who was representative of 20 controls investigated at the same period of time. The PRP platelet count was 268 x109/l for the patient and 331 x109/l for the control. (B) Luminometry-based ATP secretion assay. ATP secretion was measured in two FLI1 variant carriers (F1-II2 and F1-III1) and unrelated controls after 100 µM TRAP-6 stimulation in 100 μl of diluted PRP (107 platelet / ml). *P<0.05 vs. controls (Mann-Whitney test). (C) Flow cytometric mepacrine uptake and release assay in two FLI1 variant carriers (F1-II2 and F2-II4) and unrelated controls. The platelets were incubated with 1.1 or 2.4 μM mepacrine and stimulated with 40 μM TRAP-14 to evaluate the mepacrine release. The mepacrine uptake was defined as the MFI ratio of platelets incubated with mepacrine to platelets incubated without mepacrine, and the mepacrine release was defined as the MFI ratio of resting platelets to stimulated platelets. ADP: adenosine diphosphate; ATP: adenosine triphosphate; TRAP: thrombin receptor-activating peptide; Coll: collagen; MFI: mean fluorescence intensity.

tions of agonists (5 μM ADP, 10 μg/ml collagen, 50 μM TRAP-6) induced normal platelet aggregation. The ATP secretion upon TRAP-6 stimulation was 2.5-fold decreased in patients F1-II2 and F1-III1 as compared to controls (Figure 2B). The mepacrine uptake and release upon TRAP-14 stimulation was reduced (Figure 2C). Among FLI1 variant carriers, the platelet serotonin level was low but the serum plasminogen activator inhibitor-1 (PAI-1) antigen level (an α-granule biomarker) was normal or high (Table 1).

Increased MYH10 expression in the platelets of affected members The expression of MYH10 in platelets has been recently proposed as a biomarker for RUNX1 and FLI1 alterations.7,8,12 Intraplatelet flow-cytometry showed an almost 4-fold overexpression of MYH10 among the FLI1 variant carriers as compared to controls, using a PRP-based and a whole blood-based experimental procedure (Figure 3A and Online Supplementary Figure S1, respectively). The MYH10 mean fluorescence intensity (MFI) was expressed as a function of MPV in 13 control individuals (MPV range 7.2 - 11.4 fl) and two FLI1 patients (MPV 10.7 and 13 fl), demonstrating that the increased MPV could not explain the specific increase in MYH10 MFI observed in the FLI1 patients (Online Supplementary Figure S1). Western blot analysis confirmed the increased expression of MYH10 in the platelet lysates of the FLI1 variant carriers (Figure 3B).

Functional assessment of the FLI1 variants Western blot analysis did not reveal an alteration in FLI1 protein expression in the platelets of affected members compared with healthy controls (Figure 4A). Accordingly, protein expression of wild-type (WT) and both FLI1 variants in transfected cells were equivalent, thereby suggest1009


P. Saultier et al.

ing that these variants do not influence protein expression or stability (Figure 4B). To investigate the effect of the p.R337Q and p.K345E variants on FLI1 transcriptional activity, we investigated the ability of the recombinant FLI1 variants to regulate transcriptional activity using a dual-luciferase reporter assay. Co-transfection of the reporter plasmid containing the ETS-binding site and a plasmid encoding WT FLI1 resulted in almost 70% inhibition of luciferase activity. There was a loss in the ability of the p.R337Q and p.K345E variants to repress luciferase activity compared with that of WT FLI1 (P<0.0001, Figure 4C). Reduced transcriptional activity was also observed in this model for the FLI1 p.R324Q variant previously reported by Stevenson et al.8 The co-transfection of WT and variant FLI1 led to normal transcriptional activity (Figure 4C). The R337 and K345 residues were predicted to be part of a nuclear localization signal sequence using several prediction tools (cNLS mapper and NLStradamus; data not shown). To examine whether the p.R337Q and p.K345E

variants alter the subcellular localization of FLI1 protein, we performed cell fractionation assay and immunofluorescence staining of cells overexpressing WT or variant FLI1. Western blot analysis of subcellular fractions showed increased FLI1 protein level in the cytoplasmic fraction and decreased protein level in the nuclear fraction in cells expressing the p.R337Q and p.K345E FLI1 variants compared with cells expressing the WT protein (Figure 4D). Concordant with the western blot data, fluorescence microscopy showed that WT FLI1 concentrated primarily in cell nuclei. In contrast, both FLI1 variants exhibited predominant cytoplasmic localization (Figure 4E).

FLI1 variants are associated with reduced CD34+derived megakaryocyte differentiation and proplatelet formation in vitro We generated MKs from peripheral blood CD34+ cells in the presence of thrombopoietin (TPO) and stem cell factor (SCF) in liquid culture. At day 11, the percentage of

Table 2. FIB-SEM analysis of platelet ultrastructure.

Patient ID

FLI1 variants

Control F1-II2 F2-II4

c.1010 G>A (p.R337Q) c.1033 A>G (p.K345E)

Platelet volume (fl)

α-granule volume (fl)

7.6±0.5

0.018±0.003

12.6±0.1 12.8±2.0

Giant α-granule volume (fl)

Giant α-granules (%)

Proportion of platelets displaying Autophagosome-like Glycogen structures accumulation (%) (%)

0.014±0.005

None detected 0.22±0.02

None detected 25

None detected 3

86

None detected 9

0.017±0.005

0.08±0.02

29

None detected

51

7

8

Vacuoles (%)

Giant α-granules were defined by a diameter of >400 nm.Values are shown as the mean ± SEM as quantified for n=38, 52 or 45 randomly selected platelets for control, F1-II2 and F2-II4, respectively.

A

B

1010

Figure 3. Quantification of platelet MYH10 expression. (A) Flow cytometry-based quantitative detection of intraplatelet MYH10 expression (PRP-based assay). Left: representative overlay of histograms of intraplatelet MYH10 expression in an affected member (F1-II2) and a healthy control. The non-specific staining (irrelevant IgG) is only presented for the affected member, which can be superimposed for that of the control. Right: MYH10 mean fluorescence intensity (MYH10 MFI – irrelevant IgG MFI) in affected members (n=3) or unrelated controls (n=8) from three independent experiments. The results are expressed as fold change relative to corresponding controls; **P<0.01 vs. controls (Mann-Whitney test). (B) Left: representative western blot analysis of MYH10 expression in platelets from the affected members (F1-II2, F1-III1, F2-II4), one unaffected member (F1-II3) and two unrelated controls. GAPDH and actin were used as a protein loading control. Right: the results of densitometric analysis were normalized to actin and expressed as mean ± SEM; **P<0.01 vs. controls (Mann-Whitney test). Three independent experiments were performed. GAPDH: glyceraldehyde 3-phosphate dehydrogenase; IgG: immunoglobulin G; a.u.: arbitrary units.

haematologica | 2017; 102(6)


FLI1-associated thrombocytopenia

mature CD41hiCD42ahi MKs was strikingly reduced, while the percentage of CD41lowCD42– and CD41–CD42– cells was increased in F1-II2 compared with the control (Figure 5A). The percentage of high-ploidy cells (≥8n) was reduced among FLI1 variant carriers at day 12 (11.9, 8.2

A

C

and 5.8% for the control, the F1-II2 and the F1-III1 affected members, respectively) and day 14 of maturation (11.3, 6.2 and 4.5% for the control, the F1-II2 and the F1-III1 affected members, respectively) (Figure 5B). At days 12-13, the percentage of proplatelet (PPT)-forming MKs was sig-

B

D

E

Figure 4. Functional characterization of the FLI1 variants. (A) Left: representative western blot analysis of FLI1 expression in platelets from the affected members (F1-II2, F1-III1) and three control individuals. GAPDH and actin were used as a protein loading control. Right: the results of the densitometric analysis were normalized to actin and expressed as mean ± SEM from three independent experiments. (B) Left: representative western blot analysis of FLI1 expression in GripTite 293 MSR cells transfected with an empty vector, wild-type (WT) or variant FLI1 constructs using an anti-HA antibody. GAPDH was used as a protein loading control. Right: the results of the densitometric analysis are expressed as mean ± SEM. Three independent experiments were performed. (C) GripTite 293 MSR cells were co-transfected with an empty vector, WT or variant FLI1 constructs including the c.970C.T FLI1 variant previously reported by Stevenson et al.8 (p.R324W) along with the luciferase reporter plasmid containing three tandem copies of the ETS-binding site upstream of the HSV tk promoter (E743tk80Luc) and pGL4.73 Renilla luciferase control vector. Firefly to renilla luminescence ratios (Fluc/Rluc) were calculated to compensate for transfection efficiency and expressed as fold change relative to empty vector. The data represent the mean ± SEM of three independent experiments; *P<0.05, ****P<0.0001 vs. WT (one-way ANOVA with Dunnett’s post hoc test). (D) Western blot analysis of WT and variant FLI1 subcellular localization. GripTite 293 MSR cells were transfected with an empty vector, WT or variant FLI1 constructs. The laminB1 and GAPDH expression were used as nuclear and cytoplasmic markers, respectively. The data are the mean ± SEM of four independent experiments; *P<0.05 vs. WT (Mann-Whitney test). (E) Left: representative immunofluorescence microscopy images of H9C2 cells transfected with WT or variant FLI1 constructs visualized using bright field illumination and immunofluorescence after FLI1 and DAPI staining; scale bar, 5 μm. Right: quantification of the nuclear and cytoplasmic integrated density of fluorescence. The data are expressed as mean ± SEM of the nucleo-cytoplasmic ratio of fluorescence integrated density from two independent experiments (≥ 30 total cells were assessed for each condition); ****P<0.0001 vs. WT (Mann-Whitney test). GAPDH: glyceraldehyde 3-phosphate dehydrogenase; HA: hemagglutinin; BF: bright field.

haematologica | 2017; 102(6)

1011


P. Saultier et al.

nificantly reduced in the affected members F1-II2 and F2II4 compared with three controls (16 ± 1% vs. 3 ± 1%, P<0.05, Figure 5C). MKs from patients were smaller and formed very few PPTs, which displayed reduced extensions and branching (Figure 5C).

Platelets from affected members exhibit giant α-granules and an absence of dense granules

May-Grünwald-Giemsa staining of peripheral blood smears of patient F1-II2 revealed enlarged platelets with giant α-granules (Figure 6A). Accordingly, TEM analysis showed the presence of large fused α-granules in the carriers’ platelets (Figure 6B). The mean diameter of the αgranules was significantly increased (Figure 6C). The mean

A

α-granule number per μm2 of platelet section was slightly reduced in the FLI1 p.R337Q variant carrier but not in the p.K345E variant carrier. We quantified the platelet surface on sections (Figure 6C), which confirmed the increased platelet size associated with FLI1 variants (Table 1). Whole mount electron microscopy, FIB-SEM and TEM was used to quantify the dense granules, which were nearly absent in carriers’ platelets (Figure 6B,C and Online Supplementary Figure S2), thereby indicating a dense granule storage pool deficiency. Precise platelet content analysis and 3D reconstruction of the α- and dense granules within platelets was performed using the FIB-SEM technique. Among affected members, the proportion of platelets displaying giant αgranules (diameter > 400nm) was 25-29% (Table 2). The

B

C

Figure 5. Megakaryocyte differentiation and proplatelet formation. Circulating CD34+ progenitors from affected members or controls were isolated and cultured in the presence of TPO and SCF to induce megakaryocytic commitment. (A) MK differentiation was monitored using flow cytometry. The density plots represent CD41 and CD42a expression in Hoechst+ cells from an affected member (F1-II2) and an unrelated control at day 11 of culture. The ellipse gates show the populations CD41–CD42a–, CD41lowCD42a– cells and mature CD41hiCD42ahi MKs. (B) Ploidy level was monitored by flow cytometry. The histograms represent frequency distribution of Hoechst levels among the CD34+-derived cells from two affected members (F1-II2 and F1-III1) and an unrelated control at day 6, 12 and 14 of culture. (C) Representative microscopic images of proplatelet (PPT) formation after 13 days of culture. An expanded view of PPT formation is shown at the bottom. PPT formation was quantified in two affected members (F1-II2 and F2-II4) and three unrelated controls at culture days 12 and 13 from three independent experiments. The PPTforming MK are indicated with triangle markers; scale bar, 10 μm. The percentage of PPT-forming MKs was estimated by counting MKs harboring ≥ 1 cytoplasmic process with areas of constriction; ≥ 180 total cells were assessed for each individual. The results are expressed as mean ± SEM; *P<0.05 (Mann-Whitney test). MK: megakaryocyte.

1012

haematologica | 2017; 102(6)


FLI1-associated thrombocytopenia giant α-granules had a mean volume of almost 8-fold that of normal α-granules. While dense granules were easily distinguished through their typical dark central core and spherical geometry in control platelets (8.6 ± 0.7 dense granules per platelet; n=12), they were almost absent

within platelets of FLI1 variant carriers (1.1 ± 0.3, 1.1 ± 0.2 dense granules per platelet for F1-II2 and F2-II4 affected members, respectively; n=12) (Figure 6C). Empty dense granule membrane structures were rarely detected (3 ± 0.5, 1 ± 0.3 and 0.5 ± 0.2 copies for F1-II2, F2-II4 and con-

A

B

C

D

Figure 6. Platelet structure defects associated with FLI1 variants. (A) Representative May-Grünwald-Giemsa stained blood smears showing enlarged platelets with giant α-granules (patient F1-III3). Scale bar, 5 μm. (B) Platelet ultrastructural analysis from FLI1 variant carriers (F1-II2 and F2-II4) and an unrelated control. Representative electron microscopy images of platelets transmission electron microscopy (TEM) ultrathin sections (top) and whole mount (middle). 3D reconstruction of platelet α- and dense granules from focused ion beam scanning electron microscopy (FIB-SEM) images (bottom); Giant α-granules are indicated with triangle markers; δ: dense granule, α: alpha-granule; Scale bar, 1 μm. (C) Measurement of the α-granules diameter and number / μm2 of platelet section, platelet surface, number of dense granules and number of empty granules. Values are shown as the mean ± SEM as quantified for ≥100 randomly selected platelets for TEM and whole mount and 12 randomly selected platelets for FIB-SEM; *P<0.05, ***P<0.001, ****P<0.0001 vs. controls (Mann-Whitney test). (D) Representative ultrastructural image of double membrane structures resembling autophagosome (far left panel with enlarged view middle left panel; black triangle marker: α-granule; white triangle marker: mitochondrion), area of glycogen accumulation (middle right panel; black triangle marker: glycogen), and vacuoles (far right panel; black triangle marker: vacuole) in the platelet of the affected member F1-II2.

haematologica | 2017; 102(6)

1013


P. Saultier et al.

trol platelets, respectively; n=12; Figure 6C). Double membrane structures resembling autophagosomes were detected in 3% of the platelets from the affected member F1-II2 (Table 2; Figure 6D, left panel), while these were absent in the affected member F2-II4. Accumulation of glycogen was found in the cytoplasm of many platelets from the affected members (86% and 51% in F1-II2 and F2-II4 platelets, respectively; Table 2 and Figure 6D, middle right panel) and vacuoles containing glycogen were also observed in 9% and 7% of F1-II2 and F2-II4 platelets, respectively (Table 2 and Figure 6D, far right panel). FIBSEM image stacks in F1-II2 and a control allowed the 3D observation of the platelet morphological defects associated with FLI1 variants (Online Supplementary Videos S1 and S2).

Discussion In the present study, we identified two novel FLI1 variants in patients from two unrelated families displaying CMTP with marked bleeding diathesis in one of them. The reported FLI1 variants were associated with impaired megakaryocytic differentiation and PPT formation in vitro and with a defect in the protein subcellular localization, thereby providing further insights into the pathophysiology of FLI1-associated thrombocytopenia. Remarkably, we demonstrated that dense granules were nearly absent in the carriers’ platelets. In addition, we showed that intraplatelet flow cytometry efficiently detected the biomarker MYH10 in FLI1 variant carriers. FLI1 belongs to the ETS-domain transcription factor family, in which members bind a specific DNA consensus sequence to control the expression of genes that are essential in the regulation of cellular proliferation, differentiation and programmed cell death.13 FLI1 is mainly expressed in hematopoietic and vascular endothelial cells and plays a role in the regulation of essential MK genes.7,8,14-23 In humans, a role for FLI1 in megakaryopoiesis was highlighted by studies of Paris-Trousseau syndrome (PTS)24-26 in which the FLI1 locus is hemizygously deleted, and more recently by two articles reporting FLI1 variants associated with congenital thrombocytopenia.7,8 The affected members carried heterozygous variants, with a dominant mode of inheritance in the F1 family. The thrombocytopenia associated with the FLI1 variants reported by Stockley et al.7 and the PTS24,25 were also shown to be transmitted in a dominant manner. Inversely, Stevenson et al. have described a family displaying inherited thrombocytopenia caused by a recessive FLI1 variant.8 We hypothesize that the mode of inheritance could rely on the degree of the protein functional impairment, as suggested by the moderately reduced transcriptional activity of the p.R324W variant, associated with the recessive inheritance pattern. Intriguingly, in mouse models of a targeted null Fli1 variant, the mice carrying a heterozygous deletion of Fli1 are phenotypically indistinguishable from the WT mice. We hypothesize that these mouse models do not undergo silencing of one FLI1 allele during the early stages of megakaryopoiesis, contrary to that observed in humans.26 The bleeding phenotype of F2-II4 was much more severe than that of the affected members of the family F1 (ISTH BAT score 16 vs. 1). An additional bleeding disorder (affecting coagulation, von Willebrand factor or fibrinoly1014

sis) was ruled out. A combination of genetic, environmental and lifestyle factors may interfere in the genotype-phenotype relationship. This difference is not likely to be directly related to platelet count or patient age; however, this could be explained in part by the sex of the patient. Indeed, the individual with a high bleeding score (F2-II4) predominantly suffered from gynecological and obstetrical bleeding. This may reflect the limitation of the ISTH BAT score to evaluate the bleeding risk in males who have not been challenged enough as compared to females. In PTS, the platelet count seems to be as low as 20 x 109/l early in life and subsequently increases to near or low normal levels.25,27 Similar to that shown in a previous report of FLI1 variant carriers,7 the older patient in the F1 family exhibited a higher platelet count than the younger one. The platelet counts and the increased MPV of the affected members were consistent with that observed in most PTS cases and FLI1 variant-associated thrombocytopenia patients.7,25 As noted in some PTS patients, the bleeding phenotype of patient F2-II4 was more severe than that expected with a mild thrombocytopenia, thereby suggesting associated abnormal platelet function.25,27 In the present study, the affected members exhibited an aggregation defect upon low dose ADP, collagen, and TRAP-6 stimulation, which is a hallmark of dense granule storage pool deficiency. Data concerning aggregation assays has varied widely in the literature. Breton-Gorius et al. have reported normal aggregation curves in PTS patients.24 Regarding FLI1 variant-associated thrombocytopenia, Stevenson et al. have reported reduced aggregation to collagen and ADP.8 Finally, in a murine model expressing a C-terminal truncated Fli1 protein, aggregation tests revealed reduced ADP- and thrombin-induced aggregation.28 These discrepancies could be in part due to differences regarding agonist concentrations or sources. Furthermore, storage pool diseases are known to be associated with variable aggregation profiles.29 In the study herein, NGS-based techniques led to the identification of the FLI1 variants. These strategies require careful interpretation as they lead to the identification of a high number of non-pathogenic variants. However, clinical phenotype and screening laboratory tests are not specific and often fail to yield a specific diagnosis. Identifying biomarkers could facilitate the diagnostic process of inherited platelet disorders. In this regard, platelet-specific expression of MYH10, whose silencing is required for the switch from mitosis to endomitosis during MK maturation,30 has been shown to constitute a biomarker for the inherited platelet disorders associated with FLI1 and RUNX1 alterations.7,8,12 Platelet MYH10 expression is usually detected via western blot. This assay is complex and time-consuming, and the results must be carefully interpreted as control platelets exhibit low expression of MYH10. In this study, we developed a quantitative assay to detect MYH10 overexpression in FLI1 variant carriers’ platelets using intracellular flow cytometry. This test enabled rapid and accurate discrimination between the affected members and controls using only a small volume of PRP or whole blood. α- and dense granule defects were identified as the main platelet defect in the affected members. It is important to study the mechanisms of this functional impairment, as bleeding diathesis may still occur in FLI1-associated thrombocytopenia patients with subnormal or normal platelet counts.27 We report that up to 29% of the patients’ haematologica | 2017; 102(6)


FLI1-associated thrombocytopenia platelets displayed giant α-granules, which is a characteristic feature of PTS24,25 and have been previously reported in FLI1 variant-associated thrombocytopenia,8 albeit in a much smaller proportion of platelets (4-15%). This discrepancy can be attributed to differences in the analysis methods. Contrary to classic TEM, FIB-SEM enables evaluation of the whole granule content of a platelet. In the present study, normal or even increased serum PAI-1 levels suggest normal α-granule content and release.31 Taken together, these results indicate that the vesicle trafficking or the α-granule proteins packaging but not the storage mechanisms may be impaired. Regarding dense granules, we demonstrated for the first time using electron microscopy (notably FIB-SEM), ATP secretion, mepacrine uptake and release, CD63 expression upon stimulation and serotonin measurement, that dense granules were nearly absent in the carriers’ platelets and that empty dense granule membrane structures were rarely detected. This suggests that the platelet dense granule deficiency is due to an impaired biogenesis. A previous study on Jacobsen syndrome, a variant of PTS, has reported a dense granule storage pool deficiency in six patients32 using whole mount electron microscopy, which only allows the evaluation of the electron opaque dense granules content. More recently, Stockley et al. showed that the platelet of FLI1 variants carriers exhibited an ATP secretion defect7 but have not evaluated the dense granule content. Interestingly, in a recent study reporting ChIP-seq data in murine MK,33 FLI1 was shown to interact with HPS4 and RAB27B genes, which are known regulators of dense granules biogenesis.34,35 Autophagosome-like structures were noted in 3% of platelets from one affected member. Autophagy is known to be important in both megakaryopoiesis36 and platelet function.37 This observation thus suggests that autophagy may play a role in the pathophysiology of FLI1-associated thrombocytopenia. Glycogen accumulations were noted in the platelets of most of the affected members. The significance of this abnormal feature is unclear, although it may suggest that FLI1 plays a role in the regulation of platelet intracellular metabolic processes. The p.R337Q and p.K345E variants are located within the ETS-domain, similar to the previously reported variants.7,8 The pathogenic effect of these novel FLI1 variants was confirmed by the significant reduction in transcriptional activity, as with all other FLI1 variants reported in the literature.7,8 Furthermore, we show that both p.R337Q and p.K345E variants displayed significantly altered cellular localization with a reduced nuclear accumulation. In a previous study based on directed mutagenesis, these residues were found to be important for the DNA-bound conformation of FLI1 and its nuclear accumulation.38 Additionally, the p.R337A variant abolished ETS DNA-

References 1. Lambert MP. Update on the inherited platelet disorders. Curr Opin Hematol. 2015;22(5):460466. 2. Favier R, Raslova H. Progress in understanding the diagnosis and molecular genetics of macrothrombocytopenias. Br J Haematol. 2015;170(5):626-639. 3. Songdej N, Rao AK. Inherited platelet dys-

haematologica | 2017; 102(6)

binding activity.38 Overall, these results suggest that both nuclear accumulation and DNA-binding capacity could interfere with FLI1 regulation of MK-specific genes. Other mechanisms may also cause this impairment, such as abnormal recruitment of co-regulators or altered activation of histone deacetylases or acetylases.39 In vitro, we noted a marked diminution of mature CD41hiCD42ahi MKs, along with an increase in CD41lowCD42a- and CD41-CD42- cells among FLI1 variant carriers. Even if we cannot formally exclude an impaired expression of CD41 and CD42a by FLI1 variants, the reduced percentage of high-ploidy cells (≥8n) confirmed the MK differentiation defect. Furthermore, the platelet from the FLI1 variant carriers displayed normal expression of CD42a (data not shown) and CD41. Consistent with our findings, data from PTS patients has shown that the MK transition from CD42a- to CD42a+ was especially sensitive to a reduction in FLI1 dosage.26 In a murine model with complete Fli1 deficiency, the authors reported that megakaryocytic progenitors displayed an early blockage in MK differentiation.40 Additionally, in a murine model expressing a truncated Fli1 protein lacking the C-terminal transcriptional activation domain, bone marrow cell analysis revealed a 2-fold increase in CD34+CD41+CD42- cells.28 The study herein provides new insights into the mechanisms that drive FLI1-associated thrombocytopenia and highlights its function in normal and pathologic megakaryopoiesis and in platelet granule biogenesis. TEM and FIBSEM enabled a more detailed description of the ultrastructural features associated with this disease, demonstrating that dense granules were nearly absent in the carriers’ platelets. This work also shows that the increased expression of MYH10 in the platelets of FLI1 variant carriers was efficiently detected by flow cytometry. Funding The study was funded by the “Fondation pour la Recherche Médicale” (grant to PS: FDM20150633607). The authors acknowledge Christian Cambillau (Architecture et Fonction des Macromolécules Biologiques, UMR 7257, Centre National de la Recherche Scientifique, Marseille, France) for the structural modelling, the members of the French reference center on hereditary platelet disorders (CRPP) for clinical analysis of patients, Monique Verdier, Odile Georgelin and Jean-Yves Rinckel for experimental assistance, Marc Delépine and Céline Baulard for genetic study and Sandra Moore and Delphine Bastelica for revision of the paper. Bioinformatic analysis of the exome data were performed using the C2BIG computing cluster funded by the Region Ile de France, the Pierre and Marie Curie University and the ICAN Institute for Cardiometabolism and Nutrition (ANR10-IAHU-05). The authors would also like to thank the patients and their families.

function and hematopoietic transcription factor mutations. Platelets. 2017;28(1):20-26. 4. Nichols KE, Crispino JD, Poncz M, et al. Familial dyserythropoietic anaemia and thrombocytopenia due to an inherited mutation in GATA1. Nat Genet. 2000; 24(3):266-270. 5. Freson K, Devriendt K, Matthijs G, et al. Platelet characteristics in patients with Xlinked macrothrombocytopenia because of

a novel GATA1 mutation. Blood. 2001; 98(1):85-92. 6. Kitamura K, Okuno Y, Yoshida K, et al. Functional characterization of a novel GFI1B mutation causing congenital macrothrombocytopenia. J Thromb Haemost. 2016;14(7):1462-1469. 7. Stockley J, Morgan NV, Bem D, et al. Enrichment of FLI1 and RUNX1 mutations in families with excessive bleeding and

1015


P. Saultier et al.

8.

9.

10.

11. 12.

13. 14.

15.

16.

17.

18.

1016

platelet dense granule secretion defects. Blood. 2013;122(25):4090-4093. Stevenson WS, Rabbolini DJ, Beutler L, et al. Paris-Trousseau thrombocytopenia is phenocopied by the autosomal recessive inheritance of a DNA-binding domain mutation in FLI1. Blood. 2015; 126(17):2027-2030. Stevenson WS, Morel-Kopp M-C, Chen Q, et al. GFI1B mutation causes a bleeding disorder with abnormal platelet function. J Thromb Haemost. 2013;11(11):2039-2047. Monteferrario D, Bolar NA, Marneth AE, et al. A dominant-negative GFI1B mutation in the gray platelet syndrome. N Engl J Med. 2014;370(3):245-253. White JG. Electron opaque structures in human platelets: which are or are not dense bodies? Platelets. 2008;19(6):455-466. Antony-Debré I, Bluteau D, Itzykson R, et al. MYH10 protein expression in platelets as a biomarker of RUNX1 and FLI1 alterations. Blood. 2012;120(13):2719-2722. Sharrocks AD. The ETS-domain transcription factor family. Nat Rev Mol Cell Biol. 2001;2(11):827-837. Lemarchandel V, Ghysdael J, Mignotte V, Rahuel C, Roméo PH. GATA and Ets cis-acting sequences mediate megakaryocyte-specific expression. Mol Cell Biol. 1993; 13(1):668-676. Zhang L, Lemarchandel V, Romeo PH, BenDavid Y, Greer P, Bernstein A. The Fli-1 proto-oncogene, involved in erythroleukemia and Ewing’s sarcoma, encodes a transcriptional activator with DNA-binding specificities distinct from other Ets family members. Oncogene. 1993;8(6):16211630. Wang X, Crispino JD, Letting DL, Nakazawa M, Poncz M, Blobel GA. Control of megakaryocyte-specific gene expression by GATA-1 and FOG-1: role of Ets transcription factors. EMBO J. 2002; 21(19):5225-5234. Hashimoto Y, Ware J. Identification of essential GATA and Ets binding motifs within the promoter of the platelet glycoprotein Ib alpha gene. J Biol Chem. 1995;270(41):24532-24539. Bastian LS, Kwiatkowski BA, Breininger J, Danner S, Roth G. Regulation of the megakaryocytic glycoprotein IX promoter by the oncogenic Ets transcription factor Fli1. Blood. 1999;93(8):2637-2644.

19. Deveaux S, Filipe A, Lemarchandel V, Ghysdael J, Roméo PH, Mignotte V. Analysis of the thrombopoietin receptor (MPL) promoter implicates GATA and Ets proteins in the coregulation of megakaryocyte-specific genes. Blood. 1996; 87(11):4678-4685. 20. Furihata K, Kunicki TJ. Characterization of human glycoprotein VI gene 5’ regulatory and promoter regions. Arterioscler Thromb Vasc Biol. 2002;22(10):1733-1739. 21. Watson DK, Smyth FE, Thompson DM, et al. The ERGB/Fli-1 gene: isolation and characterization of a new member of the family of human ETS transcription factors. Cell Growth Differ. 1992;3(10):705-713. 22. Seth A, Robinson L, Thompson DM, Watson DK, Papas TS. Transactivation of GATA-1 promoter with ETS1, ETS2 and ERGB/Hu-FLI-1 proteins: stabilization of the ETS1 protein binding on GATA-1 promoter sequences by monoclonal antibody. Oncogene. 1993;8(7):1783-1790. 23. Pang L, Xue H-H, Szalai G, et al. Maturation stage-specific regulation of megakaryopoiesis by pointed-domain Ets proteins. Blood. 2006;108(7):2198-2206. 24. Breton-Gorius J, Favier R, Guichard J, et al. A new congenital dysmegakaryopoietic thrombocytopenia (Paris-Trousseau) associated with giant platelet alpha-granules and chromosome 11 deletion at 11q23. Blood. 1995;85(7):1805-1814. 25. Favier R, Jondeau K, Boutard P, et al. ParisTrousseau syndrome : clinical, hematological, molecular data of ten new cases. Thromb Haemost. 2003;90(5):893-897. 26. Raslova H, Komura E, Le Couédic JP, et al. FLI1 monoallelic expression combined with its hemizygous loss underlies ParisTrousseau/Jacobsen thrombopenia. J Clin Invest. 2004;114(1):77-84. 27. Grossfeld PD, Mattina T, Lai Z, et al. The 11q terminal deletion disorder: a prospective study of 110 cases. Am J Med Genet A. 2004;129A(1):51-61. 28. Moussa O, LaRue AC, Abangan RS, et al. Thrombocytopenia in mice lacking the carboxy-terminal regulatory domain of the Ets transcription factor Fli1. Mol Cell Biol. 2010;30(21):5194-5206. 29. Nieuwenhuis HK, Akkerman JW, Sixma JJ. Patients with a prolonged bleeding time and

30.

31.

32. 33.

34.

35.

36.

37. 38.

39.

40.

normal aggregation tests may have storage pool deficiency: studies on one hundred six patients. Blood. 1987; 70(3):620-623. Lordier L, Bluteau D, Jalil A, et al. RUNX1induced silencing of non-muscle myosin heavy chain IIB contributes to megakaryocyte polyploidization. Nat Commun. 2012;3:717. Booth NA, Simpson AJ, Croll A, Bennett B, MacGregor IR. Plasminogen activator inhibitor (PAI-1) in plasma and platelets. Br J Haematol. 1988;70(3):327-333. White JG. Platelet storage pool deficiency in Jacobsen syndrome. Platelets. 2007; 18(7):522-527. Zang C, Luyten A, Chen J, Liu XS, Shivdasani RA. NF-E2, FLI1 and RUNX1 collaborate at areas of dynamic chromatin to activate transcription in mature mouse megakaryocytes. Sci Rep. 2016;6:30255. Suzuki T, Li W, Zhang Q, et al. HermanskyPudlak syndrome is caused by mutations in HPS4, the human homolog of the mouse light-ear gene. Nat Genet. 2002;30(3):321324. Tolmachova T, Abrink M, Futter CE, Authi KS, Seabra MC. Rab27b regulates number and secretion of platelet dense granules. Proc Natl Acad Sci USA. 2007; 104(14):58725877. Cao Y, Cai J, Zhang S, et al. Loss of autophagy leads to failure in megakaryopoiesis, megakaryocyte differentiation, and thrombopoiesis in mice. Exp Hematol. 2015;43(6):488-494. Feng W, Chang C, Luo D, et al. Dissection of autophagy in human platelets. Autophagy. 2014;10(4):642-651. Hu W, Philips AS, Kwok JC, Eisbacher M, Chong BH. Identification of nuclear import and export signals within Fli-1: roles of the nuclear import signals in Fli-1-dependent activation of megakaryocyte-specific promoters. Mol Cell Biol. 2005;25(8):3087-3108. Sakimura R, Tanaka K, Nakatani F, et al. Antitumor effects of histone deacetylase inhibitor on Ewing’s family tumors. Int J Cancer. 2005;116(5):784-792. Hart A, Melet F, Grossfeld P, et al. Fli-1 is required for murine vascular and megakaryocytic development and is hemizygously deleted in patients with thrombocytopenia. Immunity. 2000; 13(2):167-177.

haematologica | 2017; 102(6)


ARTICLE

Bone Marrow Failure

An abnormal bone marrow microenvironment contributes to hematopoietic dysfunction in Fanconi anemia Yuan Zhou,1,2,3§ Yongzheng He,2,3§ Wen Xing,1,2,3 Peng Zhang,4,5 Hui Shi,1,4,5 Shi Chen,4,5 Jun Shi,1 Jie Bai,1 Steven D. Rhodes,2,3 Fengqui Zhang,1 Jin Yuan,2,3 Xianlin Yang,2,3 Xiaofan Zhu,1 Yan Li,2,3 Helmut Hanenberg,2,3,6 Mingjiang Xu,4,5 Kent A. Robertson,2,3 Weiping Yuan,1 Grzegorz Nalepa,2,3 Tao Cheng,1 D. Wade Clapp2,3 and Feng-Chun Yang4,5 §

These authors contributed equally to this work.

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2017 Volume 102(6):1017-1027

State Key Laboratory of Experimental Hematology, Institute of Hematology and Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China; 2Herman B Wells Center for Pediatric Research, Indianapolis, IN, USA; 3Department of Pediatrics, Indiana University School of Medicine, Indianapolis, IN, USA; 4Sylvester Comprehensive Cancer Center, Miami, FL, USA; 5Department of Biochemistry and Molecular Biology, University of Miami Miller School of Medicine, Miami, FL, USA and 6Department of Otorhinolaryngology and Head/Neck Surgery, Heinrich Heine University, Düsseldorf, Germany 1

ABSTRACT

F

anconi anemia is a complex heterogeneous genetic disorder with a high incidence of bone marrow failure, clonal evolution to acute myeloid leukemia and mesenchymal-derived congenital anomalies. Increasing evidence in Fanconi anemia and other genetic disorders points towards an interdependence of skeletal and hematopoietic development, yet the impact of the marrow microenvironment in the pathogenesis of the bone marrow failure in Fanconi anemia remains unclear. Here we demonstrated that mice with double knockout of both Fancc and Fancg genes had decreased bone formation at least partially due to impaired osteoblast differentiation from mesenchymal stem/progenitor cells. Mesenchymal stem/progenitor cells from the double knockout mice showed impaired hematopoietic supportive activity. Mesenchymal stem/progenitor cells of patients with Fanconi anemia exhibited similar cellular deficits, including increased senescence, reduced proliferation, impaired osteoblast differentiation and defective hematopoietic stem/progenitor cell supportive activity. Collectively, these studies provide unique insights into the physiological significance of mesenchymal stem/progenitor cells in supporting the marrow microenvironment, which is potentially of broad relevance in hematopoietic stem cell transplantation. Introduction The hematopoietic system is built upon the ordered self-renewal and differentiation of hematopoietic stem cells (HSC) within the bone marrow (BM). This process involves intrinsic and extrinsic cues including both cellular and humoral regulatory signals generated by the HSC microenvironment, termed as “niche”. The cellular composition of this “niche” is heterogeneous, including endothelial cells,1 osteoblasts,2 adipocytes, and mesenchymal stem/progenitor cells (MSPC), a common progenitor for many of the cell lineages comprising the HSC niche.3-5 For fate decisions, regulatory signals from the BM microenvironment are transmitted to HSC through intercellular interactions within the proximity of the endosteal surface, the perivascular space, soluble factors, and the extracellular matrix.6 These cellular and humoral regulatory signals dictate the fates of HSC, including self-renewal, proliferation, differentiation, and apoptosis.7 In addition, there is increasing evidence suggesting a role of the hematopoietic microenvironment in hematopoietic disorders, such as myeloproliferative neoplasms8,9 and myelodysplastic syndrome.10 Fanconi anemia (FA) is a complex inherited disorder caused by germline mutations in at least one of 16 genes including FANCA, -B, -C, -D1, -D2, -E, -F, -G, -I, haematologica | 2017; 102(6)

Correspondence: fxy37@med.miami.edu or dclapp@iupui.edu

Received: October 22, 2016. Accepted: March 20, 2017. Pre-published: March 24, 2017. doi:10.3324/haematol.2016.158717 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1017 ©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.

1017


Y. Zhou et al.

J, -L, -M, -N, -O, –P, and –Q.11-19 Clinically, FA is a chromosomal fragility disorder characterized by progressive BM failure (BMF), variable developmental anomalies, and a strong propensity to develop cancer. The risk of developing BMF by 40 years of age is as high as 90%, and the cumulative incidence of hematologic and non-hematologic malignancies was reported to be as high as 33% and 28%, respectively.20 In the natural course of the disease, FA patients develop progressive pancytopenia, indicating that the defect occurs at the level of HSC.21-23 FA patients have a high incidence of inherited skeletal malformations and osteoporosis,20,24 suggesting a role of FA proteins in osteogenesis and bone maintenance. Despite these clinical observations of multiple mesenchymal defects in FA and our increasing awareness of the interdependence between the BM niche and hematopoiesis, relatively little attention has been directed to investigating the putative association between abnormal HSC function and the BM niche in FA. MSPC are a major component of the hematopoietic niche and have been shown to serve a critical function as hematopoiesis-supporting stromal cells.25 Here, we report that defective MSPC are pivotal mediators in the pathogenesis of hematopoietic defects in Fancc-/-;Fancg-/- double knockout (DKO) mice. Our studies provide detailed cellular and molecular evidence implicating mesenchymal cells as contributory to the BMF in FA, indicating the potential utility of MSPC/HSC co-transplantation, which may improve treatment of BMF in FA.

Methods Animals and reagents The Fancc and Fancg double heterozygous mice used in this study have been described previously.26-28 These mice were backcrossed into a C57BL/6J strain and were then bred to produce Fancc-/-;Fancg-/- (DKO) and wild-type (WT) mice. Age- and gendermatched DKO and WT mice were used for all experiments. All protocols were approved by the Institutional Animal Care and Use Committee at Indiana University School of Medicine. Chemicals were obtained from Sigma (St. Louis, MO, USA) unless otherwise indicated.

Isolation and expansion of mesenchymal stem/progenitor cells MSPC from mice were generated as previously described.29 Briefly, BM mononuclear cells (BMMNC) were separated by lowdensity gradient centrifugation from 6- to 8-week-old, age- and gender-matched WT and DKO mice, then cultured in complete mouse MesenCult medium (Stem Cell Technologies Inc, Vancouver, Canada) at 37°C in 5% CO2. MSPC between passage five to ten were used for the following experiments. The phenotypic analyses of MSPC were performed by evaluating the expression of surface markers including CD44, CD105, CD146, CD29 on a FACS Calibur flow cytometer as previously described.30 For human MSPC isolation, whole BM cells from FA patients and healthy donors were cultured in Dulbecco modified Eagle medium (DMEM)/F12 (Gibco, Carlsbad, USA), containing 10% fetal bovine serum (Hyclone, South Logan, USA), 1x Insulin transferrin selenium-A (Life Technologies, Carlsbad, USA), 10 ng/mL human epidermal growth factor (Peprotech, Rocky Hill, NJ, USA), and 10 ng/mL human platelet-derived growth factor-BB (Peprotech) at 37°C in 5% CO2 and 5% O2 in a fully humidified atmosphere. MSPC at passage three to five were used for the following experiments. 1018

Micro-computed tomography To evaluate trabecular microarchitecture in the distal femoral metaphysis, fixed femora were scanned using a high-resolution desktop micro-computed tomography imaging system (μCT-20; Scanco Medical AG, Basserdorf, Switzerland). The region of interest was defined as 15% of the total femur length measured from the tip of the femoral condyle and extending proximally for 200 slices with an increment of 9 µm, and was subsequently reconstructed, filtered (σ= 0.8 and support = 1.0), and thresholded (at 22% of the possible gray scale value) for analysis, as described elsewhere.31 Trabecular bone was contoured manually within the trabecular compartment, excluding the cortical shell. The parameter of micro-architecture for bone volume fraction (BV/TV, %) was measured.

Histomorphometric measurements Upon sacrifice, the isolated bones were fixed in 10% neutral buffered formalin for 48 h, dehydrated in graded ethanol, and embedded undecalcified in methyl methacrylate. Sagittal sections (5 µm thick) were cut from the middle of the femur. Tartrate-resistant acid phosphatase (TRAP) staining was performed using a leukocyte acid phosphatase kit (Sigma Diagnostics, St. Louis, MO, USA) and McNeal staining was performed using the McNeal tetrachromat kit (Polysciences, Warrington, PA, USA), both according to the manufacturers’ protocols. One section per femur was viewed at 100x magnification on a Leitz DMRXE microscope (Leica Mikroskopie und System GmbH, Wetzlar, Germany). Images were captured using a QImaging camera and QCapturePro software (Fryer Company Inc., Cincinnati, OH, USA). The region of interest for the metaphysis was defined by a rectangular area, which begins 0.5 mm proximal to the midpoint of the growth plate, non-inclusive of cortical bone, and extends proximally for a total area of approximately 2.8 mm2.

Bone remodeling measurement Fluorochrome labeling of the bones was performed by intraperitoneal injections of calcein (20 mg/kg, 8 days before sacrifice) and alizarin (20 mg/kg, 4 days before sacrifice), as previously described.32 Trabecular bone turnover was assessed by measuring the extent of single-labeled surface (sLS), doublelabeled surface (dLS) and the surface of the bone (BS) between the calcein and alizarin labels using Image Pro Plus version 4.1 software (Media Cybernetics, Silver Spring, MD, USA). Derived histomorphometric parameters included: (i) mineralizing surface (MS/BS), a measure of active bone-forming surface, calculated as (dLS+sLS/2)/BS; (ii) mineral apposition rate (MAR, μm/ day), a measure of the rate of radial expansion of new bone, calculated as Thickness/4 day; and (iii) bone formation rate, an overall measure of bone formation that combines MS/BS and MAR, calculated as MS/BS × MAR.

Other methods Methods for the clonogenic assay, bone mineral density quantification, annexin V/propidium iodide staining, senescence assay, thymidine incorporation assay, purification of HSPC, long-term culture of HSPC on MSPC monolayers, detection of reactive oxygen species (ROS), osteoblast and adipocyte differentiation of MSPC, reciprocal transplantation and co-transplantation of FA BMMNC and MSPC in NS2 mice are described in detail in the Online Supplementary Methods.

Statistics Survival curves were compared using the log-rank test. Differences between two groups with equal variances were assessed by two-tailed Student t-tests. Multiple comparisons were haematologica | 2017; 102(6)


Microenvironment and Fanconi anemia

A

B

C

D

E

Figure 1. Poor survival and hematopoietic defects in double knockout mice. (A) Survival curves of WT and DKO recipients following transplantation with either WT (n = 11) or DKO (n = 11) BM cells were monitored over a duration of 20 months. (P<0.01, log-rank test). (B) Total number of colony-forming unit-cells (CFU-C) in five sets of DKO as compared to WT recipient mice transplanted with either WT or DKO BM cells; data are presented as mean ± SEM, **P<0.01, ***P<0.001, two-way ANOVA followed by the Bonferroni test. (C) Representative photomicrographs demonstrated BM histology of recipient mice 15 months post-transplantation at low (10×, a-d) and high (40×, e-h) magnification. Scale bar: 100 μm for a-d, and 10 μm for e-h. (D) Representative H&E-stained sections of spleens from transplanted recipients are shown at low (10×, a-d) and high (40×, e-h) magnification. Scale bar: 100 μm for a-d, and 10 μm for e-h. (E) Representative May-Grünwald-Giemsa stained peripheral blood smears of transplanted recipient mice are shown. The peripheral blood smear of DKO recipient mice with WT or DKO BM cells showed dysplastic features including monocytes (f-h, green arrows) and bilobed neutrophils (b-d, red arrows) which were consistent with pseudo Pelger-Huët cells. Scale bar: 10 μm.

haematologica | 2017; 102(6)

1019


Y. Zhou et al.

B

A

E

D

F

C

G

H

conducted with one- or two-way analysis of variance (ANOVA) followed by an appropriate post-hoc correction. P values less than 0.05 were considered statistically significant. Data are presented as mean ± standard error of mean (SEM). Statistical analyses were performed with Prism 5.0 software (GraphPad Software Inc., La Jolla, CA, USA).

Study approval FA MSPC were generated from BM cells of five FA patients from the Institute of Hematology and Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College and Indiana University School of Medicine. Written informed consent was obtained from all patients. The study was approved by the Ethics Committee of the Institute of Hematology and Blood Diseases Hospital, Chinese Academy of Medical Sciences, and Indiana University School of Medicine according to guidelines of the 1975 Helsinki Declaration.

Results Hematopoietic defects in double knockout mice are associated with a dysfunctional bone marrow microenvironment We have previously reported that Fancc/g DKO mice develop progressive hematologic abnormalities, including 1020

I

Figure 2. Double knockout mice had retarded growth and impaired bone mineralization. (A) DKO mice had decreased whole body bone mineral density (BMD) (n=20 mice per genotype) as compared to WT mice. Data are presented as mean ± SEM, **P<0.01, two-tailed Student t-test. (B) DKO mice had decreased BMD as compared to WT littermates at varying ages. Data are presented as mean ± SEM, *P<0.05, **P<0.01, ***P<0.001, two-tailed Student t-test. (C) Micro-computed tomography demonstrated that DKO mice had reduced femoral trabecular bone volume as compared to WT mice (n=8 mice per genotype). Data are presented as mean ± SEM of bone volume per tissue volume (BV/TV), *P<0.05, two-tailed Student t-test. (D) Representative micro-computed tomography reconstructions of WT and DKO mouse femora. Scale bar: 1 mm. (E) Representative H&E (a-d) and McNeal staining (e-h) analysis of the femora of WT and DKO mice. Red arrows indicate the osteoblasts on the bone surface. Scale bar: 200 μm for a, c, e, g and 50 μm for b, d, f, h. (F) DKO mice had reduced numbers of osteoblasts on the trabecular bone surface (Ob.No./BS) as compared to WT mice (n=20 mice per genotype). Data are presented as mean ± SEM, *P<0.05, two-tailed Student ttest. (G) DKO mice had increased osteoclasts along the trabecular bone surface (Oc.S/BS) as compared to WT mice (n=8 mice per genotype). Data are presented as mean ± SEM, *P<0.05, twotailed Student t-test. (H, I) Bone remodeling studies demonstrated that DKO mice had reduced MS/BS and MAR as compared to WT mice (n=3 mice per genotype). Data are presented as mean ± SEM, *P<0.05, two-tailed Student ttest.

BMF, acute myeloid leukemia, and myelodysplastic syndrome, which closely mimic hematopoietic disorders observed in FA patients.27 To assess the contribution of the microenvironment to the hematopoietic phenotype of DKO mice, we performed reciprocal transplantation experiments whereby hematopoietic cells from WT and DKO mice were transplanted into lethally irradiated WT or DKO recipients. The survival rate was lowest in the cohort of DKO mice reconstituted with DKO BM cells among the four groups (Figure 1A). DKO BM transplanted into WT recipients also caused reduced survival as compared to WT BM transplanted into WT recipients, due to the occurrence of BM dysplasia and BMF. Intriguingly, DKO recipient mice transplanted with WT BM cells also exhibited impaired survival and BM dysplastic phenotypes indicated by hematologic analysis, suggesting a putative role for an impaired marrow niche in DKO mice. In addition, an expanded granulocyte-macrophage progenitor compartment was observed in DKO recipient mice transplanted with WT BM cells compared to WT recipients (Online Supplementary Figure S1A). Furthermore, we observed a 25% or 60% reduction in the total number of colony-forming unit-cells (CFU-C) in DKO as compared to WT recipient mice transplanted with WT or DKO BM cells, respectively (Figure 1B). Consistently, a significantly hypoplastic BM was observed in WT or DKO recipients haematologica | 2017; 102(6)


Microenvironment and Fanconi anemia

A

B

C

D

E

transplanted with either DKO or WT BM cells, respectively, with the most severe BM hypoplasia occurring in DKO recipients reconstituted with DKO BM cells (Figure 1C). The histology of the spleens of DKO recipients with either WT or DKO BM cells revealed disrupted architecture. Lymphoid aggregates in the white pulp (disrupted architecture) of DKO recipient spleens were also smaller than those of WT recipients (Figure 1D). May-GrünwaldGiemsa stained peripheral blood smears prepared from DKO recipients showed dysplastic features, including hyposegmented (bilobed) neutrophils with fine nuclear bridging consistent with pseudo-Pelger-Huët cells (Figure 1E, b-d, red arrows, Online Supplementary Figure S1B) and monocytes (Figure 1E, f-h, green arrows), whereas blasts were rare. In addition, dysplastic megakaryocytes (multinuclear megakaryocytes and hyposegmented megakaryohaematologica | 2017; 102(6)

Figure 3. In vitro analysis of mesenchymal stem progenitor cell frequency and osteoblast differentiation. (A) The frequency of CFU-F per 4×106 BMMNC from WT and DKO mice is shown (n=9 mice per genotype). Data are presented as mean ± SEM, ***P<0.001, two-tailed Student t-test. (B) Significant reduction of CFU-F in DKO mice at different cell passages (n=9 mice per genotype). Data are presented as mean ± SEM, ***P<0.001, two-tailed Student t-test. (C) The frequency of CFU-osteoblasts in WT and DKO mice is shown (n=5 mice per genotype). Alkaline phosphatase staining of WT and DKO BMMNC cultured in osteogenic medium. Data are presented as mean ± SEM, ***P<0.001, two-tailed Student t-test. (D) The ratio of Oil Red O-positive adipocytes to total colonies demonstrated enhanced adipocyte differentiation in DKO mice compared to WT ones (n=5 mice per genotype). Data are presented as mean ± SEM, **P<0.01, two-tailed Student t-test. Scale bar: 100 μm (E) Significantly reduced Runx2, Cdh2 and increased Ppar-γ gene expression in DKO MSPC as compared to WT controls (n=5 mice per genotype). Data are presented as mean ± SEM, *P<0.05, two-tailed Student t-test.

cytes) were observed in the BM of DKO recipients transplanted with WT BM cells, but not in WT recipient mice (Online Supplementary Figure S1C). These data suggest that DKO recipient mice transplanted with WT or DKO BM cells develop a myelodysplastic syndrome-like disease33 and provide strong in vivo evidence that the niche plays a cooperative role in the pathogenesis of impaired marrow engraftment in the DKO FA mouse model.

Impaired skeletal development and bone mass deficits in double knockout mice Skeletal anomalies including short stature and osteopenia/osteoporosis are widespread among the FA population.34 We, therefore, sought to ascertain the impact of Fancc/g genetic inactivation on skeletal development. The body length and body weight (Online Supplementary Figure 1021


Y. Zhou et al.

S2A,B) of DKO mice were significantly reduced as compared to those of age- and sex-matched WT littermates. Whole body bone mineral density, determined by pDEXA, was also reduced in DKO mice as compared to WT controls (Figure 2A), with an even more substantial reduction in femoral bone mineral density of the DKO mice versus WT controls (Figure 2B). Consistent with the decreased bone mass in DKO mice determined by bone mineral density analysis, micro-computed tomography analysis of the animals at 6 months of age revealed decreased bone volume in the mid-shaft of DKO femora compared to WT controls (Figure 2C,D). Quantitative histomorphometric analysis also revealed significantly reduced bone volume in DKO femora versus WT ones (Online Supplementary Figure S2C). Alterations in skeletal homeostasis can occur secondary to imbalances between bone-forming osteoblast activity

A

B

C

D

E

1022

and osteoclast-mediated bone resorption. To further assess osteoblast and osteoclast development in vivo, quantitative histomorphometry was performed on histological sections from the distal femoral metaphysis stained with hematoxylin and eosin (H&E), McNeal, and the osteoclast enzyme TRAP (Figure 2E,F, Online Supplementary Figure S2D,E). Sections stained with H&E revealed a marked reduction in trabecular and cortical bone in DKO femora as compared to WT control femora (Figure 2E, a-d). Manually counting osteoblasts in the femoral trabecular bone on McNeal-stained sections revealed that the number of osteoblasts was singnificantly lower in DKO mice than in WT controls (Figure 2E, e-h, Figure 2F). In addition, a significantly increased osteoclast surface to bone surface ratio was observed in DKO mice than in WT controls, as assessed by scoring the TRAP-positive staining osteoclast surface in the femoral trabeculae normalized to the trabec-

F

Figure 4. Co-culture of double knockout mesenchymal stem/progenitor cells with hematopoietic stem/progenitor cells showed the decreased hematopoietic supportive activity. (A, B) DKO MSPC had reduced hematopoietic supportive activity as compared with WT MSPC (n=3 mice per genotype). Data are presented as mean ± SEM, *P<0.05, **P<0.01, ***P<0.001, two-way ANOVA followed by the Bonferroni test. (C, D) DKO MSPC supported HSC cultures contained higher percentages of the Gr1+/Mac1+ population and apoptotic CD45+ cells (n=3 mice per genotype). Data are presented as mean ± SEM, *P<0.05, **P<0.01, ***P<0.001, two-way ANOVA followed by the Bonferroni test. (E) MSPC-conditioned medium, adipocyte-conditioned medium, and the serum of DKO mice contained significantly increased concentrations of tumor necrosis-α (TNF-α) as compared with WT mice (n=3 mice per genotype). Data are presented as mean ± SEM, *P<0.05, **P<0.01, two-tailed Student ttest. (F) A marked reduction of the interleukin-6 (IL6) level was observed in DKO MSPC conditioned medium (n=3 mice per genotype). Data are presented as mean ± SEM, ***P<0.001, two-tailed Student ttest.

haematologica | 2017; 102(6)


Microenvironment and Fanconi anemia

ular bone surface (Figure 2G, Online Supplementary Figure S2D). To study dynamic changes in bone remodeling, WT and DKO mice were injected with fluorochrome markers to label the bone surface (Online Supplementary Figure S2E).35 A 23% reduction in the mineralizing surface (MS/BS, Figure 2H, Online Supplementary Table S1), a 17% reduction in mineral apposition rate (MAR, Figure 2I, Online Supplementary Table S1), and a 36% reduction in bone formation rate (BFR)/BS were observed in DKO mice as compared to age- and sex-matched WT controls (Online Supplementary Figure S2F, Online Supplementary Table S1). Collectively, these data suggest that abnormal osteoblastand osteoclast-mediated bone turnover in DKO mice leads to pathological bone remodeling.

Fancc/g genetic ablation alters mesenchymal stem/progenitor cell fates, favoring adipogenic versus osteoblastic differentiation As osteoblasts, the principal cells mediating bone formation were deficient in DKO mice (Figure 2F), we hypothesized that Fancc/g deficiency alters the proliferative and/or differentiative capacity of MSPC, which give rise to mature osteoblasts and their precursors. We, therefore, performed colony-forming unit-fibroblast (CFU-F) assays on BM cells of the mice to determine the frequency of MSPC in WT versus DKO mice in vivo. DKO BM exhibited a significant reduction in the number of CFU-F compared to the marrow of WT littermates (Figure 3A,B). Consistently, flow cytometric analysis showed that the frequency of CD45-CD146+Nestin+CD105+ MSPC was significantly decreased in the BM of DKO mice compared to WT controls (Online Supplementary Figure S3A). Phenotypically defined MSPC from BM of DKO and WT mice were used to conduct the following experiments (Online Supplementary Figure S3B), and a significant reduction of CD146 expression was observed in DKO MSPC compared to WT MSPC. A thymidine incorporation assay demonstrated that DKO MSPC had significantly less proliferative potential compared to WT MSPC (Online Supplementary Figure S3C). As one of the fundamental properties of MSPC is their capacity to differentiate into multiple lineages under specific culture conditions,36 we next determined whether Fancc/g deletion altered MSPC lineage commitment by performing osteoblast and adipogenic differentiation assays. Alkaline phosphatase activity is an indicator of successful differentiation of MSPC into osteoblasts.37 Compared to WT MSPC, DKO MSPC exhibited markedly reduced alkaline phosphatase staining following incubation in osteogenic differentiation medium, indicating impaired osteoblast differentiation (Figure 3C). In contrast, when MSPC were cultured in adipogenic medium for 14 days, a significantly increased Oil Red O-positive area was observed in DKO cultures compared to WT controls (Figure 3D), suggesting that DKO MSPC had an increased capacity of adipocyte differentiation. These results indicate that Fancc/g deficiency leads to impaired MSPC proliferation (as determined by CFU-F) and lineage skewing, favoring adipocyte commitment over osteoblast differentiation. To delineate the molecular basis of impaired cell fate determination in DKO MSPC, we used quantitative polymerase chain reaction analysis to examine the expression of critical genes governing lineage commitment of MSPC, including osteoblasts, and adipocytes in WT versus DKO haematologica | 2017; 102(6)

cells. Our results indicated that the expression of genes controlling osteoblast differentiation, such as Runx2 and N-cadherin (Cdh2), were significantly reduced in DKO MSPC compared to WT controls (Figure 3E). In contrast, a marked increase in the expression of the adipogenic transcription factor, Ppar-Îł, was observed in DKO MSPC compared to WT MSPC (Figure 3E). These data indicate that Fancc/g deletion alters gene expression programs governing lineage commitment of MSPC, leading to deregulated adipocyte and osteoblast lineage commitment. Osteoclasts are specialized cells derived from the monocyte/macrophage hematopoietic lineage which adhere to the bone surface, secreting acid and lytic enzymes that degrade the bone matrix. To determine whether genetic ablation of Fancc/g alters osteoclast development, we established osteoclast cultures from BMMNC in the presence of the osteoclast differentiating cytokines M-CSF and RANK-L. DKO BMMNC exhibited a significantly increased propensity to osteoclast differentiation compared to WT BMMNC, as quantified by TRAP staining (Online Supplementary Figure S3D). Collectively, these data indicate that functional imbalances between osteoblast and osteoclast differentiation in the context of Fancc/g deficiency might cooperate to alter bone remodeling in vivo, contributing to short stature and osteoporosis in DKO mice as shown in Figure 2.

Double knockout mesenchymal stem/progenitor cells exhibit defective hematopoietic supportive activity in vitro MSPC and their progeny, such as osteoblasts and adipocytes, are widely recognized to play a critical role in supporting hematopoietic cells within the BM niche.38,39 Given that DKO MSPC exhibit impaired expansion and differentiation, we sought to explore further the role of Fancc/g in maintaining MSPC hematopoietic supportive activity. We began by performing cobblestone area-forming cell (CAFC) assays to evaluate the hematopoietic supportive activity of DKO MSPC. When hematopoietic progenitors (LK cells) were co-cultured for 4 weeks on MSPC feeder layers, significantly reduced CAFC numbers were observed when using DKO MSPC compared to WT MSPC, suggesting impaired hematopoietic supportive activity by DKO MSPC (Figure 4A,B). To further assess whether Fancc/g deficiency alters the capacity of MSPC to maintain hematopoietic cell differentiation, the percentage of Gr1+/Mac1+ cells following 4 weeks of co-culture was determined by flow cytometry. As shown in Figure 4C, significantly increased percentages of Gr1+/Mac1+ cells were observed in DKO MSPC-supported WT and DKO LK hematopoietic cell cultures compared to WT MSPC, indicating an enhancement of myeloid differentiation. Consistently, a significantly increased percentage of Gr1+/Mac1+ cells was observed in the peripheral blood of DKO recipients transplanted with either WT or DKO BM cells, compared with the WT recipients (Online Supplementary Figure S4A). In addition, DKO MSPC-supported cultures contained increased percentages of apoptotic CD45+ cells (Figure 4D). Collectively, these findings indicate that Fancc/g-deleted MSPC increase myeloid cell differentiation in vitro compared to WT MSPC. Secretion of trophic and paracrine factors within the BM niche is regarded to be a central mechanism by which MSPC function to maintain hematopoiesis. We, therefore, hypothesized that deregulated secretion of paracrine fac1023


Y. Zhou et al.

tors might account for the impaired hematopoietic supportive activity of DKO MSPC. Tumor necrosis factoralpha (TNF-α) is an inflammatory cytokine, which has been shown to preferentially induce apoptosis in FA hematopoietic cells.40,41 An enzyme-linked immunosorbent assay showed a significantly increased level of TNFα in DKO MSPC supernatants (Figure 4E). In addition to MSPC, adipocytes are known to be another major source of TNF-α production.42,43 Consistent with these data, we observed a 4.5-fold increase in the concentration of TNFα in conditioned media collected from DKO adipocyte cultures compared to WT controls (Figure 4E). As further validation, significantly increased concentrations of TNFα were also observed in the serum of DKO mice as compared to that of WT mice (Figure 4E). By contrast, levels of the hematopoietic supportive cytokine interleukin-6 were found to be markedly reduced in DKO MSPC conditioned medium (Figure 4F). HSC can lose stem cell capacity and die after exposure to ROS; DKO MSPC exposed to H2O2

A

C

produced increased ROS compared to their WT counterpart (Online Supplementary Figure S4B). This enhanced ROS production has also been identified along with the senescence and adipocyte differentiation of MSPC,44,45 which is consistent with the characteristics of the DKO MSPC. Taken together, these data suggest that loss of Fancc/g in MSPC alters the production of multiple cytokines and ROS, which may serve to perpetuate dysfunctional hematopoiesis in DKO mice.

Human Fanconi anemia patient-derived and double knockout mesenchymal stem/progenitor cells exhibit similar cellular phenotypes To examine whether MSPC derived from FA patients exhibit similar phenotypes to those observed in DKO mice, MSPC were isolated from four patients with a clinical diagnosis of FA (Online Supplementary Table S2) and healthy volunteers by culturing BM cells in human MesenCult medium and phenotypically validating the

B

D

E

F Figure 5. Impaired cellular functions of Fanconi anemia patient-derived mesenchymal stem/progenitor cells. (A) FA MSPC were more sensitive to mitomycin C (MMC). Data are presented as mean ± SEM and represent one of four independent experiments. (B) FA MSPC had an increased rate of senescence as compared to control MSPC. Data are shown as mean ± SEM from triplicate wells (5 fields/well) and represent one of four independent experiments. ***P<0.001, two-tailed Student t-test. (C) Representative images demonstrating alkaline phosphatase (ALP) staining of a healthy donor and FA MSPC cultured in the osteogenic medium surface. (D) FA MSPC had impaired osteoblast differentiation. Data are presented as mean ± SEM from triplicate wells (6 fields/well) and represent one of four independent experiments. ***P<0.001, two-tailed Student t-test. (E, F) FA MSPC had impaired hematopoietic supportive activity. Data are presented as mean ± SEM from triplicate wells and represent one of four independent experiments. *** P<0.001, two-tailed Student t-test. Each experiment was performed with a different MSPC culture isolated from the individual patient.

1024

haematologica | 2017; 102(6)


Microenvironment and Fanconi anemia

cells by flow cytometry (Online Supplementary Figure S5A). Sensitivity to mitomycin C was determined as previously described30 and was greater in FA MSPC than in healthy control MSPC (Figure 5A). Cellular senescence is a key pathophysiological phenomenon characterized by cell cycle arrest and upregulation of senescence-associated βgalactosidase activity. A 3-fold increase in the percentage of senescent cells was observed in FA MSPC compared to control MSPC (Figure 5B, Online Supplementary Figure S5B). Like DKO MSPC, human FA MSPC exhibited markedly defective osteoblast differentiation (Figure 5C,D) and increased adipocyte differentiation (Online Supplementary Figure S5C). Furthermore, osteoblast numbers were significantly reduced in BM biopsy sections from FA patients compared to those from healthy controls (Online Supplementary Figure S5D,E). To evaluate the hematopoietic supportive activity of these MSPC, MSPC from FA patients were co-cultured with cord blood CD34+ cells. After 5 weeks of co-culture, CAFC were counted, and CFU-C assays were performed. Significantly lower numbers of CAFC (Figure 5E) and CFU-C (Figure 5F) were observed in the co-cultures of FA MSPC with CD34+ cells than in those of healthy control MSPC with CD34+ cells. These data suggest that MSPC derived from FA patients exhibit impaired HSPC supportive activity, which is consistent with the findings in MSPC from DKO mice. To test whether MSPC derived from healthy donor BM, compared to FANCG-deficient MSPC, would enhance the engraftment of human FANCG-deficient BM cells in vivo, MSPC were injected intra-tibially into sub-lethally irradiated NOD.Cg-Prkdcscid IL2rgtm1Wjl/Sz (NS2) recipient mice. Twenty-four hours later, BMMNC from a human FANCGdeficient patient were delivered via tail vein injection. Four months following co-transplantation, human (h) CD45+ cell engraftment in the BM of recipient mice was analyzed by flow cytometry. Injection of healthy MSPC dramatically enhanced FANCG BMMNC engraftment (19% of hCD45+ cells, Online Supplementary Figure S5F, right panel), while the percentage of hCD45+ cells in the mice that received FANCG MSPC and FANCG BMMNC was only 0.9% (Online Supplementary Figure S5F, left panel).

Discussion MSPC act as an essential component of the BM hematopoietic microenvironment and have been proven to be involved in the pathogenesis of several hematologic malignancies.46,47 A recent translational study by Dong et al. found that Ptpn11-activating mutations in the BM MSPC and osteoprogenitors cause a juvenile myelomonocytic leukemia-like cancer in mice through profound, detrimental effects on HSC.48 Several previous studies indicated that MSPC from FA patients display reduced long-term proliferation ability and spontaneous chromosome breakages.49-51 In addition, we have previously reported that MSPC from the murine Fancg-/- model exhibit impaired proliferative capacity.30 Amarachintha et al. also reported that MSC from Fanca-/- or Fancd2-/- mice impaired WT HSPC self-renewal and induced myeloid expansion.52 Although another study showed that BM MSPC did not have impaired function and contribute to the pathogenesis of the disease in acquired BMF such as aplastic anemia,53 recent studies by Zambetti et al. demonstrated that meshaematologica | 2017; 102(6)

enchymal niche-derived inflammatory signaling induces oxidative and genotoxic stress in HSPC in ShwachmanDiamond syndrome, a rare inherited BMF syndrome.54 Whether defects of BM MSPC are involved in the pathophysiology of FA deserves further in vivo investigation. FA is caused by a mutation in genes encoding proteins required for the FA pathway. Although FA patients are clinically characterized by congenital mesenchymal anomalies, and a uniformly progressive and fatal BMF which begins in infancy or childhood,20,55-58 the impact of the loss of FA genes on other stem cell compartments and the role of the BM niche in the pathogenesis of FA-dependent BMF have received limited attention. To date, more than ten FA genes have been deleted or mutated in the mouse, but none of these mouse models with single FA gene deficiency spontaneously develops severe hematologic abnormalities like FA patients.59,60 We have previously reported that Fancc/g DKO mice spontaneously develop more aggressive hematopoietic deficits including BMF, acute myeloid leukemia, and myelodysplastic syndrome.27 Here, using the DKO murine model, we provided evidence that Fanc/g-deficient MSPC within the hematopoietic microenvironment cooperate to engender dysfunctional hematopoiesis. These results provide new insights regarding the fundamental mechanisms by which the BM “niche” contributes to the pathogenesis of FA-dependent BMF. Furthermore, we demonstrated that Fancc/g-deficient MSPC lead to impaired osteoblast differentiation with concomitant lineage skewing toward adipocyte commitment. Genetic ablation of Fancc/g also altered the production of critical inflammatory cytokines including interleukin-6 and TNF-α, the latter of which is known to induce HSPC apoptosis and contribute to BMF. Collectively, this study reveals an intimate relationship between FA HSPC and the BM niche in the pathogenesis of hematopoietic deficits. Our study provides strong evidence that Fancc/g deficiency results in multiple skeletal pathologies, including reduced body size and low bone mass phenotypes. These phenotypes in mice recapitulate the clinical features of short stature and osteoporosis common in FA patients.61 Bone remodeling is a dynamic process controlled by the coordinated activity of osteoblast-mediated bone formation and osteoclast-mediated bone resorption. We attribute the bone mass deficits in DKO mice in part to reduced osteoblast activity, as evidenced by the reduced osteoblast numbers and impaired bone remodeling in vivo. Like HSC, MSPC have the potential to differentiate into multiple lineages, including osteoblasts, adipocytes, and chondrocytes. The balance between osteogenesis and adipocyte formation is required for normal niche activity to maintain hematopoiesis.62-65 Our in vitro and in vivo studies indicate that Fancc/g deficiency impairs the differentiation of MSPC into osteoblasts while favoring adipocyte differentiation. Meanwhile, DKO mice exhibit reduced MSPC numbers and impaired self-renewal, suggesting an association between dysfunctional DKO MSPC and abnormal skeletal development/homeostasis in the DKO mouse model. DKO MSPC displayed dysregulated expression of multiple key genes controlling osteoblast versus adipocyte lineage commitment including Runx2, Cdh2, and Ppar-γ. Cdh2 has been identified as a negative regulator of adipogenesis.66 Therefore, skewed lineage commitment of DKO MSPC away from osteoblast differentiation and toward adipocyte commitment may be associated 1025


Y. Zhou et al.

with reduced Cdh2 expression. In addition, we observed a significant reduction of CD146 expression in DKO MSPC compared to WT MSPC. Since CD146 has been reported to be a marker for multilineage differentiation capacity,67 the fewer CD146+ cells in DKO MSPC might be associated with the defective MSPC functions. Consistent with murine data, MSPC obtained from BM biopsies of FA patients also revealed impaired proliferation and osteoblast differentiation capacity. MSPC-HSPC co-culture assays further revealed that MSPC from DKO mice and FA patients have defective hematopoietic supportive activity, as evidenced by reduced CAFC and CFU-C formation. Dysregulated skeletal remodeling and impaired osteoblast differentiation in DKO mice may thus contribute to the defective BM microenvironment, which is unable to sustain adequate HSPC numbers. These results are consistent with findings by Morad et al. who showed that the hematopoietic supportive activity of MSPC is influenced by lineage determination.68 It has been previously demonstrated that FA HSPC are hypersensitive to TNF-α induced apoptosis and increased levels of TNF-α and other inflammatory cytokines have been observed in FA patients.26,69,70 Here we observed reduced levels of interleukin-6 and significantly higher levels of inflammatory cytokines, including TNF-α, in the DKO model. Nagajyothi et al. previously reported that adipocytes are the major source of circulating inflammatory cytokines.71,72 It is also known that adipocytes negatively regulate hematopoietic activity and that TNF-α plays an important role in the pathogenesis of BMF.38,73,74 In addition, we observed that co-transplantation of healthy donor MSPC enhanced the engraftment and expansion of human BMMNC from an FANCG patient in

References 1. 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. 2. Zhang J, Niu C, Ye L, et al. Identification of the haematopoietic stem cell niche and control of the niche size. Nature. 2003;425(6960):836-841. 3. Sacchetti B, Funari A, Michienzi S, et al. Selfrenewing osteoprogenitors in bone marrow sinusoids can organize a hematopoietic microenvironment. Cell. 2007;131(2):324336. 4. Devine SM, Hoffman R. Role of mesenchymal stem cells in hematopoietic stem cell transplantation. Curr Opin Hematol. 2000;7 (6):358-363. 5. Dazzi F, Ramasamy R, Glennie S, Jones SP, Roberts I. The role of mesenchymal stem cells in haemopoiesis. Blood Rev. 2006;20(3):161-171. 6. Schepers K, Campbell TB, Passegue E. Normal and leukemic stem cell niches: insights and therapeutic opportunities. Cell Stem Cell. 2015;16(3):254-267. 7. Williams DA, Cancelas JA. Leukaemia: niche retreats for stem cells. Nature. 2006;444(7121):827-828. 8. Walkley CR, Olsen GH, Dworkin S, et al. A

1026

9.

10.

11. 12.

13.

14.

15.

vivo in NS2 mice while FA MSPC failed to support FA BM cell reconstitution. Nevertheless, studies in a larger cohort of NSG mice with the co-transplantation of MSPC and FA BM cells are warranted in the future. Collectively, these data suggest that the pathogenesis of hematopoietic deficits in FA is complex and likely depends on the interplay between abnormal hematopoietic cells and a dysfunctional BM niche. Our study indicates that FA MSPC have skewed differentiation capacity, altered cytokine secretion and defective hematopoietic supportive activity, providing scientific insight into the role of FA mutations in impairing the hematopoietic niche and their involvement in the pathogenesis of BMF. Clinically, a more detailed understanding of the disease might lead to rethink about the niche for the development of new therapies for FA. While MSPC are increasingly appreciated to be a critical part of the niche, due to the anatomical structure of the niche, future studies to dissect the role of other niche components, e.g., the endothelial niche or MSPC progenies, with specific deletion of FA genes using Cre/loxP technology are warranted. Acknowledgments The authors thank Heather Daniel for administrative support. Funding This work was supported in part by the Leukemia Lymphoma Society (LLS 6234-12 to FCY), NIH (R01CA155294-05 to DWC, F-C Y and HH), Ministry of Science and Technology of China (2016YFA0100600) and National Natural Science Foundation of China (81570113, 81270575, 81421002). SDR was supported in part by pre-doctoral training grants from the Indiana CTSI (NCCR 5TL1RR025759-03) and Children’s Tumor Foundation.

microenvironment-induced myeloproliferative syndrome caused by retinoic acid receptor gamma deficiency. Cell. 2007;129(6): 1097-1110. Kim YW, Koo BK, Jeong HW, et al. Defective Notch activation in microenvironment leads to myeloproliferative disease. Blood. 2008;112(12):4628-4638. Raaijmakers MH, Mukherjee S, Guo S, et al. Bone progenitor dysfunction induces myelodysplasia and secondary leukaemia. Nature. 2010;464(7290):852-857. Garaycoechea JI, Patel KJ. Why does the bone marrow fail in Fanconi anemia? Blood. 2014;123(1):26-34. Reid S, Renwick A, Seal S, et al. Biallelic BRCA2 mutations are associated with multiple malignancies in childhood including familial Wilms tumour. J Med Genet. 2005;42(2):147-151. Sims AE, Spiteri E, Sims RJ, 3rd, et al. FANCI is a second monoubiquitinated member of the Fanconi anemia pathway. Nat Struct Mol Biol. 2007;14(6):564-567. Smogorzewska A, Matsuoka S, Vinciguerra P, et al. Identification of the FANCI protein, a monoubiquitinated FANCD2 paralog required for DNA repair. Cell. 2007;129 (2):289-301. Dorsman JC, Levitus M, Rockx D, et al. Identification of the Fanconi anemia complementation group I gene, FANCI. Cell

Oncol. 2007;29(3):211-218. 16. Wu Y, Shin-ya K, Brosh RM Jr. FANCJ helicase defective in Fanconia anemia and breast cancer unwinds G-quadruplex DNA to defend genomic stability. Mol Cell Biol. 2008;28(12):4116-4128. 17. Vaz F, Hanenberg H, Schuster B, et al. Mutation of the RAD51C gene in a Fanconi anemia-like disorder. Nat Genet. 2010;42(5): 406-409. 18. Kim Y, Lach FP, Desetty R, et al. Mutations of the SLX4 gene in Fanconi anemia. Nat Genet. 2011;43(2):142-146. 19. Bogliolo M, Schuster B, Stoepker C, et al. Mutations in ERCC4, encoding the DNArepair endonuclease XPF, cause Fanconi anemia. Am J Hum Genet. 2013;92(5):800-806. 20. Kutler DI, Singh B, Satagopan J, et al. A 20year perspective on the International Fanconi Anemia Registry (IFAR). Blood. 2003;101(4):1249-1256. 21. Ceccaldi R, Parmar K, Mouly E, et al. Bone marrow failure in Fanconi anemia is triggered by an exacerbated p53/p21 DNA damage response that impairs hematopoietic stem and progenitor cells. Cell Stem Cell. 2012;11(1):36-49. 22. Geiselhart A, Lier A, Walter D, Milsom MD. Disrupted signaling through the Fanconi anemia pathway leads to dysfunctional hematopoietic stem cell biology: underlying mechanisms and potential therapeutic

haematologica | 2017; 102(6)


Microenvironment and Fanconi anemia

strategies. Anemia. 2012;2012:265790. 23. Kamimae-Lanning AN, Goloviznina NA, Kurre P. Fetal origins of hematopoietic failure in a murine model of Fanconi anemia. Blood. 2013;121(11):2008-2012. 24. Wajnrajch MP, Gertner JM, Huma Z, et al. Evaluation of growth and hormonal status in patients referred to the International Fanconi Anemia Registry. Pediatrics. 2001; 107(4):744-754. 25. Prockop DJ. Marrow stromal cells as stem cells for nonhematopoietic tissues. Science. 1997;276(5309):71-74. 26. Si Y, Ciccone S, Yang FC, et al. Continuous in vivo infusion of interferon-gamma (IFNgamma) enhances engraftment of syngeneic wild-type cells in Fanca-/- and Fancg-/- mice. Blood. 2006;108(13):4283-4287. 27. Pulliam-Leath AC, Ciccone SL, Nalepa G, et al. Genetic disruption of both Fancc and Fancg in mice recapitulates the hematopoietic manifestations of Fanconi anemia. Blood. 2010;116(16):2915-2920. 28. Yang Y, Kuang Y, Montes De Oca R, et al. Targeted disruption of the murine Fanconi anemia gene, Fancg/Xrcc9. Blood. 2001;98(12):3435-3440. 29. Wu X, Estwick SA, Chen S, et al. Neurofibromin plays a critical role in modulating osteoblast differentiation of mesenchymal stem/progenitor cells. Hum Mol Genet. 2006;15(19):2837-2845. 30. Li Y, Chen S, Yuan J, et al. Mesenchymal stem/progenitor cells promote the reconstitution of exogenous hematopoietic stem cells in Fancg-/- mice in vivo. Blood. 2009;113(10):2342-2351. 31. Munugalavadla V, Vemula S, Sims EC, et al. The p85alpha subunit of class IA phosphatidylinositol 3-kinase regulates the expression of multiple genes involved in osteoclast maturation and migration. Mol Cell Biol. 2008;28(23):7182-7198. 32. O'Brien CA, Plotkin LI, Galli C, et al. Control of bone mass and remodeling by PTH receptor signaling in osteocytes. PLoS One. 2008;3(8):e2942. 33. 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. 34. Giri N, Batista DL, Alter BP, Stratakis CA. Endocrine abnormalities in patients with Fanconi anemia. J Clin Endocrinol Metab. 2007;92(7):2624-2631. 35. Warden SJ, Robling AG, Sanders MS, Bliziotes MM, Turner CH. Inhibition of the serotonin (5-hydroxytryptamine) transporter reduces bone accrual during growth. Endocrinology. 2005;146(2):685-693. 36. Lee HS, Huang GT, Chiang H, et al. Multipotential mesenchymal stem cells from femoral bone marrow near the site of osteonecrosis. Stem Cells. 2003;21(2):190199. 37. Sudo H, Kodama HA, Amagai Y, Yamamoto S, Kasai S. In vitro differentiation and calcification in a new clonal osteogenic cell line derived from newborn mouse calvaria. J Cell Biol. 1983;96(1):191-198. 38. Naveiras O, Nardi V, Wenzel PL, et al. Bonemarrow adipocytes as negative regulators of the haematopoietic microenvironment. Nature. 2009;460(7252):259-263. 39. Calvi LM, Adams GB, Weibrecht KW, et al. Osteoblastic cells regulate the haematopoietic stem cell niche. Nature. 2003;425(6960): 841-846. 40. Koh PS, Hughes GC, Faulkner GR, Keeble

haematologica | 2017; 102(6)

41.

42. 43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55. 56.

WW, Bagby GC. The Fanconi anemia group C gene product modulates apoptotic responses to tumor necrosis factor-alpha and Fas ligand but does not suppress expression of receptors of the tumor necrosis factor receptor superfamily. Exp Hematol. 1999;27 (1):1-8. Dufour C, Corcione A, Svahn J, et al. TNFalpha and IFN-gamma are overexpressed in the bone marrow of Fanconi anemia patients and TNF-alpha suppresses erythropoiesis in vitro. Blood. 2003;102(6):2053-2059. Cawthorn WP, Sethi JK. TNF-alpha and adipocyte biology. FEBS Lett. 2008;582(1): 117-131. Hoareau L, Bencharif K, Rondeau P, et al. Signaling pathways involved in LPS induced TNFalpha production in human adipocytes. J Inflamm (Lond). 2010;7:1. Zhang DY, Pan Y, Zhang C, et al. Wnt/betacatenin signaling induces the aging of mesenchymal stem cells through promoting the ROS production. Mol Cell Biochem. 2013;374(1-2):13-20. Kanda Y, Hinata T, Kang SW, Watanabe Y. Reactive oxygen species mediate adipocyte differentiation in mesenchymal stem cells. Life Sci. 2011;89(7-8):250-258. Schepers K, Pietras EM, Reynaud D, et al. Myeloproliferative neoplasia remodels the endosteal bone marrow niche into a selfreinforcing leukemic niche. Cell Stem Cell. 2013;13(3):285-299. Menendez P, Catalina P, Rodriguez R, et al. Bone marrow mesenchymal stem cells from infants with MLL-AF4+ acute leukemia harbor and express the MLL-AF4 fusion gene. J Exp Med. 2009;206(13):3131-3141. Dong L, Yu WM, Zheng H, et al. Leukaemogenic effects of Ptpn11 activating mutations in the stem cell microenvironment. Nature. 2016;539(7628):304-308. Mantelli M, Avanzini MA, Rosti V, et al. Comprehensive characterization of mesenchymal stromal cells from patients with Fanconi anaemia. Br J Haematol. 2015;170 (6):826-836. Lecourt S, Vanneaux V, Leblanc T, et al. Bone marrow microenvironment in fanconi anemia: a prospective functional study in a cohort of fanconi anemia patients. Stem Cells Dev. 2010;19(2):203-208. Barroca V, Mouthon MA, Lewandowski D, et al. Impaired functionality and homing of Fancg-deficient hematopoietic stem cells. Hum Mol Genet. 2012;21(1):121-135. Amarachintha S, Sertorio M, Wilson A, Li X, Pang Q. Fanconi anemia mesenchymal stromal cells-derived glycerophospholipids skew hematopoietic stem cell differentiation through toll-Like receptor signaling. Stem Cells. 2015;33(11):3382-3396. Bueno C, Roldan M, Anguita E, et al. Bone marrow mesenchymal stem cells from patients with aplastic anemia maintain functional and immune properties and do not contribute to the pathogenesis of the disease. Haematologica. 2014;99(7):1168-1175. Zambetti NA, Ping Z, Chen S, et al. Mesenchymal inflammation drives genotoxic stress in hematopoietic stem cells and predicts disease evolution in human preleukemia. Cell Stem Cell. 2016;19(5):613627. Kook H. Fanconi anemia: current management. Hematology. 2005;10(Suppl 1):108110. Bagby GC Jr., Segal GM, Auerbach AD, et al. Constitutive and induced expression of hematopoietic growth factor genes by

57.

58. 59. 60.

61. 62.

63.

64. 65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

fibroblasts from children with Fanconi anemia. Exp Hematol. 1993;21(11):1419-1426. Auerbach AD, Buchwald M, Joenje H. Fanconi anemia. In: Vogelstein B, Kinzler KW, eds. The Genetic Bases of Human Cancer. 2002; 2nd Edition (New York: McGraw-Hill, Inc.):289-306. Mathew CG. Fanconi anaemia genes and susceptibility to cancer. Oncogene. 2006;25 (43):5875-5884. Parmar K, D'Andrea A, Niedernhofer LJ. Mouse models of Fanconi anemia. Mutat Res. 2009;668(1-2):133-140. Bakker ST, de Winter JP, te Riele H. Learning from a paradox: recent insights into Fanconi anaemia through studying mouse models. Dis Model Mech. 2013;6(1):40-47. Glanz A, Fraser FC. Spectrum of anomalies in Fanconi anaemia. J Med Genet. 1982;19(6):412-416. Omatsu Y, Sugiyama T, Kohara H, et al. The essential functions of adipo-osteogenic progenitors as the hematopoietic stem and progenitor cell niche. Immunity. 2010;33(3): 387-399. Mendez-Ferrer S, Michurina TV, Ferraro F, et al. Mesenchymal and haematopoietic stem cells form a unique bone marrow niche. Nature. 2010;466(7308):829-834. Bianco P. Bone and the hematopoietic niche: a tale of two stem cells. Blood. 2011;117 (20):5281-5288. Visnjic D, Kalajzic Z, Rowe DW, et al. Hematopoiesis is severely altered in mice with an induced osteoblast deficiency. Blood. 2004;103(9):3258-3264. van Oostrom AJ, van Wijk JP, Sijmonsma TP, Rabelink TJ, Castro Cabezas M. Increased expression of activation markers on monocytes and neutrophils in type 2 diabetes. Neth J Med. 2004;62(9):320-325. Xu J, Wang W, Kapila Y, Lotz J, Kapila S. Multiple differentiation capacity of STRO1+/CD146+ PDL mesenchymal progenitor cells. Stem Cells Dev. 2009;18(3):487-496. Morad V, Pevsner-Fischer M, Barnees S, et al. The myelopoietic supportive capacity of mesenchymal stromal cells is uncoupled from multipotency and is influenced by lineage determination and interference with glycosylation. Stem Cells. 2008;26(9):22752286. Li J, Sejas DP, Zhang X, et al. TNF-alpha induces leukemic clonal evolution ex vivo in Fanconi anemia group C murine stem cells. J Clin Invest. 2007;117(11):3283-3295. Pang Q, Keeble W, Christianson TA, Faulkner GR, Bagby GC. FANCC interacts with Hsp70 to protect hematopoietic cells from IFNgamma/TNF-alpha-mediated cytotoxicity. EMBO J. 2001;20(16):4478-4489. Nagajyothi F, Desruisseaux MS, Thiruvur N, et al. Trypanosoma cruzi infection of cultured adipocytes results in an inflammatory phenotype. Obesity (Silver Spring). 2008;16(9):1992-1997. Bastard JP, Maachi M, Lagathu C, et al. Recent advances in the relationship between obesity, inflammation, and insulin resistance. Eur Cytokine Netw. 2006;17(1):4-12. Touw I, Lowenberg B. No stimulative effect of adipocytes on hematopoiesis in long-term human bone marrow cultures. Blood. 1983;61(4):770-774. Yokota T, Oritani K, Takahashi I, et al. Adiponectin, a new member of the family of soluble defense collagens, negatively regulates the growth of myelomonocytic progenitors and the functions of macrophages. Blood. 2000;96(5):1723-1732.

1027


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Myelodysplastic Syndromes

Ferrata Storti Foundation

Molecular features of early onset adult myelodysplastic syndrome

Cassandra M. Hirsch,1,2 Bartlomiej P. Przychodzen,1 Tomas Radivoyevitch,3 Bhumika Patel,4 Swapna Thota,4 Michael J. Clemente,1 Yasunobu Nagata,1 Thomas LaFramboise,2 Hetty E. Carraway,4 Aziz Nazha,4 Mikkael A. Sekeres,4 Hideki Makishima1,5 and Jaroslaw P. Maciejewski1,4

Department of Translational Hematology and Oncology Research, Taussig Cancer Institute, Cleveland Clinic, OH, USA; 2Department of Genetics and Genome Science, Case Western Reserve University, Cleveland, OH, USA; 3Department of Quantitative Health Sciences, Cleveland Clinic Lerner Research Institute, OH, USA; 4Leukemia Program, Taussig Cancer Institute, Cleveland Clinic, OH, USA and 5Department of Pathology and Tumor Biology, Graduate School of Medicine, Kyoto University, Japan 1

Haematologica 2017 Volume 102(6):1028-1034

ABSTRACT

M

Correspondence: maciejj@ccf.org

Received: November 11, 2016. Accepted: February 28, 2017. Pre-published: March 2, 2017. doi:10.3324/haematol.2016.159772 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1028 ©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.

1028

yelodysplastic syndromes are typically diseases of older adults. Patients in whom the onset is early may have distinct molecular and clinical features or reflect a demographic continuum. The identification of differences between “early onset” patients and those diagnosed at a traditional age has the potential to advance understanding of the pathogenesis of myelodysplasia and may lead to formation of distinct morphological subcategories. We studied a cohort of 634 patients with various subcategories of myelodysplastic syndrome and secondary acute myeloid leukemia, stratifying them based on age at presentation and clinical parameters. We then characterized molecular abnormalities detected by next-generation deep sequencing of 60 genes that are commonly mutated in myeloid malignancies. The number of mutations increased linearly with age and on average, patients >50 years of age had more mutations. TET2, SRSF2, and DNMT3A were more commonly mutated in patients >50 years old compared to patients ≤50 years old. In general, patients >50 years of age also had more mutations in spliceosomal, epigenetic modifier, and RAS gene families. Although there are age-related differences in molecular features among patients with myelodysplasia, most notably in the incidence of SRSF2 mutations, our results suggest that patients ≤50 years old belong to a disease continuum with a distinct pattern of early onset ancestral events. Introduction With the exception of treatment-related myelodysplastic syndromes (MDS), de novo MDS and the closely related secondary acute myeloid leukemia evolving from an antecedent MDS are diseases of the elderly. The median age at presentation is 71 years1,2 and the yearly incidence rate increases from 2.5/105 in the sixth decade to 30/105 in the eighth decade of life to as high as 50/105 (females) to 100/105 (males) in patients >80 years of age.1 These numbers may be underestimated, as MDS is likely underdiagnosed in the elderly.3 In children, refractory cytopenia of childhood and juvenile myelomonocytic leukemia are considered distinct entities more related to congenital bone marrow failure and familial leukemia syndromes than to adult MDS.4 Adult patients less than 50 years of age are sporadically affected by MDS, reflecting either an extreme polarity of age distribution of the disease, or, perhaps, similar to pediatric forms, constituting a separate pathological process. To determine whether MDS in younger adults should be considered a distinct disease subentity, we compared a cohort of what could be considered “early onset” MDS to a cohort of those diagnosed with MDS at a traditional age by contrasting specific clinical characteristics and molecular features. haematologica | 2017; 102(6)


Molecular features of early onset, adult MDS

Methods Patients’ samples After obtaining informed consent according to protocols approved by the Cleveland Clinic Institution Review Board, marrow and blood samples were collected from patients (2003-2016) classified according to World Health Organization 2008 criteria,5 and whole exome sequencing (n=95), and/or multi-amplicon deep sequencing (n=539) was performed. Tumor DNA was extracted from the patients’ marrow and wherever possible germline DNA was obtained from selected CD3+ T cells. Cases of clonal hematopoiesis of indeterminate potential (CHIP) were excluded from this study; all patients had to have dysplasia, increased blasts, or abnormal cytogenetics in order to be diagnosed with MDS.6

Next-generation sequencing Whole exome capture was performed according to the manufacturer’s protocol [SureSelect®, ver. 4 (Agilent Technology)].7,8 The captured targets were subjected to massive parallel sequencing using the HiSeq 2000. Multi-amplicon-based, targeted, deep sequencing was performed for a panel of 64 genes, most commonly somatically mutated in MDS (Online Supplementary Table S1)8,9 and those known to be affected by germline mutations (Online Supplementary Table S2).10 Customized probe sets amplified exons of target genes. Sequencing libraries were generated according to an Illumina paired-end library protocol and subjected to deep sequencing on MiSeq (Illumina) sequencers according to the standard protocol. A bio-analytic pipeline, devised in-house, as previously described,11 was applied to identify somatic mutations and (where appropriate) germline variants by comparison with germline controls, mutational databases (Entrez Gene12 the Ensembl Genome Browser,13 COSMIC14), and sequenced controls, Exome Aggregation Consortium (ExAC).15 Variant allelic frequencies of mutations were adjusted according to the zygosity and copy number based on single nucleotide polymorphism results. Serial samples and variant allelic frequencies were compared to determine clonal hierarchy, i.e. dominant and secondary mutations.

Statistical analysis Wilcoxon tests were performed for pairwise comparisons between continuous variables and the Fisher exact test was applied for categorical variables. Poisson regression was used to find the linear slope of mutations rates versus age. All P values were two-sided and values less than 0.05 were considered to be statistically significant. Analyses were performed using the R statistical program.

Results Demographic features of adults with “early onset myelodysplastic syndrome” We analyzed 634 patients with primary MDS (excluding treatment-related or secondary MDS) with a median presentation age of 68 years (range 20-94) (Online Supplementary Table S3). When patients were age-ranked in 5-year increments (Figure 1A), a unimodal age distribution was obtained, allowing the cohort to be empirically split into two groups: ≤50 years (n=65, with “early onset adult MDS”), and >50 years (n=569, with “traditional age of diagnosis MDS”), corresponding to 10% and 90% of patients, respectively. The split at the age of 50 was further justified by the separate clusters that were present haematologica | 2017; 102(6)

when survival by age group was analyzed (Online Supplementary Figure S1A). Additionally, 15% of patients were over 79 years old and analyzed in contrast to the other two groups as “late onset MDS.” Accordingly, the median age of the early onset MDS group was 44 years (range, 20-50), that of the group between 50 and 80 years was 70 years (range, 51-79), and the median age of the late onset group was 83 years (range, 80-94). Women were over-represented among early onset patients, whereas the converse was true among MDS patients diagnosed at a traditional age (60% versus 32%, P<0.0001) (Figure 1A). There was no significant difference between the numbers of females with 5q deletion in these groups. The Surveillance, Epidemiology and End Results (SEER) age-atdiagnosis distributions were comparable to those of our cohort18 (Figure 1B).

Clinical features of “early onset myelodysplastic syndrome” Higher-risk MDS (refractory anemia with excess blasts1/2) was more common among early onset MDS patients (35% versus 24%, P=0.048) while lower-risk MDS (refractory cytopenia with multilineage dysplasia, refractory anemia with ringed sideroblasts, refractory cytopenia with unilineage dysplasia, refractory anemia) predominated in MDS patients >50 years of age (28% versus 41%, P=0.042). Additionally, 17% of patients ≤50 years old had MDS/myeloproliferative neoplasms (including chronic myelomonocytic leukemia), and 20% had secondary acute myeloid leukemia compared to 21% and 14%, respectively, in patients >50 years of age (Online Supplementary Table S4). When grouped according to cytogenetic risk categories, survival differences were seen between patients ≤50 and >50 years old (Online Supplementary Figure 1B). There was no difference between patients ≤50 and >50 years old with regards to a family history of cancer, including leukemia, or bone marrow failure (51% versus 54%, P=0.51; Online Supplementary Table S4). Family history of cancer was defined as a selfreported presence of one or more first-degree relative (parent, sibling, or child) with cancer, hematologic neoplasia, or bone marrow failure. When family history was restricted to only hematologic neoplasia or bone marrow failure, still no difference was seen between patients ≤50 and >50 years old.

Patterns of molecular lesions When patients were screened for somatic mutations, more mutations were detected in patients >50 years old by whole exome and targeted sequencing overall (P=0.05 and P=0.02, respectively) (Figure 1C), as well as within molecular subtypes (Figure 1D). The average number of mutations per case increased in a linear fashion with age (R2=0.947, Figure 1E). Considering the subset of patients analyzed by whole exome sequencing, no difference in the pattern of transversions or transitions was identified in the mutational signatures of patients ≤50 and >50 years old (Online Supplementary Figure S2A). Focusing on specific lesions, ASXL1 (9%), TET2 (9%), TP53 (9%), and RUNX1 (8%) were the most frequently mutated genes in patients with early onset MDS (Figure 2A). However, SRSF2 and TET2 mutations were less common in patients ≤50 years old (2% versus 10%, P=0.013 and 9% versus 19%, P=0.060; respectively (Figure 2A). When categorized based on functional properties of 1029


C.M. Hirsch et al.

affected genes, patients >50 years old were found to have more spliceosomal gene mutations (P=0.025), epigenetic modifier mutations (P=0.007), and genes in the RAS family (P=0.08) (Figure 2A) when compared to patients ≤50 years old. Normal karyotype was present in about onehalf of all patients but no major differences were found in distribution of individual lesions, including complex karyotype (14% versus 8%, P=0.13) (Figure 2B). In the linear fits of the average number of mutations versus age using Poisson regression, TET2 and SRSF2 mutation rates increased with age (P=0.001 and P=0.035, respectively (Figure 2C). This trend was recapitulated by applying ageadjusted frequencies of TET2 mutants to SEER MDS demographics (Figure 2D), and parallels the trend seen in healthy controls (Figure 2E). Previously, it was suggested that hydroxymethylation may prevent C->T transitions

by decreasing MeCpG levels. Thus, more C->T transitions via MeCpG deamination would be expected in cases with founder TET2 mutations.17,21 However, when molecular signatures of cases with and without TET2 mutants were analyzed, age-related C->T transitions were similar (Online Supplementary Figure S2B). When compared to all other patients, TET2 mutant cases tended to display a higher number of additional mutations (P<0.001; targeted, P=0.074 whole exome) (Online Supplementary Figure S2C). However, among patients without a TET2 mutation, there was no difference in the number of mutations between patients ≤50 and >50 years old (Online Supplementary Figure S2D). Analysis of clonal architecture revealed that RUNX1, SF3B1, and TP53 were the most common dominant mutations in patients with early onset MDS. In contrast, TET2, SF3B1 and STAG2 were the most common

A

B

C

D

E

Figure 1. Clinical features and demographics of patients with “early onset” myelodysplastic syndrome. (A) Distribution of Cleveland Clinic MDS patients by age and sex-related trends. (B) Distribution of age and sex in MDS SEER patients. (C) Mutational frequencies in patients ≤50, 51-79, and ≥80 years old detected by targeted deep sequencing and whole exome sequencing. (D) Average number of mutations in each MDS subtype in MDS patients ≤50 and >50 years old. (E) Age-related increase in number of mutations.

1030

haematologica | 2017; 102(6)


Molecular features of early onset, adult MDS

dominant mutations in patients >50 years of age. Overall, there was no difference in the distribution of variant allele frequencies between the groups of patients (Figure 3A). Mutations were similarly distributed across functional gene families in MDS patients ≤50 and >50 years old (Figure 3B, C).

Known familial mutations In an attempt to explain the early occurrence of MDS, the cohort of patients with early onset disease was analyzed for the presence of congenital mutations known to be associated with MDS, with a focus on mutations in genes frequently associated with a familial leukemia or

C

A

D

E

B

Figure 2. Typical somatic defects in myelodysplastic syndrome patients ≤50 and >50 years old. (A) Frequencies of somatic mutations in the most recurrently mutated genes in patients ≤50 and >50 years of age. (B) Frequencies of cytogenetic abnormalities in patients ≤ 50 and >50 years of age. (C) Percentages of somatic mutations according to age in selected genes. P-values correspond to the linear slope of mutation rates vs. age found using Poisson regression. Only the genes found to be significant are shown here. Epigenetic modifier genes include DNMT3A, EZH2, KDM6A, IDH1/2, and TET2. Spliceosomal gene mutations include SRSF2, SF3B1, LUC7L2, U2AF1, ZRSRS, PRPF8, and DDX41. (D) TET2 mutational frequency per age group in MDS as predicted by SEER data and actual TET2 mutation frequency in the Cleveland Clinic cohort. (E) Frequency of TET2 mutations in healthy individuals and MDS patients by age.

haematologica | 2017; 102(6)

1031


C.M. Hirsch et al.

marrow failure as previously described10 (the panel of genes tested is shown in Online Supplementary Table S2). Familial mutations were found in 12% of patients with early onset MDS, but a higher incidence of telomerase complex or Fanconi anemia gene variants was not found (versus 6% P=0.04) (Online Supplementary Table S2).

Discussion Investigations of disease demographics may reveal clues to pathogenic mechanisms, provide insight to correct diagnosis, and help identify disease variants or even new nosological entities. For instance, aplastic anemia shows a

A

B

C

Figure 3. Molecular characteristics of patients with myelodysplastic syndrome. (A) Comparison of variant allele frequencies across patients ≤50 vs. >50 years old. (B) Distribution of dominant mutations in patients ≤50 vs. >50 years old. (C) Distribution of all mutations in patients ≤50 vs. >50 years old.

1032

haematologica | 2017; 102(6)


Molecular features of early onset, adult MDS

bimodal age distribution, with its first peak in children and young adults representing typical idiopathic disease, and the peak later in adults corresponding to a possible admix of patients with MDS, in particular hypocellular MDS.4 In contrast, de novo MDS tends to be a disease of older adults. Our cohort demonstrated a unimodal age distribution similar to that of SEER data.18 Patients with early onset MDS identified in this cohort could constitute an extreme outlier group, roughly 10% of the age continuum, or they could represent a separate disease sub-entity that pathogenically belongs to a different form of MDS such as childhood MDS or possibly juvenile myelomonocytic leukemia. With these hypotheses in mind, we analyzed the molecular profiles of MDS patients ≤50 and >50 years old, incorporating chromosomal abnormalities and somatic mutations identified through either exome or targeted sequencing approaches with a rationally selected panel of the most common mutations associations with myeloid neoplasms.9 With age, the number of somatic events increased (as detected by both exome and targeted sequencing) while the percentage of patients with normal cytogenetics remained constant.17,19,20 This observation suggests age-related accumulation of mutational events and a higher molecular complexity the later a patient is diagnosed with MDS. Analysis of mutations identified by exome sequencing in patients ≤50 and >50 years old did not reveal a higher rate of C to T transitions in older patients as previously described.21 Furthermore, the clonal burden did not differ between patients ≤50 and >50 years old, suggesting that MDS is fully clonal at presentation. Comparison of chromosomal and mutational patterns revealed several discrete differences suggesting that atypical patients with early onset MDS likely constitute an extreme group of a continuum rather than a separate entity. TET2 and SRSF2 were found to be more commonly mutated in the group of MDS patients >50 years old. Supporting this finding, control cohorts document an increase in the incidence of asymptomatic TET2 mutations with age, raising the possibility that such mutations represent pre-leukemic founder lesions with a long latency period before disease manifestation since the presence of sub-clonal events was associated with a subsequent risk of malignancies.6 Indeed, TET2, RUNX1, and TP53 mutations were as frequent as in the Cancer Genomic Atlas data for a younger cohort of patients with acute myeloid leukemia.22 In a recent study of elderly patients with acute

References 5. 1. Ma X, Does M, Raza A, Mayne ST. Myelodysplastic syndromes: incidence and survival in the United States. Cancer. 2007;109(8):1536-1542. 2. Sekeres MA, Schoonen WM, Kantarjian H, et al. Characteristics of US patients with myelodysplastic syndromes: results of six cross-sectional physician surveys. J Natl Cancer Inst. 2008;100(21):1542-1551. 3. Burgstaller S, Wiesinger P, Stauder R. Myelodysplastic syndromes in the elderly: treatment options and personalized management. Drugs Aging. 2015;32(11): 891905. 4. Niemeyer CM, Baumann I. Classification of childhood aplastic anemia and myelodys-

haematologica | 2017; 102(6)

6.

7.

8.

myeloid leukemia, genes in the spliceosomal complex as well as TET2 were found to be altered more frequently than in younger cases, similar to what is seen in MDS.23 While there are marked differences in the mutational spectrum between early onset MDS and MDS patients diagnosed at a traditional age, overall our results suggest that patients with MDS ≤50 years old constitute part of a continuum rather than a specific group with a distinct molecular pathogenesis. Increased frequency of TET2 mutations parallels the trend seen in healthy controls17,20 with increasing frequency of TET2 mutations with aging suggesting a disease-initiating role of these mutations in MDS that is consistent with increased MDS risk in asymptomatic mutant carriers. Previously, somatic mutations of SRSF2 and other spliceosomal genes were found exclusively in patients >70 years old and, therefore, associated with age-related clonal hematopoiesis.17,19,20 In pediatric disease, including refractory cytopenia of childhood and juvenile myelomonocytic leukemia, we and others have also found a low rate of TET2 and spliceosomal mutations.4 Similarly, the incidence of observed TET2 mutations in this population followed the trend observed among healthy controls as previously reported.17,20 However, in contrast to juvenile myelomonocytic leukemia, RAS gene family mutations were not predominant in patients with early onset adult MDS.24 We expect that early manifestation of MDS will be associated with familial disease, disease with a strong family history.10,25,26 Furthermore, RUNX1 was the most common dominant mutation in patients ≤50 years old, and two out of five cases were confirmed to be germline. As expected, the most common mutation in patients >50 years of age was TET2, well documented as a mutation associated with aging.17,19,20 Overall, our study suggests that the molecular underpinnings of early onset adult MDS, while distinct from juvenile forms of the disease, do not differ enough from MDS diagnosed at a traditional age to warrant a separate categorization. Acknowledgments This work was supported by grants NIH/NHLBI, R01 HL118281, NIH/NHLBI, R01 HL123904, NIH/NHLBI, R01 HL128425, Edward P. Evans Foundation, 2015 Basic Research Grant, and American Cancer Society Research Scholar Grant 123436-RSG-12-159-01-DMC.

plastic syndrome. Hematol Am Soc Hematol Educ Program. 2011;2011:84-89. Swerdlow SH. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. International Agency for Research on Cancer, 2008. 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. Thota S, Viny AD, Makishima H, et al. Genetic alterations of the cohesin complex genes in myeloid malignancies. Blood. 2014;124(11):1790-1798. Yoshida K, Sanada M, Shiraishi Y, et al. Frequent pathway mutations of splicing machinery in myelodysplasia. Nature. 2011;478(7367):64-69.

9. Haferlach T, Nagata Y, Grossmann V, et al. Landscape of genetic lesions in 944 patients with myelodysplastic syndromes. Leukemia. 2014;28(2):241-247. 10. 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. 11. Makishima H, Yoshida K, Nguyen N, et al. Somatic SETBP1 mutations in myeloid malignancies. Nat Genet. 2013;45(8):942-946. 12. Maglott D, Ostell J, Pruitt KD, Tatusova T. Entrez Gene: gene-centered information at NCBI. Nucleic Acids Res. 2005;33 (Database issue):D54-58. 13. Yates A, Akanni W, Amode MR, et al. Ensembl 2016. Nucleic Acids Res. 2016;44 (D1):D710-716.

1033


C.M. Hirsch et al. 14. Forbes SA, Beare D, Gunasekaran P, et al. COSMIC: exploring the world's knowledge of somatic mutations in human cancer. Nucleic Acids Res. 2015;43(Database issue):D805-811. 15. Lek M, Karczewski K, Minikel E, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016;536(7616):285-291. 16. Makishima H, Visconte V, Sakaguchi H, et al. Mutations in the spliceosome machinery, a novel and ubiquitous pathway in leukemogenesis. Blood. 2012;119(14):32033210. 17. Jaiswal S, Fontanillas P, Flannick J, et al. Age-related clonal hematopoiesis associated with adverse outcomes. N Engl J Med. 2014;371(26):2488-2498. 18. Surveillance, Epidemiology, and End Results (SEER) Program Research Data. In:

1034

19.

20.

21.

22.

National Cancer Institute D, Surveillance Resarch Program, Surveillance Systems Branch ed, 1973-2012. Mason CC, Khorashad JS, Tantravahi SK, et al. Age-related mutations and chronic myelomonocytic leukemia. Leukemia. 2016;30(4):906-913. McKerrell T, Park N, Moreno T, et al. Leukemia-associated somatic mutations drive distinct patterns of age-related clonal hemopoiesis. Cell Rep. 2015;10(8):12391245. Alexandrov LB, Nik-Zainal S, Wedge DC, et al. Signatures of mutational processes in human cancer. Nature. 2013;500(7463): 415-421. 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.

23. Silva P, Neumann M, Vosberg S, et al. Acute myeloid leukemia in the elderly is characterized by a distinct genetic landscape. Blood. 2015;126(23):804. 24. Chang TY, Dvorak CC, Loh ML. Bedside to bench in juvenile myelomonocytic leukemia: insights into leukemogenesis from a rare pediatric leukemia. Blood. 2014;124(16):2487-2497. 25. Hamadou WS, Bourdon V, Gaildrat P, et al. Mutational analysis of JAK2, CBL, RUNX1, and NPM1 genes in familial aggregation of hematological malignancies. Ann Hematol. 2016;95(7):1043-1050. 26. Astuto LM, Bork JM, Weston MD, et al. CDH23 mutation and phenotype heterogeneity: a profile of 107 diverse families with Usher syndrome and nonsyndromic deafness. Am J Hum Genet. 2002;71(2): 262-275.

haematologica | 2017; 102(6)


ARTICLE

Myeloproliferative Disorders

The clinical and molecular diversity of mast cell leukemia with or without associated hematologic neoplasm

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Mohamad Jawhar,1,2 Juliana Schwaab,1,2 Manja Meggendorfer,3 Nicole Naumann,1,2 Hans-Peter Horny,4 Karl Sotlar,5 Torsten Haferlach,3 Karla Schmitt,6 Alice Fabarius,1,2 Peter Valent,7 Wolf-Karsten Hofmann,1,2 Nicholas C.P. Cross,8,9 Georgia Metzgeroth1,2 and Andreas Reiter1,2

Department of Hematology and Oncology, University Medical Centre Mannheim, Germany; 2Medical Faculty Mannheim, University of Heidelberg, Germany; 3Munich Leukemia Laboratory, Germany; 4Institute of Pathology, Ludwig-Maximilians-University, Munich, Germany; 5University Institute of Pathology, University Hospital Salzburg, Paracelsus Medical University, Salzburg, Austria; 6Department of Hematology and Oncology, University Hospital Aachen, Germany; 7Department of Internal Medicine I, Division of Hematology and Ludwig Boltzmann Cluster Oncology, Medical University of Vienna, Austria; 8Wessex Regional Genetics Laboratory, Salisbury, UK and 9Faculty of Medicine, University of Southampton, UK 1

Haematologica 2017 Volume 102(6):1035-1043

ABSTRACT

M

ast cell leukemia is a rare variant of advanced systemic mastocytosis characterized by at least 20% of mast cells in a bone marrow smear. We evaluated clinical and molecular characteristics of 28 patients with (n=20, 71%) or without an associated hematologic neoplasm. De novo mast cell leukemia was diagnosed in 16 of 28 (57%) patients and secondary mast cell leukemia evolving from other advanced systemic mastocytosis subtypes in 12 of 28 (43%) patients, of which 7 patients progressed while on cytoreductive treatment. Median bone marrow mast cell infiltration was 65% and median serum tryptase was 520 μg/L. C-findings were identified in 26 of 28 (93%) patients. Mutations in KIT (D816V, n=19; D816H/Y, n=5; F522C, n=1) were detected in 25 of 28 (89%) patients and prognostically relevant additio-nal mutations in SRSF2, ASXL1 or RUNX1 (S/A/Rpos) in 13 of 25 (52%) patients. Overall response rate in 18 treatment-naïve patients was 5 of 12 (42%) on midostaurin and 1 of 6 (17%) on cladribine, and after switch 1 of 4 (25%) on midostaurin and 0 of 3 on cladribine, respectively. S/A/Rpos adversely affected response to treatment and progression to secondary mast cell leukemia (n=6) or acute myeloid leukemia (n=3) while on treatment (P<0.05). The median overall survival from mast cell leukemia diagnosis was 17 months as compared to 44 months in a control group of 124 patients with advanced systemic mastocytosis but without mast cell leukemia (P=0.03). In multivariate analyses, S/A/Rpos remained the only independent poor prognostic variable predicting overall survival (P=0.007). In conclusion, the molecular signature should be determined in all patients with mast cell leukemia because of its significant clinical and prognostic relevance.

Correspondence: andreas.reiter@medma.uni-heidelberg.de

Received: January 6, 2017. Accepted: February 28, 2017. Pre-published: March 2, 2017. doi:10.3324/haematol.2017.163964 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1035 ©2017 Ferrata Storti Foundation

Introduction According to the World Health Organization (WHO) classification, advanced systemic mastocytosis (SM) comprises patients with aggressive SM (ASM), SM with an associated hematologic neoplasm (SM-AHN), and mast cell leukemia (MCL).1-3 MCL is characterized by at least 20% of mast cells (MC) in a bone marrow (BM) smear and a particularly poor prognosis with median survival time of less than six months. In the majority of patients with MCL, the aleukemic variant (MCs represent less than 10% of all blood leukocytes) is diagnosed.2,4-6 Systemic mastocytosis is characterized by somatically acquired, activating mutations in the gene encoding the receptor tyrosine kinase KIT, most commonly haematologica | 2017; 102(6)

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.

1035


M. Jawhar et al.

KIT D816V which is identified in over 80% of all patients with SM.7 Recent data, however, have highlighted that the molecular pathogenesis of advanced SM is complex with one or more additional mutations, eg. in ASXL1, TET2, JAK2 V617F, RUNX1, SRSF2, present in over 60% of advanced SM patients.8-10 These additional mutations are usually acquired by neoplastic (stem) cells prior to KIT D816V, thereby indicating a multi-mutated stem cell disease and a step-wise process of oncogenesis.11 Molecular aberrations, eg. in SRSF2, ASXL1, or RUNX1 (S/A/R gene panel) have a strong adverse impact on disease phenotype and prognosis.9,12 A new risk score was therefore proposed for patients with SM. This score includes clinical variables, eg. splenomegaly and elevated alkaline phosphatase (AP), as well as molecular markers, eg. mutations in the S/A/R gene panel (S/A/Rpos).13 Our current knowledge of MCL, including clinical and molecular characteristics, treatment options, survival, and prognostic factors is limited to case reports, caseseries and/or literature reviews.14-16 We here report on a relatively large group of patients with recently diagnosed MCL. The aim of the present study was: a) to characterize the clinical and molecular features in patients with MCL; b) to evaluate response and resistance to various treatments; and c) to define prognostic factors.

Methods Diagnosis of mast cell leukemia The diagnosis of SM was established according to the WHO classification.2,5,6,17 BM biopsies and BM smears were evaluated by reference pathologists of the European Competence Network on Mastocytosis, ECNM (H-PH and KS). Diagnosis of MCL was based on the presence of at least 20% MCs in BM smears (Figure 1). Twenty-eight patients were diagnosed between 2008 and 2016. The study design adhered to the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of the Medical Faculty of Mannheim, University of Heidelberg, as part of the â&#x20AC;&#x2DC;German Registry on Disorders of Eosinophils and Mast Cellsâ&#x20AC;&#x2122;. All patients gave written informed consent.

Assessment of KIT D816V Qualitative and quantitative assessment of the KIT D816V expressed allele burden was performed using allele-specific quantitative real-time reverse transcriptase polymerase chain reaction analyses (qRT-PCR) as previously described.12 Molecular analyses were performed in all patients at baseline. In KIT D816V negative patients, exon 17 was sequenced by conventional Sanger sequencing. In selected patients, targeted next-generation deep amplicon sequencing (NGS) was used for identification of alternative mutations in all KIT exons.

A

C

E

B

D

F

Figure 1. Bone marrow morphology and phenotype of 3 patients with mast cell leukemia (MCL). Patients #25: MCL, #28: MCL, #20: MCL with an associated hematologic neoplasm (AHN). (A, C and E) Bone marrow (BM) smears show an abundance of pleomorphic mast cells (MCs) exhibiting metachromatic granules. Quantitative criteria for diagnosis of MCL are completely fulfilled since MC make up more than 90% of all nucleated cells. Note the bizarre giant metachromatic MCs in (A), the prominent hemophagocytic activity of MC in (C), and the marked cytological atypia of MCs with pronounced hypogranulation in (E). Case (E) also exhibits immature atypical eosinophils enabling the diagnosis of an AHN, probably myelodysplastic/myeloproliferative neoplasm with eosinophilia. (B, D and F) BM sections show extreme hypercellularity and packed MC infiltrates. Fat cells and normal blood cell precursors are subtotally depleted. The cytomorphological aspects are fully reflected by the histomorphological findings. Note the extreme siderosis in (D) and the clear-cell aspect of atypical MC in (F) with the possibility of a misdiagnosis (hairy cell leukemia, histiocytosis or even metastatic infiltrates of a renal cell carcinoma) unless appropriate immunohistochemistry is performed. (A-F) Wright-Giemsa staining.

1036

haematologica | 2017; 102(6)


Clinical and molecular diversity of MCL

Targeted next-generation sequencing analysis Next-generation sequencing analysis was either performed by 454 FLX amplicon chemistry (Roche, Penzberg, Germany) or library preparation based on the TruSeq Custom Amplicon Low Input protocol (Illumina, San Diego, CA, USA) and sequencing on the MiSeq instrument (Illumina, San Diego, CA, USA) to

investigate the KIT gene and the following 18 genes: ASXL1, CBL, ETV6, EZH2, IDH1, IDH2, JAK2, KRAS, NPM1, NRAS, RUNX1, SETBP1, SF3B1, SRSF2, TET2, TP53, U2AF1, ZRSR2.8 Gene mutations were annotated compared to the reference sequence based on the Ensembl Transcript ID (Ensembl release 85: July 2016).

Table 1. Clinical, laboratory, molecular and treatment characteristics and outcome of 28 patients with de novo mast cell leukemia (MCL, n=16) or secondary MCL (n=12).

Variables N. of patients (n) Age in years, median (range) Males, n (%) Diagnosis MCL, n (%) MCL-AHN, n (%) CMML, n (%) MDS, n (%) MDS/MPNu, n (%) CEL, n (%) Diagnosis prior to MCL ASM, n (%) SM-AHN, n (%) Transformation to MCL in months, median (range) C-findings Hemoglobin, g/dL; median (range) Platelets, x 109/L; median (range) Cytopenia,* n (%) Ascites, n (%) Hypoalbuminemia (<35 g/L), n (%) Weight loss (>10% in 6 months), n (%) Other relevant findings MC infiltration in BM, %; median (range) MC in BM smear %; median (range) Serum tryptase, μg/L; median (range) >200 μg/L, n (%) KIT D816V EAB in PB, %; median (range) Splenomegaly Spleen volume, mL; median (range) (n=15) Marked splenomegaly (≥ 1200 mL), n (%) Alkaline phosphatase, U/L; median (range) >150 U/L, n (%) Treatment modalities Midostaurin, n (%) Cladribine, n (%) Midostaurin and cladribine (vice versa), n (%) Molecular profile KIT mutations D816V Other None Additional mutations S/A/Rpos, n (%) S/A/Rneg, n (%) Outcome Death, n (%)

MCL

de novo MCL

Secondary MCL

28 67 (45-82) 16 (57)

16 69 (47-82) 10 (63)

12 65 (45-73) 6 (50)

8 (29) 20 (71) 8 (40) 5 (25) 5 (25) 2 (10)

6 (38) 10 (62) 4 (40) 0 5 (50) 1 (10)

2 (17) 10 (83) 4 (40) 5 (50) 0 1 (10)

-

-

2 (17) 10 (83) 18 (4-71)

8.9 (7.9-14.3) 69 (21-795) 26 (93) 13 (46) 11 (39) 12 (43)

9.2 (7.9-13.3) 64 (21-795) 15 (94) 7 (44) 5 (31) 8 (50)

8.7 (7.9-14.3) 86 (26-331) 11 (92) 6 (50) 6 (50) 4 (33)

65 (20-95) 25 (20-95) 520 (157-1854) 26 (93) 43 (20-98) 28 (100) 1079 (283-2442) 8 (53) 181 (59-548) 20 (71)

60 (20-95) 25 (20-100) 520 (157-1854) 15 (94) 43 (21-54) 16 (100) 1058 (758-2058) 4 (50) 161 (59-548) 10 (62)

65 (30-95) 20 (20-50) 544 (160-1250) 11 (92) 42 (20-98) 12 (100) 1240 (539-2442) 4 (57) 186 (103-424) 10 (83)

13 (52) 2 (7) 10 (36)

8 (57) 1 (7) 5 (36)

5 (45) 1 (8) 5 (42)

19 (68) 6 (21) 3 (11)

10 (63) 4 (25) 3 (19)

9 (75) 2 (17) 0

13 (52) 12 (48)

7 (50) 7 (50)

6 (55) 5 (45)

18 (64)

10 (63)

8 (67)

*Cytopenia: hemoglobin <10 g/dL and/or platelets <100 x 109/L. AHN; associated hematologic neoplasm; ASM: aggressive systemic mastocytosis; BM: bone marrow; CEL: chronic eosinophilic leukemia; CMML: chronic myelomonocytic leukemia; MDS/MPNu: myelodysplastic syndrome/myeloproliferative neoplasm unclassified; EAB: expressed allele burden; MC: mast cells; MCL: mast cell leukemia; n: number; OS: overall survival; PB: peripheral blood; S/A/Rpos: mutation(s) in the SRSF2/ASXL1/RUNX1 gene panel SM; S/A/Rneg: no mutation in the S/A/R gene panel; SM: systemic mastocytosis.

haematologica | 2017; 102(6)

1037


M. Jawhar et al.

Cytogenetic analysis and fluorescence in situ hybridization analysis

Results

Cytogenetic analyses of at least 20 Giemsa-banded bone marrow (BM) metaphases (24 hour and/or 48 hour culture) were analyzed and interpreted according to the International System for Human Cytogenetic Nomenclature.18 If necessary, chromosome banding analysis was combined with fluorescence in situ hybridization (FISH) analysis according to the manufacturer’s instructions (Metasystems, Altlussheim, Germany).19

Clinical and morphological characteristics

Statistical analyses Statistical analyses considered clinical, laboratory or molecular parameters obtained at time of MCL diagnosis. Overall survival (OS) analysis was determined as time from date of diagnosis to date of death or last visit. For analysis of survival differences between the various subtypes of advanced SM, we used the current MCL patient cohort and a patient cohort with advanced SM but without MCL (n=124), who were enrolled within the ‘German Registry on Disorders of Eosinophils and Mast Cells’. Pearson correlation analysis was performed for the correlation between two parameters. Differences in the distribution of continuous variables between categories were analyzed by Mann-Whitney test (for comparison of two groups). For categorical variables, Fishers’s exact test was used. OS probabilities were estimated with the Kaplan-Meier method and compared by the log-rank test in univariate analysis. For the estimation of hazard ratios (HRs) and multivariate analysis, the Cox proportional hazard regression model was used. P<0.05 (two-sided) was considered significant. There was no adjustment for multiple testing as all analyses were explorative. SPSS version 22.0.0 (IBM Corporation, Armonk, NY, USA) was used for statistical analysis.

A

The median age of the 28 patients was 67 years (range 45-82; male 57%). The median percentage of MCs in BM smears and trephine biopsies was 25% (range 20-95) and 65% (range 20-95; 82% of patients >50%), respectively. MCs in peripheral blood (PB) 10% or more (leukemic MCL) were only seen in 2 of 28 (7%) patients. MCL-AHN was diagnosed in 20 of 28 (71%) patients: chronic myelomonocytic leukemia (CMML, n=8), myelodysplastic myeloproliferative neoplasm unclassifiable (MDS/MPNu, n=5), MDS (n=5) or chronic eosinophilic leukemia (CEL, n=2). Blood parameters analyzed in this study included: leukocytes (median 6.1x109/L, range 1.347.6), hemoglobin (median 8.9 g/dL, range 7.9-14.3; <10 in 71% of patients), platelets (median 69x109/L, range 21795; <100 in 68% of patients), eosinophils (0.1x109/L, range 0-12.9; >1.0 in 14%), and monocytes (0.6x109/L, range 0.1-2.9; >1.0 in 25%). Hematologic C-findings such as hemoglobin less than 10 g/dL and/or platelets less than 100x109/L were identified in 26 of 28 (93%) patients. Median serum tryptase level (normal value <11.4 μg/L) was 520 μg/L (range 157-1854; 93% ≥200) and median KIT D816V expressed allele burden was 43% (range 2098) in PB. Signs of non-hematologic organ dysfunction included elevated AP (20 of 28, 71%, median 181; range 59-548) and splenomegaly in 28 of 28 (100%) patients. Spleen volumetry results obtained by magnetic resonance imaging were available in 16 patients and showed marked splenomegaly (≥1200 mL) in 8 of 16 cases (50%) (Table 1).

B

Figure 2. Mutational profile of 28 patients with mast cell leukemia (MCL). (A) Alignment of gene mutations in 28 patients with MCL. Each column represents an individual patient. Results of genetic analyses are depicted in different colors and (B) relative frequency distribution of KIT mutations, additional mutations and mutations in the SRSF2/ASXL1/RUNX1 (S/A/R) gene panel.

1038

haematologica | 2017; 102(6)


Clinical and molecular diversity of MCL

De novo mast cell leukemia and secondary mast cell leukemia De novo MCL was diagnosed in 16 of 28 (57%) patients. Secondary MCL (sMCL) evolving from other advanced SM subtypes (SM-AHN, n=10; ASM, n=2) was observed in 12 of 28 (43%) patients (Table 1) with a median time to progression of 18 months (range 4-71). Of note, 3 of 28 patients (11%) progressed into secondary acute myeloid leukemia (MCL-sAML) 18, 28 and 34 months, respectively, after the diagnosis of MCL-CMML had been established. Overall, no statistically significant differences were observed between de novo MCL and sMCL regarding clinical, morphological or molecular characteristics (Table 1).

KIT mutations Overall, 25 of 28 (89%) patients had mutations in KIT (Figure 2). KIT D816V was identified in 19 of 28 patients (68%), and alternative KIT mutations (D816H, n=3; D816Y, n=2; F522C, n=1) in 6 of 28 (21%) patients by Sanger sequencing or targeted NGS (KIT F522C), respectively.

Additional mutations Data on additional mutations were available in 25 of 28 (89%) patients. Seventeen of 25 (68%) patients with MCL had at least one additional mutation (median 2, range 1-4) in 10 of 18 analyzed genes, and 8 of 25 (32%) patients were negative for additional mutations (Figure 2). The most frequently affected genes were SRSF2 (n=10, 40% of patients), TET2 (n=8, 32%), ASXL1 (n=5, 20%), N/KRAS

(n=4, 16%), CBL (n=3, 12%) IDH1/2 (n=2, 8%), and RUNX1 (n=2, 8%). EZH2, JAK2 and SF3B1 were less frequently affected (<5%). Only the concurrent presence of CBL and TET2 achieved significance (P<0.05). At least one mutation in the S/A/R gene panel (S/A/Rpos) was identified in 13 of 25 (52%) patients. Clinical and morphological features in S/A/Rpos and S/A/Rneg patients with MCL were significantly different. In particular, S/A/Rpos patients had lower median hemoglobin levels, a higher median spleen volume, and higher AP levels (Table 2).

Cytogenetic analysis A karyotype was available in 24 of 28 (86%) patients with MCL which was normal in 19 of 24 (79%) and aberrant in 5 of 24 (21%) patients; 3 patients had a complex karyotype (≥ 3 aberrations) and 2 patients a del(5q) or del(12p), respectively. According to these findings, there is no recurrent cytogenetic abnormality in patients with MCL.

Response and resistance to treatment Twenty-five of 28 (89%) patients were treated with various cytoreductive drugs (Table 3). Three patients were not treated because of death at diagnosis (n=1) or comorbidity (n=2). Overall, 23 of 25 (92%) patients were treated with midostaurin [international phase II study on the efficacy and safety of midostaurin in advanced SM (clinicaltrial.gov identifier: 00782067), n=9; compassionate use program by Novartis Pharmaceuticals, n=14]20 at any time prior to or after diagnosis of MCL. Twelve of 25 (48%) patients received cladribine.

Table 2. Clinical, laboratory characteristics and outcome of 25 patients with mast cell leukemia (MCL) depending on the mutational status in the SRSF2/ASXL1/RUNX1 (S/A/R) gene panel.

Variables N. of patients (n) Diagnosis MCL, n (%) MCL-AHN, n (%) sAML, n (%) C-findings Hemoglobin, g/dL; median (range) Platelets, x 109/L; median (range) Cytopenia,*** n (%) Other relevant findings MC infiltration in BM, %; median (range) MC in BM smear %; median (range) Serum tryptase, μg/L; median (range) >200 µg/L, n (%) KIT D816V+ EAB in PB, %; median (range) Spleen volume, mL; median (range), n=15 Marked splenomegaly (≥ 1200 mL) Alkaline phosphatase, U/L; median (range) >150 U/L, n (%) Outcome Overall survival, months; median (95% CI) Death, n (%)

P*

MCL and S/A/Rpos

MCL and S/A/Rneg

13

12

0 13 (100) 3 (23)

6 (50) 6 (50) 0

0.005 0.005

8.7 (7.9-13.3) 71 (26-795) 13 (100)

9.8 (7.9-14.3) 95 (21-331) 10 (83)

0.04 n.s. n.s.

60 (20-80) 25 (20-50) 392 (248-1854) 13 (100) 37 (20-98) 1324 (716-2442) 7 (70) 262 (105-548) 12 (92)

80 (50-95) 20 (20-95) 600 (157-1690) 11 (92) 47 (30-55) 770 (538-1240) 1 (20) 148 (59-232) 6 (50)

n.s. n.s. n.s. n.s. n.s. 0.02 n.s. 0.003 0.03

13 (8-18) 12 (92)

n.r. 3 (25)

0.007**

*P-values refer to the Mann-Whitney U test, Fisher’s exact test or **log-rank tests comparing S/A/Rpos vs. S/A/Rneg patients. ***Cytopenia: hemoglobin <10 g/dL and/or platelets <100 x 109/L. AHN: associated hematologic neoplasm; BM: bone marrow; EAB: expressed allele burden; MC: mast cells; MCL: mast cell leukemia; OS: overall survival; PB: peripheral blood; n.r.: not reached; n.s.: not significant; S/A/Rpos: mutation(s) in the SRSF2/ASXL1/ RUNX1 gene panel SM; S/A/Rneg: no mutation in the S/A/R gene panel; sAML: secondary acute myeloid leukemia; SM: systemic mastocytosis.

haematologica | 2017; 102(6)

1039


M. Jawhar et al.

In 18 treatment-naïve MCL patients, the overall response rate (ORR) according to IWG-MRT and ECNM criteria was 5 of 12 (42%) on midostaurin and 1 of 6 (17%) on cladribine, respectively, and after switching, 1 of 4 (25%) on midostaurin and 0 of 3 on cladribine, respectively. Independently of midostaurin (n=16) or cladribine (n=13), there was no significant difference in ORR during those 29 individual treatment periods between MCLAHN and MCL patients [6 of 20 (30%) vs. 4 of 9 (44%); P=0.2]. In contrast, there was a significant difference in ORR in 27 individual treatment periods (the S/A/R status was unknown in 2 patients) between S/A/Rpos and S/A/Rneg patients [2 of 13 (15%) vs. 8 of 14 (57%); P<0.05] (Table 3). Overall, progression while on treatment (n=9) occurred predominantly in patients with AHN (8 of 9 patients) after a median of 15 months (range 1-58) from start of treatment and was significantly associated with S/A/Rpos (6 of 8 patients; P<0.05).

(1.1-3.1); P=0.03] (Figure 3). These data confirm the particularly poor prognosis of patients with MCL. In univariate analyses of multiple clinical, morphological and molecular variables only bicytopenia [hemoglobin <10 g/dL and platelets <100x109/L, n=13 vs. hemoglobin ≥10 g/dL or platelets ≥100x109/L, n=13; P=0.02, HR 3.2 (1.2-8.9)], elevated AP [P=0.009, HR 3.3 (1.3-8.3)] and S/A/Rpos [P=0.007, HR 5.0 (1.8-18.1)] were identified as poor prognostic variables regarding OS. In multivariate analyses, S/A/Rpos remained the only independent poor risk factor concerning OS (median 13 months, 95%CI: 818 vs. median not reached) (Figure 4). There was no difference in OS between de novo MCL vs. sMCL (P=0.9) or MCL-AHN vs. MCL (P=0.2). In treatment-naïve MCL patients, no difference was observed between midostaurin monotherapy (n=9) and sequential midostaurin/cladribine treatment or vice versa (n=7) (P=0.9) (Figure 4).

Prognostic factors and overall survival The median OS of all 28 MCL patients from diagnosis was 17 months [95% confidence interval (CI): 10-24] with a 2-year OS probability of 24%. The median OS of 124 patients with advanced SM (ASM and SM-AHN) but without MCL was 44 months [95%CI: 31.3-56.7; HR 1.8

Discussion In addition to the confirmation of some known clinical and molecular characteristics of MCL, we identified several new and important aspects in this heterogeneous disor-

Table 3. Treatment modalities and responses in 25 patients with mast cell leukemia (MCL).

Treatment (month or cycles)* Case# Age AHN S/A/R First

Response**

Second

Response**

Treatment prior to (first) and after (second, third) diagnosis of MCL 1 74 yes pos midostaurin (22) PR 2 67 yes pos midostaurin (07) PD 2-CDA (08) 5 70 yes pos 2-CDA (05) SD midostaurin (30) 14 62 yes pos midostaurin (34) CI FLAG (02) 19 58 yes nk midostaurin (22) PD + 5-aza (04) 21 52 yes neg midostaurin (01) PD 2-CDA (04) 16 54 yes neg midostaurin (86) CI Treatment after (first, second, third) diagnosis of MCL 8 57 yes pos midostaurin (13) PR 12 66 yes pos midostaurin (15) PD 3 71 yes pos midostaurin (11) SD 4 76 yes pos 2-CDA (02) SD midostaurin (02) 6 46 yes pos 2-CDA (06) PD midostaurin (01) 7 77 yes pos midostaurin (07) PD 2-CDA (01) 9 74 yes pos midostaurin (04) CI 10 67 yes neg midostaurin (72) PR 11 67 yes neg 2-CDA (02) SD 15 74 no neg 2-CDA (06) CI midostaurin (09) 17 76 no neg midostaurin (70) PR 18 47 no nk 2-CDA (03) PD 20 45 yes neg 2-CDA (04) PD midostaurin (03) 23 70 yes pos midostaurin (05) SD 24 62 no neg midostaurin (05) CI 2-CDA (01) 25 72 no nk midostaurin (01) PD 26 82 yes neg midostaurin (07) SD 28 64 no neg midostaurin (09) SD 2-CDA (04)

Third

PD CI PD

decitabine (02)

PR

allo SCT planned allo SCT planned

DA (01)

Response**

PD

SD PD PD

sMCL (08) sMCL (06) sMCL (24) sMCL-sAML (34) sMCL (20) sMCL (01) sMCL (58)

death (22) (DR) death (21) (DR) death (40) ( DR) death (43) (DR) death (22) (DR) alive (06) alive (86)

sAML (12) sAML (15) sMCLx

death (15) ( DR) alive (17) death (27) ( DR) death (06) ( DR) death (23) ( DR) death (10) ( DR) death (04) ( DR) alive (72) alive (06) alive (43) alive (86) death (17) (DR) alive (07) death (08) (DR) death (09) (DR) death (01) ( DR) alive (07) alive (25)

sMCLx sMCLx allo SCT planned

PD SD

Outcome (month from start first treatment) (cause of death)

sMCLx

PR SD

Progression (month from start first treatment)

allo SCT

PR

*Midostaurin treatment in months; cladribine (2-CDA), 5-azacythidine (5-aza), decitabine, FLAG (fludarabine, cytarabin), daunorubicine/cytarabin (DA) treatment in cycle number. **Response by month 6 or at the end of treatment due to progression/death prior to month 6. xProgression to secondary mast cell leukemia (sMCL) before start of first treatment. AHN: associated hematologic neoplasm; CI: clinical improvement; DR: disease related; neg: negative; nk: not known; PD: progressive disease; pos: positive; PR: partial response; sAML: secondary acute myeloid leukemia; allo SCT: allogeneic stem cell transplantation; SD: stable disease.

1040

haematologica | 2017; 102(6)


Clinical and molecular diversity of MCL

der, which may improve understanding of the clinical phenotype, prediction of response to treatment and prognosis. Confirmatory results included the prognostic relevance of the defining criterion of at least 20% of MCs in a BM smear and the presence of the KIT D816V mutation in approximately 60-70% of patients, which is significantly lower than other SM subtypes. However, alternative mutations at position 816, eg. D816H or D816Y, were identified in a significant proportion of KIT D816V negative patients leading to an overall incidence of KIT D816 mutations in approximately 90% of patients with MCL. We therefore recommend that KIT exon 17 and, if negative, other exons of KIT gene should be sequenced in KIT D816V negative patients with advanced SM and particularly MCL.14,21,22 In our series, the proposed subtypes of MCL without Cfindings (chronic MCL) and leukemic MCL were only observed in 2 patients each, in line with previously published series,14,15 the latter clearly demonstrating that presence or absence of MCs in PB is only of minor importance for the diagnosis of MCL. We have, however, observed a much higher relative frequency of MCL-AHN than that reported in the literature (71% vs. 10%),14 which is most likely due to the fact that BM aspirates (cytology) and BM sections (histology and immunohistochemistry) of all our cases were centrally reviewed. Because of the high MC burden (eg. MC infiltration in BM, serum tryptase level, KIT D816 allele burden) in MCL and the multilineage involvement of the KIT D816V mutation,11 MCL and AHN almost always derive from the same myeloid progenitors, and the question as to whether C-findings are attributable to MCL or to AHN is not relevant. Moreover, there are distinct differences between de novo MCL and sMCL evolving from other advanced SM subtypes.14 In our series, approximately one-third of patients had sMCL, in all cases progressing from other advanced SM subtypes, with 40% of those patients progressing while on cytoreductive treatment, predominantly midostaurin. We did not observe direct evolution from indolent SM (ISM) to MCL, confirming previous reports that progression usually occurs from SM-AHN or ASM.4

Importantly, there was no difference in OS between de novo MCL and sMCL. Consistent with previous reports, cladribine has only shown a little activity in our patients with MCL,14,21,23-25 with only 2 cases showing durable partial response or clinical improvement. On the other hand, the median OS of 17 months is certainly better than the survival reported for historical controls.4,14,26 Because the vast majority of patients (>90%) have also been treated with midostaurin, either prior to or following cladribine, the relatively high ORR of midostaurin compared with cladribine is likely to be associated with the improved survival. Most of the previously reported MCL cases were frequently treated with imatinib, dasatinib or masitinib, which are definitely without any substantial efficacy in KIT D816V advanced SM.14 At the moment, there are no distinct parameters to predict the response to treatment and the risk for progression to sMCL or MCL-sAML. Overall, the most significant prognostic factor identified for MCL patients was the presence or absence of mutations in S/A/R. In particular, S/A/Rpos patients with MCL had a more aggressive phenotpye, a lower response rate, more (intrinsic) resistance to disparate treatment modalities, and a poor survival compared to S/A/Rneg MCL. The patients with progression to sMCL or to MCL-AML while on treatment were predominantly S/A/Rpos. Moreover, durable remissions and longterm OS were exclusively observed in S/A/Rneg patients. This also explains why no significant differences were observed between MCL and MCL-AHN, because MCLAHN was also associated with other, potentially more favorable mutations than in S/A/R (eg. TET2, CBL, JAK2). It is, therefore, not the presence of AHN per se, but the molecular background of AHN that impacts on prognosis. Because of resistant and rapidly progressive disease followed by poor OS, particularly in S/A/Rpos patients with MCL, intensive chemotherapy and allogeneic stem cell transplantation (SCT) is recommended as the only curative treatment option for young and fit patients.23 Allogeneic SCT should be performed early because responses on midostaurin, cladribine and/or myeloid-type

Advanced SM (without MCL) (n=124) MCL (n=28)

P=0.03 HR 1.8 [1.1-3.1]

Figure 3. Overall survival (OS). Kaplan-Meier estimates of OS depending on comparisons between patients with mast cell leukemia (MCL, n=28) and a control group of patients with advanced systemic mastocytosis (without MCL) (n=124) enrolled within the â&#x20AC;&#x2DC;German Registry on Disorders of Eosinophils and Mast Cellsâ&#x20AC;&#x2122;. CI: confidence interval; HR: hazard ratio.

haematologica | 2017; 102(6)

1041


M. Jawhar et al. A

B

P=0.9

C

P=0.2

D

P=0.9 P=0.07 HR 5.0 [1.8-18.1]

Figure 4. Kaplan-Meier estimates of overall survival (OS) depending on comparisons between multiple variables. (A) De novo mast cell leukemia (MCL) and secondary MCL (sMCL), (B) MCL with or without an associated hematologic neoplasm (AHN), (C) midostaurin treatment or cladribine (2-CDA) and/or midostaurin vice versa and (D) with mutations in the SRSF2/ASXL1/RUNX1 (S/A/Rpos) or without mutation in S/A/R (S/A/Rneg) gene panel. CI: confidence interval; HR: hazard ratio; NR: not reached.

induction chemotherapy may only be short-lived.26,27 However, only a minority of patients are eligible for allogeneic SCT because of advanced age, SM-related organ damage (eg. liver cirrhosis), comorbidity, and frequently poor response or progression on various treatment options, including intensive chemotherapy. In addition, Ustun et al. have reported a 5-year survival rate of only 14% for patients with MCL, which may have been influenced by several negative factors, including the S/A/R mutation status and a poor remission status prior to transplant. Clinicial trials are warranted to define the optimal intensity, duration, and efficacy of combined strategies with midostaurin and chemotherapy for treatment of MCL in general, and also for debulking and bridging prior to allogeneic SCT. In this respect, new drugs such as brentuximab-vedotin in CD30 positive advanced SM,28 gemtuzumab ozogamicin29 and possibly also hypomethylating agents30 should be investigated. We conclude that: a) MCL has the worst prognosis of

References 1. Valent P, Akin C, Escribano L, et al. Standards and standardization in mastocytosis: consensus statements on diagnostics, treatment recommendations and response

1042

the various advanced SM subtypes; b) leukemic MCL and MCL without C-findings are rare; c) sMCL is frequent and evolves from other advanced SM subtypes but not directly from ISM; d) KIT D816 mutations are more frequent than previously reported and KIT D816V negative patients should be tested for other KIT mutations; e) mutations in the S/A/R gene panel are present in approximately 50% of patients with MCL and are adversely associated with phenotype, response to treatment, progression, and OS; and f) the median OS of approximately 1.5 years, achieved predominantly in midostaurin-treated patients, is superior compared to historical controls. Funding This work was supported by the ‘Deutsche José Carreras Leukämie-Stiftung e.V.’ (grant n. DJCLS R 13/05) and by the SEED program of the Medical Faculty Mannheim, Heidelberg University, and by the Austrian Science Fund (FWF) SFB project F4704-B20.

criteria. Eur J Clin Invest. 2007;37(6):435453. 2. Valent P, Horny HP, Escribano L, et al. Diagnostic criteria and classification of mastocytosis: a consensus proposal. Leuk Res. 2001;25(7):603-625. 3. Theoharides TC, Valent P, Akin C. Mast

Cells, Mastocytosis, and Related Disorders. N Engl J Med. 2015;373(2):163-172. 4. Lim KH, Tefferi A, Lasho TL, et al. Systemic mastocytosis in 342 consecutive adults: survival studies and prognostic factors. Blood. 2009;113(23):5727-5736. 5. Horny HP AC, Metcalfe DD, et al.

haematologica | 2017; 102(6)


Clinical and molecular diversity of MCL

6.

7.

8.

9.

10.

11.

12.

Swerdlow SH, Campo E, Harris NL, et al. World Health Organization (WHO) Classification of Tumours. Mastocytosis (Mast cell disease). Pathology & Genetics. Tumours of Haematopoietic and Lymphoid Tissues, vol. 2: Lyon, France: IARC Press, 2008, p. 54-63. Pardanani A. Systemic mastocytosis in adults: 2017 update on diagnosis, risk stratification and management. Am J Hematol. 2016;91(11):1146-1159. Gleixner KV, Mayerhofer M, CernyReiterer S, et al. KIT-D816V-independent oncogenic signaling in neoplastic cells in systemic mastocytosis: role of Lyn and Btk activation and disruption by dasatinib and bosutinib. Blood. 2011;118(7):1885-1898. Schwaab J, Schnittger S, Sotlar K, et al. Comprehensive mutational profiling in advanced systemic mastocytosis. Blood. 2013;122(14):2460-2466. Jawhar M, Schwaab J, Schnittger S, et al. Additional mutations in SRSF2, ASXL1 and/or RUNX1 identify a high-risk group of patients with KIT D816V(+) advanced systemic mastocytosis. Leukemia. 2016;30(1):136-143. Pardanani AD, Lasho TL, Finke C, et al. ASXL1 and CBL mutations are independently predictive of inferior survival in advanced systemic mastocytosis. Br J Haematol. 2016;175(3):534-536. Jawhar M, Schwaab J, Schnittger S, et al. Molecular profiling of myeloid progenitor cells in multi-mutated advanced systemic mastocytosis identifies KIT D816V as a distinct and late event. Leukemia. 2015;29(5):1115-1122. Erben P, Schwaab J, Metzgeroth G, et al. The KIT D816V expressed allele burden for diagnosis and disease monitoring of systemic mastocytosis. Ann Hematol. 2014;93(1):81-88.

haematologica | 2017; 102(6)

13. Jawhar M, Schwaab J, Hausmann D, et al. Splenomegaly, elevated alkaline phosphatase and mutations in the SRSF2/ASXL1/RUNX1 gene panel are strong adverse prognostic markers in patients with systemic mastocytosis. Leukemia. 2016;30(12):2342-2350. 14. Georgin-Lavialle S, Lhermitte L, Dubreuil P, et al. Mast cell leukemia. Blood. 2013; 121(8):1285-1295. 15. Valent P, Berger J, Cerny-Reiterer S, et al. Chronic mast cell leukemia (MCL) with KIT S476I: a rare entity defined by leukemic expansion of mature mast cells and absence of organ damage. Ann Hematol. 2015;94(2):223-231. 16. Valent P, Blatt K, Eisenwort G, et al. FLAGinduced remission in a patient with acute mast cell leukemia (MCL) exhibiting t(7;10)(q22;q26) and KIT D816H. Leuk Res Rep. 2014;3(1):8-13. 17. 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. 18. 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. 19. Schoch C, Schnittger S, Bursch S, et al. Comparison of chromosome banding analysis, interphase- and hypermetaphaseFISH, qualitative and quantitative PCR for diagnosis and for follow-up in chronic myeloid leukemia: a study on 350 cases. Leukemia. 2002;16(1):53-59. 20. Gotlib J, Kluin-Nelemans HC, George TI, et al. Efficacy and Safety of Midostaurin in Advanced Systemic Mastocytosis. N Engl J Med. 2016;374(26):2530-2541. 21. Mital A, Piskorz A, Lewandowski K, et al. A case of mast cell leukaemia with exon 9

22.

23.

24.

25. 26.

27.

28.

29.

30.

KIT mutation and good response to imatinib. Eur J Haematol. 2011;86(6):531-535. Alvarez-Twose I, Matito A, Morgado JM, et al. Imatinib in systemic mastocytosis: a phase IV clinical trial in patients lacking exon 17 KIT mutations and review of the literature. Oncotarget. 2016 Jul 19. [Epub ahead of print] Kluin-Nelemans HC, Oldhoff JM, Van Doormaal JJ, et al. Cladribine therapy for systemic mastocytosis. Blood. 2003; 102(13):4270-4276. Pardanani A, Hoffbrand AV, Butterfield JH, Tefferi A. Treatment of systemic mast cell disease with 2-chlorodeoxyadenosine. Leuk Res. 2004;28(2):127-131. Valentini CG, Rondoni M, Pogliani EM, et al. Mast cell leukemia: a report of ten cases. Ann Hematol. 2008;87(6):505-508. Ustun C, Reiter A, Scott BL, et al. Hematopoietic stem-cell transplantation for advanced systemic mastocytosis. J Clin Oncol. 2014;32(29):3264-3274. Ustun C, Gotlib J, Popat U, et al. Consensus Opinion on Allogeneic Hematopoietic Cell Transplantation in Advanced Systemic Mastocytosis. Biol Blood Marrow Transplant. 2016;22(8):1348-1356. Blatt K, Cerny-Reiterer S, Schwaab J, et al. Identification of the Ki-1 antigen (CD30) as a novel therapeutic target in systemic mastocytosis. Blood. 2015;126(26):2832-2841. Alvarez-Twose I, Martinez-Barranco P, Gotlib J, et al. Complete response to gemtuzumab ozogamicin in a patient with refractory mast cell leukemia. Leukemia. 2016;30(8):1753-1756. Ghanim V, Herrmann H, Heller G, et al. 5azacytidine and decitabine exert proapoptotic effects on neoplastic mast cells: role of FAS-demethylation and FAS re-expression, and synergism with FAS-ligand. Blood. 2012;119(18):4242-4252.

1043


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Myeloid Leukemia

Ferrata Storti Foundation

Haematologica 2017 Volume 102(6):1044-1053

Higher HOPX expression is associated with distinct clinical and biological features and predicts poor prognosis in de novo acute myeloid leukemia Chien-Chin Lin,1,2,3 Yueh-Chwen Hsu,3 Yi-Hung Li,2 Yuan-Yeh Kuo,4 Hsin-An Hou,2 Keng-Hsueh Lan,5 Tsung-Chih Chen,2 Yi-Shiuan Tzeng,4 Yi-Yi Kuo,2 Chein-Jun Kao,2 Po-Han Chuang,2 Mei-Hsuan Tseng,2 Yu-Chiao Chiu,6 Wen-Chien Chou1,2 and Hwei-Fang Tien2

Department of Laboratory Medicine; 2Division of Hematology and Department of Internal Medicine; 3Graduate Institute of Clinical Medicine; 4Graduate Institute of Oncology, College of Medicine; 5Division of Radiation Oncology and Department of Oncology and 6Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei, Taiwan 1

ABSTRACT

H

Correspondence: wchou@ntu.edu.tw or f99945006@ntu.edu.tw or hftien@ntu.edu.tw Received: November 30, 2016. Accepted: March 17, 2017. Pre-published: March 24, 2017. doi:10.3324/haematol.2016.161257 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1044 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.

1044

omeodomain-only protein homeobox (HOPX) is the smallest homeodomain protein. It was regarded as a stem cell marker in several non-hematopoietic systems. While the prototypic homeobox genes such as the HOX family have been well characterized in acute myeloid leukemia (AML), the clinical and biological implications of HOPX in the disease remain unknown. Thus we analyzed HOPX and global gene expression patterns in 347 newly diagnosed de novo AML patients in our institute. We found that higher HOPX expression was closely associated with older age, higher platelet counts, lower white blood cell counts, lower lactate dehydrogenase levels, and mutations in RUNX1, IDH2, ASXL1, and DNMT3A, but negatively associated with acute promyelocytic leukemia, favorable karyotypes, CEBPA double mutations and NPM1 mutation. Patients with higher HOPX expression had a lower complete remission rate and shorter survival. The finding was validated in two independent cohorts. Multivariate analysis revealed that higher HOPX expression was an independent unfavorable prognostic factor irrespective of other known prognostic parameters and gene signatures derived from multiple cohorts. Gene set enrichment analysis showed higher HOPX expression was associated with both hematopoietic and leukemia stem cell signatures. While HOPX and HOX family genes showed concordant expression patterns in normal hematopoietic stem/progenitor cells, their expression patterns and associated clinical and biological features were distinctive in AML settings, demonstrating HOPX to be a unique homeobox gene. Therefore, HOPX is a distinctive homeobox gene with characteristic clinical and biological implications and its expression is a powerful predictor of prognosis in AML patients.

Introduction HOPX was first identified in the expression sequence tag database for transcripts encoding proteins related to the development of the heart in mice and zebrafish.1,2 Human HOPX, located in chromosome 4q12, has five isoforms. The predominant one encodes a short protein of 73 amino acids with a molecular weight of 12 kd. This is by far the smallest homeodomain protein, with conservation of the homeodomain of 60 amino acids. But the difference between this and other homeodomain proteins is that the HOPX protein does not bind DNA directly. Rather, it exerts its transcriptional inhibition through sequestration of serum responsive factor haematologica | 2017; 102(6)


HOPX in AML

A

C

E

B

D

F

Figure 1. HOPX expression levels and acute myleoid leukemia (AML) patientsâ&#x20AC;&#x2122; survival. (A and B) In NTUH cohort, the overall survival (OS) and disease-free survival (DFS) of AML patients with higher HOPX expression are significantly shorter than those with lower expression: median OS 23.7 versus 116.8 months, P<0.001; median DFS 5.9 months versus not reached (NR); P<0.001. (C and D) The observation is validated by TCGA (median OS 11.2 months vs. 24.4 months; P<0.001) and GSE12417 (median OS 7.8 months vs. 33.3 months; P<0.001) cohorts. (E and F) When we restrict the analysis in non-acute promyelocytic leukemia (APL) patients or patients with a normal karyotype in NTUH cohort, HOPX levels still significantly correlate with OS: median OS 23.7 versus 108.1 months (P=0.0003) and median OS 39.1 versus 116.8 months (P=0.004), respectively. Green line: higher HOPX expression group; blue line: lower HOPX expression group.

by physical interaction and by recruitment of histone deacetylase.3 Recently, the HOPX gene has been regarded as a stem cell marker in intestine, hair follicles, and pulmonary alveolar cells.4-7 Some studies suggested a role for HOPX in tumorigenesis with clinical implications. HOPX has been suggested to be a tumor suppressor gene in lung, colon, esophagus, pancreas, uterine and stomach cancers.8-13 Most of the studies showed silencing of HOPX through hypermethylation of the promoter as a mechanism of its downregulation in cancer cells.9-13 However, the mechanisms for the tumor suppression remain largely unknown. The HOX family and HOPX belong to homeobox genes. However, while HOX family genes have been well studied in acute myeloid leukemia (AML),14 the clinical and biological significance of HOPX in human hematopoiesis remains undefined. We are interested in exploring the roles of HOPX in AML patients, as well as comparing HOX and HOPX in the pathophysiology of malignant hematopoiesis. In this study, we compared clinical and biological characteristics between de novo AML patients with higher and lower HOPX expression and found that higher HOPX expression was strictly correlated with unfavorable prognosis of AML patients in ours and the other two independent cohorts. Multivariate analysis revealed higher HOPX expression as an independent unfavorable prognostic factor, and independent of several published gene signatures derived from multiple cohorts. Using bioinformatics approaches, we found that HOPX expression was closely associated with known hematopoietic stem cell (HSC) signatures. While HOPX haematologica | 2017; 102(6)

and HOX family genes were highly expressed in normal hematopoietic stem/progenitor cells (HSPC), the expression patterns and associated clinical and biological features between these two classes of homeobox genes differed dramatically in AML settings. Taken together, our study suggests that HOPX has a significant impact on various clinical and biological aspects of AML, and that there is a distinction between HOX family genes and HOPX in the AML setting.

Methods Patients A total of 347 adult patients diagnosed with de novo AML according to the 2008 World Health Organization classification in the National Taiwan University Hospital (NTUH) who had cryopreserved bone marrow (BM) cells and complete clinical and laboratory data available for analysis were retrospectively enrolled. Among them, 227 patients received standard induction chemotherapy. Non-M3 (acute promyelocytic leukemia, APL) patients received idarubicin 12 mg/m2 per day for 2-3 days and cytarabine 100 mg/m2 per day for 5-7 days, as described previously.15 APL patients received concurrent all-trans retinoic acid and idarubicin. The remaining 120 patients received palliative therapy with supportive care or low-dose chemotherapy due to underlying comorbidity or in accordance with patient decision. We also prospectively enrolled another cohort of 56 newly diagnosed adult de novo AML patients with adequate BM samples for more detailed studies of the HOPX gene, including expression pattern of HOPX isoforms in AML. The study was approved by the Research Ethics Committee of the NTUH. 1045


C.-C. Lin et al. Table 1. Comparison of clinical manifestations between acute myeloid leukemia patients with higher and lower HOPX expression.

Variables

Total (n=347)

Sex† Male Female Age (years)‡ Lab data‡ WBC (×109/L) Hb (g/dL) Platelet (×109/L) Blast (×109/L) LDH (U/L) FAB* M0 M1 M2 M3 M4 M5 M6 Undetermined Induction response** CR PR+refractory Induction death

Higher HOPX expression (n=174)

Lower HOPX expression (n=173)

P 0.914

196 151

99 75 60 (15-91)

97 76 53 (18-88)

14.5 (0.6-341.4) 8.2 (3.3-13.0) 55.5 (6-655) 6.5 (0.0-283.2) 794 (202-7734)

25.1 (0.4-423.0) 8.0 (3.7-16.2) 41.0 (2-412) 10.8 (0.0-369.1) 1042 (242-13130)

5 (83.3) 42 (62.7) 48 (44.0) 4 (14.3) 58 (56.3) 4 (20.0) 7 (87.5) 6 103 60 (58.3) 35 (34.0) 8 (7.8)

1 (16.7) 25 (37.3) 61 (56.0) 24 (85.7) 45 (43.7) 16 (80.0) 1 (12.5) 0 124 106 (85.5) 10 (8.1) 8 (6.5)

6 67 109 28 103 20 8 6 227 166 (73.1) 45 (19.8) 16 (7.0)

0.023 0.011 0.911 0.008 0.182 <0.001 <0.001 0.099 0.021 0.104 <0.001 0.126 0.006 0.032 <0.001 <0.001 <0.001 0.702

Number of patients.‡Median (range). *Number of patients (% with higher or lower HOPX expression in the AML subtype). **Number of patients (% in the total patients or subgroup of patients with higher or lower HOPX expression). LDH: lactate dehydrogenase; CR: complete remission; PR: partial remission.

Cytogenetic and mutation analysis 16

Chromosomal abnormalities and mutation analyses were performed as previously described.15-20

Gene expression microarray datasets and data analysis We profiled global gene expression of BM mononuclear cells from the 347 patients (NTUH dataset) using Illumina HumanHT12 v.4 Expression BeadChip (Illumina, San Diego, CA, USA) (GSE68469 and GSE71014).21-23 Two large microarray datasets of AML with overall survival (OS) data, including The Cancer Genome Atlas (TCGA) dataset (n=186)24 and GSE12417 [all with cytogenetically normal (CN) AML; n=162],25 were utilized to validate the prognostic significance of HOPX. We used TCGA-normalized level-2 intensity and GSE12417 GPL96 data (profiled with Affymetrix Human Genome U133A Array), normalized as described by Metzeler et al.25 Gene expression profiles GSE12662 (n=91),26 GSE24006 (n=54),27 and GSE24759 (n=211)28 were also included to investigate the gene expression patterns in normal hematopoiesis.

Analysis of gene expression in next-generation sequencing datasets To investigate the absolute levels of gene expression in AML, we analyzed expression data of 179 AML samples profiled with Illumina Genome Analyzer RNA Sequencing in TCGA dataset.24 Reads per kilobase per million mapped reads (RPKM) levels of gene expression were extracted from TCGA database.24

www.broadinstitute.org/gsea/index.jsp)29 and as detailed in the Online Supplementary Appendix. In order to examine whether genes are involved in HSC quiescence, we employed another gene set enrichment scoring method that averages z-values of all involved genes.30

Methylation microarray datasets and analysis DNA methylation data from Illumina Infinium HumanMethylation450 BeadChips of AML (n=194) were downloaded from the TCGA database.24 We transformed methylation beta-values to normally distributed M-values for further analysis.31

Expression of HOPX isoforms Human HOPX has five isoforms including HOPXa (NM_032495), HOPXb (three variants including NM_139212, NM_139211 and NM_001145459; abbreviated hereafter as b1, b2, and b3, respectively), and HOPXc (NM_001145460) (UCSC genomic database; www.genome.ucsc.edu) (Online Supplementary Figure S1). Analysis of HOPX isoform expression was performed by quantitative real time-polymerase chain reaction as detailed in the Online Supplementary Appendix, Online Supplementary Table S1 and Online Supplementary Figure S1.

Bisulfite treatment and methylation analysis of HOPX We interrogated the methylation status of the CpG islands of HOPX-b2 isoform NM_139211 from -15 to +109 bp region around the transcription start site (TSS).10 Methods are described in the Online Supplementary Appendix.

Gene signature analysis The association of HOPX gene with stem cell characteristics was analyzed by the Gene Set Enrichment Analysis (GSEA; a Java application that can be down-loaded at http:// 1046

Statistical analysis Statistical analysis was carried out as described previously;21-23 a brief description is available in the Online Supplementary Appendix. haematologica | 2017; 102(6)


HOPX in AML

Table 2. Association of HOPX expression levels with other genetic alterations.

Mutation

FLT3/ITD FLT3/TKD N-RAS K-RAS PTPN11 KIT DNMT3A WTI NPM1 CEBPAdouble mutation RUNX1 MLL/PTD ASXL1 IDH1 IDH2 TP53 TET2

Whole cohort (n=347)

N. patients with alteration (%) Higher HOPX expression (n=174)

Lower HOPX expression (n=173)

P

84/347 (24.2) 32/347 (9.2) 59/347 (17.0) 15/347 (4.3) 22/347 (6.3) 15/347 (4.3) 66/347 (19.0) 26/347 (7.5) 99/347 (28.5) 27/347 (7.8) 50/347 (14.4) 13/346 (3.8) 52/347 (15.0) 20/347 (5.8) 51/347 (14.7) 16/346 (4.6) 56/347 (16.1)

38/174 (21.8) 13/174 (7.5) 27/174 (15.5) 5/174 (2.9) 11/174 (6.3) 4/174 (2.3) 41/174 (23.6) 12/174 (6.9) 41/174 (23.6) 5/174 (2.9) 39/174 (22.4) 5/173 (2.9) 34/174 (19.5) 10/174 (5.7) 37/174 (21.3) 10/173 (5.8) 23/174 (13.2)

46/173 (26.6) 19/173 (11.0) 32/173 (18.5) 10/173 (5.8) 11/173 (6.4) 11/173 (6.4) 25/173 (14.5) 14/173 (8.1) 58/173 (33.5) 22/173 (12.7) 11/173 (6.4) 8/173 (4.6) 18/173 (10.4) 10/173 (5.8) 14/173 (8.1) 6/173 (3.5) 33/173 (19.1)

0.302 0.258 0.460 0.183 0.989 0.063 0.031 0.672 0.040 0.001 <0.001 0.424 0.017 0.989 0.001 0.306 0.138

Results

ble mutations (P=0.001) and NPM1 mutation (P=0.040) (Table 2).

Correlation of HOPX expression with clinical features The 347 AML patients were divided into two groups based on the HOPX expression levels above (higher expression group) or below (lower expression group) the median level of HOPX expression on the arrays. Higher HOPX expression was associated with older age (P=0.023), higher platelet counts (P=0.008), lower white blood cell (WBC) counts (P=0.011), and lower lactate dehydrogenase (LDH) levels (P<0.001) at diagnosis (Table 1). Patients with M1 and M6 according to the FrenchAmerican-British (FAB) classification more frequently had higher HOPX expression (P=0.021 and P=0.032, respectively), while those with M3 and M5 had significantly lower levels of HOPX expression (P<0.001 and P<0.006, respectively). The comparison of clinical features between higher and lower HOPX expression groups in those receiving standard chemotherapy (n=227) is shown in the Online Supplementary Table S2. The association of higher HOPX expression with higher platelet counts, lower LDH levels, and FAB subtypes remained the same in this group of patients as that of the total cohort.

Correlation of HOPX expression with cytogenetics and molecular alterations Chromosome data were available in 325 patients at diagnosis (Online Supplementary Table S3). Higher HOPX expression was negatively associated with favorable karyotypes, including t(8;21) and t(15;17) (both P<0.001). We also analyzed the mutation status of 16 genes and found that the patients with higher HOPX expression had significantly higher incidences of mutations in RUNX1 (P<0.001), IDH2 (P=0.001), ASXL1 (P=0.017), and DNMT3A (P=0.031), but less frequently had CEBPA douhaematologica | 2017; 102(6)

Higher HOPX expression predicts poor clinical outcome in de novo AML patients Among the 227 patients who received standard chemotherapy, those with higher HOPX expression had a lower complete remission (CR) rate (58.3% vs. 85.5%; P<0.001) (Table 1), shorter OS (median 23.7 months vs. 116.8 months; log-rank P<0.001) and disease-free survival (DFS) (median 5.9 months vs. not reached; log-rank P<0.001) than those with lower HOPX expression after a median follow up of 57.0 months (Figure 1A and B). Univariate Cox proportional hazards analysis confirmed the prognostic value of HOPX expression as a continuous variable in predicting patientsâ&#x20AC;&#x2122; OS [Hazard Ratio (HR): 1.44; 95%CI: 1.21-1.71; P<0.001] and DFS (HR: 1.55; 95%CI: 1.31-1.82; P<0.001). The prognostic significance of HOPX expression could be validated in another two independent cohorts: TCGA 24 and GSE1241725 (Figure 1C and D). The unfavorable prognostic effects of higher HOPX expression were also seen in the subgroup of patients with AML other than APL (median OS 23.7 vs. 108.1 months; P<0.001) and those with a normal karyotype (median OS 39.1 vs. 116.8 months; P=0.004) (Figure 1E and F). The results could also be validated by the TCGA cohort (Online Supplementary Figure S2A and B). By univariate analysis, HOPX expression levels and several parameters exhibited a significant impact on OS (Online Supplementary Table S4). When we combined all these prognostic factors together in a multivariate analysis, higher expression of HOPX remained a poor prognostic factor for OS (P=0.005) (Table 3), independent of age, WBC counts, karyotypes, mutation statuses of FLT3, 1047


C.-C. Lin et al. A

B

C

D

E

F

Figure 2. HOPX expression levels, its correlation with treatment response, and its role as a stem cell marker by gene set enrichment analysis. (A) There was a significant difference in HOPX levels between patients with (n=166) and without (n=61) complete remission (CR) after induction chemotherapy. (B) HOPX expression is also much higher in non-APL patients. (C) Real-time PCR of HOPX in the cohort of 56 acute myeloid leukemia (AML) patients prospectively recruited showing predominant expression of isoform HOPXb2 (NM_139211) (left, isoforms a+b1+c; middle, b2; right, b3). (D, E, and F) GSEA plots of curated HSC and LSC signatures in the NTUH dataset.35,36 Red-to-indigo bars denote the genome-wide gene list ranked based on their P-values (t-test) between samples with high (top quartile) and low (bottom quartile) expression of HOPX. Significant positive GSEA enrichment scores indicate that HOPX expression is positively associated with HSC and LSC signatures.

Table 3. Multivariate analysis (Cox regression) on overall survival.*

CEBPA, MLL, TP53, WT1, and RUNX1 and expression levels of HOXA9. Further analysis showed much higher HOPX expression in those patients who failed to achieve CR than those who obtained a CR (by array signal intensity; P<0.001) (Figure 2A) suggesting a tight association of higher HOPX expression and drug resistance. Furthermore, HOPX expression was lower in APL, which consisted mainly of leukemic cells that are blocked at the differentiation stage of promyelocytes, indicating a possible relationship between HOPX expression and maturation stages of AML cells (Figure 2B).

Comparisons between HOPX expression and published prognostic gene signatures in predicting prognosis Several gene expression-based prognostic predictors have been developed from various study designs in AML. To compare the performance of prognostic prediction of HOPX expression with those published predictors, we performed pairwise multivariate Cox analysis between HOPX expression and each of the 3-gene, 7-gene, 11-gene, and 24-gene predictors in three datasets.21,32-34 Remarkably, the HOPX expression remained independent (with Cox multivariate analysis P<0.05) in most of the comparison settings (11 of 12 comparisons) (Table 4). Our data suggest HOPX to be a simple and powerful alternative for prognostication in AML.

The expression pattern and promoter methylation of HOPX isoforms in AML patients The pattern of expression of the 5 isoforms of HOPX 1048

Variables HR Age WBC/1000 Karyotype FLT3-ITD CEBPAdouble mutation RUNX1 MLL-PTD WT1 TP53 HOPX HOXA9

1.017 1.004 3.725 1.522 0.299 1.542 3.150 1.804 3.085 1.172 1.142

Overall survival 95% CI Lower Upper Total cohort (n=227) 1.002 1.001 2.273 0.968 0.114 0.849 1.438 0.993 1.151 1.050 0.815

1.031 1.006 6.105 2.391 0.785 2.800 6.902 3.278 8.267 1.307 1.600

P 0.021 0.012 <0.001 0.069 0.014 0.155 0.004 0.053 0.025 0.005 0.441

*The model was generated from a stepwise Cox regression model that included age, WBC, karyotype (unfavorable cytogenetics vs. others), gene mutations of FLT3, WT1, CEBPA, RUNX1, MLL, TP53 and expression level of HOXA9 and HOPX. HR: Hazard Ratio; CI: Confidence Interval.

and CpG methylation status in primary AML are still unknown. Because of the low levels of expression and the impossibility of separating isoforms a, b1, and c, we merged these three together for quantification. We quantified the expression levels of these isoforms in prospectively recruited AML patientsâ&#x20AC;&#x2122; marrow by real time-PCR and found that HOPXb2 (NM_139211) was the predominant isoform in human AML cells; the other four variants haematologica | 2017; 102(6)


HOPX in AML

A

B

Figure 3. Comparison of expression patterns, clinical and biological features among subgroups of NTUH AML patients based on hierarchical clustering of HOPX and HOX family genes. (A) Heatmap of HOPX and HOX family genes in NTUH data. We identify 4 groups of patients based on a 2-step hierarchical clustering: HOXhigh/HOPXlow by HOPXhigh/HOPXlow. Molecular and clinical variables including gene mutations, cytogenetic abnormalities, and leukemia classifications are compared among the clusters. (B) Average z-scores of HOPX and HOX genes in each group. Whiskers denote averageÂąstandard deviation of the z-scores. Statistical significance of the z-scores against zero is assessed by the 1-sample t-test (****P<0.0001). (C) Overall survival of the 4 groups of patients. HOXlow/HOPXlow patients have the best overall survival (OS), followed by HOXlow/HOPXhigh, HOXhigh/HOPXlow, and HOXhigh/HOPXhigh (P<0.001).

were markedly under-represented (Figure 2C). In addition, quantification of the CpG methylation in HOPXb2 promoter regions revealed low levels of methylation in most AML patients (Online Supplementary Figure S3). This finding seemed to differ from that in some studies of solid cancers in which hypermethyation of HOPX in this region caused gene silencing and was associated with poor prognosis.9-13 Further studies are needed to confirm this hypothesis.

Correlation of HOPX expression with stem cell signatures We curated xenotransplantation-derived HSC and leukemia stem cell (LSC) gene signatures and a recently up-dated LSC signature from previous reports35,36 and employed the gene set enrichment analysis (GSEA) method29 to analyze their associations with HOPX expression. GSEA tests the enrichment of each gene signature in the list of global genes ranked by HOPX-associated differential expression. Higher HOPX expression was associated with upregulation of HSC and LSC genes in the NTUH dataset (enrichment scores, 0.79, 0.61, and 0.89; P<0.0005, P=0.006, and P<0.0005, respectively) (Figure 2D-F). Concordant significant enrichments were identified in TCGA and GSE12417 AML datasets (all P-values â&#x2030;¤0.015) (Online Supplementary Figure S4). Seventeen and five genes appeared as leading-edge genes (ie. a subset of coreenrichment genes) of the HSC and LSC signatures (Figure 2D and E), respectively, in all the three cohorts (Online Supplementary Table S5). Interestingly, an ATP-bindingcassette (ABC) transporter gene, ABCB1, was a common leading-edge gene of HSC signature (Online Supplementary Table S5). ABC transporter genes were reported to be assohaematologica | 2017; 102(6)

ciated with chemoresistance in AML, with higher ABCB1, ABCG1, ABCG2 expression levels being independently poor prognostic factors.37 Expression of these three ABC genes was significantly higher in samples with higher HOPX expression (all P-values <0.001) (Online Supplementary Table S6), but the mechanistic link between ABC and HOPX expression still has to be explored in further studies.

Expression patterns of HOPX and HOX genes in normal hematopoietic cells HOPX and HOX family genes all encode homeodomain proteins and HOX genes are well-known HSC markers.3840 To further delineate the similarities and the distinctions between HOPX and the HOX gene family, we first analyzed their expression patterns in normal hematopoietic cells using arrays derived from public data. We curated three public gene expression datasets derived from normal hematopoietic cells26-28 and we chose 12 HOX genes with at least moderate expression levels (RPKM> 5 according to TCGA RNA seq data) for further analysis.24 In GSE2400627 and GSE12662,26 HOPX and HOX genes were generally expressed in a concordant manner (mean correlation coefficient 0.37 and 0.35, respectively) (left panels, Online Supplementary Figure S5A and B). We further analyzed a dataset of 9 distinct normal hematopoietic cell populations (GSE24759) (Online Supplementary Figure S6).28 HOPX and HOX family genes were all highly expressed in normal CD34+ hematopoietic cells (average z-values = 0.82 and 0.75; 1-sample t-test both P<0.001) (Online Supplementary Figure S6). The concordant expression patterns between HOPX and HOX family suggest their shared roles in normal hematopoiesis. 1049


C.-C. Lin et al.

Expression patterns of HOPX and HOX family genes in AML We investigated the absolute gene expression levels of HOPX and the HOX family in AML from the TCGA RNA sequencing dataset. Among them, HOPX was the second highest expressed gene (average RPKM = 25.6 in TCGA RNA sequencing dataset; n=179), after the most abundant gene HOXA9 (RPKM = 43.3). We then compared the expression patterns between HOPX and HOX family genes in AML cells. The concordance of expression patterns shown in normal hematopoietic cells were no longer present in AML cells (correlation in GSE24006 and GSE12662 -0.31 and 0.07, respectively) (Online Supplementary Figure S5A and B). We sought to investigate the similarities/distinctions between HOPX and HOX family genes by clustering of AML patients in our dataset (NTUH) according to their expression levels. Because of the unequal numbers of genes between HOPX and HOX family (1 vs. 12), we performed a 2-step hierarchical clustering to balance the potential bias in unsupervised clustering. Briefly, patients were first clustered only by the 12 HOX family genes. Subsequently, each cluster was subject to the second round of clustering with inclusion of HOPX. As a result, we were able to identify and focus on 4 distinct groups of patients for further analysis (HOXhigh/HOXlow by HOPXhigh/HOPXlow) (Figure 3A) in whom the high/low expressions of HOX and HOPX were confirmed significant in each cluster (comparisons of average z-scores against zero; 1-sample t-test P<0.0001) (Figure 3B). The 4 groups also showed a significantly different prognosis: HOXlow/HOPXlow patients had the longest OS, while HOXhigh/HOPXhigh patients had the poorest outcome (P<0.0001) (Figure 3C). To further compare the clinical and biological characteristics among AML patients with different expression levels of HOPX/HOX family genes, we analyzed patients' gene mutations, cytogenetic abnormalities, and other clinical and lab parameters. WBC counts and LDH levels varied significantly among groups (ANOVA P=0.026 and 0.0009, respectively) (Figure 4A and B). HOXlow/HOPXhigh patients had the lowest WBC counts and LDH levels (median 5765/ÎźL vs. 24,720/ÎźL and 643 U/L vs. 1027 U/L, respectively; both P<0.001). Each subgroup also had distinct biological characteristics, including CD34 expression, gene mutations of FLT3, NPM1, CEBPA, RUNX1, DNMT3A, IDH1/2 (all P<0.0001) and ASXL1 (P=0.0002) (Figure 3A and 4C). FLT3-ITD and mutations in NPM1 and DNMT3A are more common in HOXhigh patients regardless of HOPX expression levels; RUNX1 mutation is more frequent in HOPXhigh regardless of HOX expression levels; CEBPA double mutation is predominantly seen in HOXlow/HOPXlow patients; ASXL1 mutation is mainly present in HOXlow/HOPXhigh subgroup; IDH1/2 mutations are particularly rare in HOXlow/HOPXlow patients; CD34+ blasts are low in HOXhigh/HOPXlow patients. Compared with other patients, HOXhigh/HOPXhigh patients had higher incidences of FAB M0 (4 of 76 vs. 1 of 185; P=0.011), CD34 expression on leukemic cells (64 of 71 vs. 104 of 175; P<0.001), and mutations in ASXL1 (23 of 76 vs. 16 of 185; P<0.001), RUNX1 (21 of 76 vs. 15 of 185; P<0.001), and IDH1/2 (25 of 76 vs. 30 of 185; P=0.003), while HOXhigh/HOPXlow patients, when compared with others, had more FAB M5 (8 of 57 vs. 2 of 204; P<0.001) and mutations in NPM1 (47 of 57 vs. 19 of 204; P<0.001), FLT3 (FLT3-ITD) (26 of 57 vs. 38 of 204; P<0.001), MLL (MLL-PTD) (5 of 57 vs. 2 of 203; P=0.001), PTPN11 (8 of 1050

Table 4. Comparisons of HOPX to published prognostic gene signatures.

Predictor HOPX

NTUH (n=227)

<0.001* (1.39;1.17-1.65) 3-gene score 0.001 (Wilop et al.32) (1.46;1.17-1.82) HOPX 0.048 (1.22;1.00-1.49) 7-gene score 0.001 (Marcucci et al.33) (1.23;1.09-1.40) HOPX 0.086 (1.20;0.98-1.47) 11-gene score 0.001 (Chuang et al.21) (1.06;1.03-1.10) HOPX 0.011 (1.27;1.06-1.52) 24-gene score <0.001 (Li et al.34) (1.10;1.05-1.16)

TCGA (n=186)

GSE12417 (n=162)

0.003 (1.34;1.11-1.61) 0.005 (1.34;1.10-1.65) 0.017 (1.30;1.05-1.62) 0.305 (1.07;0.94-1.20) 0.014 (1.33;1.06-1.68) 0.597 (1.01;0.96-1.07) 0.032 (1.22;1.02-1.47) <0.001 (1.12;1.06-1.19)

<0.001 (1.52;1.22-1.89) 0.601 (1.06;0.85-1.33) 0.002 (1.43;1.14-1.78) 0.132 (1.13;0.96-1.33) 0.009 (1.35;1.08-1.68) 0.010 (1.05;1.01-1.10) <0.001 (1.47;1.21-1.80) 0.051 (1.04;1.00-1.08)

*Multivariate P-value [Hazard Ratio (HR); 95% Confidence Interval of HR] comparing HOPX expression levels and the respective gene scoring system.

57 vs. 8 of 203; P=0.005), and WT1 (8 of 57 vs. 11 of 204; P=0.026) (Figure 3A and data not shown). These results demonstrated marked distinctions between HOPX and HOX family genes in their association with genetic alterations and clinical features in AML.

Distinct associated HSC gene signatures between HOPX and HOX family genes Although all HOPX and HOX family genes are stem cell markers, our data showed that the two were associated with distinct features in AML. Stem cell signatures could be divided into two groups with either quiescence or proliferation characteristics.41 The low LDH levels and WBC counts in the AML patients with high HOPX and low HOX family gene expression (Figure 4A and B) raises the possibility that expression of HOPX may favor quiescence of stem cells. To test this hypothesis, we examined the expression profiles of each subgroup of our AML patients by a gene set scoring30 based on a known quiescence signature in HSC.41 A positive signature score denotes a tendency toward a quiescent HSC state. The significance level of a score against zero (representing no tendency) was tested by 1-sample t-test. Patients with high expression of HOX family genes did not exhibit a significant tendency toward the quiescence state regardless of the abundance of HOPX expression (P=0.22 and 0.45 for HOXhigh/HOPXhigh and HOXhigh/HOPXlow, respectively) (Figure 4D). However, when HOX expression is low, HOPXhigh and HOPXlow were significantly associated with quiescence and non-quiescence, respectively (P=0.0007 and 0.0003, respectively) (Figure 4D). Overall, our data suggested a fundamental difference between HOPX and HOX family genes in their stem cell properties in the AML setting.

Comparison of methylation patterns between HOPX and HOX family genes in AML Besides the different expression patterns and associated haematologica | 2017; 102(6)


HOPX in AML

A

B

D

C

Figure 4. The distinct clinical and genetic features among acute myeloid leukemia (AML) patients stratified by expression of HOPX and HOX family genes. (A and B) Box plots of white blood cell (WBC) counts and serum lactate dehydrogenase (LDH) levels. There is a significant difference in both variables among the 4 clusters. The highest median of WBC count appears in HOXhigh/HOPXlow group, while that of HOXlow/HOPXhigh falls to the bottom of all clusters. Similar trends are seen in LDH levels. Statistical significance were tested by ANOVA tests. (C) The 4 clusters are associated with distinctive gene mutations. FLT3-ITD and mutations in NPM1 and DNMT3A are more common in HOXhigh patients regardless of HOPX expression levels; the RUNX1 mutation is more frequent in HOPXhigh regardless of HOX expression levels; the CEBPA double mutation is predominantly seen in HOXlow/HOPXlow patients; the ASXL1 mutation is mainly present in HOXlow/HOPXhigh subgroup; IDH1/2 mutations are particularly rare in HOXlow/HOPXlow patients. CD34+ blasts are low in HOXhigh/HOPXlow patients. (D) Association of quiescence HSC signature with expression of HOPX and HOX family genes. We employ a gene set enrichment scoring to quantify the overall activity of the gene signature in each sample; a positive/negative score represents a tendency toward quiescent/proliferative HSC state. Inter-group changes are tested by an ANOVA test. Asterisks denote significant differences from zero assessed by one-sample t-tests (***P<0.001).

clinical and biological features in HOPX compared with HOX family genes, we also sought to find out whether there was any difference in their methylation patterns by analyzing the TCGA epigenome-wide microarray dataset (n=194). Generally, these genes formed 3 clusters according to the methylation patterns (Online Supplementary Figure S7A). HOPX was largely unmethylated in most AML patients (mean methylation M-value -1.63; 1-sample t-test against zero P<0.001; area under curve with negative M 69.05%) (Online Supplementary Figure S7B). Methylation levels of HOXA3, HOXA4, HOXA5, and HOXB3 were generally high, while other HOX genes, including HOXA7, HOXA9, and HOXB4, were uniformly hypomethylated (Online Supplementary Figure S7A and B). Taken together, our data highlighted the different molecular and clinical features that distinguish between HOPX and HOX family genes in AML.

Discussion To our knowledge, this is the first report regarding the prognostic significance of HOPX expression in de novo AML patients and the direct comparison between HOPX and the HOX family in normal and malignant hematopoiesis. The prognostic significance of HOPX expression is independent of common known clinical and haematologica | 2017; 102(6)

genetic factors as well as several published gene signatures. We also showed that the promoter region was barely methylated in leukemic cells from AML patients, in contrast to heavy methylation in solid cancers,8-10,12,13,42 indicating that CpG methylation is not one of the main mechanisms of regulation of HOPX gene expression in primary human AML cells. Finally, HOPX appeared to be a distinct homeobox gene in AML cells when compared with HOX family genes. Studies have shown that HOPX is a stem cell marker of hair follicle, intestine, and lung alveolar cells.5-7 Through functional annotation, our current study showed that HOPX expression was associated with HSC and LSC signatures in AML cells from our cohort and also two other validation cohorts, indicating that HOPX was an LSC marker in AML. Stemness is an established property pertaining to drug resistance and poor prognosis in cancer patients.43 LSC signature is associated with unfavorable prognosis in AML patients. The underlying mechanisms by which stem cell signatures in AML cells predict poor treatment outcome have been postulated to be related to their association with chemotherapy resistance,27 probably due to the quiescent nature of these cells. The tight association between HOPX expression and stem cell properties is likely a major reason for the unfavorable prognosis in AML patients with higher HOPX expression shown in this study. In addition, higher HOPX expression was sig1051


C.-C. Lin et al.

nificantly associated with the expression of some ABC transporters, a family of proteins that bind ATP as energy source to transport the endogenous or exogenous molecules through the cell membranes.44,45 They are abundant in stem cells, including HSCs and LSCs, and are responsible for multidrug resistance in cancer treatment.46 Therefore, leukemia patients with higher HOPX are less likely to obtain CR after induction chemotherapy. Further functional studies are needed to throw light on its significance in leukemia stemness and drug-resistance. We showed that HOPX had distinct expression pattern and associated clinical and biological features when compared with other homeobox genes such as the HOX family in the AML setting. While they were both enriched in normal CD34+ HSPCs, their expression in AML was asynchronous. The findings that higher HOPX expression, accompanied with lower HOX expression, was closely associated with FAB M0 subtype, CD34 expression on leukemic cells, lower WBC counts, LDH levels, and quiescence stem cell signature indicates its relationship with more immature and quiescent stem cell characters. Our study was mainly based on a retrospective cohort

References 1. Shin CH, Liu ZP, Passier R, et al. Modulation of cardiac growth and development by HOP, an unusual homeodomain protein. Cell 2002;110(6):725-735. 2. Chen F, Kook H, Milewski R, et al. Hop is an unusual homeobox gene that modulates cardiac development. Cell. 2002; 110(6):713-723. 3. Kook H, Yung WW, Simpson RJ, et al. Analysis of the structure and function of the transcriptional coregulator HOP. Biochemistry. 2006;45(35):10584-10590. 4. Takeda N, Jain R, Leboeuf MR, et al. Hopx expression defines a subset of multipotent hair follicle stem cells and a progenitor population primed to give rise to K6+ niche cells. Development. 2013;140(8):1655-1664. 5. Takeda N, Jain R, LeBoeuf MR, Wang Q, Lu MM, Epstein JA. Interconversion between intestinal stem cell populations in distinct niches. Science. 2011;334(6061):1420-1424. 6. Munoz J, Stange DE, Schepers AG, et al. The Lgr5 intestinal stem cell signature: robust expression of proposed quiescent '+4' cell markers. EMBO J. 2012; 31(14):3079-3091. 7. Jain R, Barkauskas CE, Takeda N, et al. Plasticity of Hopx(+) type I alveolar cells to regenerate type II cells in the lung. Nat Commun. 2015;6:6727. 8. Chen Y, Pacyna-Gengelbach M, Deutschmann N, Niesporek S, Petersen I. Homeobox gene HOP has a potential tumor suppressive activity in human lung cancer. Int J Cancer. 2007;121(5):1021-1027. 9. Harada Y, Kijima K, Shinmura K, et al. Methylation of the homeobox gene, HOPX, is frequently detected in poorly differentiated colorectal cancer. Anticancer Res. 2011;31(9):2889-2892. 10. Yamashita K, Kim MS, Park HL, et al. HOP/OB1/NECC1 promoter DNA is frequently hypermethylated and involved in tumorigenic ability in esophageal squamous cell carcinoma. Mol Cancer Res.

1052

although we validated our results from other public array cohorts and 56 prospectively enrolled patients. Further studies in large prospective cohorts are warranted to confirm our observations. Moreover, in vivo studies are necessary to delineate the pathophysiological effects of HOPX in hematopoiesis and leukemogenesis. Acknowledgments The authors would like to thank the FACS Core of National Taiwan University Hospital and Kai-Ting Yang, the sorting technician, for performing cell sorting and FACS analysis. Funding The study was supported by a National Taiwan University Hospitalâ&#x2C6;&#x2019;National Taiwan University joint research grant (UN103-051), Ministry of Science and Technology of Taiwan (MOST102-2325-B-002-028, 103-2314-B-002-130-MY3, 103-2314-B-002-131MY3 and 104-2923-B-002-001), Far Eastern Hospital and NTUH joint grant 105-FTN24, NTUH and NTUMC joint grant UN106-024, and Ministry of Health and Welfare of Taiwan (MOHW102-TD-C-111-001 and MOHW103-TD-B-111-04).

2008;6(1):31-41. 11. Katoh H, Yamashita K, Waraya M, et al. Epigenetic silencing of HOPX promotes cancer progression in colorectal cancer. Neoplasia. 2012;14(7):559-571. 12. Yamaguchi S, Asanoma K, Takao T, Kato K, Wake N. Homeobox gene HOPX is epigenetically silenced in human uterine endometrial cancer and suppresses estrogen-stimulated proliferation of cancer cells by inhibiting serum response factor. Int J Cancer. 2009;124(11):2577-2588. 13. Ooki A, Yamashita K, Kikuchi S, et al. Potential utility of HOP homeobox gene promoter methylation as a marker of tumor aggressiveness in gastric cancer. Oncogene. 2010;29(22):3263-3275. 14. Alharbi RA, Pettengell R, Pandha HS, Morgan R. The role of HOX genes in normal hematopoiesis and acute leukemia. Leukemia. 2013;27(5):1000-1008. 15. Tang JL, Hou HA, Chen CY, et al. AML1/RUNX1 mutations in 470 adult patients with de novo acute myeloid leukemia: prognostic implication and interaction with other gene alterations. Blood. 2009;114(26):5352-5361. 16. Chou WC, Chou SC, Liu CY, et al. TET2 mutation is an unfavorable prognostic factor in acute myeloid leukemia patients with intermediate-risk cytogenetics. Blood. 2011;118(14):3803-3810. 17. Chou WC, Tang JL, Lin LI, et al. Nucleophosmin mutations in de novo acute myeloid leukemia: the age-dependent incidences and the stability during disease evolution. Cancer Res. 2006; 66(6):3310-3316. 18. Hou HA, Lin CC, Chou WC, et al. Integration of cytogenetic and molecular alterations in risk stratification of 318 patients with de novo non-M3 acute myeloid leukemia. Leukemia. 2014; 28(1):50-58. 19. Chou WC, Huang HH, Hou HA, et al. Distinct clinical and biological features of de novo acute myeloid leukemia with addi-

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

tional sex comb-like 1 (ASXL1) mutations. Blood. 2010;116(20):4086-4094. Chou WC, Hou HA, Chen CY, et al. Distinct clinical and biologic characteristics in adult acute myeloid leukemia bearing the isocitrate dehydrogenase 1 mutation. Blood. 2010;115(14):2749-2754. Chuang MK, Chiu YC, Chou WC, et al. An mRNA expression signature for prognostication in de novo acute myeloid leukemia patients with normal karyotype. Oncotarget. 2015;6(36):39098-39110. Chuang MK, Chiu YC, Chou WC, Hou HA, Chuang EY, Tien HF. A 3-microRNA scoring system for prognostication in de novo acute myeloid leukemia patients. Leukemia. 2015;29(5):1051-1059. Chiu YC, Tsai MH, Chou WC, et al. Prognostic significance of NPM1 mutationmodulated microRNA-mRNA regulation in acute myeloid leukemia. Leukemia. 2016;30(2):274-284. Network CGAR. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Eng J Med. 2013;368(22):20592074. Metzeler KH, Hummel M, Bloomfield CD, et al. An 86-probe-set gene-expression signature predicts survival in cytogenetically normal acute myeloid leukemia. Blood. 2008;112(10):4193-4201. Payton JE, Grieselhuber NR, Chang LW, et al. High throughput digital quantification of mRNA abundance in primary human acute myeloid leukemia samples. J Clin Invest. 2009;119(6):1714-1726. Gentles AJ, Plevritis SK, Majeti R, Alizadeh AA. Association of a leukemic stem cell gene expression signature with clinical outcomes in acute myeloid leukemia. JAMA. 2010;304(24):2706-2715. Novershtern N, Subramanian A, Lawton LN, et al. Densely interconnected transcriptional circuits control cell states in human hematopoiesis. Cell. 2011;144(2):296-309. Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowl-

haematologica | 2017; 102(6)


HOPX in AML

30.

31.

32.

33.

34.

edge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA. 2005;102(43):15545-15550. Hsiao TH, Chiu YC, Hsu PY, et al. Differential network analysis reveals the genome-wide landscape of estrogen receptor modulation in hormonal cancers. Sci Rep. 2016;6:23035. Du P, Zhang X, Huang CC, et al. Comparison of Beta-value and M-value methods for quantifying methylation levels by microarray analysis. BMC Bioinformatics. 2010;11:587. Wilop S, Chou WC, Jost E, et al. A threegene expression-based risk score can refine the European LeukemiaNet AML classification. J Hematol Oncol. 2016;9(1):78. Marcucci G, Yan P, Maharry K, et al. Epigenetics meets genetics in acute myeloid leukemia: clinical impact of a novel seven-gene score. J Clin Oncol. 2014; 32(6):548-556. Li Z, Herold T, He C, et al. Identification of a 24-gene prognostic signature that improves the European LeukemiaNet risk classification of acute myeloid leukemia: an international collaborative study. J Clin Oncol. 2013;31(9):1172-1181.

haematologica | 2017; 102(6)

35. Eppert K, Takenaka K, Lechman ER, et al. Stem cell gene expression programs influence clinical outcome in human leukemia. Nat Med. 2011;17(9):1086-1093. 36. Ng SW, Mitchell A, Kennedy JA, et al. A 17gene stemness score for rapid determination of risk in acute leukaemia. Nature. 2016;540(7633):433-437. 37. Marzac C, Garrido E, Tang R, et al. ATP Binding Cassette transporters associated with chemoresistance: transcriptional profiling in extreme cohorts and their prognostic impact in a cohort of 281 acute myeloid leukemia patients. Haematologica. 2011; 96(9):1293-1301. 38. Argiropoulos B, Humphries RK. Hox genes in hematopoiesis and leukemogenesis. Oncogene. 2007;26(47):6766-6776. 39. Abramovich C, Pineault N, Ohta H, Humphries RK. Hox genes: from leukemia to hematopoietic stem cell expansion. Ann N Y Acad Sci. 2005;1044:109-116. 40. Owens BM, Hawley RG. HOX and nonHOX homeobox genes in leukemic hematopoiesis. Stem Cells. 2002;20(5):364379. 41. Venezia TA, Merchant AA, Ramos CA, et al. Molecular signatures of proliferation and

42.

43. 44.

45.

46.

quiescence in hematopoietic stem cells. PLoS Biol. 2004;2(10):e301. Waraya M, Yamashita K, Katoh H, et al. Cancer specific promoter CpG Islands hypermethylation of HOP homeobox (HOPX) gene and its potential tumor suppressive role in pancreatic carcinogenesis. BMC Cancer. 2012;12:397. Dean M, Fojo T, Bates S. Tumour stem cells and drug resistance. Nat Rev Cancer. 2005; 5(4):275-284. Costello RT, Mallet F, Gaugler B, et al. Human acute myeloid leukemia CD34+/CD38- progenitor cells have decreased sensitivity to chemotherapy and Fas-induced apoptosis, reduced immunogenicity, and impaired dendritic cell transformation capacities. Cancer Res. 2000; 60(16):4403-4411. Dean M, Rzhetsky A, Allikmets R. The human ATP-binding cassette (ABC) transporter superfamily. Genome Res. 2001; 11(7):1156-1166. Ishikawa F, Yoshida S, Saito Y, et al. Chemotherapy-resistant human AML stem cells home to and engraft within the bonemarrow endosteal region. Nat Biotech. 2007;25(11):1315-1321.

1053


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Myeloid Leukemia

Ferrata Storti Foundation

Targeted therapy for a subset of acute myeloid leukemias that lack expression of aldehyde dehydrogenase 1A1

Maura Gasparetto,1 Shanshan Pei,1 Mohammad Minhajuddin,1 Nabilah Khan,1 Daniel A. Pollyea,1 Jason R. Myers,2 John M. Ashton,2 Michael W. Becker,3 Vasilis Vasiliou,4 Keith R. Humphries,5 Craig T. Jordan1 and Clayton A. Smith1

Haematologica 2017 Volume 102(6):1054-1065

Division of Hematology, University of Colorado, Aurora, CO, USA; 2Genomics Research Center, University of Rochester, NY, USA; 3Department of Medicine, University of Rochester Medical Center, NY, USA; 4Department of Environmental Health Sciences, Yale University, New Haven, CT, USA and 5Terry Fox Laboratory, British Columbia Cancer Agency, Vancouver, BC, Canada 1

ABSTRACT

A

Correspondence: maura.gasparetto@ucdenver.edu

Received: November 2, 2016. Accepted: March 8, 2017. Pre-published: March 9, 2017.

ldehyde dehydrogenase 1A1 (ALDH1A1) activity is high in hematopoietic stem cells and functions in part to protect stem cells from reactive aldehydes and other toxic compounds. In contrast, we found that approximately 25% of all acute myeloid leukemias expressed low or undetectable levels of ALDH1A1 and that this ALDH1A1– subset of leukemias correlates with good prognosis cytogenetics. ALDH1A1– cell lines as well as primary leukemia cells were found to be sensitive to treatment with compounds that directly and indirectly generate toxic ALDH substrates including 4-hydroxynonenal and the clinically relevant compounds arsenic trioxide and 4-hydroperoxycyclophosphamide. In contrast, normal hematopoietic stem cells were relatively resistant to these compounds. Using a murine xenotransplant model to emulate a clinical treatment strategy, established ALDH1A1– leukemias were also sensitive to in vivo treatment with cyclophosphamide combined with arsenic trioxide. These results demonstrate that targeting ALDH1A1– leukemic cells with toxic ALDH1A1 substrates such as arsenic and cyclophosphamide may be a novel targeted therapeutic strategy for this subset of acute myeloid leukemias.

Introduction doi:10.3324/haematol.2016.159053 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1054 ©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.

1054

The Aldehyde Dehydrogenase (ALDHs) gene family contains at least 19 isoforms that metabolize retinoids, reactive oxygen species (ROS) and endogenous and exogenous aldehydes.1,2 High levels of the ALDH1A1 isoform are expressed in hematopoietic stem cells (HSCs) as measured by mRNA analysis and staining with the fluorescent ALDH1A1 substrate Aldefluor.3 Despite this, ALDH1A1 is dispensable in murine HSCs, as it is compensated for by increased expression of the ALDH3A1 isoform and possibly others.4,5 Murine HSCs deficient in both ALDH1A1 and ALDH3A1 (Aldh1a1/3a1 DKOs) have a blocked B-cell development and reduced numbers of HSCs.4 The early B cells in these mice have increased levels of ROS and reactive aldehydes along with abnormalities in cell cycling, intracellular signaling and gene expression. Notably, HSCs and progenitors in these mice have elevated p38MAPK activity, increased sensitivity to DNA damage, and aberrant cell cycle distribution.4 These findings suggest an important role for ALDHs in metabolizing ROS and reactive aldehydes in primitive and differentiated hematopoietic cells, and show that the ALDHs and their substrates impact a variety of cellular processes in hematopoiesis. The role that ALDHs play in normal HSCs suggests they may have important roles in leukemia as well.6,7 Several prior reports support this contention. In Fanconi anemia, loss of activity of the ALDH2 isoform in concert with FANC-D deficiency predisposes to the development of acute myeloid leukemia (AML) and bone marhaematologica | 2017; 102(6)


Aldehyde dehydrogenases in AML

row failure.8,9 ALDH activity in human leukemia also mediates resistance to a number of drugs10 and high levels of ALDH activity, as measured by Aldefluor staining, predict for a poor outcome to treatment.11-14 Leukemic stem cells (LSCs) that drive leukemia growth and disease relapse may also be identified with Aldefluor staining with either high or intermediate levels of staining.15,16 Given these prior observations, in the present study, we sought to further clarify and define the role and significance of ALDH activity in acute leukemia with a particular focus on determining whether ALDH biology could be exploited for new treatment approaches.

Methods

mCherry Vector from Clontech (Cat# 631987), using EcoR1/Xba1 cloning sites. The sequence was verified using EF1a-Fwd and IRES-Rev sequencing primers. High titer lentiviral production was performed in 293TN cells and AML cells were transduced as previously described.28

Xenotransplant studies To test the drug sensitivity of MOLM-13 and primary AML cells in vivo, NSGS mice29 were treated intraperitoneally (IP) with busulfan (30 mg/kg) then injected intravenously (IV) with 5-7x106 AML cells. After marrow engraftment was confirmed, control mice were treated IP with saline while the rest of the mice were treated IP with 5 μg/g ATO daily for four days together with 150 µg/g Cy at day 1 and day 4. The mice were sacrificed by CO2 asphyxiation and the bone marrow harvested for flow cytometry analysis. Further details are available in the Online Supplementary Methods.

Analysis of public databases Analysis of ALDH family genes for mRNA expression, DNA methylation, and survival outcomes was performed with data available from The Cancer Genome Atlas (TCGA) (cancergenome.nih.gov/cancersselected/acutemyeloidleukemia). 1 7 Analysis of the expression of ALDH genes in leukemic versus normal hematopoietic cells was performed using the GSE9476 data set.18

Statistical analysis All statistical analyses were performed using GraphPad Prism 7.0. Typically, unpaired two-tail Student t-tests were performed and P<0.05 was considered significant. Aggregate data were presented as means and standard deviations (SD).

Results Cell lines and primary specimens Acute myeloid leukemia specimens were obtained from apheresis products of patients who gave Institutional Review Board approved informed consent for sample procurement at the University of Rochester and the University of Colorado in Denver. Normal 34+ HSCs were enriched from umbilical cord blood (UCB) specimens collected and distributed by the University of Colorado Cord Blood Bank (UCCBB), which is accredited by the American Association of Blood Banks and licensed by the US Food & Drug Administration. Cell lines were commercially supplied by ATCC (Manassas, VA, USA).

Flow cytometry analysis Human samples were stained with CD34, CD38, CD123 and Aldefluor as previously described.3,19,20 AML cells were routinely gated through the blast window and it was confirmed that pheresis derived samples were not contaminated with significant numbers of normal HSCs based on staining with anti-CD123, a marker that distinguishes AML from normal HSCs together with other AML directed markers including HLA-DR, CD71 and CD7 (Figure 1C and Online Supplementary Figure S4A).21-23 Anti-mouse antibodies were as described previously.4,24 All antibodies and staining reagents were from BD Biosciences (San Jose, CA, USA). Intracellular 4HNE staining and visualization of gamma (γ) H2AX by flow cytometry were performed as previously described.4 All cell staining and analytical procedures were performed either manually or using high content semi-automated flow cytometry, as previously described.25,26

Western blot analysis Lysates were prepared from Kasumi-1 and UCB CD34+ enriched populations. Primary mouse monoclonal antibodies to 4HNE adducts (Ab48506, Abcam; Cambridge, MA, USA) were used in a 1:500 dilution. Western blots were performed as previously described.27

Overexpression studies For restoring ALDH1A1 gene function in Kasumi-1 cells, the HA-ALDH1A1 gene was excised from Addgene plasmid # 11610 using HindIII/Xba1 and sub-cloned into the pLVX-EF1a-IREShaematologica | 2017; 102(6)

ALDH isoform expression is variable between AMLs and correlates with cytogenetic subtype Variable levels of Aldefluor staining in AML have been reported and may correlate with outcomes to treatment.11,13 However, given the large number of ALDH isoforms and the findings that Aldefluor is a substrate for several of them, we sought to more precisely determine if variability in expression of specific individual ALDH isoforms occurs in AML and if so, whether expression of specific isoforms correlated with prognosis.4,30 First, we analyzed 165 leukemia specimens from The Cancer Genome Atlas (TCGA) dataset for levels of expression of all 19 human ALDH family genes.31 A 4-5 log difference in expression of the ALDH1A1 isoform was observed among this cohort of AML specimens, with 39.4% (65 of 165) expressing ALDH1A1 at less than 1 RPKM (Figure 1A). None of the AMLs expressed significant levels of ALDH3A1; however, a number of other isoforms including ALDH1B1, ALDH2, ALDH3A2, ALDH3B1, ALDH4A1, ALDH5A1, ALDH6A1, ALDH9A1, ALDH16A1 and ALDH18A1 were expressed at abundant levels in a majority of AML samples (Figure 1A). To confirm these findings, qPCR for ALDH isoforms was performed on 7 primary AML samples and 3 of 7 had no detectable ALDH1A1 expression, 5 of 7 had no ALDH3A1 expression and, again, a number of other ALDH isoforms were variably expressed similar to the TCGA data (Online Supplementary Figure S1A). When ALDH1A1 expression in AML was compared to that in normal HSCs and progenitors via analysis of another publicly available dataset,18 ALDH1A1 was consistently found at high levels in normal CD34+ cells from all the bone marrow (BM) and peripheral blood (PB) samples while ALDH1A1 levels were again quite variable in AMLs with a minority having low levels of expression (Figure 1B). ALDH1A3, ALDH2, ALDH3B1 and ALDH8A1 expression was also significantly different between normal hematopoietic cells and AMLs (Online Supplementary Figure S1B). Lastly, several ALDH isoforms including ALDH3B1 and ALDH1A3 had significant nega1055


M. Gasparetto et al.

tive and positive associations, respectively, with ALDH1A1 (Online Supplementary Figure S1B). Given the utility of using ALDH functional activity as a marker of HSCs and other cancer stem cells, ALDH may be utilized as a marker of LSCs as well.11,32,33 To assess the level of ALDH1A1 expression in LSCs both at diagnosis and at relapse, AML samples from 6 patients were analyzed with RNA-Seq and the level of ALDH1A1 mRNA transcripts measured in both LSCs and non-LSCs as previously described.34 In these AMLs, ALDH1A1 was expressed at significantly lower levels than HSCs controls in both LSCs and non-LSCs and no significant differences in ALDH1A1 expression between LSCs and non-LSCs were noted (Figure 1C). In order to further test whether ALDH1A1 expression increased following treatment with standard chemotherapy as a mechanism of drug resistance, LSCs from relapse samples from the same 6 patients were similarly examined for

ALDH1A1 expression. Again, ALDH1A1 expression was significantly lower in relapse LSCs than the HSC controls, and analysis of paired diagnosis and relapse LSC samples did not show consistent or significant increases in ALDH1A1 expression (Figures 1D). In addition, no significant differences were noted in any other ALDH isoforms between diagnosis and relapse in either LSCs or non-LSCs (Online Supplementary Figure S1D). Together these data demonstrate that at least in a subset of AMLs, ALDH1A1 expression is not elevated in LSCs and does not increase following standard chemotherapy treatment. Given the significant ALDH isoform variability at the mRNA level between different AMLs observed in these studies, we next sought to investigate whether there was similar variability in ALDH enzymatic activity between AML samples. Eighteen primary AMLs were stained with

A

B

C

1056

D

Figure 1. (A) Variable expression of ALDH1A1 and other aldehyde dehydrogenase (ALDH) isoforms in human acute myeloid leukemia (AML). The AML dataset from The Cancer Genome Atlas (TCGA) was analyzed for differential expression of ALDH isoforms in human AML blasts. (B) The GSE9476 dataset was analyzed for differential expression of ALDH1A1 between normal human hematopoietic cells (HSCs) and AML samples. (C) ALDH1A1 expression based on RNA-Seq analysis in LSC (white) and nonLSC (light gray) samples at diagnosis and LSC samples at relapse (black). The control is similarly analyzed normal bone marrow CD34+ cells (NBM HSC) (dark gray) (****P<0.0001). (D) ALDH1A1 expression using RNA-Seq analysis between paired diagnostic and relapsed LSC samples. (each symbol represents a different patient).

haematologica | 2017; 102(6)


Aldehyde dehydrogenases in AML

Aldefluor dye and analyzed by flow cytometry. Consistent with the gene expression data, substantial differences in the overall content of Aldefluor staining cells between the different AML samples were observed (Figure 2A and B). As with the gene expression studies, a subset of AMLs had very low or absent numbers of cells with detectable Aldefluor staining [Aldefluor (0-0.1%)] (Figure 2A) demonstrating that they lacked ALDH1A1 and any other ALDH isoforms that may use Aldefluor as a substrate. Together, these findings demonstrate considerable variability in expression of different ALDH isoforms

between different AMLs as well as relative to the HSCs and progenitors from which they were presumably derived. The most striking finding was the low to undetectable levels of expression and activity of ALDH1A1 in any cells in a substantial subset of AMLs (25%-40%) despite its universal high-level expression in normal HSCs and progenitors (henceforth termed ALDH1A1â&#x20AC;&#x201C; AML). We next analyzed whether variability in the expression of the different ALDH isoforms, particularly ALDH1A1, had prognostic significance in predicting outcomes to treatment. When TCGA AML samples were grouped into

A

B

C

**** ****

D

haematologica | 2017; 102(6)

Figure 2. Levels of ALDH1A1 are variable between acute myeloid leukemias (AMLs) and have prognostic significance. (A) Aldefluor staining was performed to analyze single cell ALDH activity in representative samples of human primary AMLs that coexpressed CD123, a useful discriminator of AML and normal hematopoietic stem cells (HSCs). (B) Summary of Aldefluor staining in 18 different AML samples. (C) Association of ALDH1A1 expression level and poor, intermediate and good risk cytogenetic categories (TCGA dataset; ****P<0.0001). (D) Event-free survival (EFS) and overall survival (OS) Kaplan-Meier curve plots showing prognostic value of ALDH1A1 expression in all AML samples analyzed in the Cancer Genome Atlas dataset.

1057


M. Gasparetto et al.

Good, Intermediate and Poor Risk cytogenetic categories and analyzed for levels of ALDH1A1 expression, a significant association between the Good Risk cytogenetic category and lower levels of ALDH1A1 expression was observed (Figure 2C). In addition, low-level expression of

the ALDH1A1 isoform was correlated with improved event-free survival (EFS) and overall survival (OS) in the entire cohort of AML patients (Figure 2D). To evaluate whether the prognostic value of ALDH1A1 was independent from cytogenetics, intermediate risk samples were

B

A

C

D

E

F ** **

1058

Figure 3. ALDH1A1– acute myeloid leukemia (AML) cell lines are sensitive to toxic substrates of aldehyde dehydrogenase (ALDH). (A) Western Blot measurement of 4HNE protein adducts in Kasumi-1 and normal CD34+ umbilical cord blood (UCB) cells treated with 30 μM 4HNE or vehicle. Kasumi-1 cells treated with 100 μM 4HNE were included as a positive control. (B) Flow cytometric assessment of DNA damage (γH2AX) in Kasumi-1 with and without 4HNE treatment compared to normal CD34+ UCB cells. (C) Flow cytometric assessment of ROS in Kasumi-1 with and without 30-40 μM 4HNE treatment compared to normal CD34+ UCB cells. (D) 4HNE dose response curves in Kasumi-1 and normal CD34+ cells treated with 20 and 40 μM 4HNE [10 μM 4HNE was added to cell media every hour, data is mean and standard deviation (SD) of triplicate samples]. (E) Western blot analysis of 4HNE protein adducts generated by treating Kasumi-1 cells with 2 μM ATO over 12 hours and flow cytometric analysis of 4HNE adducts in Kasumi-1 induced by 2.5 μM 4HNE and 2 μM ATO treatment at 12 hours (n=3; *P<0.05). (F) Apoptosis of Kasumi-1 cells following overnight treatment with ATO (2 μM or 5 μM) or 4HC (12.5 μM) alone and in combination (n=6 replicates; **P<0.0001). (D-F) Bars represent mean±SD. (D and E) Results of technical triplicates are shown.

haematologica | 2017; 102(6)


Aldehyde dehydrogenases in AML

fractionated based on ALDH1A1 expression and analyzed for EFS and OS. A trend to an association between better EFS or OS and lower expression of ALDH1A1 was found within intermediate risk cytogenetic subgroups, although

A

B

C

D

the sample sizes were not big enough to formally confirm this association (Online Supplementary Figure S1E).

Absence of ALDH1A1 in human leukemia cell lines

F

haematologica | 2017; 102(6)

γH2AX%

γH2AX γH2AX

γH2AX

γH2AX

E

Figure 4. Overexpression of ALDH1A1+ partially rescues ALDH1A1– acute myeloid leukemia (AML) cell sensitivity to toxic substrates of ALDH. (A) ROS (Mitosox) detection in ALDH1A1+ and ALDH1A1- Kasumi-1 cells after six hours of treatment with 40 μM 4-HNE (n=3; *P<0.05). (B) Western blot analysis of 4HNE protein adducts generated by treating ALDH1A1+ and ALDH1A1– Kasumi-1 cells with different concentrations of 4HNE for 1 hour. (C) Viability of Kasumi-1 cells engineered to express ALDH1A1 through lentiviral gene transfer and treated with various combinations and doses of 4HC and ATO (n=3 replicates; ***P<0.0005, **P<0.005, *P<0.05). (D) Viability of Kasumi-3 cells treated with various combinations and doses of 4HC and ATO with (black) or without (white) the ALDH inhibitor DEAB (D) (n=3; **P<0.005, ***P<0.0005). (E) Flow cytometric assessment and summary graph of DNA damage (γH2AX) in ALDH1A1+ and ALDH1A1– Kasumi-1 cells treated with 4HC+ATO (n=3; **P<0.005, ***P<0.0005). (A-C) Bars represent mean±SD. Technical triplicates were performed. (F) In vivo treatment with Cy and ATO of NSGS mice with established ALDH1A1– MOLM-13 leukemia (n=6; ***P<0.0005).

1059


M. Gasparetto et al.

renders them sensitive to toxic ALDH substrates The finding that there is a subset of AMLs that do not express ALDH1A1 (ALDH1A1– AMLs) while normal HSCs and progenitors express high levels of ALDH1A1 suggested that exploiting low level ALDH1A1 expression may offer a new avenue to selectively target this subgroup of AMLs.35 Previously we had shown that primary murine marrow cells lacking Aldh1a1 and Aldh3a1 were sensitive to the reactive aldehyde 4-hydroxynonenal (4HNE), likely because these ALDHs are the primary metabolizers of this toxic compound.4 In addition, an in silico screen of compounds active in human leukemic stem cells revealed that 4HNE has potent human anti-leukemic activity.36 Together, these findings suggested that human ALDH1A1–

AMLs may be particularly sensitive to 4HNE treatment given their predicted limited ability to metabolize this toxic compound. To test this, first, a series of human AML cell lines were screened by Aldefluor staining along with qPCR analysis and both Kasumi-1 and MOLM-13 were identified as representative ALDH1A1– AML cell lines suitable for further analysis (Online Supplementary Figure S2A and B). At baseline, Kasumi-1 had high levels of endogenous reactive aldehyde adducts relative to normal UCB CD34+ cells (Figure 3A), which further increased after 4HNE treatment. DNA damage and ROS were also increased in Kasumi-1 cells treated with 4HNE while normal CD34+ cells were relatively resistant to these effects (Figure 3B and C). In addition, when Kasumi-1 cells were

A ***

γ H2AX%

***

B

C ****

γ H2AX%

****

D

Figure 5. Primary human ALDH1A1– acute myeloid leukemias (AMLs) are sensitive to 4HC+ATO while ALDH1A1+ AMLs are relatively resistant. (A) In vitro sensitivity of primary human AMLs to 4HC and ATO. Primary human AMLs were treated overnight with 4HC+ATO (12.5 μM+2 μM) and DNA damage and cell viability were measured (4 AMLs in triplicate; ***P<0.0005). (B) Sensitivity of Aldefluor+ and Aldefluor– fractions from 2 representative AMLs purified by FACS and then treated overnight with 4HC+ATO (12.5 μM+2 μM) (n=6; **P<0.005, ***P<0.0005). (C) DNA damage (γH2AX) of Aldefluor+ and Aldefluor– fractions from an ALDH1A1+ AML following overnight treatment with 4HC+ATO (n=3; **P<0.005, ***P<0.0005). (D) Relative sensitivity of Aldefluor+ fraction from an ALDH1A1+ AML to 4HC+ATO with and without 5 μM DEAB at 18 hours (n=3; *P<0.05). (A-D) Bars represent mean±SD. (C-D) Results of technical triplicates are shown.

1060

haematologica | 2017; 102(6)


Aldehyde dehydrogenases in AML

exposed to increasing concentrations of 4HNE, an LD50 of 4 μM was found, while in contrast, CD34+ normal cord blood cells were almost completely resistant to comparable doses of 4HNE treatment (Figure 3D). In order to begin to translate this concept into a clinically relevant treatment strategy for ALDH1A1– AML, we sought to identify known drugs that would induce intracellular 4HNE in ALDH1A1– AML cells, as 4HNE itself is not clinically useful due to non-specific toxicity and a very short half-life.37 Arsenic trioxide (ATO) is a well-characterized compound used in the treatment of acute promyelocytic leukemia that has significant pro-oxidant activity, suggesting that it could elevate intracellular reactive aldehyde levels, as these are frequent byproducts of ROS.38 Kasumi-1 cells treated with ATO had elevated intracellular 4HNE adduct levels (Figure 3E)39 and ATO induced a concentration-dependent increase of apoptosis in AML cells (Figure 3F). Next, we sought to determine whether we could further amplify the anti-leukemic effects of ATO in

ALDH1A1– AML by combining it with 4-hydroperoxycyclophosphamide (4HC), an active metabolite of cyclophosphamide (Cy) and known toxic substrate of ALDH1A1. The combination of ATO and even sublethal levels of 4HC led to higher levels of cell killing than control or either drug alone (Figure 3F).40 To determine whether these effects were dependent on ALDH1A1 expression, Kasumi-1 cells were transduced with a lentiviral vector that over-expressed ALDH1A1 (Online Supplementary Figure S2C) and tested over a wide range of 4HC and ATO concentrations, alone and together. Restoration of ALDH1A1 expression at least partially blocked the effects of 4HNE mediated generation of ROS and reactive aldehyde adducts in Kasumi-1 cells (Figure 4A and B, respectively). Expression of ALDH1A1 also protected Kasumi-1 cells to varying degrees from 4HC and ATO, depending on their concentrations (Figure 4C). Similar effects were noted in MOLM-13 cells transduced to express ALDH1A1 (Online Supplementary Figure S2D and E).

A

B

Figure 6. ALDH1A1– acute myeloid leukemias (AMLs) are sensitive to 4HC+ATO while normal HSCs are relatively resistant. (A) 4HC+ATO in vitro treatment eliminated engraftment of primary AMLs in NSGS mice. Representative flow cytometric analysis of harvested bone marrow of Aldefluor[0-0.1%] AMLs that underwent overnight ex vivo treatment with 4HC+ATO (30 μM+ 5 μM) or vehicle (left panels) and summary of xenotransplant data from the 3 AMLs treated in vitro and transplanted into NSGS mice (right panels). The data are a summary of 25 of 30 control and 25 of 30 treated mice analyzed after 12 weeks (**P<0.0005). (B) In contrast, 4HC+ATO in vitro treatment did not completely eliminate engraftment of normal CD34+UCB cells in NSGS mice. NSGS mice transplanted with 2 pooled UCBs following 4HC+ATO or vehicle treatment as above (treated: n=12; untreated: n=10). While level of engraftment was reduced, in contrast to the AML experiments above, 7 of 12 mice treated with normal CD34+UCB cells displayed more than 1% engraftment (*P<0.005). Analysis of engraftment in bone marrow was performed at 13 weeks using flow cytometry.

haematologica | 2017; 102(6)

1061


M. Gasparetto et al.

Furthermore, Kasumi-3 cells, which express high levels of ALDH1A1, were rendered more sensitive to 4HC and ATO following treatment with an ALDH inhibitor (DEAB) (Figure 4D and Online Supplementary Figure S2F). In addition, differences in DNA damage were also noted between ALDH1A1– and ALDH1A1+ Kasumi-1 cells treated in vitro with 4HC and ATO (Figure 4E). Lastly, we tested the in vivo anti-leukemia activity of Cy plus ATO on MOLM-13 AML cells. MOLM-13 were used for these studies as opposed to Kasumi-1 as they consistently engraft NSGS mice, while Kasumi-1 do not engraft well in the conditions tested. Briefly, MOLM-13 cells were transplanted into NSGS mice and then treated daily with ATO for four days followed by Cy at day 1 and day 4. When the marrow content of MOLM-13 was analyzed by flow cytometry, two weeks later, the overall content of MOLM-13 cells was dramatically reduced in Cy plus ATO treated mice relative to placebotreated controls (Figure 4F). Consequently, these results demonstrate that loss of ALDH1A1 renders AML cell lines more sensitive to 4HC and ATO.

Absence of ALDH1A1 in primary AMLs renders them sensitive to toxic ALDH substrates To determine if the results obtained from the AML cell line studies were relevant to authentic primary AMLs, first, in vitro sensitivity studies were performed. Peripheral blood samples from apheresis collections of 4 ALDH1A1– AML patients were treated overnight with 4HC and ATO, alone or in combination. The 4HC plus ATO combination generated more DNA damage and cell killing than either agent alone or controls (Figure 5A). In addition, when 3 samples known to have both ALDH1A1+ and ALDH1A1– subsets were sorted into Aldefluor+ and Aldefluor– fractions (Online Supplementary Figure S3A-C) and treated in vitro with 4HC plus ATO, the Aldefluor– cells were sensitive to treatment while the Aldefluor+ cells were relatively resistant as measured by cell killing and DNA damage (Figure 5B and C). To further determine whether ALDH1A1 expression modulates the effects of 4HC and ATO in primary AML cells, Aldefluor+ AML cells were treated with DEAB, an inhibitor of ALDH1A1, and then the 4HC/ATO drug combination was added.41 Again, the

A

B

Figure 7. ALDH1A1– acute myeloid leukemias (AMLs) are sensitive to Cy+ATO while ALDH1A1+ AMLs are relatively resistant. (A) In vivo Cy+ATO treatment reduced leukemic burden in NSGS mice engrafted with primary ALDH1A1– AML cells (treated: n=7; untreated: n=9; *P<0.005). (B) In vivo Cy+ATO treatment shows no differences in leukemic burden in NSGS mice engrafted with primary ALDH1A1+ AML cells (n=10; not significant).

1062

haematologica | 2017; 102(6)


Aldehyde dehydrogenases in AML

Aldefluor+ controls were relatively resistant to treatment with 4HC/ATO, while in contrast the DEAB-treated cells had a greater sensitivity to 4HC/ATO than control (Figure 5D). Together these data further demonstrate that ALDH1A1, at least in part, mediates resistance to these agents. Next, to test whether 4HC plus ATO was also effective against human primary AML cells capable of engrafting immunodeficient mice, 3 primary Aldefluor[0-0.1%] AMLs were treated in vitro overnight with either vehicle control or 4HC plus ATO and transplanted into busulfan-treated NSGS mice. All control-treated AMLs engrafted at levels of 20%-100% while 4HC plus ATO treated AMLs failed to engraft any of 25 mice (Figure 6A). In contrast, when normal UCB derived CD34+ cells were treated with 4HC plus ATO in a similar fashion, although engraftment was reduced in NSGS mice, it could be readily detected in 75% of recipients (Figure 6B). Lastly, to replicate the clinical scenario as closely as possible, we tested in vivo the effects of Cy and ATO on an Aldefluor[0-0.1%] AML that had been previously engrafted in NSGS mice compared to an Aldefluor[>10%] AML positive control known to express high levels of ALDH1A1. Two weeks following in vivo treatment of Aldefluor[0-0.1%] AML engrafted mice with Cy plus ATO, the overall marrow content of human AML was significantly reduced in mice relative to placebo-treated controls (Figure 7A). In contrast, the Aldefluor[>10%] AML engrafted mice showed no differences in engraftment between treated and untreated groups (Figure 7B).

Discussion In this report, we demonstrate that there is expression of a number of ALDH isoforms in AML and several have substantial variability in expression. The most prominent of these is ALDH1A1, which has low levels of expression and activity in a subset of AMLs, despite its nearly universal expression in normal HSCs and progenitors. In fact, in 25%-40% of AMLs we could find few or no cells expressing any ALDH1A1 as determined by Aldefluor staining and flow cytometry as well as qPCR analysis (termed

ALDH1A1– AML). This includes finding low to absent levels of ALDH1A1 in LSCs as well. The factors governing the variability of ALDH1A1 expression in AMLs are unknown; however, the majority of ALDH1A1– AMLs in public databases do not have obvious ALDH1A1 gene deletions (data not shown) suggesting that loss of ALDH1A1 expression may occur as a consequence of gene regulation either through epigenetic or gene expression/repression mechanisms.42 Further studying the underlying mechanisms of ALDH1A1 gene regulation in HSCs, as well as AMLs, may provide useful insights into the biology of each. We also found that loss of ALDH1A1 may have prognostic implications as ALDH1A1– AMLs correlate with good prognosis cytogenetics and there was a trend to improved outcomes in intermediate risk cytogenetic AML patients based on level of ALDH1A1 expression, although we could not formally prove an association. It would be interesting to seek to extend and confirm these observations in larger prospective studies. If this trend is confirmed, then ALDH1A1 status may become a useful tool to stratify patients to different treatment approaches. These observations also raise additional questions regarding why ALDH1A1 expression would be associated with cytogenetic status. In addition, if outcomes are indeed impacted by ALDH1A1 status, it would be important to determine whether this was because loss of ALDH1A1 renders ALDH1A1– AMLs more sensitive to chemotherapies that induce ROS and/or reactive aldehydes or whether this acts through a different mechanism. Since a substantial subgroup of AMLs is deficient in ALDH1A1 expression at different stages of disease (diagnosis and relapse), we speculated that these ALDH1A1– AMLs may specifically be more sensitive to compounds that directly or indirectly generate toxic ALDH1A1 substrates. First, 4HNE, a prototypical toxic reactive aldehyde substrate of ALDH1A1 was quite effective at killing ALDH1A1– cell lines while CD34+ UCB, which expresses high levels of ALDH1A1, was quite resistant to this compound. Of more clinical relevance, we found that ATO, a widely used drug in the treatment of APL, could induce the intracellular production of 4HNE, leading to selective killing of ALDH1A1– AMLs. This effect was at least par-

Figure 8. Model for selectively targeting ALDH1A1– acute myeloid leukemias (AMLs). In this treatment strategy, AMLs will be profiled for aldehyde dehydrogenase (ALDH) isoform expression and those lacking ALDH1A1 expression will be treated with agents that directly and indirectly generate lethal levels of ALDH substrates including ROS, reactive aldehydes and others. In contrast, normal hematopoietic stem cells express high levels of ALDH1A1 that could metabolize these compounds resulting in relative sparing.

haematologica | 2017; 102(6)

1063


M. Gasparetto et al.

tially blocked by restoration of ALDH1A1 activity through gene transfer, further supporting a role for ALDH1A1 in this effect. The finding that ALDH1A1 gene transfer did not fully block the effect of ATO is likely because ATO potentially works through a variety of mechanisms, of which generating 4HNE is only one.43 To amplify the effect of ATO in ALDH1A1– AML, it was combined with 4HC for in vitro studies and Cy for in vivo studies as these generate a toxic intermediate substrate termed aldophosphamide that is primarily metabolized by ALDH1A1. ALDH1A1– AML cell lines as well as primary ALDH1A1– AMLs were particularly sensitive to the ATO and 4HC/Cy combinations both in vitro and in vivo, while, again, normal HSCs were relatively resistant. The in vivo studies ultimately may be extended to determine mechanisms of resistance to ALDH-directed therapy. The finding that there is an apparent therapeutic window between normal HSCs and ALDH1A1– AMLs treated with toxic ALDH1A1 substrates, suggests a novel targeted therapy strategy for this newly defined AML subset (modeled in Figure 8). In this strategy, we propose that AMLs lacking ALDH1A1 are more sensitive to compounds that directly or indirectly generate toxic ALDH substrates including ROS, reactive aldehydes and others,

References 1. Vasiliou V, Thompson DC, Smith C, Fujita M, Chen Y. Aldehyde dehydrogenases: from eye crystallins to metabolic disease and cancer stem cells. Chem Biol Interact. 2013;202(1-3):2-10. 2. Singh S, Brocker C, Koppaka V, et al. Aldehyde dehydrogenases in cellular responses to oxidative/electrophilic stress. Free Radic Biol Med. 2013;56:89-101. 3. Storms RW, Trujillo AP, Springer JB, et al. Isolation of primitive human hematopoietic progenitors on the basis of aldehyde dehydrogenase activity. Proc Natl Acad Sci USA. 1999;96(16):9118-9123. 4. Gasparetto M, Sekulovic S, Brocker C, et al. Aldehyde dehydrogenases are regulators of hematopoietic stem cell numbers and B-cell development. Exp Hematol. 2012; 40(4):318-329.e312. 5. Levi BP, Yilmaz OH, Duester G, Morrison SJ. Aldehyde dehydrogenase 1a1 is dispensable for stem cell function in the mouse hematopoietic and nervous systems. Blood. 2009;113(8):1670-1680. 6. Fritz KS, Petersen DR. Exploring the biology of lipid peroxidation-derived protein carbonylation. Chem Res Toxicol. 2011;24(9):1411-1419. 7. Hole PS, Darley RL, Tonks A. Do reactive oxygen species play a role in myeloid leukemias? Blood. 2011;117(22):5816-5826. 8. Garaycoechea JI, Crossan GP, Langevin F, Daly M, Arends MJ, Patel KJ. Genotoxic consequences of endogenous aldehydes on mouse haematopoietic stem cell function. Nature. 2012;489(7417):571-575. 9. Hira A, Yabe H, Yoshida K, et al. Variant ALDH2 is associated with accelerated progression of bone marrow failure in Japanese Fanconi anemia patients. Blood. 2013; 122(18):3206-3209. 10. Koelling TM, Yeager AM, Hilton J, Haynie DT, Wiley JM. Development and character-

1064

11.

12.

13.

14.

15. 16.

17.

18.

19.

while normal HSCs with high levels of ALDH1A1 are relatively resistant. Consequently, rational drug combinations focusing on these agents may be particularly effective in ALDH1A1– AML patients. As a first step towards exploring this treatment strategy, the findings in this report provide a rationale for performing a phase I/II clinical trial of Cy plus ATO in ALDH1A1– AML patients to determine efficacy and toxicity. In addition, these observations suggest the utility of further pre-clinical studies to identify even more effective and less toxic combinations of compounds that exploit the differences in ALDH1A1 expression between normal HSCs and AMLs. Ultimately, it will be interesting to determine whether therapeutic targeting of the differences in ALDH isoform expression between malignant cells and normal tissue stem cells can be extended to other hematologic cancers such as myeloma and lymphoma, and solid tumors as well. Acknowledgments The authors gratefully acknowledge technical assistance by the Flow Cytometry Cores (CFAR and UCCC) and the Animal Housing & Care Vivarium of the University of Colorado Denver (Anschutz Medical Campus).

ization of a cyclophosphamide-resistant subline of acute myeloid leukemia in the Lewis x Brown Norway hybrid rat. Blood. 1990;76(6):1209-1213. Cheung AM, Wan TS, Leung JC, et al. Aldehyde dehydrogenase activity in leukemic blasts defines a subgroup of acute myeloid leukemia with adverse prognosis and superior NOD/SCID engrafting potential. Leukemia. 2007;21(7):1423-1430. Hoang VT, Hoffmann I, Borowski K, et al. Identification and separation of normal hematopoietic stem cells and leukemia stem cells from patients with acute myeloid leukemia. Methods Mol Biol. 2013; 1035:217-230. Ran D, Schubert M, Pietsch L, et al. Aldehyde dehydrogenase activity among primary leukemia cells is associated with stem cell features and correlates with adverse clinical outcomes. Exp Hematol. 2009;37(12):1423-1434. Schuurhuis GJ, Meel MH, Wouters F, et al. Normal hematopoietic stem cells within the AML bone marrow have a distinct and higher ALDH activity level than co-existing leukemic stem cells. PLoS One. 2013;8(11):e78897. Fleischman AG. ALDH marks leukemia stem cell. Blood. 2012;119(15):3376-3377. Gerber JM, Smith BD, Ngwang B, et al. A clinically relevant population of leukemic CD34(+)CD38(-) cells in acute myeloid leukemia. Blood. 2012;119(15):3571-3577. Cancer Genome Atlas Research N. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med. 2013;368(22):2059-2074. Stirewalt DL, Meshinchi S, Kopecky KJ, et al. Identification of genes with abnormal expression changes in acute myeloid leukemia. Genes Chromosomes Cancer. 2008;47(1):8-20. Fallon P, Gentry T, Balber AE, et al. Mobilized peripheral blood SSCloALDHbr cells have the phenotypic and functional

20.

21.

22.

23.

24.

25.

26.

27.

properties of primitive haematopoietic cells and their number correlates with engraftment following autologous transplantation. Br J Haematol. 2003;122(1):99-108. Storms RW, Green PD, Safford KM, et al. Distinct hematopoietic progenitor compartments are delineated by the expression of aldehyde dehydrogenase and CD34. Blood. 2005;106(1):95-102. Jordan CT, Upchurch D, Szilvassy SJ, et al. The interleukin-3 receptor alpha chain is a unique marker for human acute myelogenous leukemia stem cells. Leukemia. 2000; 14(10):1777-1784. Liu Q, Wang M, Hu Y, et al. Significance of CD71 expression by flow cytometry in diagnosis of acute leukemia. Leuk Lymphoma. 2014;55(4):892-898. van Rhenen A, Moshaver B, Kelder A, et al. Aberrant marker expression patterns on the CD34+CD38- stem cell compartment in acute myeloid leukemia allows to distinguish the malignant from the normal stem cell compartment both at diagnosis and in remission. Leukemia. 2007;21(8):17001707. Gasparetto M, Sekulovic S, Zakaryan A, et al. Varying levels of aldehyde dehydrogenase activity in adult murine marrow hematopoietic stem cells are associated with engraftment and cell cycle status. Exp Hematol. 2012;40(10):857-866 e855. Brinkman RR, Gasparetto M, Lee SJ, et al. High-content flow cytometry and temporal data analysis for defining a cellular signature of graft-versus-host disease. Biol Blood Marrow Transplant. 2007;13(6):691700. Gasparetto M, Gentry T, Sebti S, et al. Identification of compounds that enhance the anti-lymphoma activity of rituximab using flow cytometric high-content screening. J Immunol Methods. 2004;292(1-2):5971. Pei S, Minhajuddin M, Callahan KP, et al. Targeting aberrant glutathione metabolism

haematologica | 2017; 102(6)


Aldehyde dehydrogenases in AML

28.

29.

30. 31. 32.

33.

to eradicate human acute myelogenous leukemia cells. J Biol Chem. 2013;288(47): 33542-33558. Ashton JM, Balys M, Neering SJ, et al. Gene sets identified with oncogene cooperativity analysis regulate in vivo growth and survival of leukemia stem cells. Cell Stem Cell. 2012;11(3):359-372. Wunderlich M, Chou FS, Link KA, et al. AML xenograft efficiency is significantly improved in NOD/SCID-IL2RG mice constitutively expressing human SCF, GM-CSF and IL-3. Leukemia. 2010;24(10):17851788. Moreb JS. Aldehyde dehydrogenase as a marker for stem cells. Curr Stem Cell Res Ther. 2008;3(4):237-246. Miller CA, Wilson RK, Ley TJ. Genomic landscapes and clonality of de novo AML. N Engl J Med. 2013;369(15):1473. Pearce DJ, Taussig D, Simpson C, et al. Characterization of cells with a high aldehyde dehydrogenase activity from cord blood and acute myeloid leukemia samples. Stem Cells. 2005;23(6):752-760. Hoang VT, Buss EC, Wang W, et al. The rar-

haematologica | 2017; 102(6)

34.

35.

36.

37. 38.

ity of ALDH(+) cells is the key to separation of normal versus leukemia stem cells by ALDH activity in AML patients. Int J Cancer. 2015;137(3):525-536. Ho TC, LaMere M, Stevens BM, et al. Evolution of acute myelogenous leukemia stem cell properties after treatment and progression. Blood. 2016;128(13):16711678. Smith C, Gasparetto M, Humphries K, Pollyea DA, Vasiliou V, Jordan CT. Aldehyde dehydrogenases in acute myeloid leukemia. Ann NY Acad Sci. 2014;1310(1):58-68. Hassane DC, Guzman ML, Corbett C, et al. Discovery of agents that eradicate leukemia stem cells using an in silico screen of public gene expression data. Blood. 2008;111(12): 5654-5662. Fritz KS, Petersen DR. An overview of the chemistry and biology of reactive aldehydes. Free Radic Biol Med. 2013;59:85-91. Jeanne M, Lallemand-Breitenbach V, Ferhi O, et al. PML/RARA oxidation and arsenic binding initiate the antileukemia response of As2O3. Cancer Cell. 2010;18(1):88-98.

39. Chou WC, Jie C, Kenedy AA, Jones RJ, Trush MA, Dang CV. Role of NADPH oxidase in arsenic-induced reactive oxygen species formation and cytotoxicity in myeloid leukemia cells. Proc Natl Acad Sci USA. 2004;101(13):4578-4583. 40. Andersson BS, Mroue M, Britten RA, Farquhar D, Murray D. Mechanisms of cyclophosphamide resistance in a human myeloid leukemia cell line. Acta Oncol. 1995;34(2):247-251. 41. Morgan CA, Parajuli B, Buchman CD, Dria K, Hurley TD. N,N-diethylaminobenzaldehyde (DEAB) as a substrate and mechanism-based inhibitor for human ALDH isoenzymes. Chem Biol Interact. 2015;234 (18-28). 42. Zhao D, Mo Y, Li MT, et al. NOTCHinduced aldehyde dehydrogenase 1A1 deacetylation promotes breast cancer stem cells. J Clin Invest. 2014;124(12):5453-5465. 43. Kumar S, Yedjou CG, Tchounwou PB. Arsenic trioxide induces oxidative stress, DNA damage, and mitochondrial pathway of apoptosis in human leukemia (HL-60) cells. J Exp Clin Cancer Res. 2014;33:42.

1065


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Myeloid Leukemia

Ferrata Storti Foundation

Haematologica 2017 Volume 102(6):1066-1074

Impact of ABO incompatibility on patients’ outcome after haploidentical hematopoietic stem cell transplantation for acute myeloid leukemia - a report from the Acute Leukemia Working Party of the EBMT

Jonathan Canaani,1* Bipin N Savani,2* Myriam Labopin,3,4,5 Xiao-jun Huang,6 Fabio Ciceri,7 William Arcese,8 Johanna Tischer,9 Yener Koc,10 Benedetto Bruno,11 Zafer Gülbas,12 Didier Blaise,13 Johan Maertens,14 Gerhard Ehninger,15 Mohamad Mohty3,4,5 and Arnon Nagler1,3,5

Hematology Division, Chaim Sheba Medical Center, Tel-Hashomer, Tel Aviv University, Israel; 2Vanderbilt University Medical Center, Nashville, TN, USA; 3Acute Leukemia Working Party –EBMT and Department of Hematology and Cell Therapy, Hȏpital SaintAntoine, Paris, France; 4INSERM UMR 938, Paris, France; 5Université Pierre et Marie Curie, Paris, France; 6Peking University People’s Hospital, Institute of Haematology, Xicheng District, Beijing, China; 7Ospedale San Raffaele s.r.l., Haematology and BMT, Milano, Italy; 8Tor Vergata University of Rome, Stem Cell Transplant Unit, Policlinico Universitario Tor Vergata, Italy; 9Klinikum Grosshadern, Med. Klinik III, Munich, Germany; 10Medical Park Hospitals, Stem Cell Transplant Unit, Antalya, Turkey; 11 S.S.C.V.D Trapianto di Cellule Staminali A.O.U Citta della Salute e della Scienza di Torino, Italy; 12Anadolu Medical Center Hospital, Bone Marrow Transplantation Department, Kocaeli, Turkey; 13Programme de Transplantation & Therapie Cellulaire, Centre de Recherche en Cancérologie de Marseille, Institut Paoli Calmettes, France; 14 University Hospital Gasthuisberg, Department of Hematology, Leuven, Belgium and 15 Universitaetsklinikum Dresden Medizinische Klinik und Poliklinik I, Germany 1

*JC and BNS contributed equally to this work.

ABSTRACT

Correspondence: arnon.nagler@sheba.health.gov.il

Received: November 23, 2016. Accepted: February 20, 2017. Pre-published: March 2, 2017. doi:10.3324/haematol.2016.160804 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1066 ©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.

1066

A

significant proportion of hematopoietic stem cell transplants are performed with ABO-mismatched donors. The impact of ABO mismatch on outcome following transplantation remains controversial and there are no published data regarding the impact of ABO mismatch in acute myeloid leukemia patients receiving haploidentical transplants. Using the European Blood and Marrow Transplant Acute Leukemia Working Group registry we identified 837 patients who underwent haploidentical transplantation. Comparative analysis was performed between patients who received ABO-matched versus ABO-mismatched haploidentical transplants for common clinical outcome variables. Our cohort consisted of 522 ABO-matched patients and 315 ABO-mismatched patients including 150 with minor, 127 with major, and 38 with bi-directional ABO mismatching. There were no significant differences between ABO matched and mismatched patients in terms of baseline disease and clinical characteristics. Major ABO mismatching was associated with inferior day 100 engraftment rate whereas multivariate analysis showed that bi-directional mismatching was associated with increased risk of grade II-IV acute graft-versus-host disease [hazard ratio (HR) 2.387; 95% confidence interval (CI): 1.22-4.66; P=0.01). Non-relapse mortality, relapse incidence, leukemia-free survival, overall survival, and chronic graft-versus-host disease rates were comparable between ABO-matched and -mismatched patients. Focused analysis on stem cell source showed that patients with minor mismatching transplanted with bone marrow grafts experienced increased grade II-IV acute graftversus-host disease rates (HR 2.03; 95% CI: 1.00-4.10; P=0.04). Patients with major ABO mismatching and bone marrow grafts had decreased survival (HR=1.82; CI 95%: 1.048 – 3.18; P=0.033). In conclusion, ABO incompatibility has a marginal but significant clinical effect in acute myeloid leukemia patients undergoing haploidentical transplantation. haematologica | 2017; 102(6)


ABO incompatibility in transplants for AML

Introduction As the full potential of haploidentical hematopoietic stem cell transplantation (HCT) is gaining appreciation in the field of transplantation, and its capacity to provide an alternative donor source for a substantial segment of the population of patients lacking a matched related donor (estimated recently to be as large as 70%1) is being realized, efforts aimed at optimizing donor-recipient compatibility are gaining traction. Indeed, emerging data from patients with acute myeloid leukemia (AML) undergoing haploidentical HCT is establishing this approach as a viable option for patients lacking an HLA-matched donor.24 While the extensive applicability of haploidentical HCT was limited initially by a significant component of graftversus-host disease (GvHD) contributing to increased nonrelapse mortality,5,6 recent innovative approaches employing novel immunosuppression techniques are significantly improving patients’ outcome in this setting.7-9 Although ABO incompatibility is found in up to one-half of HLAmatched transplants,10,11 and has the potential to put the recipient at risk of significant complications, its overall effect on clinical outcome measures has been debated extensively. Publications involving multiple datasets of patients with various disease states, donor sources, and conditioning regimens have shown conflicting results in this regard.12-17 In this analysis of data in the European Society for Blood and Marrow Transplantation (EBMT) registry we set out to determine whether ABO compatibility has a significant role in influencing the outcome of AML patients undergoing haploidentical HCT.

Methods Study population This was a retrospective, multicenter analysis. Data were provided and approved for this study by the Acute Leukemia Working Party (ALWP) of the EBMT group registry. The latter is a voluntary working group of more than 500 transplant centers that are required to report all consecutive stem cell transplants and follow-ups once a year. Audits are routinely performed to determine the accuracy of the data. The study protocol was approved by the institutional review board at each site and complied with countryspecific regulatory requirements. The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. All patients provided written informed consent authorizing the use of their personal information for research purposes. Using the EBMT registry, we identified adult patients (age >18 years) with AML and the following inclusion criteria: transplanted between 2005 and 2014, and HLA haploidentical donor with bone marrow or granulocyte colony-stimulating factor-mobilized peripheral blood stem cell grafts. All donors were HLA-mismatched at least at two loci (≤8/10) (-A, -B, -C, DRB1, -DQB1). Exclusion criteria were previous allogeneic or cord blood transplantation. Major ABO incompatibility was defined as serological evidence of recipient-derived antibodies directed against donor red cells, minor ABO incompatibility was defined as serological evidence of donor-derived antibodies directed against the recipient’s red cells, while bi-directional incompatibility comprised cases with serological evidence of both donor- and recipientderived red cell directed antibodies. Engraftment was defined as sustained achievement of an absolute neutrophil count of over 0.5x109/L. Conditioning regimens were classified as myeloablative or reduced intensity based on previously published criteria.18 haematologica | 2017; 102(6)

Grading of acute and chronic GvHD was performed using established criteria.19 Chronic GvHD was classified as limited or extensive according to usual criteria.20 The list of institutions reporting data included in this study is provided in the Online Supplementary Data.

Statistical analysis Five outcomes were evaluated: (i) non-relapse mortality, defined as death without previous relapse; (ii) relapse incidence, defined on the basis of morphological evidence of leukemia in bone marrow or other extramedullary organs; (iii) leukemia-free survival, defined as the time from transplantation to first event (either relapse or death in complete remission); (iv) overall survival; and (v) GvHD-free/relapse-free survival, defined as events including grade III-IV acute GvHD, chronic GvHD requiring systemic therapy, relapse, or death in the first year following the HCT. Cumulative incidence curves were used for relapse incidence and non-relapse mortality in a setting of competing risks, since death and relapse are competing events. Probabilities of overall survival and leukemia-free survival were calculated using the Kaplan– Meier estimate. All tests were two-sided with the type I error rate fixed at 0.05. Statistical analyses were performed with SPSS 19 (SPSS Inc., Chicago, IL, USA), and R 3.0.1 (R Development Core Team, Vienna, Austria) software packages.

Results Patients’ characteristics In all, 837 patients were transplanted between 20052014 with a median follow-up period of 35 months (range, 1.2-125.4 months). The characteristics of the patients, their diseases and transplants are summarized in Table 1. There were no significant differences between ABOmatched and ABO-mismatched patients in terms of disease status at transplant, high-risk cytogenetics, donor and recipient cytomegalovirus status, conditioning intensity, graft source, and rates of T-cell depletion. As shown in Online Supplementary Table S1, leukemia, GvHD, and infection were the major causes of death across all ABO incompatibility categories.

Major ABO incompatibility is associated with decreased engraftment in haploidentical stem cell transplantation Since previous data indicated that ABO mismatching affected stem cell engraftment,11 we analyzed engraftment data per ABO category. As summarized in Table 2, day 100 engraftment rates were significantly lower in patients with major ABO incompatibility compared to those with other ABO mismatch categories. An analysis focused on graft source revealed that while the engraftment rate of peripheral blood grafts did not differ between subgroups, in bone marrow grafts major ABO incompatibility was again associated with inferior engraftment (data not shown).

Bi-directional ABO incompatibility increases the incidence of acute graft-versus-host disease in haploidentical transplants To evaluate whether ABO compatibility affects clinical outcome, a univariate analysis was initially carried out and, as shown in Online Supplementary Table S2, demonstrated that patients with bi-directional ABO mismatching had a significantly higher 3-year leukemia-free survival rate com1067


J. Canaani et al.

pared to patients with major ABO mismatching who had the lowest rate (67.2% and 40.1%, respectively). A similar finding was also observed with regard to 3-year GvHDfree/relapse-free survival rates which were increased in bidirectional mismatched patients and significantly lower in patients with a major ABO mismatch. A subsequent subgroup analysis of ABO-matched versus ABO-mismatched patients followed by focused analysis of specific mismatching patterns did not show any statistically significant differences between groups (Online Supplementary Table S2). To validate our findings we performed a multivariate analysis using the group of ABO-compatible patients as the reference group (Table 3). Interestingly, bi-directional ABO mismatching (n=38) was found to be associated with a significantly increased risk of grade II-IV acute GvHD [hazard ratio (HR)=2.38, 95% confidence interval (95% CI): 1.22 4.66; P=0.01) (Figure 1).

To ensure that T-cell graft composition, namely T-cellreplete versus T-cell-depleted grafts, was not affecting our results, separate analyses were performed for patients transplanted with T-cell-replete and T-cell-depleted grafts. As detailed in Online Supplementary Table S3, in T-cellreplete grafts, univariate analysis revealed that the 3-year leukemia-free survival and GvHD-free/relapse-free survival rates were significantly increased in bi-directional ABO-mismatched patients compared to those in ABOcompatible patients and both ABO major and minor mismatched patients. Of note, chronic GvHD rates were increased in ABO-mismatched patients compared to ABO-matched patients. However, multivariate analysis failed to corroborate a statistically significantly association between ABO mismatch status and clinical outcome. Since there were only four patients with bi-directional ABO mismatching in the T-cell-depleted cohort, these

Table 1. Baseline characteristics of the study population.

Variable Follow up duration in m, median (range) Age in years, median (range) Gender, n(%) Male Female Disease status at transplant CR1 CR2/3 Active disease CMV D-/RCMV D+/RCMV D-/R+ CMV D+/R+ T-cell depletion ex-vivo Yes No T- cell depletion in-vivo Yes No Bone marrow-derived graft Peripheral blood graft Bone marrow and peripheral blood Female donor to male recipient No female donor to male recipient Conditioning regimen Myeloablative Reduced intensity

ABO matched n=522

Minor ABO mismatch n=150

Major ABO mismatch n=127

Bi-directional ABO mismatch n=38

35.9 (1.02-116.9 ) 41.8 (18-77.8)

34.5 (0-128.5) 45 (18-72.8)

34.7 (0-119.9) 42.4 (18-71.2)

35.2 (1.9-122.8) 44.5 (20.1-66.8)

299 (57.28% ) 223 (42.72% )

91 (60.67% ) 59 (39.33% )

67 (52.76% ) 60 (47.24% )

20 (52.63% ) 18 (47.37% )

271 (51.92% ) 98 (18.77% ) 153 (29.31% ) 73 (14.2% ) 30 (5.84% ) 72 (14.01% ) 339 (65.95% )

91 (60.67% ) 21 (14% ) 38 (25.33% ) 15 (10.14% ) 7 (4.73% ) 23 (15.54% ) 103 (69.59% )

71 (55.91% ) 16 (12.6% ) 40 (31.5% ) 16 (13.11% ) 3 (2.46% ) 18 (14.75% ) 85 (69.67% )

26 (68.42% ) 6 (15.79% ) 6 (15.79% ) 2 (5.56% ) 0 (0% ) 7 (19.44% ) 27 (75% )

71 (13.6% ) 451 (86.4% )

24 (16% ) 126 (84% )

24 (18.9% ) 103 (81.1% )

4 (10.53% ) 34 (89.47% )

211 (40.5% ) 310 (59.5% ) 133 (25.48% ) 243 (46.55% ) 146 (27.97% ) 132 (25.29% ) 390 (74.71% )

54 (36% ) 96 (64% ) 38 (25.33% ) 67 (44.67% ) 45 (30% ) 38 (25.33% ) 112 (74.67% )

50 (39.37% ) 77 (60.63% ) 39 (30.71% ) 51 (40.16% ) 37 (29.13% ) 29 (22.83% ) 98 (77.17% )

16 (42.11% ) 22 (57.89% ) 9 (23.68% ) 17 (44.74% ) 12 (31.58% ) 9 (23.68% ) 29 (76.32% )

307 (58.81% ) 215 (41.19% )

82 (54.67% ) 68 (45.33% )

82 (64.57% ) 45 (35.43% )

145 (57.8%) 12 (31.58% )

P*

0.668 0.558

0.146

0.465

0.389

0.779

0.872

0.945 0.245

*P value of a test of the null hypothesis that all the groups are the same. CR1, first complete remission; CR2/3: second or third complete remission; CMV, cytomegalovirus. D: donor; R: recipient.

Table 2. Engraftment rate according to ABO incompatibility category.

ABO matched Engraftment Graft failure Missing Time to PMN>500, days (range)

481 (94.1%) 30 (5.9%) 11 16 (3-44)

Minor ABO mismatch Major ABO mismatch 143 (95.3%) 7 (4.7%) 0 17 (7-45)

111 (88.1%) 15 (11.9%) 1 16 (8-63)

Bi-directional mismatch

P

36 (97.3%) 1 (2.7%) 1 15 (10-38)

0.04

PMN, polymorphonuclear cells.

1068

haematologica | 2017; 102(6)


ABO incompatibility in transplants for AML

Table 3. Multivariable analysis per ABO mismatch category of the entire cohort.

Parameter

LFS HR (95% CI)

OS HR (95% CI)

RI HR (95% CI) NRM HR (95% CI)

Matched ABO (ref) Minor ABO mismatch

1 0.95 (0.69-1.29), P=0.74 1.17 (0.86-1.6), P=0.3 0.68 (0.35-1.31), P=0.25

1 0.98 (0.71-1.34), P=0.91 1.21 (0.88-1.67), P=0.22 0.76 (0.39-1.48), P=0.42

1 1 0.831 (0.53-1.29), 1.1 (0.72-1.68), P=0.41 P=0.63 1.3 (0.85-1.98), 1.05 (0.66-1.67), P=0.2 P=0.81 0.58 (0.2-1.63), 0.83 (0.35-1.94), P=0.3 P=0.67

Major ABO mismatch Bi-directional ABO mismatch

Acute GvHD Chronic GvHD HR grade II-IV HR (95% CI) (95% CI) 1 1.48 (0.97-2.25), P=0.06 1.39 (0.87-2.23), P=0.16 2.38 (1.22-4.66), P=0.01

1 1.37 (0.86-2.18), P=0.17 1.22 (0.71-2.08), P=0.45 0.35 (0.12-1.07), P=0.06

LFS: leukemia-free survival; HR: hazard ratio; CI: confidence interval; OS: overall survival; RI: relapse incidence; NRM: non-relapse mortality; GvHD: graft-versus-host disease.

A

C

B

D

E

Figure 1. Clinical outcome according to ABO compatibility status for the entire cohort. (A) Relapse incidence. (B) Non-relapse mortality. (C) Leukemia-free survival. (D) Acute graft-versus-host disease. (E) Overall survival.

haematologica | 2017; 102(6)

1069


J. Canaani et al.

were excluded from the analysis. Neither univariate nor multivariate analysis showed that ABO mismatching status is a significant independent predictor of patients’ outcome following haploidentical HCT.

ABO mismatching status does not affect the clinical outcome of peripheral blood-derived grafts used for haploidentical transplant

decreased overall survival (HR=1.82; 95% CI: 1.048 – 3.18; P=0.033) while there was a trend for increased 3-year grade II-IV acute GvHD rates in patients with minor ABO incompatibility (HR=2.01; 95% CI: 0.99 – 4.07; P=0.0504) (Online Supplementary Table S6).

Discussion

While previous publications indicated that clinical outcome is independent of stem cell source used for haploidentical HCT,21,22 we wondered whether ABO mismatching would have a differential effect on outcome in peripheral blood-mobilized grafts compared to bone marrow grafts. To this end, an analysis of the 378 patients transplanted with peripheral blood grafts (243 ABOmatched, 67 minor ABO-mismatched, and 68 major ABOmismatched patients) was carried out. As shown in Online Supplementary Table S4 and Table 4, there was no statistically significant association between ABO incompatibility status and clinical outcome following transplantation of peripheral blood-derived grafts (Figure 2). Since our cohort grafted with peripheral blood included only 17 patients with bi-directional ABO mismatching we repeated the univariate and multivariate analyses with exclusion of these patients, again confirming that ABO mismatching does not influence clinical outcome in peripheral blood-mobilized grafts.

ABO incompatibility affects overall survival and graftversus-host disease rates in haploidentical stem cell transplantation with bone marrow-derived grafts We then repeated the abovementioned analysis for the group of patients transplanted with bone marrow grafts (n=459). Univariate analysis (Online Supplementary Table S5) revealed that 3-year chronic GvHD rates were highest in patients with a minor ABO mismatch and lowest in ABO-matched patients (45.5% and 29.1%, respectively). Notably, in multivariate regression analysis with matched ABO patients as the reference group, minor ABO mismatching increased the risk of grade II-IV acute GvHD (HR=2.03; 95% CI: 1.007 - 4.1; P=0.047) (Table 5 and Figure 3). Subsequently the analysis was repeated with exclusion of the small group of 21 patients with bi-directional mismatching. On univariate analysis with matched ABO patients as the reference group, the chronic GvHD rate was again increased in patients with a minor ABO mismatch compared to ABO-matched patients (45.5% and 29.1%, respectively). Notably, in multivariate regression analysis with matched ABO patients as the reference group, patients with major ABO mismatching had

Haploidentical HCT is an innovative approach aimed to fill a substantial therapeutic gap for the significant population of patients without a related donor or a matched unrelated donor. Since initial experience with this approach showed that there is considerable risk of transplant-related complications,5 optimizing donor-recipient compatibility is of prime importance. In this analysis, the first of its kind for haploidentical HCT, we demonstrate that patients with major ABO incompatibility have inferior polymorphonuclear cell engraftment compared to both ABO-matched and minor-mismatched patients. Additionally, our data suggest that bi-directional ABO mismatching is associated with a significantly increased risk of grade II-IV acute GvHD. Furthermore our data indicate that patients transplanted with bone marrow grafts have an increased incidence of acute GvHD if there is minor ABO incompatibility, and decreased overall survival when major ABO incompatibility is present. Donor-recipient ABO incompatibility is nearly ubiquitous in transplantation as up to one-half of transplants involve some degree of mismatching.10,11 This places patients at an increased risk of acute and delayed hemolytic reactions, and delayed recovery of red blood cell function. While secondary clinical parameters such as gender, donor age, parity, and cytomegalovirus status are clearly minor factors in dictating patients’ outcome following transplantation in general,23 the precise role ABO incompatibility holds in this regard is unclear. In the present analysis we found that bi-directional ABO mismatching, namely the presence of antibodies directed against red blood cells in both donor and recipient, was associated with a significantly increased risk of grade II-IV acute GvHD. We do cautiously note the small number of patients with bi-directional incompatibility in our analysis (n=38) may limit the generalizability of these results to some degree. Our results are consistent with recently published data by Hefazi and colleagues12 who showed, in a cohort of 127 patients with AML or myelodysplastic syndromes (47 of whom were ABO mismatched), that the composite of major and bi-directional mismatching was also associated with a higher incidence of grade II-IV acute GvHD. However, when we analyzed the entire study

Table 4. Multivariable analysis of patients’ outcome following transplantation with peripheral blood-derived grafts.

Parameter

LFS HR (95% CI)

OS HR (95% CI)

RI HR (95% CI) NRM HR (95% CI)

Matched ABO (ref) Minor ABO mismatch

1 1.14 (0.78-1.67), P=0.46 1 (0.69-1.45), P=0.97

1 1.16 (0.79-1.71), P=0.44 1.01 (0.69-1.47), P=0.95

1 1 (0.6-1.77), P=0.88 1.25 (0.77-2.03), P=0.36

Major ABO mismatch

1 1.22 (0.73-2.02), P=0.44 0.84 (0.49-1.46), P=0.55

Acute GvHD Chronic GvHD grade II-IV HR (95% CI) HR (95% CI) 1 1.41 (0.85-2.34), P=0.17 1.53 (0.93-2.51), P=0.09

1 1.06 (0.58-1.93), P=0.84 1.09 (0.62-1.92), P=0.75

LFS: leukemia-free survival; HR: hazard ratio; CI: confidence interval; OS: overall survival; RI: relapse incidence; NRM: non-relapse mortality; GvHD: graft-versus-host disease.

1070

haematologica | 2017; 102(6)


ABO incompatibility in transplants for AML

cohort we were unable to find statistically significant associations between ABO mismatching and inferior norrelapse mortality and overall survival rates which the abovementioned group did find. These differences probably reflect differences in the populations of patients, graft sources, and possibly cohort sizes. An additional noteworthy finding in our study is the observation that patients with major ABO mismatching transplanted with bone marrow grafts had a lower overall survival rate than that of their ABO-matched counterparts. These findings are comparable with the recently published Center for International Blood and Marrow Transplant Research (IBMTR) experience with a large data set from over 5000 patients with AML or myelodys-

plastic syndromes indicating that major ABO incompatibility is associated with decreased overall survival (using related and unrelated matched donors).16 Our data also concur with their results in terms of the impact of ABO status on peripheral blood-mobilized grafts since neither analysis found any detrimental effect of ABO mismatching on clinical outcome following transplantation in this subgroup of patients. Notably, in a separate single institution (Stanford) retrospective analysis presented by the same authors,16 it was suggested that minor ABO incompatibility was closely associated with bone marrow grafts and these in turn were correlated with inferior overall survival and event-free survival, as well as increased non-relpase mortality rates. We did not find

B

A

C

D

E

Figure 2. Clinical outcome according to ABO compatibility status for patients transplanted with peripheral blood mobilized grafts. (A) Relapse incidence. (B) Non-relapse mortality. (C) Leukemia-free survival. (D) Acute graft-versus-host disease. (E) Overall survival.

haematologica | 2017; 102(6)

1071


J. Canaani et al.

minor ABO incompatibility of bone marrow grafts to be associated with these clinical outcome measures but rather we did find that minor ABO mismatching was correlated significantly with grade II-IV acute GvHD. These variances could be accounted for by considering the differences in the cohorts analyzed, ours being a uniform cohort of AML patients transplanted with haploidentical HCT while the Stanford analysis was not limited to AML

and consisted of standard matched related and unrelated donors. We were also interested in specifically examining the incidence of extensive chronic GvHD in our study as recent work from the UK in 594 patients undergoing reduced intensity conditioning with alemtuzumab suggested that the incidence of extensive chronic GvHD was increased in ABO-mismatched patients. We did not find a

Table 5. Multivariable analysis of patientsâ&#x20AC;&#x2122; outcome following transplantation with bone marrow grafts.

Parameter

LFS HR (95% CI) OS HR (95% CI)

RI HR (95% CI)

NRM HR (95% CI)

Matched ABO (ref) Minor ABO mismatch

1 0.83 (0.48-1.43), P=0.5 1.36 (0.83-2.23), P=0.21

1 0.69 (0.31-1.53), P=0.36 1.3 (0.61-2.76), P=0.49

1 0.99 (0.46-2.12), P=0.99 1.45 (0.74-2.84), P=0.27

Major ABO mismatch

1 0.89 (0.51-1.54), P=0.68 1.52 (0.92-2.51), P=0.09

Acute GvHD Chronic GvHD grade II-IV HR (95% CI) HR (95% CI) 1 2.03 (1-4.1), P=0.04 1.69 (0.85-3.33), P=0.12

1 1.88 (0.83-4.22), P=0.12 0.63 (0.21-1.85), P=0.4

LFS: leukemia-free survival; HR: hazard ratio; CI: confidence interval; OS: overall survival; RI: relapse incidence; NRM: non-relapse mortality; GvHD: graft-versus-host disease.

A

C

B

D

E

Figure 3. Clinical outcome according to ABO compatibility status for patients transplanted with bone marrow grafts. (A) Relapse incidence. (B) Non-relapse mortality. (C) Leukemia-free survival. (D) Acute graft-versus-host disease. (E) Overall survival.

1072

haematologica | 2017; 102(6)


ABO incompatibility in transplants for AML

similar association in our analysis, possibly because of the difference in patient composition between the analyses with the UK study also including patients with non-malignant conditions.17 One of the strengths of our analysis is the uniformity of the analyzed cohort since we focused our analysis solely on AML patients, differing to a significant degree from most prior publications in the field in which heterogeneous disease entities were analyzed with regard to the impact of ABO incompatibility. This may help to explain the divergence between some recent publications and ours. For example, an analysis of 414 patients with both malignant and non-malignant diagnoses using bone marrow, peripheral blood-, and cord blood-derived grafts failed to show a significant effect of ABO mismatching on patientsâ&#x20AC;&#x2122; outcome;15 in the same vein a study from Sweden looking at 310 patients with various hematologic diagnoses who underwent reduced intensity conditioning transplantation also did not show a substantial correlation between ABO status and clinical outcome.14 Interestingly, graft source may modify the influence of ABO mismatching, as emerging data with cord blood transplants in both the adult and pediatric setting also did not support a prognostic role for ABO status.13,24 To substantiate our findings we also conducted a sub-analysis of the impact of ABO status on T-cell-depleted grafts versus T-cell-replete grafts to determine whether there was a possible bias related to T-cell composition of the graft; as shown above, the T-cellrepletion status of the transplanted grafts had no effect on clinical outcome. The limitations of our study include that it is a multicenter, retrospective analysis with the inherent biases

References 1. Appelbaum FR. Pursuing the goal of a donor for everyone in need. N Engl J Med. 2012;367(16):1555-1556. 2. Wang Y, Liu QF, Xu LP, et al. Haploidentical vs identical-sibling transplant for AML in remission: a multicenter, prospective study. Blood. 2015;125(25):3956-3962. 3. Ciurea SO, Zhang MJ, Bacigalupo AA, et al. Haploidentical transplant with posttransplant cyclophosphamide vs matched unrelated donor transplant for acute myeloid leukemia. Blood. 2015;126(8):1033-1040. 4. Di Stasi A, Milton DR, Poon LM, et al. Similar transplantation outcomes for acute myeloid leukemia and myelodysplastic syndrome patients with haploidentical versus 10/10 human leukocyte antigenmatched unrelated and related donors. Biol Blood Marrow Transplant. 2014;20(12): 1975-1981. 5. Rizzieri DA, Koh LP, Long GD, et al. Partially matched, nonmyeloablative allogeneic transplantation: clinical outcomes and immune reconstitution. J Clin Oncol. 2007;25(6):690-697. 6. Ciurea SO, Saliba R, Rondon G, et al. Reduced-intensity conditioning using fludarabine, melphalan and thiotepa for adult patients undergoing haploidentical SCT. Bone Marrow Transplant. 2010;45(3):429436. 7. Bashey A, Zhang X, Sizemore CA, et al. T-

haematologica | 2017; 102(6)

8.

9.

10.

11.

12.

involved in analyzing retrospective datasets. In addition, it is conceivable that additional modifying factors which were not analyzed, such as the ABH secretor status,25 graft mononuclear cell content26 or the presence of donor-specific anti-HLA antibodies which significantly affect graft failure and rejection,27-29 mediate the effect of ABO incompatibility on the final clinical outcome of patients undergoing haploidentical HCT. Supported by a recently published clinical algorithm for donor selection in haploidentical transplantations which incorporates consideration of ABO compatibility,30 we cautiously propose that our findings may have future implications for clinical practice in terms of optimizing donor selection for AML patients undergoing haploidentical HCT, a supposition which would have to be confirmed in a controlled clinical trial. In conclusion, our findings suggest that in AML patients undergoing haploidentical HCT major ABO mismatching is associated with inferior engraftment and overall survival when bone marrow grafts are used. Additionally, patients with minor ABO mismatching may experience increased acute GvHD rates when transplanted with bone marrowderived grafts. Thus, ABO incompatibility status may hold prognostic significance and should be considered and assessed routinely during evaluation for the optimal donor prior to haploidentical HCT. Acknowledgments We thank all the European Group for Blood and Marrow Transplantation (EBMT) centers and national registries for contributing patients to the study and data managers for their excellent work. Supplementary information is available at the EBMT Web site.

cell-replete HLA-haploidentical hematopoietic transplantation for hematologic malignancies using post-transplantation cyclophosphamide results in outcomes equivalent to those of contemporaneous HLA-matched related and unrelated donor transplantation. J Clin Oncol. 2013;31(10): 1310-1316. Luo Y, Xiao H, Lai X, et al. T-cell-replete haploidentical HSCT with low-dose anti-Tlymphocyte globulin compared with matched sibling HSCT and unrelated HSCT. Blood. 2014;124(17):2735-2743. Di Bartolomeo P, Santarone S, De Angelis G, et al. Haploidentical, unmanipulated, G-CSF-primed bone marrow transplantation for patients with high-risk hematologic malignancies. Blood. 2013;121(5): 849857. Booth GS, Gehrie EA, Bolan CD, Savani BN. Clinical guide to ABO-incompatible allogeneic stem cell transplantation. Biol Blood Marrow Transplant. 2013;19(8): 1152-1158. Kimura F, Sato K, Kobayashi S, et al. Impact of AB0-blood group incompatibility on the outcome of recipients of bone marrow transplants from unrelated donors in the Japan Marrow Donor Program. Haematologica. 2008;93(11):1686-1693. Hefazi M, Litzow M, Hogan W, et al. ABO blood group incompatibility as an adverse risk factor for outcomes in patients with myelodysplastic syndromes and acute myeloid leukemia undergoing HLA-

13.

14.

15.

16.

17.

18.

matched peripheral blood hematopoietic cell transplantation after reduced-intensity conditioning. Transfusion. 2016;56(2):518527. Kudek MR, Shanley R, Zantek ND, McKenna DH, Smith AR, Miller WP. Impact of graft-recipient ABO compatibility on outcomes after umbilical cord blood transplant for nonmalignant disease. Biol Blood Marrow Transplant. 2016;22(11):2019-2024. Watz E, Remberger M, Ringden O, et al. Analysis of donor and recipient ABO incompatibility and antibody-associated complications after allogeneic stem cell transplantation with reduced-intensity conditioning. Biol Blood Marrow Transplant. 2014;20(2):264-271. Blin N, Traineau R, Houssin S, et al. Impact of donor-recipient major ABO mismatch on allogeneic transplantation outcome according to stem cell source. Biol Blood Marrow Transplant. 2010;16(9):1315-1323. Logan AC, Wang Z, Alimoghaddam K, et al. ABO mismatch is associated with increased nonrelapse mortality after allogeneic hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2015;21(4): 746-754. Brierley CK, Littlewood TJ, Peniket AJ, et al. Impact of ABO blood group mismatch in alemtuzumab-based reduced-intensity conditioned haematopoietic SCT. Bone Marrow Transplant. 2015;50(7):931-938. Giralt S, Ballen K, Rizzo D, et al. Reduced-

1073


J. Canaani et al.

19.

20.

21.

22.

1074

intensity conditioning regimen workshop: defining the dose spectrum. Report of a workshop convened by the Center for International Blood and Marrow Transplant Research. Biol Blood Marrow Transplant. 2009;15(3):367-369. Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus conference on acute GVHD grading. Bone Marrow Transplant. 1995;15(6):825-828. Filipovich AH, Weisdorf D, Pavletic S, et al. National Institutes of Health consensus development project on criteria for clinical trials in chronic graft-versus-host disease: I. Diagnosis and staging working group report. Biol Blood Marrow Transplant. 2005;11(12):945-956. Castagna L, Crocchiolo R, Furst S, et al. Bone marrow compared with peripheral blood stem cells for haploidentical transplantation with a nonmyeloablative conditioning regimen and post-transplantation cyclophosphamide. Biol Blood Marrow Transplant. 2014;20(5):724-729. Bradstock K, Bilmon I, Kwan J, et al.

23.

24.

25.

26.

Influence of stem cell source on outcomes of allogeneic reduced-intensity conditioning therapy transplants using haploidentical related donors. Biol Blood Marrow Transplant. 2015;21(9):1641-1645. Confer DL, Abress LK, Navarro W, Madrigal A. Selection of adult unrelated hematopoietic stem cell donors: beyond HLA. Biol Blood Marrow Transplant. 2010;16(1 Suppl):S8-S11. Konuma T, Kato S, Ooi J, et al. Effect of ABO blood group incompatibility on the outcome of single-unit cord blood transplantation after myeloablative conditioning. Biol Blood Marrow Transplant. 2014;20(4):577-581. Holbro A, Stern M, Infanti L, et al. Impact of recipient ABH secretor status on outcome in minor ABO-incompatible hematopoietic stem cell transplantation. Transfusion. 2015;55(1):64-69. Reshef R, Huffman AP, Gao A, et al. High graft CD8 cell dose predicts improved survival and enables better donor selection in allogeneic stem-cell transplantation with

27.

28.

29.

30.

reduced-intensity conditioning. J Clin Oncol. 2015;33(21):2392-2398. Chang YJ, Zhao XY, Xu LP, et al. Donorspecific anti-human leukocyte antigen antibodies were associated with primary graft failure after unmanipulated haploidentical blood and marrow transplantation: a prospective study with randomly assigned training and validation sets. J Hematol Oncol. 2015;8:84. Gladstone DE, Zachary AA, Fuchs EJ, et al. Partially mismatched transplantation and human leukocyte antigen donor-specific antibodies. Biol Blood Marrow Transplant. 2013;19(4):647-652. Yoshihara S, Maruya E, Taniguchi K, et al. Risk and prevention of graft failure in patients with preexisting donor-specific HLA antibodies undergoing unmanipulated haploidentical SCT. Bone Marrow Transplant. 2012;47(4):508-515. Chang YJ, Luznik L, Fuchs EJ, et al. How do we choose the best donor for T-cell-replete, HLA-haploidentical transplantation? J Hematol Oncol. 2016;9:35.

haematologica | 2017; 102(6)


ARTICLE

Acute Lymphoblastic Leukemia

Targeting the 5T4 oncofetal glycoprotein with an antibody drug conjugate (A1mcMMAF) improves survival in patient-derived xenograft models of acute lymphoblastic leukemia

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Owen J. McGinn,1 Shekhar Krishnan,2,3 Jean-Pierre Bourquin,4 Puja Sapra,5 Clare Dempsey,2 Vaskar Saha2,3 and Peter L. Stern1

1 Immunology & 2Paediatric Oncology, Division of Molecular & Clinical Cancer Sciences, University of Manchester, UK; 3Tata Translational Cancer Research Center, Tata Medical Center, Kolkata, India; 4Division of Oncology & Children's Research Center, University Children’s Hospital, University of Zurich, Switzerland and 5Pfizer Inc. Pearl River, NY10965-1299, USA

Haematologica 2017 Volume 102(6):1075-1084

ABSTRACT

O

utcome in childhood acute lymphoblastic leukemia is prognosticated from levels of minimal residual disease after remission induction therapy. Higher levels of minimal residual disease are associated with inferior results even with intensification of therapy, thus suggesting that identification and targeting of minimal residual disease cells could be a therapeutic strategy. Here we identify high expression of 5T4 in subclonal populations of patient-derived xenografts from patients with high, post-induction levels of minimal residual disease. 5T4-positive cells showed preferential ability to overcome the NOD-scidIL2Rγnull mouse xenograft barrier, migrated in vitro on a CXCL12 gradient, preferentially localized to bone marrow in vivo and displayed the ability to reconstitute the original clonal composition on limited dilution engraftment. Treatment with A1mcMMAF (a 5T4-antibody drug conjugate) significantly improved survival without overt toxicity in mice engrafted with a 5T4-positive acute lymphoblastic leukemia cell line. Mice engrafted with 5T4-positive patient-derived xenograft cells were treated with combination chemotherapy or dexamethasone alone and then given A1mcMMAF in the minimal residual disease setting. Combination chemotherapy was toxic to NOD-scidIL2Rγnull mice. While dexamethasone or A1mcMMAF alone improved outcomes, the sequential administration of dexamethasone and A1mcMMAF significantly improved survival (P=0.0006) over either monotherapy. These data show that specifically targeting minimal residual disease cells improved outcomes and support further investigation of A1mcMMAF in patients with high-risk B-cell precursor acute lymphoblastic leukemia identified by 5T4 expression at diagnosis.

Introduction Acute lymphoblastic leukemia (ALL) is the most common cancer of childhood. Intensive combination chemotherapy produces cure rates of ~90% but is associated with considerable morbidity. Recent high throughput genetic analyses have successfully identified new genetic subtypes1,2 as well as aberrant epigenetic3 and signaling pathways4,5 that are potential targets for precision therapy.6 The paradigm for targeted treatment is the subset of children with a BCR-ABL1 fusion in whom the addition of the tyrosine kinase inhibitor imatinib to intensive chemotherapy improved outcomes significantly.7,8 More recently, immunological therapy, targeting antigens expressed by B cells using monoclonal antibodies with or without payloads9 and/or activating cytotoxic T cells, is showing great promise.10 Thus we are now on the cusp of a change from iteratively derived non-specific chemotherapy to a designed, targeted approach. haematologica | 2017; 102(6)

Correspondence: Peter.Stern@cruk.manchester.ac.uk or vaskar.saha@tmckolkata.com Received: October 18, 2016. Accepted: March 15, 2017. Pre-published: March 24, 2017. doi:10.3324/haematol.2016.158485 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1075 ©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.

1075


O.J. McGinn et al.

We recently reported that the 5T4 oncofetal glycoprotein [also known as trophoblast glycoprotein (TPBG) and WNT-activated inhibitory factor 1 (WAIF1)] is upregulated in high-risk cytogenetic subgroups and overexpressed on the plasma membrane of lymphoblasts obtained at relapse, in patients with B-cell precursor (BCP) ALL.11 5T4 is a 72-kDa N-glycosylated transmembrane protein expressed by syncytiotrophoblasts in the placenta. Most mature cells, including lymphoid cells, do not express it. 5T4 is associated with differentiating embryonic stem cells,12,13 and mechanistically associated with the directional movement of cells through the regulation of epithelial facilitation of mesenchymal transition,12-14 CXCL12/CXCR4 chemotaxis15,16 and favoring non-canonical over canonical WNT/β−catenin pathway signaling.17,18 5T4 is expressed by tumor-initiating cells in human nonsmall cell carcinomas19 and by a number of carcinomas.20 The selective pattern of 5T4 tumor expression, its association with a tumor-initiating phenotype plus a mechanistic involvement with cancer spread has stimulated the development of 5T4 vaccine, 5T4 antibody targeted– superantigen and 5T4 antibody-drug conjugate (ADC) therapies through preclinical and into clinical studies.21,22 The ADC is a 5T4 humanized monoclonal antibody (A1) linked by sulfydryl-based conjugation delivering a microtubule-disrupting agent, monomethyl auristatin F (MMAF) via a maleimidocaproyl (mc) linker. A1mcMMAF has shown potent activity in a variety of solid tumor models, with induction of long-term regression after the last dose and no significant toxicity in a simian model23 and tolerable toxicity in patients with solid tumors.24 Murine models of childhood ALL suggest that minimal residual disease (MRD) after therapy is represented by a rare cell population that combines the phenotypes of bone marrow microenvironment-mediated dormancy, stemness, and in vivo drug resistance.25 We previously reported that a BCP-ALL cell line had a subpopulation of cells that expressed 5T4 (5T4+) and these cells showed migration on a CXCL12 axis and a differential dissemination and infiltration in a mouse model when compared to the 5T4-negative (5T4–) subpopulation. A 5T4 mouse antibody targeted superantigen combined with human peripheral blood mononuclear cells showed activity in vitro and in vivo.11 Here we report that 5T4+ subclones are present in expanded numbers in patient-derived xenograft (PDX) samples obtained from patients with high levels of post-induction MRD. When used in combination with chemotherapy, A1mcMMAF showed specific activity in 5T4+ PDX models of ALL.

Methods B-cell precursor acute lymphoblastic leukemia patientderived xenograft samples Two categories of PDX samples were assessed, based on the reported post-induction MRD levels in the patients from whom they were derived. For the purposes of this paper, those with MRD levels ≤10-4 (SR03, SR014, SR_M1 and SR_M2) were classified as standard risk (SR) and those with higher levels (HR08, VHR03, HR_M1 and HR_M2) as high risk (HR) (Online Supplementary Table S1).26 PDX cells (1-2x106, or fewer) obtained from splenic fractions were transplanted intravenously into nonirradiated 6- to 10-week old NOD-scidIL2Rγnull (NSG) mice (group 1076

size 6 unless stated). Engraftment was assessed in tail vein blood samples using flow cytometry. Comparative engraftment rates, sites of engraftment and tumor load were assessed by harvesting animals at various time points. Relative clonogenicity was determined by limiting dilution engraftment analysis in NSG mice as previously described.27 The proportions of 5T4 leukemic blasts were quantified by flow cytometry. Studies were approved by Thameside and Glossop Research Ethics Committees (Manchester, UK: Reference 07/Q1402/56). All procedures conformed with the regulations of the UK Animal License Act.

Flow cytometry Individual or multiplex flow cytometry studies for 5T411 and human specific CD45 (CD45 APC-Cy7 or PE-Cy7) and CD34 (FITC or PE-Cy7) (all from eBioscience; Hatfield, UK), were performed with directly conjugated isotope and appropriate controls for individual and multiplex analyses; cross-channel fluorescence was compensated using the FlowJo matrix software (FlowJo LLC, OR, USA). In vivo leukemia engraftment was analyzed by human CD45 flow cytometry using 25 μL of heparinized peripheral blood after lysis of the red blood cells (eBioscience). The overall disease burden was determined by expression of the ratio of human to mouse CD45+ blasts per sample. Analyses of peripheral blood cellular components were performed using an XE-2100 automated hematology system (Sysmex, Milton Keynes, UK).

Migration assays The migration assays were performed as previously described.11

5T4 depletion Depletion and enrichment of 5T4+ blasts from PDX samples was performed using magnetic-activated cell sorting (MACS) microbeads and columns from Miltenyi Biotec (Surrey, UK), and a 5T4-specific monoclonal antibody11 conjugated to a PE fluorochrome using the EasyLink R-Phycoerythrin Conjugation Kit from Abcam (Cambridge, UK).

Antibody-drug conjugate therapy Sup5T4 Lenti/Luc/mCherry leukemia cells11 (5x106) were given intraperitoneally and different BCP-ALL PDX samples at various doses were given intravenously to NSG mice. Mice were treated with either A1mcMMAF or control-ADC (Neg-8-8hG1mcMMAF) at a dose of 5 mg/kg intraperitoneally beginning 7 days after tumor challenge with a cycle of three or four doses of ADC given at 4-day intervals (treatment block of 12-16 days) and in some cases further ADC cycles were repeated after a gap of 1 week, or mice were given no therapy23 (Online Supplementary Table S2). Sup5T4 B-ALL intraperitoneal challenge was monitored by IVIS.11 For other PDX samples tumor engraftment was assessed by peripheral blood monitoring and, at an appropriate point, comparison of tumor burden assessed by flow cytometry in cell suspensions from spleen and bone marrow. Two weeks of a vincristine, dexamethasone and asparaginase (VXL) protocol, established in NSG mice,28 was followed by various cycles of ADC starting at different times after HR08 primagraft challenge. A similar protocol using dexamethasone alone followed by A1mcMMAF was also tested. Efficacy was determined by measurement of BCP-ALL engraftment and survival.

Statistical analysis With the exception of long-term survival analyses the data presented are representative of three or more experiments. Graph-Pad Prism software was used for individual or multiple group comparisons by a two-tailed Student t-test or ANOVA-Tukey; survival haematologica | 2017; 102(6)


Targeting residual ALL subclones

analysis was based on Kaplan-Meier plots, log-rank Mantel-Cox and regression analysis for hazard ratios.

Results 5T4-expressing patient-derived xenografts have an aggressive phenotype in NSG mice Four SR and four HR BCP-ALL PDX samples were analyzed by flow cytometry for cell surface expression of 5T4. All four HR B-ALL had significant (20-80%) proportions of 5T4+ cells while 5T4 was undetectable in all four SR PDX samples (Figure 1A). The overall leukemogenicity was generally reflected in the rate of tumor engraftment and time to morbidity as recipients of HR primagrafts with >50% 5T4 positivity showed earlier engraftment and death than those which received SR primagrafts (Figure 1B). When pooling the data, the engraftment of 5T4+ HR PDX was significantly faster than that of 5T4– SR PDX, with median survivals of 71 versus 280 days, respectively (P<0.0001).

5T4 expression correlates with leukemia engraftment and time to morbidity in vivo In this study all 5T4+ PDX belonged to the MRD-positive HR group. The ability of high MRD to overcome the xenograft barrier,26 leading to shorter times to leukemia in NSG mice, has already been described.29 To further investigate leukemic engraftment according to the 5T4 phenotype, depletion and enrichment of 5T4 were performed on HR08 PDX cells using a MACS system and a 5T4-specific monoclonal antibody. The efficacy of MACS separation of 5T4+ blasts is illustrated in Figure 2A. From a starting population of 76.8% 5T4+ leukemic cells, there was ≥96% enrichment or depletion. Transplantation of 2x106 mock, 5T4-depleted or -enriched cells led to a significant delay (P<0.05) in engraftment kinetics for 5T4-depleted blasts (Figure 2B). At morbidity, post-mortem analyses of splenic content revealed that all animals, independently of frac-

A

tionation, exhibited recapitulation of the parental heterogeneous HR08 phenotype, with approximately 75% of the blasts expressing 5T4 (Online Supplementary Figure S1). The delay in engraftment did not correlate with improved survival of the animals (Figure 2C). Such subclonal recapitulation of the original leukemic population with similar survival times has been previously described, reflecting the plasticity of the BCP-ALL cell.30 It is also possible that MACS separation was incomplete and up to 80,000 residual 5T4+ blasts persisted in the depleted challenge doses. A limited dilution experiment performed using either a 1,000- or 100-cell challenge comparing mock versus 5T4depleted engraftment showed a significant impact of 5T4 depletion on engraftment (Figure 3A). This also translated into significantly improved survival (Figure 3B). Thus in the HR08 BCP-ALL PDX, 5T4+ blasts were the most clonogenic of a heterogeneous leukemic population.

Chemotaxis of 5T4-positive patient-derived xenograft cells on a CXCL12 gradient In normal culture, SupB15 BCP-ALL 5T4+ cells differentiate to produce 5T4–cells with concurrent loss of CD34 expression and a more mature phenotype.11 These changes are concordant with reduced CXCL12-mediated chemotaxis, lower production of matrix metalloproteases and modulation of the activation state of integrins, consistent with a reduced capacity to populate extramedullary sites.11 The responses of PDX cells to a CXCL12 chemokine gradient or a positive control gradient of fetal calf serum were examined. All PDX samples migrated actively along a fetal calf serum gradient, confirming the functional viability of the cells (Figure 4A). HR but not SR PDX cells were chemotactically responsive to CXCL12; this responsiveness was inhibited by pre-incubation with 5T4-specific monoclonal antibody but not the IgG1 control (Figure 4B). Further evidence that 5T4+ blasts were the predominant responders to the chemokine gradient was provided by preferential accumulation of 5T4-expressing PDX cells in the chemoattractant chamber in response to CXCL12 but not fetal calf serum (Online Supplementary Figure S2A). NSG

B

Figure 1. Engraftment kinetics of B-cell precursor acute lymphoblastic leukemia patient-derived xenograft mice according to 5T4 phenotype. (A) The expression of 5T4 at the surface of four HR ( blue symbols) and four SR (red symbols) or relapse primagrafts was determined by flow cytometry and representative histograms plotted using FlowJo. The mean percentages of 5T4+ blasts from three separate experiments (± SEM) were plotted to compare expression across risk stratification (unpaired, 2-tailed t-test; P<0.0001). (B) Kaplan Meier plots of the individual HR (28 mice) versus SR (20 mice) transplanted animals. When the data from SR and HR challenged mice were pooled the median survival was 71 versus 280 days, respectively (P<0.0001); hazard ratio (95% confidence interval) = 8.84 (4.2-18.5).

haematologica | 2017; 102(6)

1077


O.J. McGinn et al.

mice have high expression of CXCL12 in the bone marrow.31 In the HR08 PDX model, peripheral blood and harvested spleens of NSG mice were composed of approximately 75% 5T4+ blasts, while >95% of the HR08 cells recovered from femoral flushes expressed 5T4 (Online Supplementary Figure S2B). Thus, there is both ex vivo and in vivo evidence suggesting that 5T4+ blasts respond to a CXCL12 gradient.

In vivo A1mcMMAF therapy of the Sup5T4 B-cell precursor acute lymphoblastic leukemia cell line

A

B

To assess the efficacy of A1mcMMAF, mice were engrafted intraperitoneally with Sup5T4 Lenti/Luc/mCherry11 (5x106 cells) and treatment initiated 1 week later with control-ADC or A1mcMMAF. Bioluminescent imaging used to monitor leukemic engraftment showed that both one and two cycles of

C

Figure 2. Engraftment of 5T4-depleted and -enriched HR08 blasts in NSG mice. (A) HR08 blasts were separated by surface expression of 5T4 and resultant populations were determined to be 97%-depleted and 96%-enriched, respectively. (B) Depleted, enriched and mock-depleted populations of HR08 (1x106) were transplanted into NSG mice. The rate of engraftment monitored by the detection of hCD45 cells in the peripheral blood demonstrated a significant impact of 5T4 depletion on engraftment (ANOVA-Tukey; P<0.05). (C) KaplanMeier plots showed no significant differences in time to morbidity of the groups receiving the different fractionated leukemia blasts.

A

B

Figure 3. Engraftment of 5T4- and mock-depleted HR08 blasts in NSG mice. 5T4-depleted and mock-depleted populations of HR08 (1000 and 100 cells) were transplanted into NSG mice. (A) 5T4-specific depletion had a significant impact on engraftment (hCD45 cells in the peripheral blood) with both challenge doses (ANOVA/Tukey; P<0.001). (B) Kaplan-Meier plots of time to morbidity for each fractionated population. 5T4-depleted compared to mock-treated cells showed significant differences in survival for both doses of tumor (log-rank Mantel-Cox; for 1000 cells: P=0.001; hazard ratio (95% confidence interval) = 20.1 (3.4-120.8); for 100 cells: P=0.003; hazard ratio = 17.3 (2.7-110.4).

1078

haematologica | 2017; 102(6)


Targeting residual ALL subclones

A1mcMMAF were effective in limiting tumor growth compared to control-ADC (Figure 5A,B). Control-ADC had some impact in comparison to the untreated tumors (Figure 5B), possibly through Fc binding on the blasts in the peritoneal cavity (untreated versus 1 cycle or 2 cycle control-ADC; P<0.05 and P<0.01, respectively). 5T4-targeted ADC treatments significantly prolonged survival when compared to untreated or control-ADC (P=0.04) (Figure 5C). Although the median survival of animals that received two cycles of A1mcMMAF was 133 days, compared to 113 days in those that received one cycle, this difference was not statistically significant (P=0.25). Cessation of the administration of A1mcMMAF (day 37 for second

cycle) correlated with increased tumor growth and mCherry-positive leukemia cells isolated after therapy from the ovary, a common extramedullary site of Sup5T4 tumor spread,11 retained the parental 5T4 phenotype (Online Supplementary Figure S3). This suggests the potential for continuing multiple cycles of therapy.

In vivo A1mcMMAF therapy of B-cell acute lymphoblastic leukemia patient-derived xenografts A significant therapeutic effect of A1mcMMAF on a leukemic cell line that uniformly expresses 5T4 may not be as effective in leukemia with heterogeneous 5T4 expression. To test this, mice were challenged intra-

A

B

Figure 4. CXCL12 chemotaxis of Bcell precursor acute lymphoblastic leukemia patient-derived xenograft cells. Transwell migration of two SR and two HR PDX samples in response to a 10% fetal calf serum gradient or 12.5 nM CXCL12 was monitored kinetically using a modified Boyden chamber system. (NG= no gradient). (A) All PDX samples were responsive to the serum gradient (blue lines) compared to no gradient (NG; black lines) (2-tailed t-test; P<0.001) but (B) only HR primagrafts responded to CXCL12 (2-tailed t-test; P<0.05); this chemotaxis was blocked by pre-incubating cells with a monoclonal antiboby (mAb) to 5T4 (red lines) but not an isotype IgG control (blue lines) or with NG (black lines).

haematologica | 2017; 102(6)

1079


O.J. McGinn et al.

venously with 2x106 HR08 (~75% 5T4+) cells, and 1 week later either remained untreated or received three cycles (4 doses every 4 days, repeated after 1 week, last dose on day 61) of control or A1mcMMAF. The efficacy of therapy was monitored by weekly analysis of peripheral blood for the presence of human tumor cells as a measure of leukemia engraftment. A1mcMMAF administration significantly delayed engraftment (day 66; P=0.001), but did not influence overall survival (Online Supplementary Figure S4A,B). Similar results were seen with VHR03 PDX (~56% 5T4+) given at a dose of 2x106 intravenously followed by two cycles of ADC (Online Supplementary Figure S5). Postulating that this could be due to an excess of tumor cells injected, the experiment was repeated with transplantation of 2x103 HR08 cells, followed 1 week later by three cycles of therapy. At this transplantation dose level, A1mcMMAF treatment significantly reduced engraftment (P=0.02) and prolonged survival (P=0.0014) (Figure 6A,B). No effects of the control-ADC were observed in the PDXchallenged mice.

fered rapid and unpredicted adverse reactions after the first course of VXL, prior to the administration of A1mcMMAF, requiring immediate cessation of the experiment (Online Supplementary Figure S6A). Upon autopsy organs displayed signs consistent with tumor lysis syndrome,32 and so in a subsequent experiment VXL therapy was delivered 1 week after challenge to reduce the target tumor load. Two weeks of VXL therapy was followed by three cycles (4 doses every 4 days, last day of treatment day 75) of control or A1mcMMAF treatment. At day 66, animals in the untreated group had a mean tumor engraftment of 69.2% ± 3.1 while all treated groups exhibited little if any evidence of leukemia in the peripheral blood (Online Supplementary Figure S6B). Leukemic cells were present in peripheral blood samples by day 85 in both VXL and VXL/control-ADC treated animals with mean engraftment of 18.7% ± 7.2 and 7.9% ± 4.8, respectively. No circulating blasts were detected in the peripheral blood of the mice treated with the combination therapy (Online Supplementary Figure S6C). Despite the apparent success of these therapies in reducing leukemia engraftment, no significant impact on overall survival was observed in any of the treated groups (Online Supplementary Figure S6D). Although animals in treated groups reached morbidity at the same time as untreated animals, they did so for very different reasons. The most typical post-mortem sign of leukemia engraftment, marked splenic enlargement, was

A1mcMMAF in combination with chemotherapy for B-cell acute lymphoblastic leukemia primagraft treatment Next, we investigated the impact of A1mcMMAF in the context of VXL therapy. Unexpectedly, animals transplanted with 1x106 HR08 cells and treated 4 weeks later suf-

A

C

1080

B

Figure 5. A1mcMMAF monotherapy of Sup5T4 cells in vivo. Animals were challenged with Sup5T4 cells intraperitoneally at day 0 and received either no treatment or one or two cycles of A1mcMMAF or one or two cycles of control-ADC treatment starting after 1 week. (A) IVIS images of tumor growth at day 43. Growth of tumors was quantified using log radiance (photons/sec/cm2/sr) =photons. A1mcMMAF produced significant growth control: ANOVA-Tukey: untreated versus one cycle or two cycles A1mcMMAF; P<0.0001; Control-ADC one or two cycles versus A1mcMMAF one or two cycles, respectively: P<0.05 and P<0.01. (C) Kaplan-Meier plots show that only A1mcMMAF (one or two cycles) but not the control-ADC treatments influenced the overall survival. Log-rank Mantel-Cox shows significant effects compared to untreated animals of one and two cycles of A1mcMMAF, respectively (P=0.04; hazard ratio: 6.3 (1.08-36.52) and P=0.002; hazard ratio: 24.14 (3.36-173.4) and no significant differences of control-ADC treatments. Dotted vertical lines represent timing of doses of ADC therapy (see Online Supplementary Table S2).

haematologica | 2017; 102(6)


Targeting residual ALL subclones

only evident in animals that had remained untreated. All animals that received VXL therapy reached morbidity with normal or only marginally enlarged spleens, indicating greatly reduced tumor load. Indeed by the termination of the experiment no signs of leukemia had been detected in the marrow, spleen or peripheral blood of any animal that had received VXL therapy and reached morbidity prematurely.

was noted at days 75 and 100 in untreated and dexamethasone-treated mice, respectively, and at day 108 all groups receiving dexamethasone had significantly fewer circulating blasts in peripheral blood samples compared to the untreated group (Figure 7A). By day 129, engraftment in the dexamethasone only, and dexamethasone followed by control-ADC groups had increased and was no longer statistically different from that in the untreated group while the peripheral blood of the animals treated with A1mcMMAF/dexamethasone combination therapy remained tumor-free (Figure 7B). By day ~170 engraftment levels in the dexamethasone and untreated groups were similar, being ~90%, compared to <20% in the group treated with the A1mcMMAF/dexamethasone combination (Figure 8A). The median survival times of untreated, control-ADC/dexamethasone-, dexamethasone- and A1mcMMAF/dexamethasone-treated animals were 169.5, 144.5, 220 and >350 days, respectively. While dexamethasone treatment was able to confer a significant survival advantage (P=0.006) when employed as monotherapy, the greatest impact on survival was observed when it was given in combination with A1mcMMAF (P=0.0006)

Dexamethasone and A1mcMMAF treatment To minimize the adverse reactions seen with VXL, an alternative combination protocol of dexamethasone followed by four cycles of control or A1mcMMAF was tested. Furthermore, general indicators of normal hematopoiesis, i.e. hemoglobin, total red and white blood cell and reticulocyte counts, were measured. Accordingly, 2x103 HR08 PDX cells were transplanted. A week later, dexamethasone was given daily (Monday-Friday) for 2 weeks, followed by four cycles of control or A1mcMMAF (4 doses every 4 days, last dose day 96). No toxicity or myelosuppression was observed during the course of this experiment (Online Supplementary Figure S7). Engraftment

B

A

Figure 6. A1mcMMAF monotherapy of HR08 B-cell acute lymphoblastic leukemia patient-derived xenotransplant challenge. (A) At a lower tumor challenge of 2000 HR08 cells only A1mcMMAF significantly reduced engraftment (ANOVA/Tukey; P=0.02) which corresponded with (B) a significant improvement in overall survival with A1mcMMAF treatment (log-rank Mantel-Cox; P=0.0014, hazard ratio: 21.55 (95% confidence interval 3.73-124.5). Dotted vertical lines represent timing of doses of ADC therapy (see Online Supplementary Table S2). Untreated, black symbols/line; control-ADC, red symbols/line and 5T4-ADC blue symbols/line.

B

A

haematologica | 2017; 102(6)

Figure 7. Combination dexamethasone chemotherapy and A1mcMMAF treatment of HR08 patient-derived xenograft: early engraftment. (A) The percentage of peripheral blood blasts was significantly reduced at day 108 in NSG mice engrafted with 2x104 HR08 cells administered dexamethasone (DEX) therapy 1 week after transplantation (ANOVA-Tukey; P<0.05 for DEX and DEX/CTRL-ADC, P<0.001 for DEX/A1mcMMAF). (B) At day 129 engraftment of HR08 blasts had increased in the DEX and DEX-control treated groups and were comparable to that of the untreated group. There was significantly less engraftment in the DEX/A1mcMMAF group when compared to the untreated group (P<0.001), and the other DEX-treated groups (P<0.001 vs. DEX and P<0.05 vs. DEX/CTRL-ADC).

1081


O.J. McGinn et al.

Our data show that development of leukemia was faster with 5T4-enriched cells suggesting that cells that express this protein are better adapted to overcome the xenograft barrier. 5T4+ ALL blasts migrate on a CXCL12 axis and favor migration to the bone marrow environment. The hematopoietic stem cell compartment in NSG mice is thought to facilitate donor over host cell engraftment,34 a potential explanation for the rapid engraftment of 5T4+ cells. In murine xenotransplantation models of childhood ALL, reconstitution of leukemia recapitulates the clinical manifestations of the disease with more rapid engraftment times correlating with a higher risk of therapeutic failure.26,29 Although our sample size was small, in the model described here, PDX selection was based on the MRD response to therapy. Subclonal populations expressing significant amounts of 5T4 were seen only in the four MRDhi PDX and our previous studies linked 5T4 expression to high risk of relapse in pediatric BCP-ALL patients.11 The bone marrow microenvironment provides a protective niche for ALL cells25,35-38 and in the context of this study suggests that 5T4+ cells in the protective bone marrow microenvironment niche may survive chemotherapy and contribute significantly to the MRD population that has recently been described.25 Using the MLL primagraft with the highest proportion of 5T4+ blasts, 5T4-specific antibody/magnetic bead depletion and limiting dilution chal-

lenge in NSG mice clearly demonstrated that 5T4+ blasts are the most clonogenic in vivo and consistent with the concept of a leukemia-initiating cell.30 Concomitant with earlier reports on ALL xenografts, limiting dilution studies show that 5T4+ cells are able to recapitulate the original leukemic population demonstrating the considerable plasticity of the ALL cell.26,27 In the context of MRD, residual 5T4+ cells could therefore give rise to disease recurrence in patients. Current treatment strategies for ALL use non-specific cytotoxic drugs. Recently immunological therapy targeting antigens expressed specifically on the surface of the B cell have generated considerable interest. The target in the majority of these trials has been CD19, expressed on malignant and non-malignant B cells.39-41 As the antigens are expressed on non-malignant cells, they are also associated with prolonged B-cell suppression as well as the emergence of escape clones no longer expressing CD19. Ideally the target should be expressed selectively by leukemic subclone(s) that remain after therapy and give rise to recurrences. The problem is that not a great deal is known about surface antigens associated with a resistant phenotype. Recently, a chimeric antigen receptor T-cell approach was shown to be effective in targeting the thymic stromal lymphopoietic receptor (TLSPR).42 Like 5T4, TLSPR is expressed primarily by malignant subclones that persist after therapy. In this study, we modeled this by first transplanting 5T4+ PDX, treating with chemotherapy and then, in the MRD setting in which blasts were not detectable in peripheral tail vein bleeds, administering A1mcMMAF. A1mcMMAF significantly delayed the emergence of leukemic blasts in both VXL- and dexamethasone-treated mice. Although we previously reported that HR08 is resistant to steroids ex vivo,26 dexamethasonetreated PDX mice showed an initial response, followed by disease kinetics similar to that of untreated mice. In contrast, mice receiving 4 weeks of 5T4 showed a significantly longer latency in disease recurrence and a significant survival advantage, suggesting there is benefit from using

A

B

(Figure 8B). The outcomes of dexamethasone/controlADC-treated mice appeared to be inferior to those of untreated animals (P=0.012). We speculate that this is a consequence of dexamethasone-induced upregulation of Fc receptors33 mediating a non-specific uptake of the microtubule-disrupting agent, leading to increased toxicity and death. The specificity of A1mcMMAF targets the drug to leukemic cells, presumably avoiding such toxicity in the animals treated with dexamethasone and A1mcMMAF.

Discussion

Figure 8. Combination dexamethasone chemotherapy and A1mcMMAF treatment of HR08 patient-derived xenograft: engraftment kinetics and survival. (A) The engraftment of HR08 blasts remained significantly lower (P<0.001) in the group treated with dexamethasone (DEX)/A1mcMMAF than in all other groups for at least 172 days. (B) A significant improvement in overall survival compared to untreated animals (log-rank Mantel-Cox) was seen with DEX: P=0.006; hazard ratio (95% confidence interval): 9.6 (1.9-47.6); DEX/A1mcMMAF P=0.0006; hazard ratio (95% confidence interval): 21.55 (3.7-124.5). DEX/control-ADC treated animals showed slightly poorer survival than untreated animals: P=0.012; hazard ratio (95% confidence interval): 0.13 (0.03-0.63). Hatched areas and dotted vertical lines represent timing of chemotherapy and doses of ADC therapy, respectively (see Online Supplementary Table S2). Black circles/line, untreated; purple triangles/line, DEX alone; red diamonds/line DEX-control-ADC; blue circles/line DEX-5T4-ADC.

1082

haematologica | 2017; 102(6)


Targeting residual ALL subclones

repeated cycles. In fact a mcMMAF-conjugated anti-CD19 (denintuzomab mafodotin) ADC, given at 3-week intervals has produced remission rates of 35% with acceptable toxicities in an ongoing phase I study.43 While NOD-SCID and NSG mice are now the xenografts of choice for leukemia PDX models, our results suggest that there are drawbacks that may require some caution in the interpretation of the effects of drugs. An increased tumor load in NSG mice led to tumor lysis, requiring an altered cell dose and treatment schedule. Clearly engraftment kinetics differs with patients’ samples and passages, and needs to be established before the efficacy of therapy can be assessed. In leukemia–engrafted NSG mice, treatment with VXL therapy was associated with mortality despite absence of post-mortem leukemia. In contrast to previous reports,28 morbidity and mortality in these mice was neither due to central nervous system infiltration nor attributable to impaired murine haematopoiesis. In general our experience is that NSG mice are less tolerant of physical handling compared to the NOD-SCID strain. Children treated with combination chemotherapy for ALL require supportive therapy to avoid treatment-related mortality. We speculate that a combination of frequent intraperitoneal administrations, influences of multiple cytotoxic drugs, the aggressive expansion and subsequent necrosis of malignant cells combine to produce the morbidity observed. 5T4+ blasts home toward CXCL12 in vitro and we speculate that this is a contributing factor to their engraftment capacity, as demonstrated by the enrichment of 5T4+ blasts in NSG mouse femora. Furthermore, a specific monoclonal antibody to 5T4 was shown to interfere with CXCL12 chemotaxis of HR B-ALL patient-derived primagraft cells. This may be of clinical relevance when considering ways to increase the exposure of leukemia cells to cytotoxic drugs. A CXCR4 inhibitor, AMD3100, has been used as a means of mobilizing leukemic blasts from the bone marrow systemically to increase the relative bioavailability of chemotherapy.44 A limitation of such therapy is that CXCR4 is a chemokine receptor widely expressed by many cell lineages. Since normal tissue levels of 5T4 are low, if its influence on chemotaxis could be specifically targeted it

References 6. 1. Zhang J, Ding L, Holmfeldt L, et al. The genetic basis of early T-cell precursor acute lymphoblastic leukaemia. Nature. 2012;481(7380):157-163. 2. Den Boer ML, van Slegtenhorst M, De Menezes RX, et al. A subtype of childhood acute lymphoblastic leukaemia with poor treatment outcome: a genome-wide classification study. Lancet Oncol. 2009;10(2): 125-134. 3. Figueroa ME, Chen SC, Andersson AK, et al. Integrated genetic and epigenetic analysis of childhood acute lymphoblastic leukemia. J Clin Invest. 2013;123(7):30993111. 4. Roberts KG, Li Y, Payne-Turner D, et al. Targetable kinase-activating lesions in Phlike acute lymphoblastic leukemia. N Engl J Med. 2014;371(11):1005-1015. 5. Chen Z, Shojaee S, Buchner M, et al. Signalling thresholds and negative B-cell

haematologica | 2017; 102(6)

7.

8.

9.

10.

might allow disruption of CXCR4 function confined more specifically to malignant hematopoietic cells. The bulk of ALL cells are chemosensitive and morphological remission is achieved using a combination of three or four drugs in ~98% of cases. The current strategy for childhood ALL consists of intensifying therapy for those with high MRD levels, continuing treatment for 2-3 years and, in some cases, performing allogeneic stem cell transplantation. Thus conventional therapy, though effective, is complicated, expensive and toxic. CD19-targeted therapy has already been shown to work optimally in the MRD setting,43 as treating overt disease with targeted immunological agents is associated with toxicity. Ideally such therapy should be specific to the cells that comprise the MRD population to avoid unwanted side effects. Our experiments provide evidence that immunological targeting of antigens specific to resistant leukemic subclones in the MRD setting offers a novel adjunct to current therapeutic strategies. It is possible that combinations of antibodies along with other targeted approaches may gradually change the way we treat ALL.45 Now that we have an extensive understanding of CD antigens expressed on the B-cell plasma membrane,46 this needs to be extended to biologically relevant markers expressed by resistant leukemic blasts. As denintuzomab mafodotin, which uses the same payload, is already showing promising results, our data suggest that A1mcMMAF could be safely and efficaciously employed in either induction or consolidation therapy regimens in BCP-ALL patients identified to be 5T4+ by flow cytometry prior to starting induction chemotherapy. Acknowledgments The authors would like to thank Bloodwise (Leukemia and Lymphoma Research), for grants 12054 and 13067 to VS and PLS and Cancer Research UK, for grant C8230 to VS. A grant to J-PB from the Swiss National Science Foundation (310030133108) supported establishment of leukemia xenografts. VS is a Wellcome-DBT India Alliance Margdarshi Fellow. We thank the staff of the BRU, in particular Diane Beeston, and other members of the Children’s Cancer Group for help and advice. Finally we are indebted to the children with ALL, their families and treating clinicians.

selection in acute lymphoblastic leukaemia. Nature. 2015;521(7552):357-361. Hunger SP, Mullighan CG. Redefining ALL classification: toward detecting high-risk ALL and implementing precision medicine. Blood. 2015;125(26):3977-3987. Schultz KR, Bowman WP, Aledo A, et al. Improved early event-free survival with imatinib in Philadelphia chromosome-positive acute lymphoblastic leukemia: a Children’s Oncology Group study. J Clin Oncol. 2009;27(31):5175-5181. Biondi A, Schrappe M, De Lorenzo P, et al. Imatinib after induction for treatment of children and adolescents with Philadelphiachromosome-positive acute lymphoblastic leukaemia (EsPhALL): a randomised, openlabel, intergroup study. Lancet Oncol. 2012;13(9):936-945. Jabbour E, O'Brien S, Ravandi F, Kantarjian H. Monoclonal antibodies in acute lymphoblastic leukemia. Blood. 2015;125(26): 4010-4016. Ruella M, Gill S. How to train your T cell:

11.

12.

13.

14.

genetically engineered chimeric antigen receptor T cells versus bispecific T-cell engagers to target CD19 in B acute lymphoblastic leukemia. Expert Opin Biol Ther. 2015;15(6):761-766. Castro FV, McGinn OJ, Krishnan S, et al. 5T4 oncofetal antigen is expressed in high risk of relapse childhood pre-B acute lymphoblastic leukemia and is associated with a more invasive and chemotactic phenotype. Leukemia. 2012;26(7):1487-1498. Eastham AM, Spencer H, Soncin F, et al. Epithelial-mesenchymal transition events during human embryonic stem cell differentiation. Cancer Res. 2007;67(23):1125411262. Spencer HL, Eastham AM, Merry CL, et al. E-cadherin inhibits cell surface localization of the pro-migratory 5T4 oncofetal antigen in mouse embryonic stem cells. Mol Biol Cell. 2007;18(8):2838-2851. Ward CM, Barrow K, Woods AM, Stern PL. The 5T4 oncofoetal antigen is an early differentiation marker of mouse ES cells and

1083


O.J. McGinn et al.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

1084

its absence is a useful means to assess pluripotency. J Cell Sci. 2003;116(Pt 22):4533-4542. McGinn OJ, Marinov G, Sawan S, Stern PL. CXCL12 receptor preference, signal transduction, biological response and the expression of 5T4 oncofoetal glycoprotein. J Cell Sci. 2012;125(Pt 22):5467-5478. Southgate TD, McGinn OJ, Castro FV, et al. CXCR4 mediated chemotaxis is regulated by 5T4 oncofetal glycoprotein in mouse embryonic cells. PLoS One. 2010;5(4): e9982. Stern PL, Brazzatti J, Sawan S, McGinn OJ. Understanding and exploiting 5T4 oncofoetal glycoprotein expression. Semin Cancer Biol. 2014;29:13-20. Kagermeier-Schenk B, Wehner D, OzhanKizil G, et al. Waif1/5T4 inhibits Wnt/betacatenin signaling and activates noncanonical Wnt pathways by modifying LRP6 subcellular localization. Dev Cell. 2011;21(6):1129-1143. Damelin M, Geles KG, Follettie MT, et al. Delineation of a cellular hierarchy in lung cancer reveals an oncofetal antigen expressed on tumor-initiating cells. Cancer Res. 2011;71(12):4236-4246. Southall PJ, Boxer GM, Bagshawe KD, Hole N, Bromley M, Stern PL. Immunohistological distribution of 5T4 antigen in normal and malignant tissues. Br J Cancer. 1990;61(1):89-95. Stern P. Immunotherapies targeting a tumour associated antigen, 5T4 oncofoetal glycoprotein. In: Resaei N, ed. Cancer Immunology: Translational Medicine from Bench to Bedside Berlin Springer-Verlag, 2015:409-421. Stern PL, Harrop R. 5T4 oncofoetal antigen: an attractive target for immune intervention in cancer. Cancer Immunol Immunother. 2017;66(4):415-426. Sapra P, Damelin M, Dijoseph J, et al. Longterm tumor regression induced by an antibody-drug conjugate that targets 5T4, an oncofetal antigen expressed on tumor-initiating cells. Mol Cancer Ther. 2013;12(1):3847. Shapiro GI, Vaishampayan UN, LoRusso P, et al. First-in-human trial of an anti-5T4 antibody-monomethylauristatin conjugate, PF-06263507, in patients with advanced solid tumors. Invest New Drugs. 2017 Jan 9. [Epub ahead of print] Ebinger S, Ozdemir EZ, Ziegenhain C, et al. Characterization of rare, dormant, and therapy-resistant cells in acute lymphoblas-

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

tic leukemia. Cancer Cell. 2016;30(6):849862. Schmitz M, Breithaupt P, Scheidegger N, et al. Xenografts of highly resistant leukemia recapitulate the clonal composition of the leukemogenic compartment. Blood. 2011;118(7):1854-1864. Rehe K, Wilson K, Bomken S, et al. Acute B lymphoblastic leukaemia-propagating cells are present at high frequency in diverse lymphoblast populations. EMBO Mol Med. 2013;5(1):38-51. Carol H, Szymanska B, Evans K, et al. The anti-CD19 antibody-drug conjugate SAR3419 prevents hematolymphoid relapse postinduction therapy in preclinical models of pediatric acute lymphoblastic leukemia. Clin Cancer Res. 2013;19(7): 1795-1805. Meyer LH, Eckhoff SM, Queudeville M, et al. Early relapse in ALL is identified by time to leukemia in NOD/SCID mice and is characterized by a gene signature involving survival pathways. Cancer Cell. 2011;19(2):206-217. le Viseur C, Hotfilder M, Bomken S, et al. In childhood acute lymphoblastic leukemia, blasts at different stages of immunophenotypic maturation have stem cell properties. Cancer Cell. 2008;14(1):47-58. Leng Q, Nie Y, Zou Y, Chen J. Elevated CXCL12 expression in the bone marrow of NOD mice is associated with altered T cell and stem cell trafficking and diabetes development. BMC Immunol. 2008;9(1):1-12. Treuting PM, Albertson TM, Preston BD. Case series: acute tumor lysis syndrome in mutator mice with disseminated lymphoblastic lymphoma. Toxicol Pathol. 2010;38(3):476-485. Zuckerman SH, Evans GF, Bryan N. Chronic administration of dexamethasone results in Fc receptor up-regulation and inhibition of class I antigen expression on macrophages from MRL/lpr autoimmune mice. Clin Diagn Lab Immunol. 1997;4(5):572-578. Qing Y, Lin Y, Gerson SL. An intrinsic BM hematopoietic niche occupancy defect of HSC in SCID mice facilitates exogenous HSC engraftment. Blood. 2012;119(7): 1768-1771. Boutter J, Huang Y, Marovca B, et al. Imagebased RNA interference screening reveals an individual dependence of acute lymphoblastic leukemia on stromal cysteine support. Oncotarget. 2014;5(22):1150111512.

36. Johnson SM, Dempsey C, Chadwick A, et al. Metabolic reprogramming of bone marrow stromal cells by leukemic extracellular vesicles in acute lymphoblastic leukemia. Blood. 2016;128(3):453-456. 37. Polak R, de Rooij B, Pieters R, den Boer ML. B-cell precursor acute lymphoblastic leukemia cells use tunneling nanotubes to orchestrate their microenvironment. Blood. 2015;126(21):2404-2414. 38. Liu J, Masurekar A, Johnson S, et al. Stromal cell-mediated mitochondrial redox adaptation regulates drug resistance in childhood acute lymphoblastic leukemia. Oncotarget. 2015;6(40):43048-43064. 39. Topp MS, Gokbuget N, Stein AS, et al. Safety and activity of blinatumomab for adult patients with relapsed or refractory Bprecursor acute lymphoblastic leukaemia: a multicentre, single-arm, phase 2 study. Lancet Oncol. 2015;16(1):57-66. 40. Lee DW, Kochenderfer JN, StetlerStevenson M, et al. T cells expressing CD19 chimeric antigen receptors for acute lymphoblastic leukaemia in children and young adults: a phase 1 dose-escalation trial. Lancet. 2015;385(9967):517-528. 41. Davila ML, Riviere I, Wang X, et al. Efficacy and toxicity management of 19-28z CAR T cell therapy in B cell acute lymphoblastic leukemia. Sci Transl Med. 2014;6(224): 224ra25. 42. Qin H, Cho M, Haso W, et al. Eradication of B-ALL using chimeric antigen receptorexpressing T cells targeting the TSLPR oncoprotein. Blood. 2015;126(5):629-639. 43. Fathi AT, Borate U, DeAngelo DJ, et al. A phase 1 study of denintuzumab mafodotin (SGN-CD19A) in adults with relapsed or refractory B-lineage acute leukemia (B-ALL) and highly aggressive lymphoma. Blood. 2015;126(23):1328. 44. Welschinger R, Liedtke F, Basnett J, et al. Plerixafor (AMD3100) induces prolonged mobilization of acute lymphoblastic leukemia cells and increases the proportion of cycling cells in the blood in mice. Exp Hematol. 2013;41(3):293-302.e1. 45. d'Argouges S, Wissing S, Brandl C, et al. Combination of rituximab with blinatumomab (MT103/MEDI-538), a T cellengaging CD19-/CD3-bispecific antibody, for highly efficient lysis of human B lymphoma cells. Leuk Res. 2009;33(3):465-473. 46. Mirkowska P, Hofmann A, Sedek L, et al. Leukemia surfaceome analysis reveals new disease-associated features. Blood. 2013; 121(25):e149-159.

haematologica | 2017; 102(6)


ARTICLE

Chronic Lymphocytic Leukemia

Distinct molecular genetics of chronic lymphocytic leukemia in Taiwan: clinical and pathogenetic implications

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Shang-Ju Wu,1 Chien-Ting Lin,1,2 Andreas Agathangelidis,3 Liang-In Lin,4 Yuan-Yeh Kuo,5 Hwei-Fang Tien1 and Paolo Ghia3

Division of Hematology, Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan; 2Tai-Cheng Stem Cell Therapy Center, National Taiwan University, Taipei, Taiwan; 3Strategic Research Program on CLL and B-cell Neoplasia Unit, Division of Experimental Oncology, Vita-Salute San Raffaele University and IRCCS San Raffaele Scientific Institute, Milan, Italy; 4Department of Clinical Laboratory Science and Medical Technology, College of Medicine, National Taiwan University, Taipei, Taiwan and 5Graduate Institution of Oncology, College of Medicine, National Taiwan University, Taipei, Taiwan 1

Haematologica 2017 Volume 102(6):1085-1090

ABSTRACT

D

ifferences in chronic lymphocytic leukemia between the Asian and the Western population are widely known. To further clarify these ethnic differences, we profiled the molecular genetics in a cohort of 83 newly diagnosed patients from Taiwan. In detail, we assessed: (i) the usage and the mutational status of the clonotypic immunoglobulin heavy-chain variable region (IgHV) genes, (ii) the presence of VH CDR3 stereotypes, and (iii) TP53, NOTCH1, SF3B1, BIRC3, and MYD88 mutations. The IgHV gene repertoire was biased and distinct from that observed in the West with the most common IgHV genes being IgHV3-23, IgHV3-7, and IgHV3-48. In terms of IgHV gene mutational status, 63.8% of patients carried mutated rearrangements, whereas 22.4% of patients were assigned to stereotyped subsets (6.9% to major subsets and 15.5% to minor ones). The frequencies of NOTCH1, SF3B1, BIRC3 and MYD88 mutations were 9.6%, 7.2%, 1.2%, and 2.4%, respectively; however, the frequency of TP53 mutations was significantly higher (20.5%). Patients with TP53 mutations or del(17p), SF3B1 mutations and unmutated IgHV had a worse outcome compared to the other patients. In conclusion, the differences observed in IgHV properties suggest different pathogenetic factors implicated in the development of chronic lymphocytic leukemia, while the high frequency of TP53 mutations could in part explain the dismal outcome of these patients in Taiwan Introduction Chronic lymphocytic leukemia (CLL) is the most prevalent adult leukemia in the Western world, accounting for 5-11% of non-Hodgkin lymphomas (NHL).1 CLL belongs to indolent NHL, yet its clinical course varies widely.2 In the past 3 decades, many clinical and molecular features have been identified as predictors of outcome or response to therapy. Chromosome aberrations are important prognostic markers. With traditional cytotoxic agents and rituximab as the backbone of therapies, patients harboring 17p13 deletions have a very short survival with a median of less than 3 years, whereas those with 13q14 deletions, trisomy 12, or normal karyotype have much better outcomes, with a median survival up to 10 years or more, and those carrying 11q22-q23 deletions have a survival time in between the aforementioned.3 In addition, the mutational status of the immunoglobulin (IG) gene is one of the most robust prognostic factors; mutated IG is associated with better outcomes, compared to unmutated IG.4 With the advance of next-generation sequencing technology, some recurrent genetic mutations are found to possess prognostic significance, such as the negative survival impacts of TP53, NOTCH1, SF3B1 or haematologica | 2017; 102(6)

Correspondence: hftien@ntu.edu.tw or ghia.paolo@hsr.it Received: October 3, 2016. Accepted: February 16, 2017. Pre-published: March 2, 2017. doi:10.3324/haematol.2016.157552 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1085 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.

1085


S-J. Wu et al.

BIRC3 mutations and the favorable survival impact of MYD88 mutations.5,6 These molecular and clinical markers of CLL greatly assist in our understanding of disease biology and, perhaps most importantly, in the optimization of individual patient’s management options. CLL in Asia differs in terms not only of a lower prevalence7,8 but also of different treatment outcomes.9 Up to now, CLL has mostly been viewed as a “Western disease”, and the majority of knowledge is derived from studies in Western populations. This is especially true in terms of novel prognostic factors, which have not been thoroughly studied among Asian patients. Such a gap in knowledge becomes even more compelling since studies concerning the clinical course and the epidemiology of CLL have reported different outcomes in Taiwan and Japan compared to the West.9,10 Thus, any differences in the biological characteristics that may underlie the disparities in prognosis of CLL in Asian populations would be of great interest and value. The study herein was aimed at characterizing and validating the clinical implications of CLL immunogenetic and genetic features in a cohort of Taiwanese CLL patients in order to highlight any potential geographical differences that may underlie the outcome disparities. This knowledge could lead to the refinement of management strategies for Taiwanese CLL patients.

IG heavy-chain genes usage and somatic hypermutation status. The clusters of sequences with common heavy-chain complementarity determining region 3 (HCDR3) motifs, described as “stereotyped” B-cell receptors, were identified by the methods described by Bystry, Agathangelidis and Bikos et al.18 The results were compared with the data derived from Western cohorts in plots to delineate the differences and similarities between the populations.4,19,20

Conventional cytogenetic analysis and fluorescence in situ hybridization (FISH) Chromosomal and FISH analyses were performed in 80 patients as previously described.21,22 The FISH panels, including the probes for the centromere of chromosome 12 (CEP12), 13q14.3 (LSI D13S319), 17p13 (p53), and 11q22.3 (ATM), were all from Vysis Inc. (Downers Grove, IL, USA).

Statistical analysis

c2 or Fisher’s exact tests were used for the between-group comparison of discrete variables. Two-sample t-test was used for the between-group comparison of the means. Kaplan–Meier survival curves were used for the estimation of overall survival (OS). Logrank test was used to test for the differences in the OS between groups. All the directional P-values were two-tailed and a P-value of <0.05 was considered as significant. All analyses were performed by using the PASW 18 software (SPSS, Inc., Chicago, IL, USA).

Methods Patients, treatments, and survival In total, 83 consecutive patients with newly diagnosed CLL at the National Taiwan University Hospital (NTUH) between February 1994 and December 2006 constituted the study cohort. The clinical characteristics and cytogenetic profiles in 80 of them have been reported previously.11 The diagnosis was made through the findings of blood cell counts, classification, morphology, and immunophenotying, all of which were carried out at the Specialized Hematology Laboratory in NTUH, according to the World Health Organization (WHO) classification.12 The patients’ demographics is summarized in the Online Supplementary Table S1. The study was approved by the Institutional Review Board of the National Taiwan University Hospital; written informed consent in accordance with the Declaration of Helsinki was obtained from all participants.

SF3B1, NOTCH1, MYD88, BIRC3 and TP53 mutations Gene mutations were determined by polymerase chain reaction (PCR) of genomic DNA extracted from bone marrow samples, followed by direct Sanger sequencing in 83 patients. The PCR protocols of gene mutation analyses for MYD88 (exon 5),13 NOTCH1 (exon 34),14 SF3B1 (exons 14, 15, and 16),15 and TP53 (exons 3 to exon 9)16 were adapted from prior reports, and that for BIRC3 (exons 7, 8, and 9) was designed in-house. The primer sequences are summarized in the Online Supplementary Table S2. The results were compared with the data derived from Western cohorts in plots to delineate the differences and similarities between the populations.5,6

Amplification and analysis of Immunoglobulin (IG) gene rearrangements Analysis of Immunoglobulin (IG) gene rearrangements was successfully performed in 58 patients. The methods for the amplification of IgHV-IgHD-IgHJ rearrangements sequences were adapted from Ghia et al.17 and Hamblin et al.4 Nucleotide sequences were aligned to the IMGT/V-QUEST database in order to determine the 1086

Results CLL IGH gene rearrangement analysis Using 2% difference from the closest germline gene as the cutoff, IG heavy-chain rearrangements were classified as mutated in 37 of the 58 subjects (63.8%). The median difference from the closest IgHV germline gene was 7.7% (range 2.1–11.9%). The IgHV gene repertoire in Taiwanese patients is summarized in Figure 1A: IgHV3-23, IgHV3-7, and IgHV3-48 genes were those most frequently expressed (17.2%, 15.5%, and 13.8%, respectively). All 10 IgHV3-23 gene rearrangements were mutated, whereas IgHV1-69 and IgHV3-15 rearrangements were unmutated (Figure 1A). IgHV3-23 and IgHV3-7 genes were more common in Taiwan as compared to the West (Figure 1B), whereas IgHV4 genes, especially IgHV4-34, one of the most overexpressed IgHV genes in Western CLL patients,4,19 and IgHV1-69, the most common IgHV gene in European patients,20 were underrepresented in the current study. The frequency of IgHV3-21 was 3.4%, similar to that reported in Ghia et al.20 Among the analyzed IG gene rearrangements, 13 (22.4%) were assigned to stereotyped subsets based on a robust bioinformatics algorithm.23,24 Nine cases (15.5%) were assigned to minor subsets, and 4 (6.9%) to major subsets: 2 to subset #1, 1 to subset #8 and 1 to subset #77. The assignment to major subsets was further confirmed through the use of AssignSubsets, a novel major subset assignment tool.18 The clinical course and molecular profile of patients assigned to major subsets are summarized in Table 1. Similar to the report regarding prognostic impact,18 patients in aggressive subsets #1 and #8 had concurrent poor prognostic profiles and much shorter survival, whereas the patient in indolent subset #77 had a more favorable prognostic profile and longer survival. haematologica | 2017; 102(6)


Distinct Molecular genetics of CLL in Taiwan

Discussion

Gene mutations in CLL patients The profiles of gene mutations are summarized in Figure 2A and the Online Supplementary Table S3. Mutations in TP53, NOTCH1, MYD88, SF3B1 or BIRC3 were identified in 30 of the 83 patients (36.1%): 17 patients (20.5%) with TP53 mutations, 8 (9.6%) with NOTCH1 mutations, 6 (7.2%) with SF3B1 mutations, 2 (2.4%) with MYD88 mutations, and 1 (1.2%) with a BIRC3 mutation. Three patients had concurrent mutations in TP53 and NOTCH1, and 1 patient in TP53 and SF3B1. The summary of the gene mutations and cytogenetic changes is shown in Figure 2A. Compared with the data derived from Western populations,5 the frequency of TP53 mutation among newly diagnosed patients was dramatically higher in our cohort, whereas those of other mutations were quite similar (Figure 2B). In addition, a novel TP53 mutation, V31I, was noted in 3 cases in the current cohort. Cytogenetic changes detected by either conventional cytogenetics or FISH were available in 80 patients. Correlations of gene mutations with cytogenetic abnormalities demonstrated that TP53 mutations were closely associated with del(17p); 5 (71.4%) of the 7 patients with del(17p) had TP53 mutations while 12 (16.4%) of the 73 patients without del(17p) had TP53 mutations (P=0.004). On the other hand, SF3B1 mutations were correlated with del(11q); 5 (55.6%) of the 9 patients with del(11q) had SF3B1 mutations while only 1 (1.4%) of the 71 patients without del(11q) had SF3B1 mutations (P<0.001) (Online Supplementary Table S4).

It is well known that the incidence and prevalence of cases of CLL are much lower in Eastern countries.7,8 In addition to this disparity, quite a few studies have demonstrated that the cytogenetic profile and outcome of CLL in this area are different from that observed in the West.7,9,11 As for the molecular genetics of CLL, ethnical/geographical differences between populations have also been reported.8,25 In the study herein, we studied a series of immunogenetic and genetic features which are commonly evaluated at CLL diagnosis in the West, in a cohort of Taiwanese patients. The study results demonstrated that

A

Mutated IgHV Unmutated IgHV

Associations between IgHV mutational status and genetic alterations The mutational status of the rearranged IgHV gene correlated closely with the genetic profiles (Table 2). In more detail, patients with unmutated IgHV genes were more likely to have concurrent TP53, NOTCH1, or SF3B1 mutations, compared to those with mutated IgHV genes (45.0% vs. 5.3%, P=0.001; 20.0% vs. 0%, P=0.011; and 20.0% vs. 2.6%, P=0.044, respectively). Furthermore, the unmutated IgHV gene was correlated with del(17p) and del(11q) (26.3% vs. 5.6%, P=0.041; and 26.3% vs. 2.8%, P=0.015, respectively). Lastly, patients with mutated IgHV genes were diagnosed with early disease stage in comparison with unmutated cases (63.2% vs. 25.0% at Binet stage A, P=0.017).

B

The prognostic impact of gene mutations Novel gene mutations correlated with the patients OS: patients with either del(17p) or TP53 mutations had significantly shorter OS compared to those with neither aberrations (41.8 months vs. 86.8 months, P=0.048, Figure 3A). SF3B1 mutations were also associated with shorter OS (11.1 months vs. 86.8 months, P<0.001, Figure 3B), whereas patients with mutated IgHV genes had longer OS (89.2 months vs. 40.0 months, P=0.001, Figure 3C).

Figure 1. IgHV gene usage. (A) Frequency of the IgHV subgroups and somatic hypermutation status in Taiwanese CLL patients. (B) Comparison between Western and Taiwanese CLL patients regarding the frequency of IgHV gene usage. The data of these common IgHV genes, derived in the West for references, are from the study by Ghia et al.20

Table 1. Brief clinical profiles of patients in major CDR3 stereotype subsets.

Subset

UPN

IgHV

Gene mutations

Cytogenetics

Binet

Course

#1 #1 #8 #77

026 079 012 017

IgHV1-3, unmutated IgHV1-18, unmutated IgHV4-39, unmutated IgHV4-59, mutated

TP53 mutation(+) TP53 mutation(+) None None

17p deletion 17p deletion Trisomy 12 None

3 2 1 1

Dead; OS 3 months Dead; OS 31 months Dead; OS 64 months Alive; OS 148 months

OS: overall survival; UPN: unique patient number.

haematologica | 2017; 102(6)

1087


S-J. Wu et al.

CLL in Taiwan is highly distinct compared to Western CLL in terms of molecular genetic profiles. With regard to immunogenetics, 63.8% of patients in the current cohort had mutated IG rearrangements, a rate similar to that in the West where 50~60% of patients show somatic hypermutations.4,26 Associations between IgHV gene usage and somatic hypermutation status reported in the West were also evident in Taiwanese patients, with IgHV1-69 gene rearrangements belonging to the unmutated CLL fraction and IgHV3-23 to the mutated one (Figure 1A). On the contrary, the IgHV gene usage distribution in Taiwanese CLL patients was quite different from that in the West, characterized by the underrepresentation of the IgHV1-69 gene and the absence of IgHV4-34 gene rearrangements. A similar profile of IgHV subgroup gene usage was observed between our cohort and mainland Chinese patients, characterized by a higher frequency of the IgHV3 subgroup and a lower frequency of the IgHV4 subgroup compared to the West. However, the distribution of individual IgHV genes in Taiwanese patients was different from that in Western as well as Chinese patients, with the most characteristic example being that of the IgHV4-34 gene, which was the most common gene in Chinese patients,27 yet it was completely absent in our cohort. Since people from mainland China and Taiwan are expected to be similar in terms of genetic background, such differences in IgHV gene usage support the notion that environmental variables with population-specific antigenic selection

may play a role in the development of CLL. Although IgHV3-7, IgHV3-26, and IgHV4-34 have been correlated with antibodies against influenza A, influenza B, and I Table 2. The correlations of IgHV mutation status and other molecular genetics, cytogenetics, and clinical stages.

IgHV mutation(-) IgHV mutation(+)

P value

Molecular genetics TP53 mutation (+) NOTCH1 mutation (+) SF3B1 mutation (+) MYD88 mutation (+) BIRC3 mutation (+)

(n=20) 9 (45.0%) 4 (20.0%) 4 (20.0%) 0 (0%) 0 (0%)

(n=38) 2 (5.3%) 0 (0%) 1 (2.6%) 1 (2.6%) 1 (2.6%)

0.001 0.011 0.044 N.S. N.S.

FISH 17p deletion 11q deletion 13q deletion Trisomy 12 Normal

(n=19) 5 (26.3%) 5 (26.3%) 6 (31.6%) 5 (26.3%) 2 (10.5%)

(n=36) 2 (5.6%) 1 (2.8%) 19 (52.8%) 5 (13.9%) 14 (38.9%)

0.041 0.015 N.S. N.S. 0.033

Binet stage A:B:C

(n=20) (n=38) 5(25.0%):9(45.0%): 24(63.2%): 6(30.0%) 7(18.4%):7(18.4%)

0.017

N.S.: not significant; FISH: fluorescence in situ hybridization.

A

B

Figure 2. Gene mutation and cytogenetic status. (A) The mutation landscape including IgHV mutation, gene mutation, and cytogenetic aberrancy in the current study. Each column represents an untreated CLL case. (B) The frequencies of common gene mutations in Taiwanese CLL patients and Western CLL patients. The Western data for references are derived from the review by FoĂ et al.5 and Martinez-Trillos et al.6

1088

haematologica | 2017; 102(6)


Distinct Molecular genetics of CLL in Taiwan

antigen in cold agglutinin disease,28-32 respectively, no apparent racial differences in these diseases have been reported to support the possible associations between chronic stimulation resulting from these antigens or pathogens and CLL development in this region. Regarding stereotyped VH CDR3, although only a few patients were assigned to major subsets, the associations between stereotypes and clinical characteristics held true in the current cohort (Table 1) with subset #1 and #8 patients exhibiting aggressive disease behavior in contrast to the subset #77 patient who showed an indolent disease course. Frequencies of SF3B1, NOTCH1, MYD88, and BIRC3 mutations in the current cohort were similar to those reported in the West. The only exception was the TP53 gene whose mutation rate at diagnosis was twice as high compared to the West (20.5% vs. 5~10%, Figure 2B),5 which is in line with previous studies in Korea.33 In this context, the fact that TP53 mutations frequently coincide with del(17p) together with the finding that del(17p) is also more common in Taiwan and other Asian cohorts11,34-36 corroborate our findings. Since TP53 mutations can be acquired during disease progression, one might question that this high frequency of TP53 mutations could result from diagnosis delay. However, the mutation analysis was performed on samples obtained at the time of diagnosis of all patients, and the mutation rate in the current Taiwanese cohort was even higher than that of patients after first treatment in the West (8~11%).5 In addition, SF3B1 mutations, which are also frequently acquired during disease progression,5 were not more common in Taiwan. With these observations, delayed diagnosis per se does not appear to be able to explain such a huge disparity of TP53 mutation rates. Since TP53 mutation/del(17p) is one of the strongest poor prognostic factors in CLL,5 the higher mutation rate might partially explain the dismal outcomes of CLL patients in Taiwan. A surprising observation was that the V31I mutation in exon 3 of TP53, which was recurrently identified in our cohort, has not been reported in Western CLL populations. According to the literature, this mutation could be of significance given the fact that it has been found in other hematological malignancies and solid cancers.16,37-42 In addition, findings from a functional study suggested that this mutation correlated with lower transcriptional activity and lower cell proliferation suppressing activity compared to wild-type TP53.43 The mechanisms and effects of this mutation in CLL pathogenesis need to be further clarified. Of note, it has been reported in the literature that around 3% to 12% of cases without del(17p) harbored TP53 mutations,5,44-46 but the mutation rate was higher, up to 16.4% (12 of 73 cases, Online Supplementary Table S4) in the current study. Such a high mutation rate in this population suggests that TP53 mutation analysis is of great importance in assessing the risk and prognosis in patients without del(17p) in Taiwan. In summary, the distinct usage of IgHV genes in Taiwanese CLL patients suggests different pathogenetic mechanisms with distinct antigenic elements being implicated in the development of the disease. In addition, we demonstrated herein a significantly higher TP53 mutation frequency, a finding compatible with the overall dismal outcome of CLL patients in Taiwan. Further studies are needed to better characterize and understand haematologica | 2017; 102(6)

A

P=0.048

B

P<0.001

C

P<0.001

IgHV: mutated IgHV: unmutated

IgHV: mutated IgHV: unmutated Figure 3. The impact of molecular genetics on CLL patients overall survival in Taiwan. (A) The OS for patients with or without 17p deletion or TP53 mutation. (B) The OS for patients with or without SF3B1 mutation. (C) The OS for patients with or without IgHV gene mutation.

the actual impact of these differences in the development and progression of CLL among different ethnic and geographical groups. Acknowledgements This work was supported by a grant from the Ministry of Science and Technology (104WFA0150879), Taiwan, ROC, and a grant from the Hematology Society of Taiwan, Taiwan, ROC. We thank the staff of the Third Core Lab, Department of Medical Research, National Taiwan University Hospital for technical support. 1089


S-J. Wu et al.

References 1. Rozman C, Montserrat E. Chronic lymphocytic leukemia. N Engl J Med. 1995; 333(16):1052-1057. 2. Nabhan C, Rosen ST. Chronic lymphocytic leukemia: a clinical review. JAMA. 2014; 312(21):2265-2276. 3. Dohner H, Stilgenbauer S, Benner A, et al. Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med. 2000; 343(26):1910-1916. 4. Hamblin TJ, Davis Z, Gardiner A, Oscier DG, Stevenson FK. Unmutated Ig V(H) genes are associated with a more aggressive form of chronic lymphocytic leukemia. Blood. 1999;94(6):1848-1854. 5. Foa R, Del Giudice I, Guarini A, Rossi D, Gaidano G. Clinical implications of the molecular genetics of chronic lymphocytic leukemia. Haematologica. 2013;98(5):675685. 6. Martinez-Trillos A, Pinyol M, Navarro A, et al. Mutations in TLR/MYD88 pathway identify a subset of young chronic lymphocytic leukemia patients with favorable outcome. Blood. 2014;123(24):3790-3796. 7. Wu SJ, Huang SY, Lin CT, Lin YJ, Chang CJ, Tien HF. The incidence of chronic lymphocytic leukemia in Taiwan, 1986-2005: a distinct increasing trend with birth-cohort effect. Blood. 2010;116(22):4430-4435. 8. Wu SJ, Chiang CJ, Lin CT, Tien HF, Lai MS. A nationwide population-based cross-sectional comparison of hematological malignancies incidences between Taiwan and the United States of America. Ann Hematol. 2016;95(1):165-167. 9. Wu SJ, Chiang CJ, Lin CT, Tien HF, Lai MS. Improving but inferior survival in patients with chronic lymphocytic leukemia in taiwan: a population-based study, 1990-2004. PloS One. 2013;8(4):e62930. 10. Tomomatsu J, Isobe Y, Oshimi K, et al. Chronic lymphocytic leukemia in a Japanese population: varied immunophenotypic profile, distinctive usage of frequently mutated IGH gene, and indolent clinical behavior. Leuk Lymphoma. 2010;51(12):2230-2239. 11. Wu SJ, Lin CT, Huang SY, et al. Chromosomal abnormalities by conventional cytogenetics and interphase fluorescence in situ hybridization in chronic lymphocytic leukemia in Taiwan, an area with low incidence--clinical implication and comparison between the West and the East. Ann Hematol. 2013;92(6):799-806. 12. Jaffe ES HN, Stein H, Vardiman JW. Pathology and genetics of tumours of haematopoietic and lymphoid tissues: WHO/IARC classification of tumours, 3rd edition, 2001. 13. Varettoni M, Arcaini L, Zibellini S, et al. Prevalence and clinical significance of the MYD88 (L265P) somatic mutation in Waldenstrom's macroglobulinemia and related lymphoid neoplasms. Blood. 2013; 121(13):2522-2528. 14. Weng AP, Ferrando AA, Lee W, et al. Activating mutations of NOTCH1 in human T cell acute lymphoblastic leukemia. Science. 2004;306(5694):269-271. 15. Hou HA, Liu CY, Kuo YY, et al. Splicing factor mutations predict poor prognosis in patients with de novo acute myeloid leukemia. Oncotarget. 2016;7(8):9084-9101. 16. Hou HA, Chou WC, Kuo YY, et al. TP53

1090

17.

18.

19.

20.

21.

22.

23. 24.

25.

26.

27.

28.

29.

30.

mutations in de novo acute myeloid leukemia patients: longitudinal follow-ups show the mutation is stable during disease evolution. Blood Cancer J. 2015;5:e331. Ghia P, Stamatopoulos K, Belessi C, et al. ERIC recommendations on IGHV gene mutational status analysis in chronic lymphocytic leukemia. Leukemia. 2007;21(1):13. Bystry V, Agathangelidis A, Bikos V, et al. ARResT/AssignSubsets: a novel application for robust subclassification of chronic lymphocytic leukemia based on B cell receptor IG stereotypy. Bioinformatics. 2015;31(23):3844-3846. Fais F, Ghiotto F, Hashimoto S, et al. Chronic lymphocytic leukemia B cells express restricted sets of mutated and unmutated antigen receptors. J Clin Invest. 1998;102(8):1515-1525. Ghia P, Stamatopoulos K, Belessi C, et al. Geographic patterns and pathogenetic implications of IGHV gene usage in chronic lymphocytic leukemia: the lesson of the IGHV3-21 gene. Blood. 2005;105(4):16781685. Ko BS, Tang JL, Lee FY, et al. Additional chromosomal abnormalities and variability of BCR breakpoints in Philadelphia chromosome/BCR-ABL-positive acute lymphoblastic leukemia in Taiwan. Am J Hematol. 2002;71(4):291-299. Huang SY, Yao M, Tang JL, et al. Clinical significance of cytogenetics and interphase fluorescence in situ hybridization analysis in newly diagnosed multiple myeloma in Taiwan. Ann Oncol. 2005;16(9):1530-1538. Carstensen B PM, Laara E and Hills M. Epi: a package for statistical analysis in epidemiology. R package version 1.1.40 ed. 2012. Darzentas N, Hadzidimitriou A, Murray F, et al. A different ontogenesis for chronic lymphocytic leukemia cases carrying stereotyped antigen receptors: molecular and computational evidence. Leukemia. 2010; 24(1):125-132. Marinelli M, Ilari C, Xia Y, et al. Immunoglobulin gene rearrangements in Chinese and Italian patients with chronic lymphocytic leukemia. Oncotarget. 2016; 7(15):20520-20531. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38 expression as novel prognostic indicators in chronic lymphocytic leukemia. Blood. 1999;94(6):18401847. Xia Y, Fan L, Wang L, et al. Frequencies of SF3B1, NOTCH1, MYD88, BIRC3 and IGHV mutations and TP53 disruptions in Chinese with chronic lymphocytic leukemia: disparities with Europeans. Oncotarget. 2015;6(7):5426-5434. Pascual V, Victor K, Spellerberg M, Hamblin TJ, Stevenson FK, Capra JD. VH restriction among human cold agglutinins. The VH4-21 gene segment is required to encode anti-I and anti-i specificities. J Immunol. 1992;149(7):2337-2344. Schutte ME, van Es JH, Silberstein LE, Logtenberg T. VH4.21-encoded natural autoantibodies with anti-i specificity mirror those associated with cold hemagglutinin disease. J Immunol. 1993;151(11):65696576. Krause JC, Tsibane T, Tumpey TM, et al. Epitope-specific human influenza antibody repertoires diversify by B cell intraclonal sequence divergence and interclonal conver-

gence. J Immunol. 2011;187(7):3704-3711. 31. Lucas AH, Reason DC. Polysaccharide vaccines as probes of antibody repertoires in man. Immunol Rev. 1999;171(1):89-104. 32. Xu GJ, Kula T, Xu Q, et al. Viral immunology. Comprehensive serological profiling of human populations using a synthetic human virome. Science. 2015;348(6239):aaa0698. 33. Kim J-A, Hwang B, Park SN, et al. Ethnic difference in genomic profiles of chronic lymphocytic leukemia in Korea: targeted exome sequencing and molecular cytogenetics. Blood. 2015;126(23):1726-1726. 34. Irons RD, Le A, Bao L, et al. Characterization of chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) in Shanghai, China: molecular and cytogenetic characteristics, IgV gene restriction and hypermutation patterns. Leuk Res. 2009;33(12):1599-1603. 35. Qiu HX, Xu W, Cao XS, et al. Cytogenetic characterisation in Chinese patients with chronic lymphocytic leukemia: a prospective, multicenter study on 143 cases analysed with interphase fluorescence in situ hybridisation. Leuk Lymphoma. 2008;49(10):1887-1892. 36. Xu W, Li JY, Pan JL, et al. Interphase fluorescence in situ hybridization detection of cytogenetic abnormalities in B-cell chronic lymphocytic leukemia. Int J Hematol. 2007;85(5):430-436. 37. Kishimoto Y, Murakami Y, Shiraishi M, Hayashi K, Sekiya T. Aberrations of the p53 tumor suppressor gene in human non-small cell carcinomas of the lung. Cancer Res. 1992;52(17):4799-4804. 38. Mori S, Ito G, Usami N, et al. p53 apoptotic pathway molecules are frequently and simultaneously altered in nonsmall cell lung carcinoma. Cancer. 2004;100(8):1673-1682. 39. Jeck WR, Parker J, Carson CC, et al. Targeted next generation sequencing identifies clinically actionable mutations in patients with melanoma. Pigment Cell Melanoma Res. 2014;27(4):653-663. 40. Chen CH, Dickman KG, Moriya M, et al. Aristolochic acid-associated urothelial cancer in Taiwan. Proc Natl Acad Sci USA. 2012;109(21):8241-8246. 41. Wang L, Yamaguchi S, Burstein MD, et al. Novel somatic and germline mutations in intracranial germ cell tumours. Nature. 2014; 511(7508):241-245. 42. Xu L, Gu ZH, Li Y, et al. Genomic landscape of CD34+ hematopoietic cells in myelodysplastic syndrome and gene mutation profiles as prognostic markers. Proc Natl Acad Sci USA. 2014;111(23):8589-8594. 43. Yamada H, Shinmura K, Okudela K, et al. Identification and characterization of a novel germ line p53 mutation in familial gastric cancer in the Japanese population. Carcinogenesis. 2007;28(9):2013-2018. 44. Zenz T, Vollmer D, Trbusek M, et al. TP53 mutation profile in chronic lymphocytic leukemia: evidence for a disease specific profile from a comprehensive analysis of 268 mutations. Leukemia. 2010;24(12):20722079. 45. Zenz T, Eichhorst B, Busch R, et al. TP53 mutation and survival in chronic lymphocytic leukemia. J Clin Oncol. 2010;28(29):4473-4479. 46. Yu L, Kim HT, Kasar SN, et al. Survival of del17p CLL depends on genomic complexity and somatic mutation. Clin Cancer Res. 2017;23(3):735-745.

haematologica | 2017; 102(6)


ARTICLE

Non-Hodgkin Lymphoma

Relevance of ID3-TCF3-CCND3 pathway mutations in pediatric aggressive B-cell lymphoma treated according to the non-Hodgkin Lymphoma Berlin-FrankfurtMünster protocols

Marius Rohde,1 Bettina R. Bonn,1 Martin Zimmermann,1 Jonas Lange,2,3 Anja Möricke,4 Wolfram Klapper,5 Ilske Oschlies,5 Monika Szczepanowski,5 Inga Nagel,6 Martin Schrappe,4 MMML-MYC-SYS Project,7 ICGC MMML-Seq Project,6,8 Markus Loeffler,7 Reiner Siebert,6,8 Alfred Reiter1 and Birgit Burkhardt2

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2017 Volume 102(6):1091-1098

Department of Pediatric Hematology and Oncology, Justus-Liebig-University Giessen; Pediatric Hematology and Oncology, University Hospital Münster; 3Translational Oncology, Department of Medicine A, University Hospital Münster; Cluster of Excellence EXC 1003, Cells in Motion, Münster; 4Pediatric Hematology and Oncology, University Medical Center Schleswig-Holstein, Campus Kiel; 5Department of Pathology, Hematopathology Section and Lymph Node Registry, University Hospital Schleswig-Holstein, Campus Kiel/ChristianAlbrecht University, Kiel; 6Institute of Human Genetics, Christian-Albrechts-University Kiel & University Hospital Schleswig-Holstein, Campus Kiel; 7Institute for Medical Informatics Statistics and Epidemiology, University Leipzig and 8Institute of Human Genetics, University of Ulm and University Medical Center Ulm, Germany 1 2

ABSTRACT

M

ature B-cell non-Hodgkin lymphoma is the most common subtype of non-Hodgkin lymphoma in childhood and adolescence. B-cell non-Hodgkin lymphomas are further classified into histological subtypes, with Burkitt lymphoma and Diffuse large B-cell lymphoma being the most common subgroups in pediatric patients. Translocations involving the MYC oncogene are known as relevant but not sufficient for Burkitt lymphoma pathogenesis. Recently published large-scale next-generation sequencing studies unveiled sets of additional recurrently mutated genes in samples of pediatric and adult B-cell nonHodgkin lymphoma patients. ID3, TCF3 and CCND3 are potential drivers of Burkitt lymphomagenesis. In the study herein, frequency and clinical relevance of mutations in ID3, TCF3 and CCND3 were analyzed within a well-defined cohort of 84 uniformly diagnosed and treated pediatric B-cell non-Hodgkin lymphoma patients of the Berlin-Frankfurt-Münster group. Mutation frequency was 78% (ID3), 13% (TCF3) and 36% (CCND3) in Burkitt lymphoma (including Burkitt leukemia). ID3 and CCND3 mutations were associated with more advanced stages of the disease in MYC rearrangement positive Burkitt lymphoma. In conclusion, ID3-TCF3-CCND3 pathway genes are mutated in more than 88% of MYC-rearranged pediatric B-cell non-Hodgkin lymphoma and the pathway may represent a highly relevant second hit of Burkitt lymphoma pathogenesis, especially in children and adolescents. Introduction Non-Hodgkin lymphoma (NHL) belong to the most common hematological malignancies in childhood and adolescence.1 About two thirds of all pediatric NHL belong to mature B-cell lymphomas, with Burkitt lymphoma (BL) and Diffuse large B-cell Lymphoma (DLBCL) representing the most prevalent entities in this subgroup.2,3 During the past decades remarkable increases in patient survival were achieved by continuous advancement of treatment approaches worldwide.4 Nowadays, risk-stratified polychemotherapy treatment reaches up to a total of 90% probability of event-free survival (pEFS) in pediatric NHL.5 More than 95% of haematologica | 2017; 102(6)

Correspondence: birgit.burkhardt@ukmuenster.de

Received: September 27, 2016. Accepted: February 7, 2017. Pre-published: February 16, 2017. doi:10.3324/haematol.2016.156885 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1091 ©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.

1091


M. Rohde et al.

all pediatric patients diagnosed with mature B-cell lymphoma in Germany are registered in the clinical trials of the non-Hodgkin Lymphoma - Berlin-Frankfurt-Münster (NHL-BFM) study group and are treated according to standardized treatment plans. Classical genetic and molecular pathological studies on the pathogenesis of B-cell lymphoma provided distinct pathogenetic features, like translocation of the tumor oncogene MYC in BL6 and specific molecular gene expression signatures in DLBCL.7 MYC translocation was shown to be involved in cell cycle regulation, cellular growth, metabolism and apoptosis.6,8,9 However, MYC translocation alone is not sufficient to initiate malignant transformation of B cells.10,11 In DLBCL multiple co-acting molecular alterations have been described. Differentiation between activated B-cell-like (ABC) and germinal center Bcell-like (GCB) by gene expression is well established, profiling revealed differences in prognosis, especially in adult patients.12 Recent next-generation sequencing (NGS) studies provided valuable insight into the landscape of genomic alterations in B-cell non-Hodgkin lymphoma (B-NHL) and independently introduced Inhibitor of DNA binding 3 (ID3) to be recurrently mutated in BL.13-15 ID3 encodes for a helix-loop-helix (HLH) protein that typically lacks a basic DNA-binding domain and therefore inhibits other HLH proteins from binding to their transcriptional target sites by heterodimerization.16-18 One such ID3-inhibited protein is Transcription Factor 3 (TCF3), which is consecutively expressed at high levels during B-cell development.19,20 TCF3 itself was also shown to be recurrently mutated in BL in the transcriptional study from Schmitz and colleagues, who additionally showed that both TCF3 and ID3 mutations resulted in increased expression of TCF3 targets,14 promoting growth and survival by activation of B-cell receptor signaling. A direct target of TCF3 is the cell cycle regulating Cyclin D3 (CCND3),21,22 which was also shown to harbor activating mutations in different subtypes of B-NHL cases.13,14 The above mentioned studies described ID3 mutations to accumulate in the HLH domain and functional analyses showed ID3 mutant proteins to be less effective or completely ineffective in inhibiting TCF3, thus forcing increased cell proliferation and survival via phosphoinositide 3-kinase (PI3K) and Cyclin D3.13-15 While TCF3 mutations also affected the basic HLH (bHLH) domain of its isoform E47, TCF3 mutant proteins did not lose their effect on downstream targets when compared to wildtype TCF3, but displayed ID3/TCF3 interaction, turning them immune to the inhibitory effect of ID3.14 CCND3 mutant proteins showed an increase in cell cycle stimulation when compared to unaffected CCND3, thereby indicating a gain-of-function.14 In summary, mutations in each of the candidate genes are thought to contribute to cellular growth, cell survival and proliferation.23,24 Within the index studies there was a large variation in the incidence of ID3 mutations in BL. The frequency of ID3 mutations varied between 34% (Love et al.), 58% (Schmitz et al.) and 68% (Richter et al.) (Online Supplementary Table S1). Schmitz et al. reported TCF3 mutations in 27% and additional CCND3 mutations in 38% of sporadic BL cases. CCND3 mutations were also analyzed in the study by Richter et al., who also reported 38% of the cases to display these aberrations.13-15 In this study we analyzed a well-defined cohort of 84 1092

pediatric B-NHL patients, diagnosed and treated according to the NHL-BFM protocols for mutations in ID3, TCF3 and CCND3 to describe the incidence and relevance of such mutations in a uniformly diagnosed and treated representative pediatric cohort. Furthermore, we analyzed samples from 96 pediatric patients diagnosed with precursor B-cell acute lymphoblastic leukemia (pB-ALL), to examine whether disruption of this pathway also occurs in this precursor B-cell malignancy.

Methods Patient samples 1117 Pediatric patients diagnosed with “Burkitt lymphoma”, “Burkitt leukemia”, “Diffuse large B-cell lymphoma” or “B-cell lymphoma, unclassifiable, with features intermediate between diffuse large B-cell lymphoma and Burkitt lymphoma” between January 2000 and December 2012 were eligible for the recruited population based study cohort. Pretreatment tumor samples from fresh frozen tissue, bone marrow or effusion samples were available for 84 patients (“study cohort”). For a more robust examination regarding the relevance of mutation status with respect to patient outcome, initial tumor samples from an additional 10 patients with a known history of relapse or progress were analyzed as the “extended cohort”. All patients analyzed were registered in the NHL-BFM data center and treated according to the NHL-BFM protocols (NHL-BFM 95 and analogous protocol B-NHL BFM 04).3 Tumor DNA samples from 96 pediatric patients diagnosed with precursor B-ALL were kindly provided by the ALL-BFM study center, University of Kiel, Germany. All patients had previously been diagnosed between 2000 and 2006 and were treated according to the ALL-BFM 2000 protocol.25 More detailed clinical characteristics of the analyzed patients can be found in the online supplement (Online Supplementary Methods). This study was approved by the Ethical Advisory Board of the University of Giessen, Germany (A89/11 Amendment 2013).

ID3, TCF3 and CCND3 mutation analysis In the study cohort the full coding region of the ID3 gene, exon 17 of the TCF3 gene and the coding region of CCND3 exon 5 were sequenced. pB-ALL samples were analyzed for ID3 mutations only. Cases presenting with mutations were confirmed within a repetition experiment. More detailed descriptions of primer pairs, sequencing modalities, reference sequence annotation and exclusion of singular nuclear polymorphisms are given in the online supplement (Online Supplementary Methods).

Statistics Statistical analysis was performed in order to identify differences in typical patient characteristics, such as sex, age, stage of disease, bone marrow (BM) involvement, central nervous system (CNS) involvement, lactate dehydrogenase (LDH) levels, diagnosis, pEFS and probability of overall survival (pOS) according to the mutational status of the analyzed candidate genes. MYC rearrangement status was available from the study database. Clinical data for each calculation referred to patients with successful investigation of the respective criteria. Differences in the distribution of individual parameters among patient subsets were analyzed using Pearson’s c2 test26 or Fisher's exact test27 where appropriate. pEFS was calculated according to Kaplan and Meier,28 taking into consideration the time between the date of diagnosis and either the date of event or date of last follow up. pOS was calculated according to Kaplan and Meier28 under consideration of the haematologica | 2017; 102(6)


ID3-TCF3-CCND3 pathway mutations in pediatric B-NHL

time between date of diagnosis and death from any cause. Survival estimates were compared by the log-rank test. 29 Significant differences were assumed when the respective P value (P) was lower than 0.05. Calculations were conducted using the SAS statistical program (SAS-PC, Version 9.3, Cary, NC, USA: SAS Institute Inc.). Fisher’s exact tests were calculated using the software Prism 6 for Mac OS X (GraphPad Software, Version 6.0c, San Diego, CA, USA). The two-tailed option was used.

Results Patient characteristics of the study cohort Characteristics of the 84 analyzed patients in the study cohort are shown in Table 1. Clinical characteristics of patients analyzed and not analyzed were similar regarding age, sex, BM involvement, CNS involvement, stage of disease and outcome. Histological subtype was BL in 64 cases (including 14 Burkitt leukemias, B-AL), DLBCL in 13 cases and B-cell lymphoma, unclassifiable, with features intermediate between diffuse large B-cell lymphoma and Burkitt lymphoma (B-NHL not further classifiable [nfc]) in 7 cases. Comparing patient characteristics of the analyzed cohort with the not analyzed patients revealed a trend towards

higher LDH levels in the study cohort and an overrepresentation of BL and B-NHL nfc over DLBCL cases. These mild differences, at least in part, might be related to the availability of tissue for molecular analysis. BM obtained in B-AL and tumor after ileocoecal resection of BL was more likely to be sent to the NHL-BFM study center for research than samples of small biopsies of e.g., cervical lymph nodes in DLBCL. These circumstances may also explain the trend towards higher LDH levels in the study cohort, as a high tumor burden is associated with higher LDH levels, which in turn is key to a larger availability of sample material in the study center. To compensate for this slight imbalance in the representation of histological subtypes, the analyses were run for the study cohort and for the subgroup of MYC rearrangement positive BL/B-AL separately. Incidence and relevance of ID3, TCF3 and CCND3 mutation status were analyzed in the study cohort. A detailed description of all genomic variants including the predicted change on protein level is presented in the Online Supplementary Table S2.

ID3, TCF3 and CCND3 sequencing results ID3 mutations were found and verified in 56 out of 84 B-NHL samples (Figure 1). Thirty-one of 56 cases showed

Table 1. Patient characteristics of the study cohort.

Characteristics Sex male female Age < 10 y 10-14 y > 14 y Stage of disease stage I stage II stage III stage IV B-AL BM involvement yes CNS involvement yes LDH < 500 U/l 500-1000 U/l > 1000 U/l Diagnosis BL B-AL DLBCL B-NHL nfc Outcome pEFS (2y) pOS (2y)

Patients not analyzed (n=1033)

Patients analyzed (n=84)

P

819 214

79% 21%

69 15

82% 18%

.53

520 346 167

50% 34% 16%

43 27 14

51% 32% 17%

0.97

97 247 412 73 165 201

10% 25% 41% 7% 17% 20%

5 14 43 3 14 14

6% 18% 55% 4% 18% 17%

0.17 0.53

102 606

10% 59%

8 39

10% 46%

155 264

15% 26%

16 29

19% 35%

0.07

576 165 265 27

56% 16% 26% 3%

50 14 13 7

60% 17% 15% 8%

0.0008

90 ± 1% 93 ± 1%

89 ± 3% 92 ± 3%

0.92

0.87 (LR) 0.83 (LR)

Y: years; BM: bone marrow; CNS: central nervous system; LDH: lactate dehydrogenase serum level; BL: Burkitt lymphoma; B-AL: Burkitt leukemia; DLBCL: Diffuse large B-cell lymphoma; B-NHL nfc: B-cell non-Hodgkin lymphoma with features intermediate between BL and DLBCL; pEFS: probability of event-free survival; pOS, probability of overall survival; LR: log-rank

haematologica | 2017; 102(6)

1093


M. Rohde et al.

Figure 1. ID3 gene plot with annotated mutations of the study cohort. ID3 coding region of exon 1 is illustrated with single base-pair substitutions on the upper and more complex alterations (insertions, deletions, InDels, duplications) on the lower site. Substitutions resulting in a nonsense mutation are depicted in red. Hatched bars delineate deletions and InDels, dotted bars characterize insertions and duplications. Each mutation is labeled with a correspondent description on the genomic and protein level, as well as the absolute number of occurrences in brackets. The functional helix-loop-helix domain is mapped according to UniProt (Q02535).

multiple ID3 mutations, 26 cases with 2 mutations, 4 cases with 3 mutations and 1 case with 4 mutations. Ten of those cases were randomly selected for cloning, and biallelic involvement was shown in all cases. With respect to hotspots, single nucleotide substitutions affecting position C190 were the most frequent (13 cases), followed by C166 (10 cases) and C241 (5 cases). Four disambiguates were genomic variants that were not predicted to result in changes on amino acid levels: 144C>T, 193A>T, 300+44T>C, and 301-23C>T. Notably, each of the cases with one of those silent mutations also harbored at least a second ID3 mutation. On the genomic level, 77 of 93 (83%) mutations directly affected the functional HLH coding region. The remaining 18 mutations were allocated either close to the splice-site of exon 1 (4 mutations), upstream or downstream of the HLH domain (13 mutations) or in the intronic region between exon 1 and 2 (1 mutation). Again, all cases with mutations not directly affecting the HLH domain or the splice-site were associated with at least a second mutation in the HLH domain. The frequency of ID3 mutations according to diagnosis was 50/64 (78%) for BL/B-AL and 2/13 (15%) for DLBCL. In the subgroup of 7 analyzed B-NHL nfc, 4 showed mutations in ID3. Mutations in TCF3 (8/84, 10%) were considerably lower compared to the high frequency of cases with ID3 and were only found in BL/B-AL cases (8/64, 13%). All 8 mutations occurred in the coding region of the bHLH binding domain of TCF3 (Online Supplementary Figure S1). Mutation 1675G>A was present in 2 cases. Twenty-six cases harbored CCND3 mutations (Online Supplementary Figure S2). Mutations affected nucleotide C811 with a cytosine duplication in 9 cases, resulting in a protein elongating frameshift. Four cases presented with T869G substitution and 3 cases showed C580T mutations. Twenty-three out of 64 BL/B-AL cases (36%) presented with CCND3 mutations. In DLBCL there were 2 out of 14 cases affected. One mutation was present in a case with a B-NHL nfc diagnosis. 1094

Mutational pattern of ID3, TCF3 and CCND3 and correlation with MYC rearrangement status In total, 63 out of 84 cases (75%) had at least 1 mutation in 1 of the investigated genes. Exclusive ID3 mutations were the most frequent (51%). This was followed by cases with concurrent ID3 and CCND3 mutation (31%). Cases 33 and 12 harbored mutations in all 3 genes. The pattern of mutations within the study cohort is depicted in Figure 2. Results of fluorescence in situ hybridization (FISH) for detection of MYC rearrangements were available for 77 cases. Fifty-eight of 65 MYC rearrangement positive cases (89%) had at least one mutation in ID3 and/or TCF3 and/or CCND3. In contrast, within 12 MYC rearrangement negative patients only 1 case was affected by ID3 mutations (P<0.0001). In this patient (case 68) ID3 mutations 20T>A and 164T>A were present.

Clinical characteristics according to mutational status in ID3, TCF3 and CCND3 Clinical characteristics and outcome regarding ID3, TCF3 and CCND3 mutational status were first analyzed in the study cohort (Table 2). ID3 mutations were positively correlated with reference diagnosis of BL/B-AL (P=0.0003), higher LDH serum levels (P=0.0038) and higher stage of disease (P=0.03) (Table 2). ID3 mutations occurred at a higher frequency in BL/B-AL cases when compared to DLBCL (P=0.0001). However, these results are strongly biased by diagnosis, as BL/B-AL patients comprised higher LDH serum levels, higher stage of disease and a frequent discovery of ID3 mutations. To investigate the actual clinical relevance of mutations we further analyzed patient characteristics within BL/B-AL MYC rearrangement positive cases (n=61) (Online Supplementary Table S3). ID3 mutated cases were still associated with higher LDH serum levels (P=0.0431). Furthermore, CCND3 mutated cases were positively associated with advanced stage of disease (P=0.0482). Regarding pEFS and pOS, there were no significant differences between wildhaematologica | 2017; 102(6)


ID3-TCF3-CCND3 pathway mutations in pediatric B-NHL Table 2. Clinical characteristics of study cohort regarding ID3, TCF3 and CCND3 mutation status. All Sex male female Age < 10 y 10-14 y > 14 y Stage of disease I II III IV B-AL BM involvement yes CNS involvement yes LDH < 500 U/l 500-1000 U/l > 1000 U/l Diagnosis BL B-AL DLBCL B-NHL nfc Outcome pEFS (2y) pOS (2y)

ID3mutated

ID3wt

n=56

n=28

P TCF3mutated n=8

47 9

84% 16%

22 6

79% 6%

31 19 9

55% 29% 16%

12 11 5

43% 39% 18%

1 8 29 1 13 13

2% 17% 52% 2% 27% 25%

4 6 14 2 1 1

6

11%

19 12 25 37 13 2 4 88 ± 4% 90 ± 4%

TCF3wt

P CCND3mutated

n=76

n=26

0.55

6 2

75% 25%

63 13

83% 17%

0.52

7 1 0

88% 12% 0%

36 26 14

47% 34% 18%

14% 21% 50% 7% 7% 7%

0.03 0.27

0 4 2 0 2 2

0% 50% 25% 0% 25% 25%

5 10 41 3 12 12

2

7%

0.60

1

13%

34% 21% 45%

20 4 4

71% 14% 14% <0.01

4 2 2

66% 23% 4% 7%

13 1 11 3

46% 4% 40% 11% <0.01

6 2 0 0

93 ± 5%.43 (LR) 93 ± 6%.26 (LR)

CCND3wt

P

n=58

0.58

21 5

81% 19%

48 17

83% 17%

0.83

0.09

16 7 3

62% 27% 12%

27 20 11

47% 35% 19%

0.43

7% 14% 58% 4% 17% 16%

0.09 .51

1 5 10 2 8 8

4% 19% 39% 8% 31% 31%

2 6 24 0 6 6

8% 17% 62% 2% 11% 10%

0.11 0.03

7

9%

0.76

5

19%

3

5%

0.06

50% 25% 25%

35 14 27

46% 18% 36%

0.81

9 5 12

35% 19% 46%

30 11 17

52% 19% 29%

0.27

75% 25% 0% 0%

44 12 13 7

58% 16% 17% 9%

0.42

15 8 2 1

58% 31% 8% 4%

36 6 11 6

60% 10% 19% 10%

0.08

88 ± 12% 88 ± 12%

89 ± 4%.82 (LR) 92 ± 3%.61 (LR)

85 ± 7% 89 ± 6%

91 ± 4% 93 ± 3%

0.38 (LR) 0.52 (LR)

Wt: wild-type; y: years; BM: bone marrow; CNS: central nervous system; LDH: lactate dehydrogenase serum level; BL: Burkitt lymphoma; B-AL: Burkitt leukemia; DLBCL: Diffuse large B-cell lymphoma; B-NHL nfc: B-cell non-Hodgkin lymphoma with features intermediate between BL and DLBCL; pEFS: probability of event-free survival; pOS: probability of overall survival; LR: log-rank

type and mutated cases either in the study cohort or in subgroup analyses. Next we analyzed for clinical relevance of certain mutational patterns. Of particular note was the combination of ID3 and/or TCF3 mutations, these cases were again associated with higher LDH levels (P=0.0023), which is also reflected in an increased frequency of these cases in higher risk groups. Also, in patients with simultaneous and exclusive ID3 and CCND3 mutations, the frequency of BM involvement (P=0.014) and as a consequence a diagnosis of B-AL (P=0.0175) was increased. Regarding ID3 mutation hotspots it is of note that mutations affecting position 241C, resulting in Q81* nonsense mutation on the protein level, accumulated in cases with B-AL (P=0.0073) compared to BL. Further investigated ID3 hotspots at 190C and 166C did not show association to any clinical criteria.

Outcome and event-free survival with respect to mutational status In the study cohort 9 out of 84 patients suffered disease progression or relapse. Detailed analyses of ID3, TCF3 and CCND3 mutation frequencies and mutational patterns were not significantly associated with pOS and pEFS (Table 2). This observation was confirmed within the analysis of haematologica | 2017; 102(6)

an additional 10 initial BL samples from patients with a medical history of subsequent disease progression or relapse, adding up to a total of 19 cases with refractory or relapsed disease compared to 75 event-free cases.

ID3 sequencing results in pB-ALL In the cohort of 96 pediatric pB-ALL patients DNA isolated from leukemic blasts was analyzed for ID3 mutations. There were no pathogenic ID3 mutations found.

Discussion Burkitt Lymphoma is the most common subtype of NHL in children. With current polychemotherapy treatment regimen event-free survival rates of 90% can be achieved. However, the outcome of patients who suffer from relapse is often fatal. Most patients do not achieve second remission despite intensive salvage treatment. Therefore, new treatment concepts are urgently needed to salvage these patients. New drugs directly targeting pathogenetic pathways of lymphoma cells might represent one possible strategy. However, despite MYC-activating translocations, detectable in the majority of cases,30 little is known about Burkitt pathogenesis and molecular-based risk factors are lacking. 1095


M. Rohde et al.

Figure 2. Overview of B-NHL patient sequencing results on ID3, TCF3 and CCND3. Reference diagnosis according to NHL-BFM study. Red block indicates a case with mutation, black block indicates wild-type. MYC rear.: MYC status as reported in the study database; +: MYC rearrangement positive; -: MYC rearrangement negative; o: MYC status unknown. ∗No reference pathology review available. Diagnosis according to study center review. †No reference diagnosis available. Diagnosis according to local pathology report. ‡MYC FISH analysis not available. However, MYC-Ig PCR report was positive for MYC rearrangement. BL: Burkitt lymphoma; B-AL: Burkitt leukemia; DLBCL: Diffuse large B-cell lymphoma; B-NHL nfc: B-NHL unclassifiable, with features intermediate between BL and DLBCL.

Therefore, the study herein aimed at identifying the frequency and clinical relevance of ID3-TCF3-CCND3 pathway mutations and presents the largest analysis of such mutations in pediatric B-NHL thus far. While the 3 NGS studies cited previously were the first to describe genomic alterations of ID3 and equally evaluated these finding as a new hallmark of BL, there were striking differences regarding the ID3 mutation frequency in the studied cohorts, ranging between 34% and 68%.13-15 Those variations point to relevant differences in inclusion criteria (i.e., histological-/morphological-/study-based vs. molecularbased definitions of BL) and clinical characteristics of the analyzed patients. This was also stressed by Havelange et al., who recently published a series of 13 pediatric and 11 adult BL patients with respect to age-related genetic differences.31 In their cohort they found 10 out of 10 evaluable ID3 mutated cases in the adult group compared to only 5 out of 13 pediatric ones and discussed a potential higher prevalence of ID3 mutations in adults. However, Richter et al. found an age-related correlation of ID3 mutations towards younger patients and another study of only adult BL patients reported a rather low mutation rate of 47%.32 With the finding of 78% ID3 mutations in BL in the current study and comparison of the age structure of the recent studies (Online Supplementary Table S1), we conclude that ID3 mutations occur at high frequency in pediatric BL patients. As patient age itself was not associated with mutation frequency within pediatric cases, these observations lead to the conclusion that ID3 mutation frequency might in fact be associated with more homogeneously presenting pediatric BL and occur less often in the more heterogeneous group of Burkitt-like adult B-cell lymphoma. This is also supported by previously found differences in molecular presentation of BL between pediatric and adult patients, with the general mutational load being significantly higher in older BL patients.33 Furthermore, BL 1096

in general is known for homogeneous gene expression profiles, especially in comparison to the related group of DLBCL.34,35 Results from whole genome sequencing of 13 pediatric BL cases of the NHL-BFM group supported these observations on the genomic level, showing a median of only 28 protein changing somatic mutations per tumor and a high frequency of recurrently affected genes, even in the small number of 13 cases.36 The frequency of TCF3 mutations in BL (13%) occurred less often in our cohort, while the incidence of CCND3 mutations in BL (36%) cases was consistent with the findings of other groups.13,14,31 Within the subgroup of 61 MYC rearrangement positive BL, ID3 mutations were significantly associated with a more disseminated presentation of disease and CCND3 showed positive correlation to an advanced stage of disease, supporting their pro-proliferative and cell cycle driving role. These effects became even more evident when evaluating cases with ID3 and/or TCF3 mutations, which can be regarded as equal with respect to the resulting functional disruption of the pathway. The frequency of simultaneous ID3 and CCND3 mutations was significantly higher in patients with BM involvement, hinting at a potential relevance in terms of blast migration. Havelange et al. found a poorer outcome for patients affected by such simultaneous mutations in ID3 and CCND3.31 In our cohort we could not confirm this finding for pediatric patients, and it is, however, difficult to compare pediatric and adult patients with respect to clinical characteristics and prognosis as treatment regimens are generally different and outcome is inferior in adult patients.37 In our study there was no association between clinical outcome and mutational status. These findings represent a contrast to previous data of Richter et al., where superior outcome for patients with ID3 mutations was reported. Again, one possible explanation haematologica | 2017; 102(6)


ID3-TCF3-CCND3 pathway mutations in pediatric B-NHL

Figure 3. ID3, TCF3, Cyclin D3 pathway with frequencies of respective mutations in MYC rearrangement positive BL. BL with positive MYC translocation had mutations in at least one of the three investigated candidate genes in 89%, representing affection of the ID3-TCF3-CCND3 pathway in the vast majority of pediatric BL cases.

are the aforementioned general differences between BL in pediatric and adult patients. The high number of ID3 mutations and recurrent involvement of its partners suggest a role of these alterations in Burkitt-lymphomagenesis, rather than a role for disease recurrence in a small subgroup of patients. In the current study, 89% of BL with positive MYC translocation had mutations in at least 1 of the 3 investigated candidate genes, representing affection of the ID3-TCF3-CCND3 pathway in the vast majority of pediatric BL cases (Figure 3). The study cohort included 7 cases of B-NHL nfc and 13 cases of DLBCL. Among them we detected 5 and 3 cases with ID3-TCF3-CCND3 pathway mutations, respectively. Information on the MYC rearrangement status was limited in some of these patients, however, the overall strong association of ID3-TCF3-CCND3 pathway mutations with MYC rearrangements in BL could similarly be observed in B-NHL nfc and DLBCL. Despite the defining histological diagnosis some of those cases show Burkittlike features with respect to the genetic findings. In many malignancies certain discrepancies between histological and molecular diagnosis have been observed after the establishment of molecular profiling. In this context, proof of ID3-TCF3-CCND3 pathway impairment might be helpful to better discriminate such borderline cases as BL in the future. This is supported by the recently updated World Health Organization (WHO) classification of lymphoid neoplasms, wherein ID3 and TCF3 mutations were added to the molecular characteristics of BL.38 The analysis of 96 samples of pB-ALL patients representing an immature B-cell malignancy did not show any pathogenic ID3 mutations, supposing their exclusive occurrence in mature B-cell lymphoma. Regarding the process of malignant transformation in mature B cells, initial studies attributed the occurrence of the pathognomonic MYC translocations in BL to altered recombination-activating gene (RAG)-mediated recombination, however, more recently it has been widely accepted that aberrant somatic hypermutation processes involving activationinduced cytidine deaminase (AID) lead to these haematologica | 2017; 102(6)

changes.39,40 Regarding ID3, it is of note that mutations were shown to recurrently occur in the RGYW-motif that is favorably affected by AID as well.13 As MYC translocation alone seems not to be sufficient to induce lymphomagenesis,10 one might speculate that subsequent impairment of the investigated pathway serves as a relevant second hit for BL development. This hypothesis is furthermore supported by the detection of ID3 mutations in mature B-cell malignancies exclusively. In addition, the lack of associations with clinical characteristics or prognosis may even imply an essential function of ID3-TCF3-CCND3 pathway disruption. Cases presenting without mutations might still be affected by focal loss of ID3 or mutations in other functional partners that are involved up- or downstream within the same pathway, likely in the B-cell receptor (BCR), PI3K and cyclin-dependent kinase (CDK)4/6 pathways and their regulators. However, additional candidates will less likely present at similar high frequencies, as NGS studies thus far should have covered most of the highly recurrent genomic events in BL.13-15,41 The overall high number of affected cases asks for therapeutic targeting of this pathway. There is initial promising evidence for successful application and efficacy of the orally available CDK4/6 inhibitor, palbociclib (PD) 0332991, as demonstrated by tumor mass reduction in a BL mouse model by Schmitz et al.14 CDK4/6 inhibitors have also recently been shown to be effective in renal cell carcinoma cell lines and breast cancer cell lines42,43 and are in preparation for clinical phase I and II studies in breast cancer patients (clinicaltrials.gov Identifier: 02297438). Further functional investigation of this pathway will shed more light on molecular processes in BL and hopefully reveal more specific therapeutic options. In the context of relatively homogeneous genomic alterations in pediatric Burkitt lymphoma, the high number of ID3 mutations found in this study of pediatric B-NHL patients suggests an essential role for this pathway with respect to lymphomagenesis and the phenotype of Burkitt lymphoma. 1097


M. Rohde et al.

Funding The research on MYC-positive lymphomas is supported by the BMBF in the framework of E:bio Network Molecular Mechanisms in Malignant Lymphoma with MYC-Deregulation (MMMLMYC-SYS) Grant 0316166. The ICGC MMML-Seq Network is supported by the grants of the BMBF (01KU1002A-J and 01KU1505G). Support of research infrastructure by the Forschungshilfe-Peiper, Giessen, is acknowledged. WK, RS, MS

References 1. Swerdlow SH, Campo E, Harris NL, et al. (eds.). WHO classification of tumours of hematopoietic and lymphoid tissue. IARC press, Lyon 2008 2. Burkhardt B, Zimmermann M, Oschlies I, et al. The impact of age and gender on biology, clinical features and treatment outcome of non-Hodgkin lymphoma in childhood and adolescence. Br J Haematol. 2005; 131(1):39-49. 3. Woessmann W, Seidemann K, Mann G, et al. The impact of the methotrexate administration schedule and dose in the treatment of children and adolescents with B-cell neoplasms: a report of the BFM Group Study NHL-BFM95. Blood. 2005;105(3):948-958. 4. Rossig C, Juergens H, Schrappe M, et al. Effective childhood cancer treatment: The impact of large scale clinical trials in Germany and Austria. Pediatr Blood Cancer. 2013;60(10):1574-81. 5. Worch J, Rohde M, Burkhardt B. Mature BCell lymphoma and leukemia in children and adolescents-review of standard chemotherapy regimen and perspectives. Pediatr Hematol Oncol. 2013;30(6):465-83. 6. Boxer LM, Dang CV. Translocations involving c-myc and c-myc function. Oncogene. 2001;20(40):5595-5610. 7. Miles RR, Raphael M, McCarthy K, et al. Pediatric diffuse large B-cell lymphoma demonstrates a high proliferation index, frequent c-Myc protein expression, and a high incidence of germinal center subtype: Report of the French-American-British (FAB) international study group. Pediatr Blood Cancer. 2008;51(3):369-374. 8. Sanchez-Beato M, Sanchez-Aguilera A, Piris MA. Cell cycle deregulation in B-cell lymphomas. Blood. 2003;101(4):1220-1235. 9. Dang CV. c-Myc target genes involved in cell growth, apoptosis, and metabolism. Mol Cell Biol. 1999;19(1):1-11. 10. Roschke V, Kopantzev E, Dertzbaugh M, Rudikoff S. Chromosomal translocations deregulating c-myc are associated with normal immune responses. Oncogene. 1997; 14(25):3011-3016. 11. Nepal RM, Zaheen A, Basit W, Li L, Berger SA, Martin A. AID and RAG1 do not contribute to lymphomagenesis in Emu c-myc transgenic mice. Oncogene. 2008;27(34):4752-4756. 12. Rosenwald A, Wright G, Chan WC, et al. The use of molecular profiling to predict survival after chemotherapy for diffuse large-Bcell lymphoma. N Engl J Med. 2002; 346(25):1937-1947. 13. Richter J, Schlesner M, Hoffmann S, et al. Recurrent mutation of the ID3 gene in Burkitt lymphoma identified by integrated genome, exome and transcriptome sequencing. Nat Genet. 2012;44(12):1316-1320. 14. Schmitz R, Young RM, Ceribelli M, et al. Burkitt lymphoma pathogenesis and therapeutic targets from structural and functional

1098

and IO are supported by the Kinderkrebsinitiative Buchholz, Holm-Seppensen (KKI). The research of JL is supported by the Clinician-Scientist Programme of the Cells-in-Motion Cluster of Excellence, Muenster, Germany. Acknowledgments We thank the patients, their parents and participating clinics of the NHL-BFM studies.

genomics. Nature. 2012;490(7418):116-120. 15. Love C, Sun Z, Jima D, et al. The genetic landscape of mutations in Burkitt lymphoma. Nat Genet. 2012;44(12):1321-1325. 16. Benezra R, Davis RL, Lassar A, et al. Id: a negative regulator of helix-loop-helix DNA binding proteins. Control of terminal myogenic differentiation. Ann N Y Acad Sci. 1990;599:1-11. 17. Benezra R, Davis RL, Lockshon D, Turner DL, Weintraub H. The protein Id: a negative regulator of helix-loop-helix DNA binding proteins. Cell. 1990;61(1):49-59. 18. Murre C. Helix-loop-helix proteins and lymphocyte development. Nat Immunol. 2005;6(11):1079-1086. 19. Hsu LY, Lauring J, Liang HE, et al. A conserved transcriptional enhancer regulates RAG gene expression in developing B cells. Immunity. 2003;19(1):105-117. 20. Beck K, Peak MM, Ota T, Nemazee D, Murre C. Distinct roles for E12 and E47 in B cell specification and the sequential rearrangement of immunoglobulin light chain loci. J Exp Med. 2009;206(10):22712284. 21. Song S, Cooperman J, Letting DL, Blobel GA, Choi JK. Identification of cyclin D3 as a direct target of E2A using DamID. Mol Cell Biol. 2004;24(19):8790-8802. 22. 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. 2011;31(1):127-137. 23. Campo E. New pathogenic mechanisms in Burkitt lymphoma. Nat Genet. 2012;44(12):1288-1289. 24. Schmitz R, Ceribelli M, Pittaluga S, Wright G, Staudt LM. Oncogenic mechanisms in Burkitt Lymphoma. Cold Spring Harb Perspect Med. 2014;4(2). 25. Moricke A, Zimmermann M, Valsecchi MG, et al. Dexamethasone vs prednisone in induction treatment of pediatric ALL: results of the randomized trial AIEOP-BFM ALL 2000. Blood. 2016;127(17):2101-2112. 26. Pearson K. X. On the criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling. Philos Mag. 1900;50(302):157-175. 27. Fisher RA. On the Interpretation of 2 from Contingency Tables, and the Calculation of P. J R Stat Soc. 1922;85(1):87-94. 28. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53(282):457-481. 29. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep. 1966;50(3):163-170. 30. Salaverria I, Martin-Guerrero I, Wagener R, et al. A recurrent 11q aberration pattern characterizes a subset of MYC-negative high-grade B-cell lymphomas resembling Burkitt lymphoma. Blood. 2014; 123(8):1187-1198.

31. Havelange V, Pepermans X, Ameye G, et al. Genetic differences between paediatric and adult Burkitt lymphomas. British journal of haematology. Br J Haematol. 2016; 173(1):137-144. 32. Forero-Castro M, Robledo C, Lumbreras E, et al. The presence of genomic imbalances is associated with poor outcome in patients with burkitt lymphoma treated with doseintensive chemotherapy including rituximab. Br J Haematol. 2016;172(3):428-438. 33. Trautmann H, Hadzidimitriou A, Darzentas N, et al. Evidence for antigen-driven development of molecularly classified Burkitt Lymphomas. Blood (ASH Annual Meeting Abstracts). 2009; 114(22):Abstract Number 317. 34. Lenze D, Leoncini L, Hummel M, et al. The different epidemiologic subtypes of Burkitt lymphoma share a homogenous micro RNA profile distinct from diffuse large B-cell lymphoma. Leukemia. 2011; 25(12):1869-1876. 35. Hummel M, Bentink S, Berger H, et al. A biologic definition of Burkitt's lymphoma from transcriptional and genomic profiling. N Engl J Med. 2006;354(23):2419-2430. 36. Rohde M, Schlesner M, Richter J, et al. Pediatric Burkitt Lymphoma of the NHLBFM group analyzed within the ICGCMMML-SEQ: Whole genome sequencing data from 12 cases. Hematol Oncol. 2013; 31 Suppl 1(Abstract number 081):96-150. 37. Perkins AS, Friedberg JW. Burkitt lymphoma in adults. Hematology Am Soc Hematol Educ Program. 2008;341-348. 38. 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. 39. Goossens T, Klein U, Kuppers R. Frequent occurrence of deletions and duplications during somatic hypermutation: implications for oncogene translocations and heavy chain disease. Proc Natl Acad Sci USA. 1998;95(5):2463-2468. 40. Casellas R, Yamane A, Kovalchuk AL, Potter M. Restricting activation-induced cytidine deaminase tumorigenic activity in B lymphocytes. Immunology. 2009; 126(3):316328. 41. Kretzmer H, Bernhart SH, Wang W, et al. DNA methylome analysis in Burkitt and follicular lymphomas identifies differentially methylated regions linked to somatic mutation and transcriptional control. Nat Genet. 2015;47(11):1316-1325. 42. Logan JE, Mostofizadeh N, Desai AJ, et al. PD-0332991, a potent and selective inhibitor of cyclin-dependent kinase 4/6, demonstrates inhibition of proliferation in renal cell carcinoma at nanomolar concentrations and molecular markers predict for sensitivity. Anticancer Res. 2013; 33(8):2997-3004. 43. Finn RS, Dering J, Conklin D, et al. PD 0332991, a selective cyclin D kinase 4/6 inhibitor, preferentially inhibits proliferation of luminal estrogen receptor-positive human breast cancer cell lines in vitro. Breast Cancer Res. 2009;11(5):R77.

haematologica | 2017; 102(6)


ARTICLE

Plasma Cell Disorders

Prognostic impact of circulating plasma cells in patients with multiple myeloma: implications for plasma cell leukemia definition

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Miquel Granell,1 Xavier Calvo,2,6 Antoni Garcia-Guiñón,3 Lourdes Escoda,4 Eugènia Abella,5 Clara Mª Martínez,1 Montserrat Teixidó,3 Mª Teresa Gimenez,4 Alicia Senín,5 Patricia Sanz,1 Desirée Campoy,3 Ana Vicent,4 Leonor Arenillas,6 Laura Rosiñol,2 Jorge Sierra,1 Joan Bladé2 and Carlos Fernández de Larrea,2 on behalf of GEMMAC (Grup per l’estudi del mieloma i l’amiloïdosi de Catalunya)

Department of Haematology, Hospital de la Santa Creu i Sant Pau, IIB-Sant Pau and Josep Carreras Leukemia Research Institutes, Universitat Autònoma de Barcelona; 2 Amyloidosis and Myeloma Unit, Department of Haematology, Hospital Clínic and IDIBAPS, Universitat de Barcelona; 3Department of Haematology, Hospital Universitari Arnau de Vilanova, Universitat de Lleida; 4Department of Haematology, Hospital Joan XXIII, Universitat Rovira i Virgili, Tarragona; 5Department of Haematology. Hospital del Mar-IMIM, Universitat Autònoma de Barcelona, and 6Laboratory of Cytology. Department of Pathology, GRETNHE, IMIM Hospital del Mar Research Institute, Barcelona, Spain 1

Haematologica 2017 Volume 102(6):1099-1104

ABSTRACT

T

he presence of circulating plasma cells in patients with multiple myeloma is considered a marker for highly proliferative disease. In the study herein, the impact of circulating plasma cells assessed by cytology on survival of patients with multiple myeloma was analyzed. Wright-Giemsa stained peripheral blood smears of 482 patients with newly diagnosed myeloma or plasma cell leukemia were reviewed and patients were classified into 4 categories according to the percentage of circulating plasma cells: 0%, 1-4%, 5-20%, and plasma cell leukemia with the following frequencies: 382 (79.2%), 83 (17.2%), 12 (2.5%) and 5 (1.0%), respectively. Median overall survival according to the circulating plasma cells group was 47, 50, 6 and 14 months, respectively. At multivariate analysis, the presence of 5 to 20% circulating plasma cells was associated with a worse overall survival (relative risk 4.9, 95%CI 2.6–9.3) independently of age, creatinine, the Durie-Salmon system stage and the International Staging System (ISS) stage. Patients with ≥5% circulating plasma cells had lower platelet counts (median 86x109/L vs. 214x109/L, P<0.0001) and higher bone marrow plasma cells (median 53% vs. 36%, P=0.004). The presence of ≥5% circulating plasma cells in patients with multiple myeloma has a similar adverse prognostic impact as plasma cell leukemia. Introduction Plasma cell leukemia (PCL) was originally defined by the presence of both >20% circulating plasma cells (PCs) and an absolute count >2 × 109/L PCs,1,2 although in many studies the presence of only 1 of the above criteria was required.3-5 PCL may be classified as primary when it presents de novo in patients without previous evidence of multiple myeloma (MM) or secondary when it is presented as a leukemic transformation of a previously recognized MM. Primary PCL is a rare entity with an incidence of 2 to 4% of MM6-8 and is associated with a worse prognosis than MM. Its median survival, in a large epidemiological study, was only 4 months.7 However, with the use of novel drugs upfront, median survival ranging from 18 to 36 months has been reported.9-15 The presence of circulating PCs, identified by cytology,16 multiparameter flow cytometry17,18 or slide-based immunofluorescence,19 is also associated with a worse prognosis in myeloma patients not fulfilling the criteria of PCL. The presence of circulating PCs is also a risk factor of progression to active disease in patients with haematologica | 2017; 102(6)

Correspondence: cfernan1@clinic.ub.es

Received: October 18, 2016. Accepted: February 17, 2017. Pre-published: March 2, 2017. doi:10.3324/haematol.2016.158303 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1099 ©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.

1099


M. Granell et al.

monoclonal gammopathy of undetermined significance20 and smoldering MM.21 It has been suggested that MM patients with circulating PCs, even below 20%, could have the same bad prognosis as patients with PCL. Indeed, a lower cutoff of ≥5% or ≥0.5 x109/L of nucleated peripheral blood cells to redefine PCL has been proposed.3 In the present study, the impact of the presence of circulating PCs assessed by cytology on the survival of patients with MM was analyzed.

Methods Requirements to enter the study were a diagnosis of symptomatic MM or primary PCL according to the International Myeloma Working Group (IMWG) criteria22 between January 2008 and December 2013 in 5 University Hospitals from Catalonia, and to have peripheral blood smears at diagnosis available for review. The study was approved by the Ethic Committee of the Hospital de la Santa Creu i Sant Pau and was conducted according to the declaration of Helsinki. Clinical data including age, sex, myeloma isotype, percentage of bone marrow plasma cells, lactate dehydrogenase (LDH), Durie-Salmon and ISS stages, as well as initial treatment and follow up, were collected from medical records. Cytogenetic analysis was performed according to local policies and patients were treated according to regional protocols. Wright-Giemsa stained peripheral blood smears were reviewed by 5 experienced hematologists on peripheral blood cytology. A minimum of 100 nucleated cells per smear were systematically counted. Each sample was analyzed by a single morphologist, with each following the same common criteria. The primary endpoint was overall survival (OS) measured from the date of diagnosis to the date of death or last follow up. Differences in demographics and baseline characteristics were compared using the two-sided Fisher's exact test for categorical variables and the Mann-Whitney U test for continuous variables. Survival analysis was performed using the Kaplan-Meier method and differences were tested for statistical significance using the log-rank test. Multivariate analysis was conducted using the Cox proportional hazards model. All calculations were performed using the software SPSS® statistics version 22.

Results Clinical characteristics The study cohort included 482 patients diagnosed with MM between January 2008 and December 2013. The median age at diagnosis was 69 years (range 28 to 92 years). Two hundred and sixty (53.9%) of the patients were males. The median follow up was 28 months for the whole cohort and 38 months for the patients alive. Two hundred and thirty-one (47.9%) patients died during follow up. First-line therapy was based on bortezomib combinations in 230 (47.7%) patients, alkylating agents with glucocorticoids in 114 (23.7%) patients, vincristine, doxorubicin (adriamycin), and dexamethasone (VAD) or VAD-like chemotherapy in 60 (12.4%), immunomodulatory-based combinations in 26 (5.4%), high-dose dexamethasone in 4 (0.8%), and only palliative care in 48 (9.9%) patients. One hundred and fifty-six (32.4%) patients received autologous stem cell transplantation as part of their first-line treatment. Twelve (2.5%) patients received allogeneic stem cell transplantation during the course of the disease. 1100

<5% PCs ≥5% PCs Figure 1. Overall survival according to the circulating plasma cells (PCs) group in patients with multiple myeloma (P<0.001).

Clinical characteristics at diagnosis according to the circulating PCs group are summarized in Table 1. Differences in age, sex, myeloma isotope, LDH, Durie-Salmon and ISS stages between the 4 groups were not statistically significant. However, patients within the 5 to 20% circulating PCs group had lower platelet counts (median 86x109/L vs. 214x109/L, P<0.0001) and a higher proportion of bone marrow PCs (median 53% vs. 36%, P=0.004). The 5 patients with >20% circulating PCs were initially treated with bortezomib and dexamethasone (2 patients); bortezomib, cyclophosphamide and dexamethasone (1 patient); melphalan and prednisone (1 patient); and bortezomib, melphalan and prednisone (1 patient). Of the 12 patients with 5-20% circulating PCs, 1 died the day after diagnosis and only received supportive care, the remaining 11 patients were initially treated as follows: bortezomib and dexamethasone (6 patients); bortezomib, thalidomide and dexamethasone (2 patients); VAD (1 patient); bortezomib, melphalan and prednisone (1 patient); and cyclophosphamide and steroids (1 patient). One patient with >20% circulating PCs and 3 patients with 5-20% circulating PCs received an autologous stem cell transplantation as consolidation.

Risk factors for overall survival in the overall cohort According to the percentage of circulating PCs, 4 groups were considered for the analysis of survival: no circulating plasma cells, 382 (79.2%) patients; 1 to 4% circulating plasma cells, 83 (17.2%) patients; 5 to 20% circulating plasma cells, 12 (2.5%) patients; and classical PCL group (>20% or >2x109/L plasma cells), 5 (1%) patients. A patient with 15% circulating PCs but an absolute circulating PC count of 2.7x109/L was included in the classical PCL group. The median OS of patients with no circulating PCs, 1 to 4%, 5 to 20% and >20% were 47 (95%CI 38.6–55.4) haematologica | 2017; 102(6)


Circulating plasma cells in myeloma

months, 50 (95%CI 31.0–68.9) months, 6 (95%CI 0.9– 11.1) months and 14 (95%CI 9.7–18.3) months, respectively (Figure 1) (P<0.001).

circulating plasma cells, age older than 65 years, DurieSalmon III, creatinine >2 mg/dL, and ISS 3 retained their significance in the multivariate analysis (Table 2).

In the univariate analysis, the other factors associated with a worse survival together with circulating PCs were age older than 65 years at diagnosis, creatinine >2 mg/dL, treatment with new drugs upfront (proteasome inhibitors or immunodulatory drugs), and Durie-Salmon and ISS advanced stages. Cytogenetic data was not included in the survival analysis due to the lack of such data in most patients. The 5 patients with classical PCL were excluded from the multivariate analysis. The finding of 5 to 20%

Risk factors for overall survival in patients treated with novel agents upfront Two hundred and sixty five of the 482 (54.9%) patients were treated upfront with proteasome inhibitors and/or inmunomodulatory drugs. Of these, 192 (72.5%) patients had 0% circulating PCs, 61 (23.0%) patients had 1 to 4% circulating PCs, 9 (3.4%) patients had 5 to 20% circulating PCs and 3 (1.1%) patients had a diagnosis of classical PCL. One hundred and ten (41.5%) patients treated with novel

Table 1. Clinical characteristics according to the number of circulating plasma cells group.

Circulating plasma cells N (% overall) Male, n (%) Age, y. median (range) Heavy chain, n (%): Ig G Ig A light chain only IgD or IgM Light chain, n (%): κ l Hemoglobin, g/L. Median (range) WBC. x109/L. Median (range) Platelets. x109/L. Median (range) Calcium, mg/dL. Median (range) Creatinine, mg/dL. Median (range) Lytic lesions. n (%): None One to 3 More than 3 D-S stage. n (%): I II III B (creat >2mg/dL) β2-m, mg/L. median (range) Albumin, g/L. median (range) ISS stage. n (%): I II III NA LDH UNL, n (%) Bone marrow PCs, %. median (range) Extramedullary disease, n (%) Cytogenetics, n/assessed (%) t(4;14) del 17p.

0%

1-4%

5-20%

>20%

382 (79.2) 218 (57.1) 70.1 (32-92)

83 (17.2) 33 (39.7) 68 (28-88)

12 (2.5) 5 (42) 65 (48-85)

5 (1.0) 4 (80) 62 (49-80)

230 (60.2) 92 (24.1) 57 (14.9) 3 (0.7)

46 (55.4) 18 (21.7) 18 (21.7) 1 (1.2)

6 (50) 2 (17) 4 (33) 0

3 (60) 0 2 (40) 0

247 (64.6) 135 (35.3) 104 (56–171) 6.2 (1.7–22.6) 214 (50–558)* 9.3 (7.2–14.8) 1.1 (0.3–11.9)

54 (65.1) 29 (34.9) 99 (68-139) 5.9 (0.9-16.2) 223 (58-493) 9.4 (7.3-14.2) 1.1 (0.5-12.2)

5 (42) 7 (58) 92 (65-144) 6.3 (3.7-7.0) 86 (24-174)* 10.1 (8.6-15.8) 1.5 (0.6-8.0)

2 (40) 3 (60) 103 (40-129) 24 (63-8) 138 (28-242) 9.4 (9.0-11.4) 1.1 (0.6-6.8)

172 (45.0) 186 (48.7) 24 (6.3)

39 (46.9) 43 (51.8) 1 (1.2)

7 (58) 4 (33) 1 (8)

3 (60) 1 (20) 1 (20)

55 (14.4) 156 (40.8) 171 (44.7) 64 (16.7) 4.6 (1–48) 37.0 (15-48.9)

7 (8.4) 41 (49.4) 35 (42.2) 18 (21.7) 5.0 (1.7-35.1) 35.7 (20-43)

1 (8) 3 (25) 8 (67) 5 (42) 5.8 (2.9-31) 31 (24-45)

2 (40) 0 3 (60)

85 (22.2) 132 (34.6) 137 (35.8) 28 (7.3) 59 (15.4) 36 (1-100)** 77 (20.1)

13 (15.7) 28 (33.7) 36 (43.4) 6 (7.2) 11 (13.2) 48 (2-100) 11 (13.2)

1 (8) 4 (33) 7 (58) 4 (33) 53 (38-90)** 3 (25)

0 1 (20) 3 (60) 1 (20) 2 (40) 43 (18-95) 1 (20)

7/164 (4.2) 17/177 (9.6)

5/31 (16.1) 3/39 (7.7)

0/4 (0) 1/8 (12)

0/3 (0) 1/3 (33)

6.7 (4.9-7) 38.1 (29-47)

*P=0.001. **P=0.006. WBC: White blood cells; D-S: Durie-Salmon; β2-m: beta 2 microglobulin; ISS: International Staging System; LDH: Lactate dehydrogenase; UNL: Upper normal limit; PCs: plasma cells.

haematologica | 2017; 102(6)

1101


M. Granell et al. Table 2. Risk factors for overall survival, overall cohort.

5-20% circulating PCs >65 years D-S stage III Creatinine >2mg/mL ISS II or III New drugs upfront LDH UNL

RR

Univariate* 95% CI

P

RR

Multivariate* 95% CI

P

4.0 2.1 1.6 1.7 2.4 1.5 1.3

2.1 – 7.3 1.6 – 2.8 1.2 – 2.0 1.3 – 2.3 1.6 – 3.5 1.1 – 2.0 0.9 – 1.9

<.001 <.001 0.001 0.001 <.001 <.001 0.07

4.9 2.0 1.7 1.5 1.7 1.6 1.4

2.6 – 9.3 1.5 – 2.8 1.2 – 2.2 1.1 – 2.1 1.1 – 2.6 1.2 – 2.2 0.9 – 1.9

<0.001 <0.001 <0.001 0.010 0.014 <0.001 0.09

*Cox model. RR: relative risk; CI: confidence interval; PCs: plasma cells; D-S: Durie-Salmon; ISS: International staging System; LDH: lactate dehydrogenase. UNL: Upper normal limit.

agents upfront died during follow up and their median OS was 50 (95% CI 38–61) months. In patients treated with novel agents upfront, the median OS in cases with 0%, 1 to 4%, 5 to 20% and PCL were 58 (95% CI NR) months, 60 (95% CI 33–86) months, 22 (95% CI 0–65) months and 14 (95% CI 1–26) months, respectively. When only 2 groups were considered, <5% and ≥5% circulating PCs, median OS were 58 (95% CI 46 – 69) months and 14 (95% CI 0.4 – 27) months (Figure 2). Together with the percentage of circulating PCs, the other factors associated with a worse OS in univariate analysis were creatinine >2mg/dL and ISS stage II or III. A trend was observed in patients >65 years old and Durie-Salmon stage III. Taking into account the aforementioned variables, having ≥5% circulating PCs alone was the only fact which maintained statistical significance in the multivariate analysis (Table 3). LDH was not included in the multivariate analysis due to the lack of statistical significance or even a trend in univariate analysis (RR 1.1, 95% CI 0.7–1.9, P=0.467).

0% PC 1-4% PCs

Discussion This study aimed to address the impact of circulating PCs on the survival of patients with MM. Seventeen per cent of patients had between 1 and 4% circulating plasma cells. That finding was not associated with other clinical characteristics and had no impact on survival. There was a completely different picture for the 2.5% of patients with 5 to 20% circulating PCs. Such patients had lower platelet counts, higher bone marrow infiltration and, importantly, a shorter survival independent of other known clinical prognostic factors. In fact, the median OS of 6 months observed in these patients is closer to that of patients with the “classical” definition of PCL. When the analysis was restricted to patients treated with novel agents upfront, the impact of circulating PCs was consistent with the whole cohort. The differences in OS observed between patients with 5-20% circulating PCs and >20% circulating PCs (6 vs. 14 months) may be explained by the low number of cases in both groups. Although the presence of circulating PCs has been previously associated with survival, conventional cytology has been used for their assessment in only one study.16 In that case, patients with circulating plasma cells constituted 14.1% of the overall series and had a median survival of 25 months. The results of the present study are consistent with the ominous prognostic impact of peripheral blood 1102

5-20% PCs PCL Figure 2. Overall survival according to the circulating plasma cells (PCs) in patients with multiple myeloma and plasma cell leukemia (PCL) treated with novel drugs upfront (P<0.001).

plasmacytosis; however, the definition of a high-risk group found was different; ≥2% circulating PCs in the study by An et al. and ≥5% circulating PCs in the study herein. Using multiparameter flow cytometry, the presence of circulating PCs has also been associated with survival.17,18,23 In the study from the Mayo Clinic,17 24% of patients had more than 400 circulating PCs. Such patients had a median OS of 32 months versus not reached in patients with 400 or less circulating PCs. In another study from the same institution, which in this case used slide-based immunofluorescence microscopy for plasma cell quantification,19 54% of patients with >4% PCs were identified. These patients had a median survival of 2.4 years compared to 4.4 years in patients with fewer circulating PCs. The aforementioned studies used much more sensitive techniques than those used in the present; this may explain the different percentages of patients with circulating PCs identified haematologica | 2017; 102(6)


Circulating plasma cells in myeloma Table 3. Factors associated with overall survival in patients treated with novel drugs upfront. ≥5% circulating PCs >65 years D-S stage III Creatinine >2mg/mL ISS II or III

RR

Univariate* 95% CI

P

RR

Multivariate* 95% CI

P

4.8 1.4 1.3 1.8 1.9

2.5 – 9.1 0.9 – 2.1 0.9 – 1.9 1.2 – 2.8 1.1 – 3.3

<0.001 0.060 0.102 0.007 0.010

4.5 1.3 1.3 1.5 1.6

2.4 – 8.8 0.8 – 1.9 0.9 – 2.0 0.9 – 2.4 0.9 – 2.7

<0.001 0.186 0.132 0.082 0.092

*Cox model. RR: relative risk; CI: confidence interval; PCs: plasma cells; D-S: Durie-Salmon; ISS: International staging System; LDH: lactate dehydrogenase.

in those studies in comparison with the present one. Despite these findings, the definition of PCL has been based on standard morphological examination of peripheral blood.3 In comparison with flow cytometry and immunofluorescence, conventional cytology identifies a smaller number of patients with an extremely poor prognosis. Additionally, conventional cytology has the advantage of being a simple and inexpensive technique that can be applied in any clinical laboratory worldwide. However, it has a limitation; conventional cytology is not able to identify the clonality of PCs, as can be achieved by flow cytometry and immunofluorescence. This may hamper the specificity of conventional cytology since polyclonal reactive PCs may be rarely detected in some patients with MM.24,25 The presence of t(4;14), del(17p), amp(1q21) and del(1p21) in malignant PCs are adverse prognostic factors in MM.26,27 Indeed, adverse cytogenetics together with ISS 3 and/or high LDH identifies a group of patients with an ultra-high risk of MM.28 Several of these genetic abnormalities, particularly del(17p)9 and chromosome 1 alterations29 are more frequent in PCL than in MM. In the series presented herein, del(17p) by fluorescence in situ hybridization (FISH) was observed in 1 of 7 patients with 5 to 20% circulating PCs and 1 of 2 patients with >20% circulating PCs. A limitation of the present study is that, due to the lack of cytogenetic data in most patients, the prognostic impact of unfavorable cytogenetic abnormalities and the revised ISS28 could not be analyzed.

References 1. Kyle RA, Maldonado JE, Bayrd ED. Plasma cell leukemia. Report on 17 cases. Arch Intern Med. 1974;133(15):813-818. 2. Noel P, Kyle RA. Plasma cell leukemia: an evaluation of response to therapy. Am J Med. 1987;83(6):1062-1068. 3. Fernández de Larrea C, Kyle RA, Durie BGM, et al. Plasma cell leukemia: consensus statement on diagnostic requirements, response criteria, and treatment recommendations by the International Myeloma Working Group. Leukemia. 2013;27(4):780791. 4. Van de Donk NW, Lokhorst HM, Anderson KC, Richardson PG. How I treat plasma cell leukemia. Blood. 2012;120(12):2376-2389. 5. Jelinek T, Kryukov F, Rihova L, Hajek R. Plasma cell leukemia: from biology to treat-

haematologica | 2017; 102(6)

As highlighted in the last consensus by IMWG,3 the diagnosis of PCL has been classically done on the basis of the presence of >20% circulating PCs and/or an absolute count >2 × 109/L PCs. However, lower peripheral blood PC counts, as showed in our study (that is, ≥5% peripheral blood plasma cells), should be considered as a diagnostic criteria of PCL (“PCL-like” myeloma or early PCL), due to the independent and strong prognostic impact. Prospective multicenter analysis with translational and correlative studies into the biology of these patients is encouraged as well as risk-oriented therapeutic strategies.30,31 Careful examination of peripheral blood by conventional microscopy should be done for all patients with MM in daily clinical practice. In conclusion, the presence of ≥5% circulating PCs by conventional cytology easily identifies a group of patients with myeloma with a prognosis as poor as that of PCL, suggesting that the diagnosis of PCL should be revisited. If confirmed in other series, especially in prospective studies of uniformly treated patients, such patients may benefit from a distinct and more intensified therapeutic approach. Funding This study was supported in part by grants AGAUR 2014SGR-1281 and 2014SGR-552 (Generalitat de Catalunya), and RD12/0036/0071, RD12/0036/0046 and PI16/00423 (Instituto de Salud Carlos III) and Fondo Europeo de Desarrollo Regional (FEDER) as well as a grant from the Cellex Research Foundation, Barcelona, Spain.

ment. Eur J Haematol. 2015;95(1):16-26. 6. Dimopoulos MA, Palumbo A, Delasalle KB, Alexanian R. Primary plasma cell leukemia. Br J Haematol. 1994;88(4):754-759. 7. Ramsingh G, Mehan P, Luo J, Vij R, Morgensztern D. Primary plasma cell leukemia: a Surveillance, Epidemiology, and End Results database analysis between 1973 and 2004. Cancer. 2009; 115(24):57345739. 8. García-Sanz R, Orfao A, González M, et al. Primary plasma cell leukemia: clinical, immunophenotipic, DNA ploidy, and cytogenetic characteristics. Blood. 1999; 93(3):1032-1037. 9. Tiedemann RE, González-Paz N, Kyle RA, et al. Genetic aberrations and survival in plasma cell leukemia. Leukemia. 2008; 22(5):1044-1052. 10. Katodritou E, Terpos E, Kelaidi C, et al. Treatment with bortezomib-based regi-

mens improves overall response and predicts for survival in patients with primary or secondary plasma cell leukemia: Analysis of the Greek myeloma study group. Am J Hematol. 2014;89(2):145-150. 11. D’Arena G, Valentini CG, Pietrantuono G, et al. Frontline chemotherapy with bortezomib-containing combinations improves response rate and survival in primary plasma cell leukemia: a retrospective study from GIMEMA Multiple Myeloma Working Party. Ann Oncol. 2012; 23(6):1499-1502. 12. Musto P, Simeon V, Martorelli MC, et al. Lenalidomide and low-dose dexamethasone for newly diagnosed primary plasma cell leukemia. Leukemia. 2014;28(1):222225. 13. Pulte D, Jansen L, Castro FA, et al. Trends in survival of multiple myeloma patients in Germany and the United States in the first

1103


M. Granell et al.

14.

15.

16.

17.

18.

19.

1104

decade of the 21st century. Br J Haematol. 2015;171(2):189-196. Gonsalves WI, Rajkumar V, Go RS, et al. Trends in survival of patients with primary plasma cell leukemia: a population-based analysis. Blood. 2014;124(6):907-912. Royer B, Minvielle S, Diouf M, et al. Bortezomib, doxorubicin, cyclophosphamide, dexamethasone induction followed by stem cell transplantation for primary plasma cell leukemia: a prospective phase II study of the intergroupe francophone du myĂŠlome. J Clin Oncol. 2016; 34(18):2125-2132. An G, Qin X, Acharya C, et al. Multiple myeloma patients with low proportion of circulating plasma cells had similar survival with primary plasma cell leukemia patients. Ann Hematol. 2015;94(2):257-264. Nowakowski GS, Witzig TE, Dingli D, et al. Circulating plasma cells detected by flow cytometry as a predictor of survival in 302 patients with newly diagnosed multiple myeloma. Blood. 2005;106(7):22762279. Gonsalves WI, Rajkumar SV, Gupta V, et al. Quantification of clonal circulating plasma cells in newly diagnosed multiple myeloma: implications for redefining high-risk myeloma. Leukemia. 2014;28(10):20602065. Witzig TE, Gertz MA, Lust JA, Kyle RA, Oâ&#x20AC;&#x2122;Fallon WM, Greipp PR. Peripheral blood

20.

21.

22.

23.

24.

25.

monoclonal plasma cells as a predictor of survival in patients with multiple myeloma. Blood. 1996;88(5):1780-1787. Kumar S, Rajkumar AV, Kyle RA, et al. Prognostic value of circulating plasma cells in monoclonal gammopathy of undetermined significance. J Clin Oncol. 2005; 23(24):5668-5674. Bianchi G, Kyle RA, Larson DR, et al. High levels of peripheral blood circulating plasma cells as a specific risk factor for progression of smoldering multiple myeloma. Leukemia. 2013;27(3):680-685. The International Myeloma Working Group. Criteria for the classification of monoclonal gammopathies, multiple myeloma and related disorders: a report of the International Myeloma Working Group. Br J Haematol. 2003;121(5):749-757. Vagnoni D, Travaglini F, Pezzoni V, et al. Circulating plasma cells in newly diagnosed symptomatic multiple myeloma as a possible prognostic marker for patients with standard-risk cytogenetics. Br J Haematol. 2015;170(4):523-531. Shtalrid M, Shvidel L, Vorst E. Polyclonal reactive peripheral blood plasmacytosis mimicking plasma cell leukemia in a patient with Staphylococcal sepsis. Leuk Lymphoma. 2003;44(2):379-380. Touzeau C, Pellat-Deceunynck C, Gastinne T, et al. Reactive plasmacytoses can mimick plasma cell leukemia: therapeutical implica-

26.

27.

28.

29.

30. 31.

tions. Leuk Lymphoma. 2007; 48(1):207208. Avet-Loiseau H, Campion L, et al. Longterm analysis of the IFM 99 trials for myeloma: cytogenetic abnormalities [t(4;14), del(17p), 1q gains] play a major role in defining long-term survival. J Clin Oncol. 2012;30(16):1949-1952. Chang H, Ning Y, Qi X, Yeung J, Xu W. Chromosome 1p21 deletion is a novel prognostic marker in patients with multiple myeloma. Br J Haematol. 2007;139(1):51-54. Moreau P, Cavo M, Sonneveld P, et al. Combination of international scoring system 3, high lactate dehydrogenase, and t(4;14) and/or del(17p) identifies patients with multiple myeloma (MM) treated with front-line autologous stem-cell transplantation at high risk of early MM progressionrelated death. J Clin Oncol. 2014; 32(20):2173-2180. Chang H, Qi X, Yeung J, Reece D, Xu W, Patterson B. Genetic aberrations including chromosome 1 abnormalities and clinical features of plasma cell leukemia. Leuk Res. 2009;33(2):259-262. Musto P. Progress in the treatment of primary plasma cell leukemia. J Clin Oncol. 2016;34(18):2082-2084. Neri A, Todoerti K, Lionetti M, et al. Primary plasma cell leukemia 2.0: advances in biology and clinical management. Expert Rev Hematol. 2016;9(11):1063-1073.

haematologica | 2017; 102(6)


ARTICLE

Plasma Cell Disorders

Monitoring multiple myeloma by next-generation sequencing of V(D)J rearrangements from circulating myeloma cells and cell-free myeloma DNA

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Anna Oberle,1 Anna Brandt,1 Minna Voigtlaender,1 Benjamin Thiele,1 Janina Radloff,1 Anita Schulenkorf,1 Malik Alawi,2 Nuray Akyüz,1 Manuela März,1 Christopher T. Ford,1 Artus Krohn-Grimberghe3,4 and Mascha Binder1

Department of Oncology and Hematology, University Medical Center HamburgEppendorf, Hamburg; 2Bioinformatics Core, University Medical Center HamburgEppendorf, Hamburg; 3LYTIQ GmbH, Paderborn and 4Analytische Informationssysteme und Business Intelligence, Universität Paderborn, Germany

1

Haematologica 2017 Volume 102(6):1105-1111

ABSTRACT

R

ecent studies suggest that circulating tumor cells and cell-free DNA may represent powerful non-invasive tools for monitoring disease in patients with solid and hematologic malignancies. Here, we conducted a pilot study in 27 myeloma patients to explore the clonotypic V(D)J rearrangement for monitoring circulating myeloma cells and cell-free myeloma DNA. Next-generation sequencing was used to define the myeloma V(D)J rearrangement and for subsequent peripheral blood tracking after treatment initiation. Positivity for circulating myeloma cells/cell-free myeloma was associated with conventional remission status (P<0.001) and 91% of non-responders/progressors versus 41% of responders had evidence of persistent circulating myeloma cells/cell-free myeloma DNA (P<0.001). About half of the partial responders showed complete clearance of circulating myeloma cells/cell-free myeloma DNA despite persistent M-protein, suggesting that these markers are less inert than the M-protein, rely more on cell turnover and, therefore, decline more rapidly after initiation of effective treatment. Positivity for circulating myeloma cells and for cell-free myeloma DNA were associated with each other (P=0.042), but discordant in 30% of cases. This indicates that cell-free myeloma DNA may not be generated entirely by circulating myeloma cells and may reflect overall tumor burden. Prospective studies need to define the predictive potential of high-sensitivity determination of circulating myeloma cells and DNA in the monitoring of multiple myeloma.

Correspondence: m.binder@uke.de

Received: November 30, 2016. Accepted: February 7, 2017. Pre-published: February 9, 2017. doi:10.3324/haematol.2016.161414 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1105

Introduction The introduction of novel proteasome inhibitors,1 immunomodulatory drugs2 and monoclonal antibodies3,4 has led to deeper and longer-lasting responses in patients with multiple myeloma.5-7 The detection of minimal residual disease has, therefore, become increasingly important for the management of this disease.8 Bone marrow minimal residual disease studies using multicolor flow cytometry9 or next-generation sequencing (NGS) of the clonotypic V(D)J immunoglobulin (Ig) rearrangement10,11 suggest that minimal residual disease predicts progression-free and overall survival. The latter technology is currently the most sensitive, with a detection rate of one per 10-6 bone marrow cells. It does, however, requires repetitive bone marrow sampling, a procedure that is painful for the majority of patients. In solid tumors but also some hematologic cancers, circulating nucleic acids (cellfree DNA) and circulating tumor cells are becoming a promising minimally-invasive tool for monitoring tumor burden and response to treatment.12-14 To date, diffusehaematologica | 2017; 102(6)

©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.

1105


A. Oberle et al.

large B-cell lymphoma is a prototype hematologic disease in which the detection of the disease-specific V(D)J rearrangement in peripheral blood cell-free DNA by NGS has been shown to predict relapses before established radiological staging methods demonstrate evidence of disease recurrence.15,16 The general applicability of this monitoring concept to multiple myeloma does, however, remain to be determined, since disease distribution, vascularity, spreading and cell turnover differ from those in lymphoma, potentially affecting the circulating malignant cell and cell-free DNA compartments. Here, we investigated the clinical utility of circulating myeloma cells and DNA in the monitoring of multiple myeloma using the clonotypic V(D)J rearrangement as a highly patient-specific detection marker.

Methods Patients’ characteristics and ethics statement A cohort of 27 patients with multiple myeloma treated at the University Medical Center Hamburg-Eppendorf was investigated as approved by the ethics committee (Ethikkommission der Landesärztekammer Hamburg) and after informed consent from the patients (Table 1). All patients underwent bone marrow sampling, performed at a time point with active disease, to determine the myeloma clonotype sequence (IGH, IGL or IGK) as well as peripheral blood sampling for analysis of V(D)J rearrangements before and during the course of treatment. Baseline clinical characteristics and remission status, according to the International Myeloma Working Group (IMWG), were assessed for all patients.

DNA preparation from peripheral blood and bone marrow leukocytes and plasma Whole blood or bone marrow aspirate was collected into tubes containing heparin and processed within 2 h. Plasma was separated by centrifugation and stored at 80°C. Leukocytes were isolated with erythrocyte lysis using a standard lysis buffer (ammonium chloride 8.29 g/L, EDTA 0.372 g/L, potassium hydrogen carbonate 1 g/L) and frozen in freezing medium (90% fetal bovine serum, Biochrom, VWR, Darmstadt, Germany; 10% dimethyl sulfoxide, Sigma-Aldrich, Taufkirchen, Germany). Genomic DNA was extracted from frozen peripheral blood or bone marrow leukocytes using a Gen Elute Mammalian Genomic DNA Miniprep Kit (SigmaAldrich, Taufkirchen, Germany). Cell-free DNA was extracted from plasma using a QIAamp Circulating Nucleic Acid Kit (QIAGEN, Hilden, Germany). DNA was quantified by absorbance (Nano Drop ND-1000, Peqlab) for genomic DNA or fluorometric methods (Qubit 3.0, Thermo Fischer Scientific) for cell-free DNA.

Next-generation sequencing of V(D)J repertoires Ig V(D)J segments were amplified with BIOMED2-FR1/FR3 (IGH), -Ig kappa (IGK) or -Ig lambda (IGL) primer pools17 containing Illumina-compatible adapters and barcodes as described previously with an input of 500 ng (75000 genomes) genomic DNA or 250 ng cell-free DNA.18 Figure 1 shows a scheme of the amplification strategy. The polymerase chain reaction (PCR) product was cleaned-up using SPRIselect reagent (Beckmann Coulter, Brea, CA, USA) and 2 μL eluted DNA were used for a second PCR, 1106

during which Illumina adapter sequences were extended and a sample-specific barcode was added. The final PCR product was size-separated with 1.5% agarose gel electrophoresis and amplicons were purified using the NucleoSpin® Gel and PCR Clean-up Kit (Macherey-Nagel, Düren, Germany). The concentration of the final PCR products was determined on Qubit 3.0 (Thermo Fischer Scientific) and amplicon purity was controlled on an Agilent 2100 Bioanalyzer (Agilent Technologies, Böblingen, Germany). NGS was performed on an Illumina MiSeq sequencer with 500 or 600 cycle single-indexed, paired-end runs.

Data analysis and statistics Demultiplexing and Fastq formatted data output was generated by the MiSeq reporter. Raw sequences were processed to Ig V(D)J clonotypes based on the MiXCR analysis tool,19 and the different sequencing samples were compared using the tcR R package for Ig analysis.20 Data were plotted using R statistical software tools as well as GraphPad Prism 5. A sample was considered positive when the clonotypic rearrangement was detected at least twice. Differences between two groups were analyzed using the two-sided Student t test and categorical data were compared by the Fisher exact test. Linear regression analyses were performed to evaluate an association between response to treatment and positivity for circulating myeloma cell V(D)J [cmc-V(D)J] and/or cell-free myeloma V(D)J [cfm-V(D)J]. Analyses were carried out using IBM SPSS version 22. A P value of <0.05 was considered statistically significant.

Results and Discussion Study design A total of 27 myeloma patients requiring myelomadirected treatment were included in this investigation. Clonotypic V(D)J rearrangements of the malignant plasma cell were determined from the bone marrow and subsequently used for clonal tracking in peripheral blood leukocyte DNA [cmc-V(D)J] and cell-free DNA [cfm-V(D)J] before and after treatment initiation at routine clinical remission assessments. Blood sampling was performed after two to four courses of the indicated treatment or within 6 months after high-dose melphalan/allogeneic stem cell transplantation, unless specified otherwise. The patients’ characteristics, treatments and sampling time points are summarized in Table 1 and Online Supplementary Table S1. A diagram of the study workflow is shown in Figure 2.

Determination of clonotypic myeloma V(D)J rearrangements In 19 of 27 patients, an unambiguous IGH VDJ rearrangement could be identified, two cases were biclonal and another two patients only showed an IGL VJ rearrangement (Table 1). The four cases without discernable rearrangements were excluded from further analysis.

Next-generation screening for circulating myeloma cell-V(D)J and cell-free myeloma-V(D)J Based on previous studies,21-23 we established optimal PCR and NGS conditions for comprehensive immune repertoire analysis and high-sensitivity detection of V(D)J haematologica | 2017; 102(6)


Myeloma V(D)J monitoring from blood

rearrangements from leukocyte DNA. A sequencing depth of 80,000 reads per sample was sufficient to comprehensively analyze the B-lineage repertoire from 500 ng genomic or 250 ng cell-free DNA, which can typically be extracted from 1-5 mL of blood (Online Supplementary

Figure S1A). To measure the sensitivity of our approach, we spiked monoclonal B-cell DNA, derived from the Burkitt lymphoma cell line DG75, into polyclonal leukocyte DNA and determined detection rates of the clonotypic DG75 V(D)J rearrangement by NGS. These experiments

Table 1. Patients’ characteristics.

Code

Sex

Age (years)

Isotype

Clonotype CDR3*

MM023

M

69

IgG κ

CACPSRYSSVWRIDYW

MM031

M

56

IgG l

CAREYYYNYVRYFDSW

MM032

M

57

IgG l

CVHRRMGQLQDWYFDLW

MM048

F

72

IgA l

CARGGYGDNPYYHYGLDVW

MM050

M

79

IgG κ

CATEISSGASVGSVKVLW

MM056

F

73

IgG l

MM059 MM060 MM062 MM082

M F F M

61 55 53 66

IgG κ IgA κ IgG l IgG l

CAREEQLFDDW** CTREGGSTDYAKNFDCW** CARDHLW CARAPAVSGPFDYW CQVSDSSSDHYVF (lambda) GARGCGSSGYY_TSFYDYVMDVW** CSTALGFWSGYRNYFDYW**

MM085

F

50

IgG l

CVRAPARWLHPVSFGYW

MM087 MM088

M M

62 57

IgG l IgG κ

CARLDGYNYYYYMDIW CARDRPLDWGSGLDFW

MM090

M

59

IgG κ

CAHTGLSVAGFHYW

MM095 MM098

F M

62 74

IgA l IgA κ

CCSYVGSYTYVF (lambda) CARIGISSPGTDYW

MM120

M

76

IgG κ

CARLGFQWFGQSIW

MM122 MM123 MM125 MM155 MM170 MM174 MM094 MM116 MM131 MM132

M F M M M M M F M M

54 61 58 56 50 75 54 60 73 70

IgA κ IgA l IgA κ IgG κ IgG κ IgA κ LC κ IgG κ IgA κ IgA κ

CARIQTWAAGWYFDLW CATNTISHAFDYW CVRVPSLYSGSYYFDYW CARDGVAADYW CIRAQLLRGYDHYYYMDVW CARDTTGGHYSGKIYPFKDW no IGH or IGK rearrangement no IGH or IGK rearrangement no IGH or IGK rearrangement no IGH or IGK rearrangement

Treatment line preceding blood sampling

Remission at sampling time point1

Btz (1st line) Len (r/r) HD-Mel (1st line) Len (r/r) Len (r/r) HD-Mel (1st line) HD-Mel (1st line) Btz (1st line) CTX (r/r) Btz (r/r) Btz/panobinostat (r/r) HD-Mel (1st line)

PR SD PD2 vgPR CR3 vgPR PD2 CR PR vgPR PR vgPR

HD-Mel (r/r) allo-HSCT (r/r) HD-Mel (r/r) CTX (r/r) HD-Mel (1st line)

PR vgPR CR SD PR

HD-Mel (1st line) Btz (1st line) HD-Mel (1st line) Btz (1st line) Btz (1st line) HD-Mel (1st line) HD-Mel (1st line) maintenance Btz (1st line) HD-Mel (1st line) Btz/Len (r/r) HD-Mel (r/r) Btz (1st line) Carf/Len (r/r) HD-Mel (1st line) Len (r/r) HD-Mel (1st line) Btz (1st line) Btz (1st line) Btz (1st line)

PD2 vgPR vgPR PR PR PD2 vgPR vgPR CR PD2 PD4 SD SD PR vgPR vgPR vgPR vgPR PR SD

Remission at sampling time point according to the International Myeloma Working Group uniform response criteria for multiple myeloma. CR: complete response; vgPR: very good partial response,; PR: partial response; SD: stable disease; PD: progressive disease; M: male; F: female; LC: light chain; κ: kappa light chain; l :lambda light chain; CDR3: complementarity-determining region; Btz: bortezomib; Len: lenalidomide; Carf: carfilzomib; HD-Mel: high-dose melphalan plus autologous hematopoietic stem cell transplantation (HSCT); allo-HSCT: allogeneic HSCT; r/r: relapsed/refractory; * IGH rearrangement, unless otherwise indicated, ** biclonal IGH rearrangement. 1sampling was generally performed after two to four cycles of the indicated treatment, or within 6 months of HD-Mel or allo-HSCT, unless otherwise specified. 2sampling at PD occurring 2-4 years after HD-Mel. 3sampling after six courses of Len. 4sampling after 15 courses of Btz/Len.

haematologica | 2017; 102(6)

1107


A. Oberle et al.

Figure 1. Scheme of V(D)J DNA amplification and next-generation sequencing from peripheral blood cellular and cell-free DNA. Illumina adapters are shown in green and blue, barcode sequences are shown in red. cfDNA: cell-free DNA; IGH: immunoglobulin heavy chain; IGK/L: immunoglobulin kappa/lambda chain; FW: forward; RV: reverse; R1/R2: read 1/2; NGS: next-generation sequencing.

Figure 2. Schematic illustration of study workflow. BMMC: bone marrow mononuclear cells; IGH: immunoglobulin heavy chain; IGK/L: immunoglobulin kappa/lambda chain; gDNA: genomic DNA; cfDNA: cell-free DNA.

1108

haematologica | 2017; 102(6)


Myeloma V(D)J monitoring from blood

showed high fidelity detection even if only low amounts of clonotypic genomes were spiked into the polyclonal background (Online Supplementary Table S2). Using these high-sensitivity detection conditions, the 23 cases with a definable myeloma V(D)J rearrangement underwent further screening of blood samples before and after initiation of treatment when routine remission evaluation was performed (Table 1). Figure 3 shows representative baseline bone marrow and peripheral blood V(D)J plots of patient MM123 with evidence of cmc-V(D)J and cfm-V(D)J. Overall, cmc-V(D)J was detectable in 71% and cfm-V(D)J in 100% of cases at the baseline screening

(Figure 4A). At the follow-up time points after treatment initiation, cmc-V(D)J was detectable in 40% and cfmV(D)J in 34% of samples (Figure 4A). For further analyses, V(D)J sampling was considered positive if cmc-V(D)J or cfm-V(D)J or both resulted positive, which was the case in 47% of follow-up samples (Figure 4B). Clear associations were observed between poor remission status (assessed by M-protein-based IMWG criteria) and positive cmcV(D)J sampling (regression coefficient 1.60; 95% CI: 0.682.50; P=0.002) (Figure 4A), evidence of cfm-V(D)J (regression coefficient 1.49; 95% CI: 0.70-2.27; P=0.001) (Figure 4A) as well as detection of V(D)J in at least one compart-

Figure 3. Representative V(D)J bone marrow and peripheral blood repertoires of patient MM123 at diagnosis. Every dot represents a clonotypic V(D)J rearrangement within the immunoglobulin repertoire. The size of each dot represents the size of the clone. The malignant plasma cell clone is highlighted in the bone marrow as well as in the cellular and cell-free peripheral blood compartments. The plot was generated using R statistical software tools. BM: bone marrow; PB: peripheral blood.

A

B

C

Figure 4. Monitoring of circulating myeloma cells [(cmc-V(D)J)] and cell-free myeloma DNA [(cfm-V(D)J)] after myeloma treatment by next-generation sequencing. (A) Positivity of patientsâ&#x20AC;&#x2122; samples for cmc-V(D)J and cfm-V(D)J at diagnosis/relapse and after treatment, respectively. Remission status is indicated according to the IMWG criteria. (B) Positivity of patientsâ&#x20AC;&#x2122; samples for V(D)J at diagnosis/relapse and after treatment. Time points were considered V(D)J-positive if the malignant clone was detectable in at least one compartment (cellular or cell-free). (C) Quantification of cmc-/cfm-V(D)J per global number of V(D)J rearrangements per compartment. Patients with PD and SD were summarized as non-responders/progressors and patients with PR, vgPR and CR were summarized as responders. PD: progressive disease; SD: stable disease; PR: partial response; vgPR: very good partial response; CR: complete response.

haematologica | 2017; 102(6)

1109


A. Oberle et al.

ment (regression coefficient 1.67; 95% CI: 0.82-2.53; P<0.001) (Figure 4B), and 91% of non-responders (patients with stable or progressive disease) remained positive for cmc-/cfm-V(D)J, compared to 41% of responders (patients with partial remission or better) (P<0.001) (Figure 4B). The percentage of clonotypic to polyclonal cellular and cellfree V(D)J DNA did not differ significantly between responders and non-responders/progressors (P=0.170) (Figure 4C). Detection of cmc-V(D)J was significantly associated with cfm-V(D)J-positivity (P=0.042). Nevertheless, polyclonal V(D)J repertoires were virtually non-overlapping and a discordance of about 30% was noted between cmcV(D)J and cfm-V(D)J positivity (example repertoire overlaps are shown in Figure 5). This suggests that circulating cellular and cell-free compartments contain complementary information and that cfm-V(D)J may not be generated entirely within the vascular space (from circulating myeloma cells), but that it reflects myeloma burden in extravascular sites such as the bone marrow or extramedullary manifestations. Concurrent investigation of both compartments does, therefore, appear reasonable. One unexpected aspect of our findings was the rather low rate of positivity for cmc-V(D)J (45%) or cfm-V(D)J (39%) or both (68%) in patients with no or incomplete Mprotein responses (very good partial response or less). Since our technical approach showed high repertoire coverage as well as high sensitivity of detection, we concluded that this reflects true absence of the clonotypic DNA in the blood-derived genomes used for sequencing library preparation and that only higher genomic input could eventually enhance the sensitivity of detection. To study whether higher genomic PCR input resulted in higher detection rates of the clonotypic V(D)J DNA, we selected two cfm-V(D)J negative cases with progressive disease and scaled up the PCR input from 250 ng (37,500 genomes) to 1250 ng (187,500 genomes), typically extractable from >10 mL of blood. As expected, repertoire diversity increased, but the clonotypic myeloma V(D)J rearrangement could still not be detected in these samples (Online Supplementary Figures S1B-D). This confirmed our previous hypothesis that a fraction of patients with no or incomplete M-protein responses may not release any myeloma DNA into the plasma and that this biomarker potentially has different biological implications than those of M-protein. Taken together, our pilot study gives valuable biological insights into the circulating cellular and cell-free compartments that can be explored by â&#x20AC;&#x153;liquid biopsyâ&#x20AC;? in multiple myeloma. It indicates that cmc-V(D)J and cfm-V(D)J may decline more promptly in response to effective treatments

References 1. Dou QP, Zonder JA. Overview of proteasome inhibitor-based anti-cancer therapies: perspective on bortezomib and second generation proteasome inhibitors versus future generation inhibitors of ubiquitin-proteasome system. Curr Cancer Drug Targets. 2014;14(6):517-536. 2. Wang Y, Yang F, Shen Y, et al. Maintenance therapy with immunomodulatory drugs in

1110

Figure 5. Overlap of V(D)J repertoires from cellular and cell-free peripheral blood compartments. Shared V(D)J clonotypes from different compartments were calculated using the tcR tool20 and overlap repertoires were plotted using the R statistical software tool. The V(D)J rearrangement of the malignant plasma cell clone [(cmc- and/or cfm-V(D)J)] is shown in red.

than the relatively inert M-protein and may, therefore, be more informative regarding cell turnover and potentially suitable for immediate estimation of treatment efficacy or even early prediction of minimal residual disease negativity, not only in patients with low- or asecretory myeloma. Due to the limitations of this study (small cohort size, heterogeneous treatments), the actual predictive significance of rapid clearance of cmc-V(D)J or cfm-V(D)J, but also its persistence in M-protein responders cannot be reliably assessed. Future prospective studies will need to address whether this noninvasive diagnostic tool is of predictive importance and therefore of additional value to the established proteinbased monitoring approach in multiple myeloma. Acknowledgments This study was supported by the Eppendorfer Krebs- und Leukämiehilfe e.V. and the Deutsche Krebshilfe (grant 110906 to MB).

multiple myeloma: a meta-analysis and systematic review. J Natl Cancer Inst. 2016;108(3). 3. Lonial S, Dimopoulos M, Palumbo A, et al. Elotuzumab therapy for relapsed or refractory multiple myeloma. N Engl J Med. 2015;373(7):621-631. 4. Lokhorst HM, Plesner T, Laubach JP, et al. Targeting CD38 with daratumumab monotherapy in multiple myeloma. N Engl J Med. 2015;373(13):1207-1219. 5. Kristinsson SY, Anderson WF, Landgren O.

Improved long-term survival in multiple myeloma up to the age of 80 years. Leukemia. 2014;28(6):1346-1348. 6. 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. 7. Howlader N, Noone AM, Krapcho M, et al. SEER Cancer Statistics Review, 1975-2013. In: NCI ed. Bethesda, MD; 2016. 8. Fulciniti M, Munshi NC, Martinez-Lopez J.

haematologica | 2017; 102(6)


Myeloma V(D)J monitoring from blood

9.

10.

11.

12.

13.

Deep response in multiple myeloma: a critical review. Biomed Res Int. 2015; 2015:832049. Rawstron AC, Gregory WM, de Tute RM, et al. Minimal residual disease in myeloma by flow cytometry: independent prediction of survival benefit per log reduction. Blood. 2015;125(12):1932-1935. Martinez-Lopez J, Lahuerta JJ, Pepin F, et al. Prognostic value of deep sequencing method for minimal residual disease detection in multiple myeloma. Blood. 2014;123(20):3073-3079. Paiva B, van Dongen JJ, Orfao A. New criteria for response assessment: role of minimal residual disease in multiple myeloma. Blood. 2015;125(20):3059-3068. Dawson SJ, Tsui DW, Murtaza M, et al. Analysis of circulating tumor DNA to monitor metastatic breast cancer. N Engl J Med. 2013;368(13):1199-1209. Armand P, Oki Y, Neuberg DS, et al. Detection of circulating tumour DNA in patients with aggressive B-cell nonHodgkin lymphoma. Br J Haematol. 2013;163(1):123-126.

haematologica | 2017; 102(6)

14. Ramirez JM, Fehm T, Orsini M, et al. Prognostic relevance of viable circulating tumor cells detected by EPISPOT in metastatic breast cancer patients. Clin Chem. 2014;60(1):214-221. 15. Roschewski M, Dunleavy K, Pittaluga S, et al. Circulating tumour DNA and CT monitoring in patients with untreated diffuse large B-cell lymphoma: a correlative biomarker study. Lancet Oncol. 2015;16(5): 541-549. 16. Kurtz DM, Green MR, Bratman SV, et al. Noninvasive monitoring of diffuse large Bcell lymphoma by immunoglobulin highthroughput sequencing. Blood. 2015;125 (24):3679-3687. 17. van Dongen JJ, Langerak AW, Bruggemann M, et al. Design and standardization of PCR primers and protocols for detection of clonal immunoglobulin and T-cell receptor gene recombinations in suspect lymphoproliferations: report of the BIOMED-2 Concerted Action BMH4-CT98-3936. Leukemia. 2003;17(12):2257-2317. 18. Schliffke S, Akyuz N, Ford CT, et al. Clinical response to ibrutinib is accompanied by normalization of the T-cell environ-

19.

20.

21.

22.

23.

ment in CLL-related autoimmune cytopenia. Leukemia. 2016;30(11):2232-2234. Bolotin DA, Poslavsky S, Mitrophanov I, et al. MiXCR: software for comprehensive adaptive immunity profiling. Nat Methods. 2015;12(5):380-381. Nazarov VI, Pogorelyy MV, Komech EA, et al. tcR: an R package for T cell receptor repertoire advanced data analysis. BMC Bioinformatics. 2015;16:175. Braig F, Marz M, Schieferdecker A, et al. Epidermal growth factor receptor mutation mediates cross-resistance to panitumumab and cetuximab in gastrointestinal cancer. Oncotarget. 2015;6(14):12035-12047. Braig F, Voigtlaender M, Schieferdecker A, et al. Liquid biopsy monitoring uncovers acquired RAS-mediated resistance to cetuximab in a substantial proportion of patients with head and neck squamous cell carcinoma. Oncotarget. 2016;7(28):42988-42995. Thiele B, Kloster M, Alawi M, et al. Nextgeneration sequencing of peripheral B-lineage cells pinpoints the circulating clonotypic cell pool in multiple myeloma. Blood. 2014;123(23):3618-3621.

1111


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Cell Therapy & Immunotherapy

Ferrata Storti Foundation

Haematologica 2017 Volume 102(6):1112-1119

A risk factor analysis of outcomes after unrelated cord blood transplantation for children with Wiskott-Aldrich syndrome

Zhanna Shekhovtsova,1,2 Carmem Bonfim,3 Annalisa Ruggeri,1,4 Samantha Nichele,3 Kristin Page,5 Amal AlSeraihy,6 Francisco Barriga,7 José Sánchez de Toledo Codina,8 Paul Veys,9 Jaap Jan Boelens,10 Karin Mellgren,11 Henrique Bittencourt,12 Tracey O’Brien,13 Peter J. Shaw,14 Alicja Chybicka,15 Fernanda Volt,1 Federica Giannotti,1,4 Eliane Gluckman,1,16 Joanne Kurtzberg,5 Andrew R. Gennery17 and Vanderson Rocha1,18 on behalf of Eurocord, Cord Blood Committee of Cellular Therapy and Immunobiology Working Party of the EBMT, Federal University of Parana, Duke University Medical Center and Inborn Errors Working Party of the EBMT

Hôpital Saint Louis, Eurocord, Paris, France; 2Dmitry Rogachev National Research Centre of Pediatric Hematology, Oncology and Immunology, Moscow, Russian Federation; 3Bone Marrow Transplantation Service, Hospital de Clínicas, Universidade Federal do Paraná, Curitiba, Brazil; 4Service d’Hematologie et Therapie Cellulaire, Hôpital Saint Antoine, Paris, France; 5Pediatric Blood and Marrow Transplantation Program, Duke University Medical Center, Durham, NC, USA; 6Section of Pediatric SCT, King Faisal Specialist Hospital & Research Centre-Riyadh, Saudi Arabia; 7Programa de Hematologia Oncologia Departamento de Pediatria, Pontificia Universidad Catolica de Chile, Santiago, Chile; 8Servicio de Hematologia y Oncologia Pediatrica, Hospital Vall d’Hebron, Barcelona, Spain; 9Great Ormond Street Hospital Children’s Charity, London, UK; 10Pediatric Blood and Marrow Transplantation Program, University Hospital Utrecht, the Netherlands; 11Department of Oncology, Hematology and Stem Cell Transplantation, The Queen Silvia Children’s Hospital Gothenburg, Sweden; 12Hematology-Oncology Division, Centre Hospitalier Universitaire Sainte-Justine, Montréal, QC, Canada; 13Sydney Children’s Hospital Kids Cancer Centre, Randwick, Australia; 14The Children’s Hospital at Westmead, Sydney, Australia; 15Wroclaw Medical University, Poland; 16Centre Scientifique de Monaco, Monaco; 17Institute of Cellular Medicine, Newcastle University, Newcastle-Upon-Tyne, UK and 18Oxford University Hospitals NHS Trust, UK 1

Correspondence: ABSTRACT

zhanna.shekhovtsova@fccho-moscow.ru

Received: October 27, 2016. Accepted: February 28, 2017. Pre-published: March 2, 2017. doi:10.3324/haematol.2016.158808 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1112 ©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.

1112

W

iskott-Aldrich syndrome is a severe X-linked recessive immune deficiency disorder. A scoring system of Wiskott-Aldrich syndrome severity (0.5-5) distinguishes two phenotypes: X-linked thrombocytopenia and classic Wiskott-Aldrich syndrome. Hematopoietic cell transplantation is curative for Wiskott-Aldrich syndrome; however, the use of unrelated umbilical cord blood transplantation has seldom been described. We analyzed umbilical cord blood transplantation outcomes for 90 patients. The median age at umbilical cord blood transplantation was 1.5 years. Patients were classified according to clinical scores [2 (23%), 3 (30%), 4 (23%) and 5 (19%)]. Most patients underwent HLA-mismatched umbilical cord blood transplantation and myeloablative conditioning with antithymocyte globulin. The cumulative incidence of neutrophil recovery at day 60 was 89% and that of grade II-IV acute graft-versus-host disease at day 100 was 38%. The use of methotrexate for graft-versus-host disease prophylaxis delayed engraftment (P=0.02), but decreased acute graft-versushost disease (P=0.03). At 5 years, overall survival and event-free survival rates were 75% and 70%, respectively. The estimated 5-year event-free survival rates were 83%, 73% and 55% for patients with a clinical score of 2, 4-5 and 3, respectively. In multivariate analysis, age <2 years at the time of the umbilical cord blood transplant and a clinical phenotype of X-linked thrombocytopenia were associated with improved event-free survival. Overall survival tended to be better in patients transplanted after 2007 (P=0.09). In conclusion, umbilical cord blood transplantation is a good alternative option for young children with Wiskott-Aldrich syndrome lacking an HLA identical stem cell donor. haematologica | 2017; 102(6)


UCBT for Wiskott-Aldrich syndrome

Introduction Wiskott-Aldrich syndrome (WAS) is a severe X-linked recessive immune deficiency disorder caused by mutations in the gene encoding for Wiskott-Aldrich syndrome protein (WASP), a key regulator of actin polymerization signaling and cytoskeletal reorganization in hematopoietic cells.1-3 A mutation in WASP results in a broad spectrum of clinical manifestations ranging from the relatively mild Xlinked thrombocytopenia (XLT) to the classic WAS phenotype characterized by microthrombocytopenia, immunodeficiency, eczema, and high susceptibility to lymphoproliferative tumors and autoimmune diseases.2,4,5 A simple scoring system on a scale from 0.5 to 5 was introduced to differentiate XLT from classic WAS patients based on the severity of the clinical phenotype (Online Supplementary Table S1).6 XLT patients (score <3) have excellent overall survival (OS), but also have a high probability of severe diseaserelated complications.7 In contrast, the classic WAS (score ≥3) usually leads to death in early childhood or adolescence, despite advances in clinical care, with a median life expectancy of only 15 years.6,8,9 Currently, the only proven curative therapy for patients with WAS is hematopoietic stem cell transplantation (HSCT).6,10 Various series of HSCT from HLA-matched related donors have consistently resulted in survival rates above 80% for patients with WAS.11-15 In the absence of a matched related donor, the OS reported after matched unrelated HSCT has been around 70%.11 In the last 20 years, unrelated donor umbilical cord blood transplantation (UCBT) has become an option for patients lacking an HLA-matched donor.16 To date, there are only few reports on outcomes after UCBT for patients with primary immune deficiencies,17-19 and they include only a few patients with WAS;18,20-22 furthermore, none of the studies has analyzed factors associated with outcomes after UCBT. We, therefore, conducted a collaborative, multicenter, retrospective risk factor analysis of patients with WAS reported to Eurocord. A total of 90 UCBT recipients met the criteria and were included in the study.

Methods Data collection This retrospective analysis is based on data reported to the European Blood and Marrow Transplantation group (EBMT) and/or Eurocord from European and non-European transplant centers through a standardized questionnaire that included information on patients, donors, diseases, and transplant outcomes. Missing information was requested in the form of a Microsoft Excel file listing transplants performed by each center along with key data extracted from the Eurocord-EBMT databases. In addition, data from Duke University Medical Center (USA), the Federal University of Parana (Brazil) and the Pontifical Catholic University of Chile were obtained from the respective centers. Recipients’ parents or legal guardians gave informed consent for HSCT according to the Declaration of Helsinki. Eurocord and the Working Party of Inborn Errors of the EBMT approved this study.

Inclusion criteria The inclusion criteria for the study were: (i) patients transplanted for WAS before December 31, 2013, and reported to Eurocord; (ii) first allogeneic unrelated HSCT. haematologica | 2017; 102(6)

Patients were excluded from the study if the diagnosis of immune deficiency was not specified or if transplants were performed with a cord blood unit that was expanded, combined with other sources of hematopoietic stem cells, or injected intrabone.

Endpoints and definitions The primary endpoint was: event-free survival (EFS), defined as survival from transplantation to last contact without any of the following events: autologous reconstitution (defined by documentation of <5% donor-derived engraftment), graft failure (defined as a lack of neutrophil recovery or transient engraftment of donor cells after transplantation, and/or a requirement for a second transplant) and death. All surviving patients were censored at the date of last contact. Other endpoints reported included: (i) OS, defined as the time from transplantation to death from any cause; (ii) cumulative incidence of neutrophil engraftment, defined as the first day of achieving a neutrophil count of ≥0.5x109/L for 3 consecutive days with evidence of donor hematopoiesis; (iii) cumulative incidence of platelet engraftment, defined as the first of 3 consecutive days after HSCT with a platelet count ≥20x109/L without platelet transfusions for at least 7 days; (iv) graft failure: primary failure defined as the neutrophil count never reaching 0.5x109/L or evidence of autologous reconstitution; secondary graft failure defined as reaching a neutrophil count of 0.5x109/L after transplantation, but experiencing a subsequent, non-transitory, decrease, or loss of donor chimerism; and (v) the incidence of acute and chronic graft-versushost disease (GvHD). Acute GvHD grade II-IV was diagnosed and graded according to published criteria.23 Chronic GvHD was also graded according to standard criteria24 and evaluated in patients who survived at least 100 days with sustained engraftment. Myeloablative conditioning was defined as conditioning including an intravenous busulfan total dose of more than 6.4 mg/kg or an oral dose greater than 8 mg/kg/day, or treosulfan >36 mg/m2 for infants (less than 12 months old) and >42 mg/m2 for others. Other regimens were considered reduced intensity conditioning. Donor-recipient HLA matching was defined considering low resolution typing for HLA class I (A and B) and high resolution typing for HLA class II (DRB1). Donor-recipient chimerism was reported on the basis of available data during the first 100 ± 30 days, 1 year ± 30 days after UCBT and at the last chimerism evaluation. Full donor chimerism was defined as the presence of ≥95% donor-derived hematopoietic cells, mixed-chimerism as 5% to 94% of these cells and autologous recovery if <5%. The patients’ immunophenotype (CD3+CD4+ and CD3+CD8+ Tlymphocytes; CD19+ B-lymphocytes) was determined 100 ± 30 days, 1 year ± 30 days and at the last assessment reported after UCBT. As normal values in childhood vary considerably with age, absolute numbers of CD4+, CD8+ and CD19+ cells were related to age-specific normal values.25,26 Immune recovery was defined as being alive with neutrophil engraftment and achieving absolute numbers of CD4+, CD8+ and CD19+ cells within the age-related normal values, as shown in Online Supplementary Table S2.

Statistical analysis To analyze risk factors for outcomes, we considered factors related to the patient (median age at diagnosis, median age at transplant, median weight at time of transplantation, gender, pretransplant cytomegalovirus serology status, Lansky score), disease [pre-transplant information on: infections, severe thrombocytopenia (platelets <20x109/L), number of platelet transfusions, platelet abnormalities, history of severe bleeding, splenectomy, presence of eczema, autoimmunity, malignancy, congenital neutropenia, clinical phenotype, median interval from date of birth to diagnosis, 1113


Z. Shekhovtsova et al.

median interval from diagnosis to transplant], cord blood unit (HLA matching, median numbers of total nucleated cells and CD34+ cells collected and infused), and the transplant (year of transplantation, use of reduced intensity or myeloablative conditioning, type of GvHD prophylaxis). Cumulative incidence curves were calculated for neutrophil and platelet engraftment, and acute and chronic GvHD in a competing risk setting, with death as a competing event.27 Gray test was used for univariate comparisons. Probabilities of EFS and OS were calculated using the Kaplan-Meier estimate; the two-sided log-rank test was used for univariate comparisons. Multivariate analyses were performed using the Cox proportional hazard regression model28 for EFS and OS, and the proportional sub-distribution hazard regression model of Fine and Gray for acute and chronic GvHD, and neutrophil and platelet engraftment. Variables that reached a P-value of 0.10 in the univariate analysis, and other relevant factors such as HLA matching and cell dose, were included in the initial models and variables were eliminated one by one in a stepwise fashion in order to retain only those variables that reached a P-value of 0.05 in the final model. P-values were twosided. Statistical analyses were performed using SPSS (Inc., Chicago, IL, USA) and S-Plus (MathSoft, Inc., Seattle, WA, USA) software packages.

Results Characteristics of the patients, donors and transplants Ninety patients with a clinical diagnosis of WAS who underwent UCBT between 1996 and 2013 in 33 centers from 20 countries met the eligibility criteria for the study. The baseline characteristics of the patients, donors and transplants are shown in Table 1. Disease severity before transplantation was expressed as a WAS score of 2 to 5. Eighteen patients (23%) had a WAS clinical score of less than 3 at the time of UCBT, indicating that they had not experienced any of the following: severe infections, difficult-to-treat eczema, autoimmunity, or malignancy. The majority of patients (n= 61, 77%) had severe clinical features of the disease at the time of transplantation, including 52 patients with a history of recurrent and/or severe infections. Seven patients had a history of autoimmune disorders and two had a history of Epstein-Barr virusassociated lymphoproliferative disease. Four patients were splenectomized before UCBT, of whom two had a platelet count <20x109/L at the time of transplantation. All of the splenectomized patients had the classic WAS clinical phenotype and three experienced severe infection before UCBT. Data on WAS gene mutations were available for 39 patients (43%). Almost equal proportions of patients carried nonsense, missense and deletion mutations. The median age at transplantaion was 1.48 years (range, 4.8 months – 14.25 years). Only nine patients were more than 5 years old at the time of UCBT. Most patients (76%) had a good performance status at the time of UCBT (Lansky score >80%). The vast majority of children were conditioned with a busulfan-containing myeloablative regimen (97%), mainly busulfan/cyclophosphamide (76%). Three patients received reduced-intensity conditioning, two of them due to severe infection at the time of UCBT. Most patients received antithymocyte globulin (n=79). All patients who received GvHD prophylaxis received a calcineurin inhibitor-containing regimen, including either cyclosporine A (in most cases) 1114

Table 1. Baseline characteristics of the patients, disease, donors and transplants.

N =90 (%*) Patients’ characteristics Gender (male/female) Weight, kg Age at diagnosis, years Time interval diagnosis- UCBT, years Age at transplantation, n. ≤5 years > 5 years CMV status before UCBT, n (%) Seropositive Seronegative

88/2 10.25 (5-51.7) 0.32 (0-8.3) 1.05 (0.06-13.05) 1.48 (0.40-14.25) 81 9 47(70) 20 (30)

Disease characteristics, n (%) WAS clinical phenotype at UCBT Severe infections Eczema Mild Moderate Severe Severe thrombocytopenia (<20x109cells/L) Microthrombocytopenia Life-threatening bleeding Malignancies Autoimmunity Congenital neutropenia WAS score 2 3 4 5 Splenectomy before UCBT Donors’ characteristics, n (%) HLA-matching 6/6 5/6 4/6 3/6 Cell dose Collected NC (x107/kg) Collected CD34+ (x105/kg) Transplant characteristics Year of transplantation Conditioning regimen Myeloablative Cy/Bu Cy/Bu+others Bu/Fluda Fluda/Treo 42 Cy/Hydroxyurea Reduced intensity Cy/Bu Fluda/Melph ± Treo 36 GvHD prophylaxis CNI/prednisolone CNI/methotrexate CNI/mycophenolate mofetil CNI CNI/others Not given Serotherapy Anti-T serotherapy (before day 0) Monoclonal antibody Not given

Median (range)

52 (68) 19 (28) 35 (51) 14 (21) 43 (70) 22 (51) 41 (59) 2 (4) 7 (13) 9 (19) 18 (23) 24 (30) 18 (23) 19 (24) 4 (5)

11 (12) 52 (58) 25 (28) 1 7.5 (0.2-3) 3.03 (0.03-35) 2007 (1996-2013) 87 (97) 67 8 6 5 1 3 1 1/1 49 (54) 17 (20) 11 (12) 9 (10) 3 (3) 1 (1) 79 (89) 2 (2) 8 (9)

*Percentage of evaluable cases. UCBT: umbilical cord blood transplantation; CMV: cytomegalovirus; WAS: Wiskott-Aldrich syndrome; NC: nucleated cells; MAC: myeloablative conditioning regiment; Cy,: cyclophosphamide; Bu: busulphan; Fluda: fludarabine; Treo: treosulphan; GvHD: graft-versus-host disease; CNI: calcineurin inhibitor.

haematologica | 2017; 102(6)


UCBT for Wiskott-Aldrich syndrome

or tacrolimus. The median numbers of total nucleated cells and CD34+ cells infused were 6.8x107 cells/kg and 3.04x105/kg (pre-cryopreservation counts), respectively.

Neutrophil and platelet recovery Eighty patients (89%) achieved neutrophil engraftment, with a median time to engraftment of 21 days (range, 9-54). The cumulative incidence of neutrophil engraftment was 70% and 89% at days 30 and 60, respectively. Ten (11%) patients did not achieve neutrophil engraftment. Of the patients who failed to engraft, four received a second transplant and were alive at last follow-up; six did not receive a second HSCT and died. Multivariate analysis showed that use of methotrexate in GvHD prophylaxis [hazard ratio (HR)=0.55, 95% confidence interval (95% CI): 0.32-0.93; P=0.02)] was associated with a lower cumulative incidence of neutrophil engraftment. At day 180, the cumulative incidence of platelet engraftment was 75%, with a median time to engraftment of 45 days (range, 11-224). Platelet engraftment in the four splenectomized patients seemed faster, occurring at a median time of 35 days. Due to the small number of splenectomized patients, it was not possible to confirm whether this procedure has a real impact on the engraftment rate. In univariate analysis, the only risk factor associated with a lower cumulative incidence of platelet engraftment was age more than 2 years (67% versus 79% for younger patients, P=0.03). Multivariate analysis confirmed that age was independently associated with platelet engraftment (HR=0.34, 95% CI: 0.16-0.73; P=0.005).

Chimerism and immune recovery Chimerism data were available for 66 (86%) out of 77 evaluable patients at 100 (±30) days after UCBT, 50 (80%) out of 63 patients at 1 year (±30 days) after UCBT, and 51 (82%) out of 62 patients at the last assessment. At day 100, 68% of patients had full donor chimerism and 32% had mixed chimerism; at 1 year, these values were 76% and 24%, respectively while at the last assessment they were 80% and 20%, respectively. In 12 cases, mixed chimerism became full donor reconstitution in a further assessment, and two patients who, initially, had full donor reconstitution became stable mixed chimera at their last assessment. Information on the absolute number of CD3+/4+, CD3+/8+ and CD19+ lymphocytes at 100 ± 30 days, 1 year ± 30 days and at the latest assessment after UCBT was available for 29, 25 and 25 of the patients who were alive with neutrophil engraftment at the specific time-points, respectively. In this subset analysis, 31 (67%) out of 46 patients achieved immune recovery. The median time between UCBT and the first reported immune recovery testing was 12 months; 11 patients achieved immune recovery within the first 12 months after transplantation; the earliest confirmed immune recovery was reported 4 months after UCBT. Fifteen patients did not achieve immune recovery. Of these, 13 patients were being treated with immunosuppressive agents due to acute GvHD; the remaining two patients, who had no history of GvHD, had mixed chimerism results.

Acute and chronic graft-versus-host disease Acute GvHD grade II-IV was observed in 35 patients: 22 had grade II (24%), 8 had grade III (8%), and 5 had grade haematologica | 2017; 102(6)

IV (6%). The cumulative incidence of acute GvHD grade II-IV at day 100 was 38%. In univariate analysis, none of the factors analyzed was significantly associated with an increased risk of grade II-IV GvHD. However, in multivariate analysis, the use of methotrexate for GvHD prophylaxis was associated with a decreased incidence of grade II-IV GvHD (22% versus 42%) (HR=0.34, CI 95%: 0.120.91; P=0.03). The cumulative incidence of chronic GvHD at 5 years was 17% (n=15; 6 extensive and 9 limited cases). In univariate analysis, the cumulative incidence of chronic GvHD decreased after 2007 (25% versus 3%, P<0.01). Chronic GvHD also decreased for patients receiving a total nucleated cell dose lower than 6.8x107/kg, (23% versus 7%, P=0.03). In multivariate analysis, none of the risk factors studied was significantly associated with an increased risk of chronic GvHD.

Overall survival, event-free survival and causes of death The probabilities of OS and EFS at 5 years were 75±5% and 70±5%, respectively. Table 2 shows the univariate analysis of risk factors for OS and EFS. The risk factors associated with worst OS in multivariate analysis (Table 3) were: age over 2 years at UCBT (HR=2.61, 95% CI: 1.16.16; P=0.02) and clinical score >2 (HR=4.49, 95% CI: 1.02-19.78; P=0.04)] (Figures 1 and 2). There was a trend to improved OS in patients transplanted after 2007 (HR=2.27, 95% CI: 0.86-5.98; P=0.09) (Figure 3). In multivariate analysis for EFS, older children (more than 2 years) at UCBT also had a significantly worse prognosis (HR=2.47, 95% CI: 1.1-5.52; P=0.02)]. Sixty-seven patients were alive at the last assessment, with a median follow-up of 5 years (range, 0.25-17). Twenty-three (25%) patients had died. Table 4 shows causes of death less than and more than 100 days after UCBT, according to WAS score. Infection-related deaths were commonly observed among patients with all disease scores, and infection was the main cause of death, especially before day 100.

Discussion This multicenter, retrospective study on UCBT recipients with WAS confirms that, for most patients, HSCT using HLA-matched or -mismatched cord blood cells can cure and prevent the long-term, life-threatening complications associated with WAS. Outcomes of patients with WAS who do not undergo HSCT remain poor, with the mean age at death being 20 years in previous reports9 and with increasing risk of malignancies with age. Several groups11,15,16 reported successful HSCT results with an OS of up to 88% when using a “gold standard donor”.14 In the absence of a matched related donor, other groups have reported the successful use of matched unrelated donors with 71% OS.11,29,30 but with higher risks of acute and chronic GvHD. Unfortunately, many patients do not have an available matched unrelated donor and, therefore, other donor sources for transplantation have been investigated, such as T-cell-depleted haploidentical HSCT and umbilical cord blood HSCT. The results after haploidentical HSCT with TcRαβ/CD19-depletion for patients with primary immunodeficiency currently seem promising.31 On the other hand, UCBT is still attractive because of the naïvety of the stem cells, the lower HLA matching requirements, and easy availability (compared to matched related 1115


Z. Shekhovtsova et al.

and matched unrelated donors) and decreased GvHD (compared to haploidentical HSCT).32 We conducted this risk factor analysis for WAS, a rare disease, using retrospective-registry-based data. The limitations of our study are mainly due to some missing data related to the disease and the long inclusion period with changes in cord blood unit selection and better supportive care in more recent years. Despite these limitations, the study remains noteworthy, as it is the largest series of children with WAS treated with UCBT. We were able to identify two main factors associated with EFS and OS after UCBT: age at UCBT and clinical disease score (Figures 1 and 2). EFS and OS were significantly improved when transplantation was performed before 2 years of age, with almost 80% of these young patients being cured. This finding supports the need for early referral for transplantation in infants diagnosed with WAS. We could speculate that older children could have more previous complications before UCBT, which, in turn, could affect outcomes. However, in patients with available data, we determined that OS and EFS were not associated with number of previous infections, Lansky score, severity of eczema, thrombocytopenia, number of previous platelet transfusions, autoimmunity or congenital neutropenia. Clinical score was also a prognostic factor. As expected, patients with XLT (score <3) had better OS and EFS than patients with other clinical phenotypes. XLT patients have, historically, excellent OS without transplantation in contrast to patients with classic WAS. However, EFS in XLT patients seems to worsen over time.33,34 Data from the XLT registry showed an EFS of 74% at 15 years, decreasing to 56% by 30 years, without subsequent transplantation. There is currently no consensus on the indications for HSCT in XLT and the decision regarding transplantation for such patients has to be made on an individual basis.7 In our cohort, 23% of the patients were classified as having score 2, and there were no patients with a score of 0.5 or 1 (Table 1), showing that some degree of severity was present to justify the transplantation. We found that patients with a clinical score of 3 seemed to have worse OS and EFS probabilities, although we were not able to draw definitive conclusions because of the small number of patients in the groups. The most frequent cause of death after UCBT was infection. The majority of patients in our series received antithymocyte globulin as part of their conditioning regimen, which may explain the high number of infection-related deaths observed.35 Infections are commonly seen after UCBT due to delayed engraftment and impaired immune recovery mainly when anti-thymocyte globulin is used before UCBT.36 We were able to analyze immune recovery in a subset of patients. We found that 67% of the 46 patients with available information achieved immune recovery. The median time between UCBT and the first test reporting immune recovery was 12 months, and 11 patients achieved immune recovery within 12 months of transplantation. These results seem comparable to those regarding immune recovery usually observed after UCBT.26 However, due to the retrospective nature of our analysis, we were unable to collect details on intravenous immunoglobulin use or vaccine-specific antibody responses; the immune recovery results reported here should, therefore, be taken with caution. Most patients who did 1116

Table 2. Probability of 5-year overall and event-free survival after umbilical cord blood transplantation for Wiskott-Aldrich syndrome.

5 years OS % (95%CI) Variable

P

N

Patients’ characteristics Age at UCBT ≤ 2 years 60 > 2 years 30 CMV status negative 19 positive 47 Weight at UCBT <10 kg 45 ≥ 10 kg 45 Splenectomy Yes 4 No 75 Lansky score <90% 19 ≥ 90% 59 Disease characteristics Clinical score 2 18 3 24 4 18 5 19 UCBT characteristics HLA-match 6/6 11 5/6 53 4/6 or 3/6 26 ABO-compatibility no incompatibility 36 minor incompatibility 17 major incompatibility 29 Year of UCBT before 2007 51 after 2007 39 Time from diagnosis to UCBT < 7 months 23 7-13 months 22 >13-19 months 22 > 19 months 22 Conditioning regimen Cy/Bu/ATG 25 others 65 GvHD prophylaxis methotrexate 18 others 72 Cell doses TNC collected 43 < 7x107/kg 42 ≥ 7x107/kg TNC infused < 6x106/kg 43 ≥ 6x106/kg 42 CD34+collected 34 < 3x106/kg 33 ≥ 3x106/kg CD34+ infused <3x105/kg 34 ≥3x105/kg 33

5 years EFS % (95%CI) P

83 (71-91) 58 (41-74)

0.027

78 (67-86) 55 (38-71)

0.05

77 (54-91) 80 (66-89)

0.9

73 (51-88) 74 (59-85)

0.9

89 (76-96) 60 (44-74)

0.002

82 (68-91) 58 (42-72)

0.02

50 (15-85) 76 (42-93)

0.2

50 (15-85) 72 (61-81)

0.33

67 (44-84) 76 (63-86)

0.63

57 (34-77) 74 (61-84)

0.2

89 (69-97) 61 (41-77) 78 (55-91) 79 (57-91)

0.03

83 (61-94) 55 (35-75) 72 (50-88) 74 (52-88)

0.18

71 (41-90) 76 (63-86) 72 (52-86)

0.9

61 (33-83) 72 (59-82) 69 (50-83)

0.75

70 (47-86) 76 (58-88) 70 (49-85)

0.8

72 (55-850 70 (47-86) 70 (51-84)

0.9

69 (54-81) 83 (66-93)

0.13

65 (51-77) 78 (62-51)

0.2

91 (73-97) 80 (56-93) 72 (50-87) 58 (37-77)

0.14

86 (68-95) 76 (53-90) 68 (47-83) 54 (34-73)

0.18

83 (64-93) 72 (60-81)

0.2

83 (60-91) 67 (55-77)

0.3

74 (63-83) 78 (54-91)

0.7

70 (57-80) 72 (48-88)

0.7

72 (57-83) 77 (61-88)

0.43

67 (52-79) 73 (58-84)

0.4

77 (61-88) 72 (57-83)

0.8

74 (59-85) 65 (49-79)

0.5

68 (51-81) 78 (59-90)

0.2

62 (46-76) 75 (57-87)

0.12

66 (52-78) 86 (64-95)

0.11

74 (59-85) 75 (57-87)

0.14

UCBT: umbilical cord blood transplantation; OS: overall survival; EFS: event-free survival; CI: confidence interval; CMV,: cytomegalovirus; Cy: cyclophosphamide; Bu: busulphan; ATG: anti-thymocyte globulin; GvHD: graft-versus-host disease; TNC: total nucleated cells.


UCBT for Wiskott-Aldrich syndrome

not achieve immune recovery were treated with immunosuppressive agents due to acute GvHD, which may explain our findings. Mixed chimerism is an undesirable outcome following HSCT for WAS since it may be associated with lymphopenia, autoimmunity and thrombocytopenia.16 In our series, chimerism data were available for 86% of the patients at 3 months, 12 months and the last assessment. Full donor chimerism was observed in 68%, 76% and 80% of the patients at these timepoints, respectively. In a recent study of chimerism in 194 HSCT recipients with WAS, 72.1% of patients achieved full and stable donor chimerism.16 In this study it was also shown that mixed chimerism affects the myeloid compartment (16.5% of

Table 3. Multivariate analysis for overall survival and event-free survival. Overall survival Older than 2 years WAS score more than 2 UCBT after 2007 Event-free survival Older than 2 years WAS score more than 2 UCBT after 2007

P

HR

95% CI

4.49 2.61 2.27

1.02-19.78 1.1-6.16 0.86-5.98

0.04 0.02 0.09

2.47 3.13 1.68

1.1-5.52 0.9-10.87 0.7-4.04

0.02 0.07 0.24

HR,: hazard ratio; CI: confidence interval; WAS,: Wiskott-Aldrich syndrome; UCBT: umbilical cord blood transplantation.

Figure 1. Probability of overall survival after umbilical cord blood transplantation for Wiskott-Aldrich syndrome according to age.

Figure 2. Probability of event-free survival after umbilical cord blood transplantation for Wiskott-Aldrich syndrome according to clinical score.

haematologica | 2017; 102(6)

1117


Z. Shekhovtsova et al.

Figure 3. Probability of overall survival after umbilical cord blood transplantation for Wiskott-Aldrich syndrome according to year of transplantation.

Table 4. Primary causes of death before and more than 100 days after umbilical cord blood transplantation for Wiskott-Aldrich syndrome.

Primary causes of death N deaths/N total Before 100 days Viral infection Fungal infection Bacterial infection Unknown infection After 100 days Infection GvHD Multiorgan failure Unknown Interstitial pneumonitis Hemorrhage

2 2/18

3 11/24

1

5 3

WAS clinical score 4 4/18

Unknown 2/11

1

1

1

Total number 5 4/19

1 1 1 6 5

4

1 2 1 1 1

3 1 2

2 1

1

7 3 2 1 1 16 7 3 2 2 1 1

WAS: Wiskott-Aldrich syndrome; GvHD: graft-versus-host disease.

cases), followed by the B-cell compartment (7.4% of cases) and uncommonly the T-cell compartment (3.2% of cases). Unfortunately, in our series, we did not have data on lineage-specific chimerism; however, our results are comparable with the 72% full donor chimerism seen with other sources of hematopietic stem cells. We were unable to identify cord blood donor-related factors associated with outcomes. Previously, the Eurocord group reported that for UCBT recipients with non-malignant disorders, a cell dose higher than 5x107/kg and 6/6 or 5/6 HLA-matched grafts are associated with decreased mortality.35 In our study, patients were transplanted with a median cell dose of 7.5x107/kg and 70% received a 6/6 or 5/6 matched cord blood graft, which is in agreement with the recommendations for cord blood selection for patients with non-malignant disorders. 1118

In conclusion, early referral for UCBT in patients with WAS is associated with better outcomes. New treatment strategies such as autologous gene-modified HSCT may overcome the disadvantages of graft rejection and GvHD after allogeneic HSCT. However, until these strategies become clinically available, UCBT remains a good alternative for patients lacking an HLA-matched donor. Acknowledgments VR was supported by NHS Blood and Transplants and funded by the NIHR Biomedical Research Centers funding scheme, Oxford, UK and by FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo) grant 2013/02162-8, São Paulo, Brazil. The authors thank the following participant centers for sharing patients’ data: Curitiba, Brazil - Bone Marrow Transplantation Service, Hospital de Clínicas, Universidade haematologica | 2017; 102(6)


UCBT for Wiskott-Aldrich syndrome

Federal do Paraná; Durham, USA - Pediatric Blood and Marrow Transplantation Program, Duke University Medical Center; Riyadh, Saudi Arabia - Section of Pediatric SCT, King Faisal Specialist Hospital & Research Centre-Riyadh; Santiago, Chile - Programa de Hematologia Oncologia Departamento de Pediatria, Pontificia Universidad Catolica de Chile; Barcelona, Spain - Servicio de Hematologia y Oncologia Pediétrica, Hospital Vall d’Hebron; London, United Kingdom Great Ormond Street Hospital Children’s Charity; Utrecht, the

References 1. Jin Y, Mazza C, Christie JR, et al. Mutations of the Wiskott-Aldrich syndrome protein (WASP): hotspots, effect on transcription, and translation and phenotype/genotype correlation. Blood. 2004;104(13):4010-4019. 2. Ochs HD. Mutations of the Wiskott-Aldrich syndrome protein affect protein expression and dictate the clinical phenotypes. Immunol Res. 2009;44(1-3):84-88. 3. Massaad MJ, Ramesh N, Geha RS. WiskottAldrich syndrome: a comprehensive review. Ann N Y Acad Sci. 2013;1285:26-43. 4. Bosticardo M, Marangoni F, Aiuti A, Villa A, Grazia Roncarolo M. Recent advances in understanding the pathophysiology of Wiskott-Aldrich syndrome. Blood. 2009;113 (25):6288-6295. 5. Imai K, Nonoyama S, Ochs HD. WASP (Wiskott-Aldrich syndrome protein) gene mutations and phenotype. Curr Opin Allergy Clin Immunol. 2003;3(6):427-436. 6. Ochs HD, Filipovich AH, Veys P, Cowan MJ, Kapoor N. Wiskott-Aldrich syndrome: diagnosis, clinical and laboratory manifestations, and treatment. Biol Blood Marrow Transplant. 2009;15(1 Suppl):84-90. 7. Albert MH, Bittner TC, Nonoyama S, et al. X-linked thrombocytopenia (XLT) due to WAS mutations: clinical characteristics, long-term outcome, and treatment options. Blood. 2010;115(16):3231-3238. 8. Notarangelo LD, Miao CH, Ochs HD. Wiskott-Aldrich syndrome. Curr Opin Hematol. 2008;15(1):30-36. 9. Sullivan KE, Mullen CA, Blaese RM, Winkelstein JA. A multiinstitutional survey of the Wiskott-Aldrich syndrome. J Pediatr. 1994;125(6 Pt 1):876-885. 10. Hacein-Bey Abina S, Gaspar HB, Blondeau J, et al. Outcomes following gene therapy in patients with severe Wiskott-Aldrich syndrome. JAMA. 2015;313(15):15501563. 11. Filipovich AH, Stone JV, Tomany SC, et al. Impact of donor type on outcome of bone marrow transplantation for Wiskott-Aldrich syndrome: collaborative study of the International Bone Marrow Transplant Registry and the National Marrow Donor Program. Blood. 2001;97(6):1598-1603. 12. Kobayashi R, Ariga T, Nonoyama S, et al. Outcome in patients with Wiskott-Aldrich syndrome following stem cell transplantation: an analysis of 57 patients in Japan. Br J Haematol. 2006;135(3):362-366. 13. Buckley RH, Schiff SE, Schiff RI, et al.

haematologica | 2017; 102(6)

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

Netherlands - University Medical Center Utrecht; Goteborg , Sweden - Deptartment of Oncology and Immunology, The Queen Silvia Children’s Hospital Center for Hematopoietic Cell Transplantation; Montréal, Canada - Haematology-Oncology Division, Centre Hospitalier Universitaire Sainte-Justine; Randwick NSW, Australia - Sydney Children’s Hospital Centre for Children’s Cancer; Sydney, Australia - The Children`s Hospital at Westmead; Wroclaw, Poland - Wroclaw Medical University.

Hematopoietic stem-cell transplantation for the treatment of severe combined immunodeficiency. N Engl J Med. 1999;340(7):508516. Ozsahin H, Cavazzana-Calvo M, Notarangelo LD, et al. Long-term outcome following hematopoietic stem-cell transplantation in Wiskott-Aldrich syndrome: collaborative study of the European Society for Immunodeficiencies and European Group for Blood and Marrow Transplantation. Blood. 2008;111(1):439445. Shin CR, Kim MO, Li D, et al. Outcomes following hematopoietic cell transplantation for Wiskott-Aldrich syndrome. Bone Marrow Transplant. 2012;47(11):1428-1435. Moratto D, Giliani S, Bonfim C, et al. Longterm outcome and lineage-specific chimerism in 194 patients with WiskottAldrich syndrome treated by hematopoietic cell transplantation in the period 1980-2009: an international collaborative study. Blood. 2011;118(6):1675-1684. Knutsen AP, Steffen M, Wassmer K, Wall DA. Umbilical cord blood transplantation in Wiskott Aldrich syndrome. J Pediatr. 2003;142(5):519-523. Knutsen AP, Wall DA. Umbilical cord blood transplantation in severe T-cell immunodeficiency disorders: two-year experience. J Clin Immunol. 2000;20(6):466-476. Diaz de Heredia C, Ortega JJ, Diaz MA, et al. Unrelated cord blood transplantation for severe combined immunodeficiency and other primary immunodeficiencies. Bone Marrow Transplant. 2008;41(7):627-633. Bhattacharya A, Slatter MA, Chapman CE, et al. Single centre experience of umbilical cord stem cell transplantation for primary immunodeficiency. Bone Marrow Transplant. 2005;36(4):295-299. Jaing TH, Tsai BY, Chen SH, Lee WI, Chang KW, Chu SM. Early transplantation of unrelated cord blood in a two-month-old infant with Wiskott-Aldrich syndrome. Pediatr Transplant. 2007;11(5):557-559. Kaneko M, Watanabe T, Watanabe H, et al. Successful unrelated cord blood transplantation in an infant with Wiskott-Aldrich syndrome following recurrent cytomegalovirus disease. Int J Hematol. 2003;78(5):457-460. Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant. 1995;15(6):825-828. Shulman HM, Sullivan KM, Weiden PL, et al. Chronic graft-versus-host syndrome in man. A long-term clinicopathologic study of

25.

26.

27.

28. 29.

30.

31.

32.

33.

34.

35.

36.

20 Seattle patients. Am J Med. 1980;69(2): 204-217. Comans-Bitter WM, de Groot R, van den Beemd R, et al. Immunophenotyping of blood lymphocytes in childhood. Reference values for lymphocyte subpopulations. J Pediatr. 1997;130(3):388-393. Niehues T, Rocha V, Filipovich AH, et al. Factors affecting lymphocyte subset reconstitution after either related or unrelated cord blood transplantation in children -- a Eurocord analysis. Br J Haematol. 2001;114 (1):42-48. Gooley TA, Leisenring W, Crowley J, Storer BE. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med. 1999;18(6):695-706. Cox. Regression models and life tables. J R Stat Soc. 1972;34:187. Filipovich AH, Shapiro RS, Ramsay NK, et al. Unrelated donor bone marrow transplantation for correction of lethal congenital immunodeficiencies. Blood. 1992;80(1):270276. Lenarsky C, Weinberg K, Kohn DB, Parkman R. Unrelated donor BMT for Wiskott-Aldrich syndrome. Bone Marrow Transplant. 1993;12(2):145-147. Balashov D, Shcherbina A, Maschan M, et al. Single-center experience of unrelated and haploidentical stem cell transplantation with TCRalphabeta and CD19 depletion in children with primary immunodeficiency syndromes. Biol Blood Marrow Transplant. 2015;21(11):1955-1962. Gluckman E, Rocha V, Arcese W, et al. Factors associated with outcomes of unrelated cord blood transplant: guidelines for donor choice. Exp Hematol. 2004;32(4):397407. Albert MH, Bittner T, Stachel D, et al. Clinical Phenotype and long term outcome in a large cohort of X-linked thrombocytopenia (XLT)/mild Wiskott-Aldrich-syndrome patients. Blood. 2008;112(11):40. Buchbinder D, Nugent DJ, Fillipovich AH. Wiskott-Aldrich syndrome: diagnosis, current management, and emerging treatments. Appl Clin Genet. 2014;7:55-66. Rocha V, Gluckman E, Eurocord-Netcord registry and European Blood and Marrow Transplant Group. Improving outcomes of cord blood transplantation: HLA matching, cell dose and other graft- and transplantation-related factors. Br J Haematol. 2009;147(2):262-274. Ballen KK. ATG for cord blood transplant: yes or no? Blood. 2014;123(1):7-8.

1119


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Complications in Hematology

Ferrata Storti Foundation

Haematologica 2017 Volume 102(6):1120-1130s

Human rhinovirus detection in the lower respiratory tract of hematopoietic cell transplant recipients: association with mortality

Sachiko Seo,1,2 Alpana Waghmare,1,3,4 Emily M Scott,5 Hu Xie,6 Jane M Kuypers,1,7 Robert C. Hackman,1 Angela P. Campbell,1,3,4,* Su-Mi Choi,8 Wendy M. Leisenring,6 Keith R. Jerome,1,7 Janet A. Englund1,3,4 and Michael Boeckh1,6,9

Vaccine and Infectious Disease Division, Fred Hutchinson Cancer Research Center, Seattle, WA, USA; 2Department of Hematology and Oncology, National Cancer Research Center East, Chiba, Japan; 3Department of Pediatrics, University of Washington, Seattle, WA, USA; 4Pediatric Infectious Disease Division, Seattle Childrenâ&#x20AC;&#x2122;s Hospital, WA, USA; 5 School of Medicine, University of Washington, Seattle, WA, USA; 6Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, WA, USA; 7Department of Laboratory Medicine, University of Washington, Seattle, WA, USA; 8Division of Infectious Diseases, Department of Internal Medicine, College of Medicine, The Catholic University of Korea, Seoul, Korea and 9Department of Medicine, University of Washington, Seattle, WA, USA 1

*Present affiliation: Centers for Disease Control and Prevention, Atlanta, GA, USA

ABSTRACT

H

Correspondence: mboeckh@fredhutch.org

Received: August 1, 2016. Accepted: January 31, 2017. Pre-published: February 9, 2017. doi:10.3324/haematol.2016.153767 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/102/6/1120 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.

1120

uman rhinoviruses are the most common respiratory viruses detected in patients after hematopoietic cell transplantation. Although rhinovirus appears to occasionally cause severe lower respiratory tract infection in immunocompromised patients, the clinical significance of rhinovirus detection in the lower respiratory tract remains unknown. We evaluated 697 recipients transplanted between 1993 and 2015 with rhinovirus in respiratory samples. As comparative cohorts, 273 recipients with lower respiratory tract infection caused by respiratory syncytial virus (N=117), parainfluenza virus (N=120), or influenza (N=36) were analyzed. Factors associated with mortality were analyzed using Cox proportional hazard models. Among 569 subjects with rhinovirus upper respiratory tract infection and 128 subjects with rhinovirus lower respiratory tract infection, probabilities of overall mortality at 90 days were 6% and 41%, respectively (P<0.001). The survival rate after lower respiratory tract infection was not affected by the presence of copathogens (55% in patients with co-pathogens, 64% in patients without, P=0.34). Low monocyte count (P=0.027), oxygen use (P=0.015), and steroid dose greater than 1 mg/kg/day (P=0.003) before diagnosis were significantly associated with mortality among patients with lower respiratory tract infection in multivariable analysis. Mortality after rhinovirus lower respiratory tract infection was similar to that after lower respiratory tract infection by respiratory syncytial virus, parainfluenza virus or influenza in an adjusted model. In summary, transplant recipients with rhinovirus detection in the lower respiratory tract had high mortality rates comparable to viral pneumonia associated with other well-established respiratory viruses. Our data suggest rhinovirus can contribute to severe pulmonary disease in immunocompromised hosts. Introduction Human rhinoviruses (HRVs) are the most common cause of respiratory virus infections in both immunocompetent and immunocompromised individuals.1-4 Although well documented cases suggested a possible role of HRV in severe disease,5 earlier cohort studies did not conclusively demonstrate the ability of HRV to haematologica | 2017; 102(6)


Human rhinovirus infections in HCT recipients

cause lower respiratory tract disease, in part because of the presence of co-pathogens as well as a lack of sensitive molecular diagnostic techniques.6,7 With the widespread adoption of multiplex molecular detection platforms, HRV RNA is now detected frequently in the bronchoalveolar lavage (BAL) fluid of immunocompromised patients undergoing evaluation for pulmonary infiltrates.3,8,9 However, the significance of HRV RNA detection in the lower respiratory tract remains poorly defined. In a recent large prospective study, detection of HRV in the upper respiratory tract of hematopoietic cell transplantation (HCT) candidates was associated with poor outcome after HCT.4 Another study also suggested poor outcomes in immunocompromised patients with HRV infection, comparable to those infected with 2009 H1N1 influenza.10 As for lower respiratory tract infection (LRI), recent studies of community-acquired pneumonia in immunocompetent patients identified HRV as the most common pathogen using molecular diagnostic techniques.1,11 HRV was also a key pathogen in immunocompromised patients originally diagnosed with idiopathic pneumonia syndrome, and its detection in the lower respiratory tract was associated with a particularly poor outcome.12 Overall, these data suggest that HRV may be a clinically significant pathogen with the potential to cause serious pulmonary disease. The purpose of this study was to determine the significance of detection of HRV RNA in the BAL fluid in HCT recipients.

lected by medical chart review. As comparative cohorts, patients with LRI caused by respiratory syncytial virus (RSV), parainfluenza virus (PIV), or influenza virus were included in the analyses.14-16 The first episode of LRI by any of the 4 viruses was selected and the cases with overlapped infections of the 4 viruses were excluded from the analyses. The study was approved by the Institutional Review Board at FHCRC.

Methods

Statistical analysis

Study design This retrospective study includes patients who were transplanted between 1993 and 2015 at the Fred Hutchinson Cancer Research Center (FHCRC) and had virologically-confirmed HRV infection following HCT at the University of Washington Virology Laboratories.13 Only an individual’s first episode of HRV infection was analyzed. HRV upper respiratory tract infection (URI) was defined as HRV detection in a nasopharyngeal sample, and LRI was defined as HRV detection in a BAL or lung biopsy sample. Patients’ demographic data and transplant information closest to the HRV infection were retrieved from the FHCRC database, and other data related to the clinical course of HRV infections were col-

A

Laboratory testing Nasopharyngeal samples were collected when HCT recipients had URI symptoms and a BAL sample was obtained when patients had lower respiratory tract symptoms and a radiographic abnormality. HRV detection was performed by conventional culture and/or reverse transcription-polymerase chain reaction (RTPCR) assay in respiratory samples. The culture was performed using 3 different culture systems (rhesus monkey kidney cells [RMK], adenocarcinomic human alveolar basal epithelial cells [A549], and human foreskin fibroblast [HFF]), and BAL samples were incubated for 10 days (an additional 11 days in HFFs culture) at 37˚C. RT-PCR to detect HRV has been used routinely since 2007 in our center. RSV, PIV or influenza virus was detected by conventional culture, direct fluorescent antibody tests and/or PCR. The viral load of HRV was roughly estimated from the PCR cycle threshold (Ct) value of the sample that provided the initial diagnosis of HRV LRI.17 All biopsy or autopsy samples were inoculated in 3 different cell lines as above and also tested for HRV using RT-PCR. In autopsy samples, 2 curls from a frozen specimen of each side of lungs were separately tested for HRV using RT-PCR.

Patients’ demographic characteristics were summarized and compared between URI and LRI or among LRI with different viruses (HRV, RSV, PIV, and Influenza virus) using chi-square or Fisher’s exact test for categorical variables and Wilcoxon rank sum test for continuous variables (as appropriate). The probability of overall survival was estimated using the Kaplan-Meier method. The probability of mortality after HRV URI was estimated by cumulative incidence curves, treating progression to LRI as a competing risk. The probability of mortality from respiratory failure was estimated by cumulative incidence curves, treating death due to other causes as a competing risk. The log-rank test was used to compare hazards of time-to-event outcomes among patients’ characteristics. Cox proportional hazards models were used to evalu-

B

Figure 1. Probability of mortality after HRV infections. (A) Cumulative incidence of overall mortality after HRV URI or LRI (N=752, P=<0.001). (B) Kaplan-Meier estimate of overall survival after HRV LRI by presence of co-pathogens (N=128, P=0.34). HRV: human rhinovirus; LRI: lower respiratory tract infection; URI: upper respiratory tract infection.

haematologica | 2017; 102(6)

1121


S. Seo et al.

ate unadjusted (uHR) and adjusted hazard ratios (aHR) of risk factors for overall mortality or respiratory mortality. Variables with P≤0.05 in the univariable models were candidates for multivari-

able models. Two-sided P values <0.05 were considered statistically significant. All statistical analyses were performed using SAS 9.3 for Windows.

Table 1. Characteristics of patients with human rhinovirus infection (N=697),

Sex Male Female Age at HCT, year ≤ 20 21-60 > 60 Disease risk at HCT Standard High Transplant year 1993-2006 2007-2015 Number of transplantation 1st 2nd 3rd or 4th Stem cell source Bone marrow Peripheral blood stem cell Cord blood Donor type Autologous Related Unrelated Conditioning regimen MA including high-dose TBI (≥ 12Gy) MA ± low-dose TBI (≤ 2Gy) Reduced intensity GvHD prophylaxis CNI + MTX CNI + MMF Others CMV serostatus Negative Positive Missing Days between HCT and infection ≤ 30 31-365 > 365 Diagnostic methods Culture±PCR PCR alone Co-pathogen* None Any pathogen Ct value at each diagnosis#

Total (n=697)

URI alone (n=569)

LRI (n=128)

420 (60) 277 (40)

337 (59) 232 (41)

83 (65) 45 (35)

163 (23) 384 (55) 150 (22)

139 (24) 318 (56) 112 (20)

24 (19) 66 (51) 38 (30)

487 (70) 210 (30)

428 (75) 141 (25)

59 (46) 69 (54)

107 (15) 590 (85)

78 (14) 491 (86)

29 (23) 99 (77)

P 0.24

0.041

<0.001

0.010

<0.001 584 (84) 100 (14) 13 (2)

491 (86) 67 (12) 11 (2)

93 (72) 33 (26) 2 (2)

120 (17) 483 (69) 94 (14)

96 (17) 401 (70) 72 (13)

24 (19) 82 (64) 22 (17)

178 (26) 176 (25) 343 (49)

160 (28) 145 (25) 264 (47)

18 (14) 31 (24) 79 (62)

177 (25) 326 (47) 194 (28)

146 (26) 293 (51) 130 (23)

31 (24) 33 (26) 64 (50)

0.28

0.001

<0.001

0.002 206 (40) 250 (48) 63 (12)

173 (42) 181 (44) 55 (14)

33 (30) 69 (63) 8 (7)

301 (43) 376 (54) 20 (3)

257 (45) 294 (52) 18 (3)

44 (34) 82 (64) 2 (2)

218 (31) 310 (45) 169 (24)

176 (31) 249 (44) 144 (25)

42 (33) 61 (48) 25 (19)

78 (11) 619 (89)

62 (11) 507 (89)

16 (12) 112 (88)

562 (81) 135 (19)

502 (88) 67 (12) 25.4 (12.2-39.8)

60 (47) 68 (53) 27.0 (13.8-39.4)

0.035

0.38

0.60

<0.001

0.15

Data are presented No. (%) unless otherwise specified. *A co-pathogen was defined as a significant bacterial, fungal, or viral pathogen detected in concurrent respiratory samples, or in a blood sample obtained within two days of diagnosis of infection.15 #Values are indicated as the median (range). URI: upper respiratory tract infection; LRI: lower respiratory tract infection; HCT: hematopoietic cell transplantation; MA: myeloablative; TBI: total body irradiation; GvHD: graft-versus-host disease; CNI: calcineurin inhibitor; MTX: methotrexate; MMF: mycophenolate mofetil; CMV: cytomegalovirus; Ct: cycle threshold; PCR: polymerase chain reaction.

1122

haematologica | 2017; 102(6)


Human rhinovirus infections in HCT recipients

Results Patient characteristics There were 697 patients diagnosed with HRV infection between 1993 and 2015; 569 (82%) and 128 (18%) had URI alone and LRI, respectively. Characteristics of each group are shown in Table 1. More than 80% of the HRV cases were diagnosed in and after 2007, reflecting the initiation of routine use of PCR to detect HRV. The median time to HRV URI and LRI following HCT was 74 days (range, 0 to 7063 days) and 87 days (range, 0 to 4309 days), respectively. Approximately half of patients with HRV LRI had co-pathogens at diagnosis, such as Aspergillus (N=11;

including 5 cases positive by galactomannan alone) and Pseudomonas aeruginosa (N=5) considered to be non-viral co-pathogens, and RSV (N=3), PIV (N=3) and adenovirus (N=3) as viral co-pathogens.

Mortality after HRV infection and risk factors There were 52 patients who died within 90 days after the onset of LRI. Forty-one patients (79%) died from pulmonary failure, and 5, 4 and 2 died from organ dysfunction, disease relapse and acute graft versus host disease (GvHD), respectively. The probabilities of overall mortality at 90 days following HRV diagnosis in patients with URI or LRI are shown in Figure 1A (6% in URI and 41%

Table 2. Risk factors for mortality from all causes or respiratory failure by day 90 after HRV LRI (N=128).

Overall mortality Lymphocyte count at diagnosis > 300 cells/ L ≤ 300 cells/ L Monocyte count at diagnosis > 300 cells/ L ≤ 300 cells/ L LRI symptoms at diagnosis No Yes Oxygen use at diagnosis No Yes Mechanical ventilation Steroid dose before diagnosis* No < 1 mg/kg ≥ 1 mg/kg Ct value at diagnosis# Mortality from respiratory failure Lymphocyte count at diagnosis > 300 cells/ L ≤ 300 cells/ L Monocyte count at diagnosis > 300 cells/ L ≤ 300 cells/ L LRI symptoms at diagnosis No Yes Oxygen use at diagnosis No Yes Mechanical ventilation Steroid dose before diagnosis* No < 1 mg/kg ≥ 1 mg/kg Ct value at diagnosis#

HR

Univariable analysis 95% CI

P

HR

Multivariable analysis 95% CI

P

1.00 2.70

1.49-4.76

0.001

1.00 1.76

0.95-3.25

0.07

1.00 2.44

1.33-4.55

0.004

1.00 2.04

1.09-3.85

0.027

1.00 3.03

1.37-6.74

0.006

1.00 1.52

0.53-4.37

0.44

1.00 2.94 10.3

1.54-5.62 4.82-22.0

0.001 <0.001

1.00 0.82 3.69 0.99

0.41-1.64 1.93-7.03 0.95-1.04

0.58 <0.001 0.70

1.00 2.50

1.30-4.76

0.006

1.00 1.62

0.82-3.21

0.17

1.00 2.44

1.22-4.76

0.012

1.00 2.05

1.00-4.19

0.049

1.00 4.28

1.52-12.0

0.006

1.00 2.60

0.75-8.96

0.13

0.90-4.92

0.09

1.38-5.32

0.004

1.00 2.79

1.22-6.37

0.015

1.35-4.50

0.003

1.00

1.00 2.72 10.3

1.31-5.66 4.47-23.8

0.007 <0.001

1.00 1.22 4.65 0.99

0.55-2.67 2.18-9.92 0.94-1.04

0.63 <0.001 0.68

2.47

1.00 2.11 1.00 2.71

*Peak steroid dose was recorded from the period within two weeks prior to LRI. #This analysis was performed as a continuous variable. All variables in Table 1 were used for the univariable analysis. Only variables with P<0.05 are shown in this table. Ct value was shown regardless of P values. HRV: human rhinovirus; LRI: lower respiratory tract infection; Ct: cycle threshold; CI: confidence interval; HR; hazard ratio.

haematologica | 2017; 102(6)

1123


S. Seo et al.

in LRI, P<0.001). The probabilities of 90-day survival after HRV LRI were 55% and 64% in patients with and without co-pathogens, respectively (P=0.34) (Figure 1B). In a multivariable analysis of risk factors for overall mortality, low monocyte count, oxygen requirement at diagnosis, and steroid dose greater than 1 mg/kg/day before diagnosis were significantly associated with higher mortality.

In the analysis of risk factors for mortality from respiratory failure, only low monocyte count at diagnosis and steroid use ≥1 mg/kg/day before diagnosis were significant factors in multivariable models (Table 2). Steroid use after diagnosis was also a significantly important factor (overall mortality: HR, 2.55; 95% confidence interval [CI], 1.39-4.69; P=0.003, mortality from respiratory failure: HR, 2.77; 95%

Table 3. Characteristics of patients with human rhinovirus infection in a restrictive cohort consisting of allogeneic transplant recipients after 2007 with LRI within 2 years after HCT (N=434).

Total (n=385) Sex Male Female Age at HCT, year ≤ 20 21-60 > 60 Disease risk at HCT Standard High Number of transplantation 1st 2nd or 3rd Stem cell source Bone marrow Peripheral blood stem cell Cord blood Donor type Related Unrelated Conditioning regimen MA including high-dose TBI (≥ 12Gy) MA ± low-dose TBI (≤ 2Gy) Reduced intensity GvHD prophylaxis CNI + MTX CNI + MMF Others CMV serostatus Negative Positive Missing Days between HCT and infection ≤ 30 31-365 > 365 Diagnostic methods Culture±PCR PCR alone Co-pathogen* None Any pathogen Ct value at each diagnosis#

URI alone (n=312)

LRI (n=73)

P 0.28

237 (62) 148 (38)

188 (60) 124 (40)

49 (67) 24 (33)

123 (32) 198 (51) 64 (17)

112 (36) 160 (51) 40 (13)

11 (15) 38 (52) 24 (33)

274 (71) 111 (29)

240 (77) 72 (23)

34 (47) 39 (53)

321 (83) 63 (17)

272 (87) 39 (13)

49 (67) 24 (33)

85 (22) 222 (58) 77 (20)

75 (24) 177 (57) 59 (19)

10 (14) 45 (62) 18 (25)

123 (32) 262 (68)

103 (33) 209 (67)

20 (27) 53 (73)

116 (30) 132 (34) 136 (35)

103 (33) 118 (38) 90 (29)

13 (18) 14 (19) 46 (63)

146 (38) 187 (49) 51 (13)

130 (42) 137 (44) 44 (14)

16 (22) 50 (68) 7 (10)

156 (41) 221 (57) 8 (2)

137 (44) 169 (54) 6 (2)

19 (26) 52 (71) 2 (3)

133 (35) 216 (56) 36 (9)

112 (36) 172 (55) 28 (9)

21 (29) 44 (60) 8 (11)

9 (2) 376 (98)

6 (2) 306 (98)

3 (4) 70 (96)

299 (78) 86 (22)

270 (87) 42 (13) 25.3 (12.2-39.2)

29 (40) 44 (60) 27.0 (13.8-39.4)

<0.001

<0.001

<0.001

0.23

0.35

<0.001

0.002

0.020

0.50

0.27

<0.001

*A co-pathogen was defined as a significant bacterial, fungal, or viral pathogen detected in concurrent respiratory samples, or in a blood sample obtained within two days of diagnosis of infection.15 #Values are indicated as the median (range). URI: upper respiratory tract infection; LRI; lower respiratory tract infection; HCT: hematopoietic cell transplantation; MA: myeloablative; TBI: total body irradiation; GvHD: graft-versus-host disease; CNI: calcineurin inhibitor; MTX: methotrexate; MMF: mycophenolate mofetil; CMV: cytomegalovirus; Ct: cycle threshold; PCR: polymerase chain reaction.

1124

haematologica | 2017; 102(6)


Human rhinovirus infections in HCT recipients

CI, 1.39-5.52; P=0.004) (Online Supplementary Table S1), but allogeneic transplantation did not reach statistical significance (overall mortality: HR, 2.09; 95% CI, 0.76-5.81; P=0.16, mortality from respiratory failure: HR, 2.18; 95% CI, 0.67-7.06; P=0.19). To examine a more homogeneous cohort, we analyzed a cohort restricted to cases with allogeneic transplantation after 2007 and HRV infections within 2 years after HCT (Table 3). Oxygen requirement at diagnosis and steroid dose greater than 1 mg/kg/day before diagnosis were important factors for mortality (Table 4).

The HRV PCR Ct value of the BAL sample obtained at diagnosis was not significantly associated with mortality after HRV LRI (Tables 2 and 4). There were 20 cases with sequential BALs within 90 days following diagnosis of LRI. Fourteen of those 20 patients had decreased viral load, of whom 6 (43%) died; the remaining 6 patients had stable or increased viral load, and all six died. There were 13 patients with sequential nasopharyngeal samples and 7 and 6 had decreased and increased viral loads, respectively. Each group had 1 patient who died.

Table 4. Risk factors for mortality from all causes or respiratory failure by day 90 after HRV LRI onset in a restricted cohort consisting of allogeneic transplant recipients after 2007 with LRI within 2 years after HCT (N=73).

HR Overall mortality Lymphocyte count at diagnosis > 300 cells/μL ≤ 300 cells/μL Monocyte count at diagnosis > 300 cells/μL ≤ 300 cells/μL LRI symptoms at diagnosis No Yes Oxygen use at diagnosis No Any Mechanical ventilation Steroid dose before diagnosis* No < 1 mg/kg ≥ 1 mg/kg Ct value at diagnosis# Mortality from respiratory failure Lymphocyte count at diagnosis > 300 cells/μL ≤ 300 cells/μL Monocyte count at diagnosis > 300 cells/μL ≤ 300 cells/μL LRI symptoms at diagnosis No Yes Oxygen use at diagnosis No Any Mechanical ventilation Steroid dose before diagnosis* No < 1 mg/kg ≥ 1 mg/kg Ct value at diagnosis#

Univariable analysis 95% CI P

1.00 2.44

1.18-5.26

0.017

1.00 1.41

0.67-2.94

0.37

1.00 2.42

0.93-6.31

0.07

1.00 2.50 8.53

1.12-5.58 3.06-23.8

0.025 <0.001

HR

Multivariable analysis 95% CI

P

1.00 1.96

0.91-4.23

0.09

1.00

1.00 0.42 2.70 0.99

0.15-1.19 1.24-5.89 0.95-1.04

0.10 0.013 0.70

1.00 2.63

1.11-6.25

0.027

1.00 1.28

0.56-2.94

0.55

1.00 3.20

0.95-10.7

0.059

1.00 2.27 8.14

0.91-5.65 2.71-26.1

0.079 <0.001

1.00 0.62 2.80 0.99

0.21-1.84 1.12-6.97 0.94-1.04

0.39 0.027 0.68

2.50

1.15-5.43

0.021

2.33

1.07-5.06

0.033

1.00 2.12

0.87-5.16

1.00

0.10

1.00 2.37

0.99-5.69

0.053

1.00 2.12

0.88-5.11

0.10

*Peak steroid dose was recorded from the period within two weeks prior to onset of LRI. #This analysis was performed as a continuous variable. All variables in Table 1 were used for the univariable analysis. Only variables with P<0.05 are shown in this table. Ct value was shown regardless of P values. LRI: lower respiratory tract infection; Ct: cycle threshold; CI: confidence interval; HR; hazard ratio.

haematologica | 2017; 102(6)

1125


S. Seo et al. A

B

C

Figure 2. Probability of overall survival in patients without copathogens by viral type. (A) Kaplan-Meier estimate of overall survival by viral type (N=222, P=0.62). (B) Kaplan-Meier estimate of overall survival in patients without oxygen requirement at diagnosis by viral type (N=97, P=0.76). (C) Kaplan-Meier estimate of overall survival in patients with oxygen requirements at diagnosis by viral type (N=125, P=0.95). HRV: human rhinovirus; RSV: respiratory syncytial virus; PIV: parainfluenza virus.

Detection of HRV in tissue samples To investigate whether HRV could be potentially responsible for lower respiratory tract disease or subsequent death, we examined the biopsy and/or autopsy samples for HRV using conventional culture and the RTPCR assay. Of 15 biopsies performed at LRI diagnosis and 7 after diagnosis, 15 (100%) and 6 (86%), respectively, were HRV positive by RT-PCR; 2 (13%) and 4 (57%), respectively, were HRV positive by culture. Among 128 patients with HRV LRI, 52 died within 90 days after onset of HRV LRI. Among them, 12 patients underwent autopsy (median days between LRI diagnosis and autopsy, 10 days; range, 2 to 67 days) and HRV was detected in 6 samples by RT-PCR. Both samples obtained from right and left lungs were positive in 5 of 6 positive cases. No samples were positive by culture.

Mortality comparison with LRI due to other respiratory viruses We compared mortality after HRV LRI with that after LRI caused by RSV (N=117), PIV (N=120), or influenza virus (N=36). Patientsâ&#x20AC;&#x2122; characteristics are shown in Table 3. We observed differences among the various virus cohorts with regard to time of onset after transplantation, cell source, and oxygen requirements at diagnosis. Overall survival by 90 days among groups without copathogens were similar (P=0.62 in Figure 2A). Since oxygen requirement at diagnosis is an important risk factor for mortality and is a reflection of the degree of lung injury, we analyzed survival by oxygen requirement. As shown in Figure 2B and 2C, the 4 groups of patients with respiratory viral LRI were comparable in overall survival (P=0.76 in Figure 2B, P=0.95 in Figure 2C). To confirm that mortality after HRV LRI is similar to that after RSV, PIV or influenza virus, we performed a multivariable analysis among the 388 total patients (Table 5). In 1126

an adjusted model, mortality in HRV LRI remained similar to that after LRI by other respiratory viruses (Table 6).

Discussion The study herein showed that patients who have HRV detected in the lower respiratory tract have similar outcomes to patients with LRI due to known pathogenic viruses such as RSV, PIV and influenza virus, even after excluding other potential bacterial, fungal and viral copathogens. Although the detection of HRV RNA by RTPCR is not proof of ongoing viral replication, we were also able to detect replicating virus from the lung tissue of a subset of patients with a fatal outcome, thus demonstrating that these viruses were viable and potentially responsible for the pulmonary disease in these patients. To prove a pathogen-disease association for a pathogen that is difficult to cultivate,18 especially in the molecular era, is difficult, as Kochâ&#x20AC;&#x2122;s postulates or other assessments of direct viral injury cannot be easily applied despite the detection of viral nucleic acid. We employed multiple strategies to examine the plausibility of HRV as a significant pathogen, including: (i) a comparison of mortality in patients with HRV detection in the upper and lower respiratory tract, (ii) a comparison between patients with and without co-pathogens in outcome after HRV LRI, (iii) an evaluation of factors associated with death following HRV detection in BAL samples, (iv) a comparison of mortality with respiratory viruses of well-established pathogenicity, and (v) tissue detection of the virus. Our primary comparison was between patients with HRV URI and those with HRV detection in the BAL (Figure 1A). While the comparison of mortality after URI and LRI alone is not conclusive evidence for a causative effect of HRV for LRI, the two curves are significantly difhaematologica | 2017; 102(6)


Human rhinovirus infections in HCT recipients

Table 5. Characteristics of patients with LRI due to respiratory virus (N=388).

Total (n=388) Transplant year 1989-1992 1993-2006 2007-2015 Stem cell source Bone marrow Peripheral blood stem cell Cord blood Donor type Autologous Related Unrelated Days between HCT and infection ≤ 30 31-365 > 365 Diagnostic methods Other methods ±PCR PCR alone Missing Co-pathogen None Any viral co-pathogen Any non-viral co-pathogen Lymphocyte count at diagnosis > 300 cells/μL ≤ 300 cells/μL Missing Monocyte count at diagnosis > 300 cells/μL ≤ 300 cells/μL Missing Oxygen use at diagnosis No Any Steroid dose before diagnosis* No < 1 mg/kg 1-2 mg/kg > 2 mg/kg Missing

HRV (n=115)

RSV (n=117)

PIV (n=120)

Influenza (n=36)

P <0.001

24 (6) 201 (52) 163 (42)

0 (0) 27 (23) 88 (77)

16 (14) 75 (64) 26 (22)

7 (6) 78 (65) 35 (29)

1 (3) 21 (58) 14 (39)

146 (38) 210 (54) 32 (8)

22 (19) 73 (63) 20 (17)

61 (52) 51 (44) 5 (4)

55 (46) 58 (48) 7 (6)

8 (22) 28 (78) 0 (0)

63 (16) 130 (34) 195 (50)

15 (13) 27 (23) 73 (64)

23 (20) 47 (40) 47 (40)

17 (14) 43 (36) 60 (50)

8 (22) 13 (36) 15 (42)

129 (33) 192 (49) 67 (17)

40 (35) 53 (46) 22 (19)

51 (44) 51 (44) 15 (13)

31 (26) 74 (62) 15 (13)

7 (19) 14 (39) 15 (42)

239 (61) 142 (37) 7 (2)

15 (13) 100 (87) 0 (0)

109 (94) 4 (3) 4 (3)

88 (73) 30 (25) 2 (2)

27 (75) 8 (22) 1 (3)

222 (57) 37 (10) 129 (33)

58 (50) 13 (12) 44 (38)

74 (63) 12 (11) 31 (26)

67 (56) 10 (8) 43 (36)

23 (64) 2 (5) 11 (31)

181 (26) 203 (52) 4 (1)

160 (28) 60 (52) 1 (1)

18 (14) 66 (56) 0 (0)

18 (14) 58 (48) 3 (3)

18 (14) 19 (53) 0 (0)

259 (67) 120 (31) 9 (2)

70 (61) 44 (38) 1 (1)

87 (74) 28 (24) 2 (2)

80 (67) 38 (32) 2 (2)

22 (61) 10 (28) 4 (11)

160 (41) 228 (59)

60 (52) 55 (48)

52 (44) 65 (56)

43 (36) 77 (64)

5 (14) 31 (86)

164 (42) 120 (31) 87 (22) 16 (4) 1 (0)

53 (46) 43 (37) 16 (14) 3 (3) 0 (0)

52 (44) 29 (25) 33 (28) 3 (3) 0 (0)

44 (37) 35 (29) 31 (26) 9 (8) 1 (1)

15 (42) 13 (36) 7 (19) 1 (3) 0 (0)

<0.001

0.022

<0.001

0.64

0.42

0.51

0.15

<0.001

0.11

All values are indicated as the number (percentage). *Peak steroid dose was recorded from the period within two weeks prior to onset of LRI. LRI: lower respiratory tract infection; HRV: human rhinovirus; RSV: respiratory syncytial virus; PIV: parainfluenza virus; HCT: hematopoietic cell transplantation; PCR: polymerase chain reaction.

ferent, and in addition, look remarkably similar to curves previously documented for other respiratory viruses, such as PIV, in HCT recipients.15 Second, the fact that mortality in patients with or without co-pathogens is similar could suggest that HRV by itself is a cause of LRI, because otherwise one would expect the outcome to be improved in patients without co-pathogens. Next, factors associated with mortality following HRV BAL positivity were similar to those for other respiratory viruses (RSV, PIV, HMPV, influenza viruses) reported by our group and others,14-16,19-23 and included oxygen use, cytopenias, and high-dose cortihaematologica | 2017; 102(6)

costeroid use (Tables 2 and 4). An important analysis was the comparison of the outcome of HRV LRI with other well-established viral pneumonias. We used 3 previously described cohorts14-16 and analyzed overall mortality in a multivariable Cox model that adjusted for key factors that are associated with poor outcome. There was no difference in mortality after LRI between HRV and the other 3 viruses, even after adjusting for several important factors, such as the presence of co-pathogens and oxygen use at the time of diagnosis (Table 6, Figure 2), providing strong evidence that HRV detection in the BAL is indeed a clini1127


S. Seo et al.

cally significant finding. Although the Cox model adjusted for oxygen use, we performed a subset analysis of patients who presented without oxygen use (Figure 2B). This group likely had minimal or no acute lung injury at the time of diagnosis and the subsequent outcome in these patients was more likely to be associated with the viral insult rather than the inflammatory changes associated with acute lung injury. Although mortality was, as expected, lower in this group, mortality rates approaching 20% by 3 months after diagnosis were documented, indicating

an important negative impact on survival. Importantly, there were no apparent differences in LRI outcomes between HRV and RSV, PIV or influenza virus. We also examined virologic factors to evaluate the role of HRV in lung disease. Tissue documentation of a pathogen is considered a key factor supporting a pathogenic role of an infectious agent.24,25 Although we were limited by the small number of tissue samples from lung biopsies and autopsies in our cohort, we were able to culture HRV from lung biopsies in 6 of 22 patients who had

Table 6. Risk factors for overall mortality comparing each respiratory virus (N=388).

HR Virus type HRV RSV PIV Influenza virus Transplant year 1989-2006 2007-2015 Stem cell source Peripheral blood stem cell Bone marrow Cord blood Donor type Autologous Allogeneic Days between HCT and infection ≤ 30 31-365 > 365 Diagnostic methods Other methods ±PCR PCR alone Co-pathogen None Any viral co-pathogen Any non-viral co-pathogen Lymphocyte count at diagnosis > 300 cells/μL ≤ 300 cells/μL Monocyte count at diagnosis > 300 cells/μL ≤ 300 cells/μL Oxygen use at diagnosis No Any Steroid dose before diagnosis* No < 1 mg/kg 1-2 mg/kg > 2 mg/kg

Univariable analysis 95% CI P

HR

Multivariable analysis 95% CI

P

1.00 1.39 1.59 1.49

0.85-2.28 0.98-2.58 0.80-2.78

0.19 0.06 0.21

1.00 0.79 0.91 0.85

0.43-1.45 0.53-1.56 0.40-1.80

0.45 0.73 0.68

1.00 0.68

0.51-0.92

0.013

1.00 1.04

0.69-1.57

0.84

1.00 1.85 0.79

1.37-2.50 0.42-1.48

<0.001 0.47

1.00 1.59 0.81

1.11-2.27 0.41-1.59

0.011 0.53

1.00 1.66

1.05-2.61

0.029

1.00 1.29

0.76-2.20

0.35

1.00 0.95 0.50

0.70-1.31 0.31-0.82

0.77 0.006

1.00 1.22 0.72

0.80-2.34 0.39-1.31

0.22 0.28

1.00 1.42

1.04-1.94

0.028

1.00 1.23

0.72-2.10

0.45

1.00 1.36 1.48

0.69-2.70 1.10-1.99

0.38 0.010

1.00 1.39 1.48

0.82-2.34 1.04-2.11

0.22 0.029

1.00 1.89

1.40-2.56

<0.001

1.00 1.16

0.80-1.67

0.43

1.11-2.78

0.016

1.00 2.57

1.76-3.75

<0.001

1.00 1.76

1.00 3.08

2.19-4.32

<0.001

1.00 3.37

2.33-4.79

<0.001

1.00 0.94 1.81 2.98

0.65-1.37 1.26-2.58 1.61-5.52

0.76 0.001 <0.001

1.00 1.28 1.35 2.35

0.81-2.02 0.86-2.12 1.17-4.70

0.23 0.20 0.016

All variables in Table 3 were used for the univariable analysis. Only variables with P<0.05 are shown in this table. *Peak steroid dose was recorded from the period within two weeks prior to the onset of LRI. HRV: human rhinovirus; RSV: respiratory syncytial virus; PIV: parainfluenza virus; HCT: hematopoietic cell transplantation; PCR: polymerase chain reaction; CI: confidence interval; HR; hazard ratio.

1128

haematologica | 2017; 102(6)


Human rhinovirus infections in HCT recipients

cultures performed. This is a remarkable result because our culture system is not optimized for HRV recovery (which in general is improved by incubation at lower temperatures and other methods, such as roller flasks, as used in previous papers),7 and suggests a high viral load may have been present. Since viral culture positivity from biopsy material is generally considered to indicate invasive viral disease for other respiratory viruses and cytomegalovirus (CMV),24,26,27 we consider this finding to be highly significant. We also detected HRV RNA by PCR in archived frozen tissue from lung biopsies and autopsy samples. The HRV PCR Ct values of the BAL samples, which provide a rough estimate of the HRV viral load, were not associated with mortality in our cohort (Tables 2 and 4). The lack of association does not rule out a pathogenic role of HRV, since studies with other well-established pathogens such as CMV, RSV, and PIV also failed to demonstrate an association of viral load in the BAL with mortality.12,15,28,29 Methodologic issues may be responsible for these results, including the inability to account for BAL dilution and the potentially differential amplification efficiency of different HRV serotypes.30 Additional studies are needed to conclusively study the role of the viral load in the BAL for HRV and other viruses. The study herein has both strengths and limitations. We examined a large number of cases of HRV RNA positive BALs and comparative cases with other viral lower respiratory diseases which allowed us to perform appropriate statistical analyses that carefully accounted for copathogens and stages of acute lung injury. Our study was not restricted to specific transplant types or time periods after HCT, however, a subgroup analysis of recent allogeneic transplant recipients showed similar results. Due to a lack of specific information about the severity of GvHD in some cases, the association between mortality after HRV LRI and GvHD remains unclear, although we used steroid dose as a surrogate marker for GvHD severity. Our largely unbiased approach of evaluating patients with pulmonary infiltrates by bronchoscopy and BAL and the availability of lung tissue are additional strengths. According to previous reports, approximately 30% of HRV URI cases were asymptomatic4 and 5% of BAL sam-

References 1. Jain S, Self WH, Wunderink RG, Team CES. Community-acquired pneumonia requiring hospitalization among U.S. adults. N Engl J Med. 2015;373(5):415-427. 2. Milano F, Campbell AP, Guthrie KA, et al. Human rhinovirus and coronavirus detection among allogeneic hematopoietic stem cell transplantation recipients. Blood. 2010;115(10):2088-2094. 3. Jacobs SE, Lamson DM, Soave R, et al. Clinical and molecular epidemiology of human rhinovirus infections in patients with hematologic malignancy. J Clin Virol. 2015;71:51-58. 4. Campbell AP, Guthrie KA, Englund JA, et al. Clinical outcomes associated with respiratory virus detection before allogeneic hematopoietic stem cell transplant. Clin Infect Dis. 2015;61(2):192-202. 5. Gutman JA, Peck AJ, Kuypers J, Boeckh M. Rhinovirus as a cause of fatal lower respiratory tract infection in adult stem cell trans-

haematologica | 2017; 102(6)

6.

7.

8.

9.

10.

ples from asymptomatic HCT recipients included HRV LRI;12 we therefore think that asymptomatic lower respiratory tract infection is only a minor factor in the present study. Other limitations include the lack of tissue immunohistochemical studies and the lack of direct viral quantitation, as well as our inability to adequately culture for HRV in autopsy or biopsy specimens. Currently available antibodies for HRV are serotype-specific.31 Because we did not identify the HRV serotypes in our tissue samples, we could not perform immunohistochemistry to confirm tissue detection of HRV. Furthermore, we acknowledge that certain HRV strains (particularly HRV-C viruses) may not have been identified earlier in this study and would be unlikely to be successfully cultivated. Thus, infections with HRV-C might be underrepresented in the patients prior to 2008. In conclusion, we provide further evidence that HRV is a serious pathogen in the lower respiratory tract of HCT recipients. This conclusion is supported by similar outcomes of patients with HRV in the BAL compared to those with well-established respiratory pathogens such as RSV and influenza virus and by the presence of HRV in lung tissue. The data are also consistent with the recent literature on HRV in other clinical settings, which supports a potential causal role in serious clinical disease,1 and agree with earlier studies in HCT recipients carried out prior to the use of molecular diagnostic techniques.6,7 However, while these data are supportive of a pathogenic role of HRV, ultimate proof could best be provided by randomized placebo-controlled trials. These data provide the rationale for the rapid development and clinical evaluation of new therapeutics. Acknowledgments The authors would like to thank Zachary Stednick for database services, Terry Stevens-Ayers for laboratory assistance, and Amanda C Moklebust for preparation of pathology samples. Funding This work was partially supported by grants [K24HL093294, K23AI114844, and CA15704] from the National Institutes of Health. SS is a recipient of a fellowship from the Joel Meyers Endowment Scholarship.

plantation patients: a report of two cases. Bone Marrow Transplant. 2007;40(8):809811. Ison MG, Hayden FG, Kaiser L, Corey L, Boeckh M. Rhinovirus infections in hematopoietic stem cell transplant recipients with pneumonia. Clin Infect Dis. 2003;36(9):1139-1143. Ghosh S, Champlin R, Couch R, et al. Rhinovirus infections in myelosuppressed adult blood and marrow transplant recipients. Clin Infect Dis. 1999;29(3):528-532. Jacobs SE, Soave R, Shore TB, et al. Human rhinovirus infections of the lower respiratory tract in hematopoietic stem cell transplant recipients. Transpl Infect Dis. 2013;15(5):474-486. Mikulska M, Del Bono V, Gandolfo N, et al. Epidemiology of viral respiratory tract infections in an outpatient haematology facility. Ann Hematol. 2014;93(4):669-676. Kraft CS, Jacob JT, Sears MH, et al. Severity of human rhinovirus infection in immunocompromised adults is similar to that of

11.

12.

13.

14.

15.

2009 H1N1 influenza. J Clin Microbiol. 2012;50(3):1061-1063. Gadsby NJ, Russell CD, McHugh MP, et al. Comprehensive molecular testing for respiratory pathogens in community-acquired pneumonia. Clin Infect Dis. 2016; 62(7):817-823. Seo S, Renaud C, Kuypers JM, et al. Idiopathic pneumonia syndrome after hematopoietic cell transplantation: evidence of occult infectious etiologies. Blood. 2015;125(24):3789-3797. Lu X, Holloway B, Dare RK, et al. Real-time reverse transcription-PCR assay for comprehensive detection of human rhinoviruses. J Clin Microbiol. 2008;46(2):533-539. Waghmare A, Campbell AP, Xie H, et al. Respiratory syncytial virus lower respiratory disease in hematopoietic cell transplant recipients: viral RNA detection in blood, antiviral treatment, and clinical outcomes. Clin Infect Dis. 2013;57(12):1731-1741. Seo S, Xie H, Campbell AP, et al. Parainfluenza virus lower respiratory tract

1129


S. Seo et al.

16.

17.

18.

19.

20.

1130

disease after hematopoietic cell transplantation: viral detection in the lung predicts outcome. Clin Infect Dis. 2014;58(10):13571368. Choi SM, Boudreault AA, Xie H, et al. Differences in clinical outcomes after 2009 influenza A/H1N1 and seasonal influenza among hematopoietic cell transplant recipients. Blood. 2011;117(19):5050-5056. Martin EK, Kuypers J, Chu HY, et al. Molecular epidemiology of human rhinovirus infections in the pediatric emergency department. J Clin Virol. 2015;62:2531. Fredricks DN, Relman DA. Infectious agents and the etiology of chronic idiopathic diseases. Curr Clin Top Infect Dis. 1998;18:180-200. Seo S, Campbell AP, Xie H, et al. Outcome of respiratory syncytial virus lower respiratory tract disease in hematopoietic cell transplant recipients receiving aerosolized ribavirin: significance of stem cell source and oxygen requirement. Biol Blood Marrow Transplant. 2013;19(4):589-596. Shah DP, Ghantoji SS, Shah JN, et al. Impact of aerosolized ribavirin on mortality in 280 allogeneic haematopoietic stem cell transplant recipients with respiratory syncytial virus infections. J Antimicrob

Chemother. 2013;68(8):1872-1880. 21. Chemaly RF, Hanmod SS, Rathod DB, et al. The characteristics and outcomes of parainfluenza virus infections in 200 patients with leukemia or recipients of hematopoietic stem cell transplantation. Blood. 2012;119(12):2738-2745. 22. Renaud C, Xie H, Seo S, et al. Mortality rates of human metapneumovirus and respiratory syncytial virus lower respiratory tract infections in hematopoietic cell transplantation recipients. Biol Blood Marrow Transplant. 2013;19(8):1220-1226. 23. Ljungman P, de la Camara R, Perez-Bercoff L, et al. Outcome of pandemic H1N1 infections in hematopoietic stem cell transplant recipients. Haematologica. 2011; 96(8):1231-1235. 24. Ljungman P, Griffiths P, Paya C. Definitions of cytomegalovirus infection and disease in transplant recipients. Clin Infect Dis. 2002;34(8):1094-1097. 25. Harrington RD, Hooton TM, Hackman RC, et al. An outbreak of respiratory syncytial virus in a bone marrow transplant center. J Infect Dis. 1992;165(6):987-993. 26. Boeckh M, Berrey MM, Bowden RA, et al. Phase 1 evaluation of the respiratory syncytial virus-specific monoclonal antibody palivizumab in recipients of hematopoietic

27.

28.

29.

30. 31.

stem cell transplants. J Infect Dis. 2001; 184(3):350-354. Erard V, Huang ML, Ferrenberg J, et al. Quantitative real-time polymerase chain reaction for detection of adenovirus after T cell-replete hematopoietic cell transplantation: viral load as a marker for invasive disease. Clin Infect Dis. 2007;45(8):958-965. Erard V, Guthrie KA, Seo S, et al. Reduced mortality of cytomegalovirus pneumonia after hematopoietic cell transplantation due to antiviral therapy and changes in transplantation practices. Clin Infect Dis. 2015; 61(1):31-39. Campbell AP, Chien JW, Kuypers J, et al. Respiratory virus pneumonia after hematopoietic cell transplantation (HCT): associations between viral load in bronchoalveolar lavage samples, viral RNA detection in serum samples, and clinical outcomes of HCT. J Infect Dis. 2010; 201(9):1404-1413. Jacobs SE, Lamson DM, St George K, Walsh TJ. Human rhinoviruses. Clin Microbiol Rev. 2013;26(1):135-162. Mosser AG, Brockman-Schneider R, Amineva S, et al. Similar frequency of rhinovirus-infectible cells in upper and lower airway epithelium. J Infect Dis. 2002; 185(6):734-743.

haematologica | 2017; 102(6)


Haematologica, volume 102, issue 6  
Read more
Read more
Similar to
Popular now
Just for you