Haematologica, volume 101, issue 11 by Haematologica

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)

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 2016 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 October 2016.

haematologica calendar of events

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

5th ESLHO educational meeting: New developments in the ESLHO networks European Scientific foundation for Laboratory Hemato Oncology (ESLHO) Chairs: J van Dongen, P Groenen, B Schäfer November 3, 2016 Prague, Czech Republic

2 MEGMA Conference on Thalassaemia and Other Haemoglobinopathies Thalassaemia International Federation (TIF) Chairs: A Taher, J Porter, A Piga, A Beshlawy November 11-12, 2016 Amman, Jordan nd

Portuguese Society of Haematology Annual Meeting Portuguese Society of Haematology (SPH) Chairs: J Guimarães, J Andrade, F Príncipe, MJ Silva November 17-19, 2-16 Espinho, Portugal

JIHEMA: XVII Conference ON INFECTIONS IN HEMATOLOGY. Clinical cases PETHEMA Chairs: J Diaz Mediavilla, A Santiago, C Vallejo November 18-19, 2016 Toledo, Spain

Highlights of Past EHA - HOPE Dubai 2016 Chairs: R Foà, M Qari November 24-26, 2016 Dubai, UAE

JACIE Inspector Training Course – Stockholm JACIE Chairs: E McGrath, P Llungman November 24-25, 2016 Märsta, Sweden

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

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

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

EHA Scientific Meeting on Aging and Hematology Chair: D Bron May 4-6, 2017 Location: TBC

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

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

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

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

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

Calendar of Events updated on October 3, 2016

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

Table of Contents Volume 101, Issue 11: November 2016 Cover Figure Graphical representation of the proteins in the plasmamembrane of a red blood cell - image accompanying the review article on page 1284. (Image created by www.somersault1824.com)

Editorials 1275

Genetic panels in young patients with bone marrow failure: are they clinically relevant? Amy E. DeZern and Robert A. Brodsky

1276

Autotransplants in older multiple myeloma patients: hype or hope in the era of novel agents? Monika Engelhardt, et al.

1279

FDG-PET as a biomarker for early response in diffuse large B-cell lymphoma as well as in Hodgkin lymphoma? Ready for implementation in clinical practice? JosĂŠe M. Zijlstra, et al.

Review Articles 1284

New insights on hereditary erythrocyte membrane defects Immacolata Andolfo, et al.

1295

Osteonecrosis in children with acute lymphoblastic leukemia Marina Kunstreich, et al.

Articles Red Cell Biology & its Disorders

1306

Gene panel sequencing improves the diagnostic work-up of patients with idiopathic erythrocytosis and identifies new mutations Carme Camps, et al.

Blood Transfusion

1319

Human neutrophil peptides and complement factor Bb in pathogenesis of acquired thrombotic thrombocytopenic purpura Wenjing Cao, et al.

Platelet Biology & its Disorders

1327

Comparison of two dosing schedules for subcutaneous injections of low-dose anti-CD20 veltuzumab in relapsed immune thrombocytopenia Howard A. Liebman, et al.

1333

Clinical and pathogenic features of ETV6-related thrombocytopenia with predisposition to acute lymphoblastic leukemia Federica Melazzini, et al.

Bone Marrow Failure

1343

Genetic features of myelodysplastic syndrome and aplastic anemia in pediatric and young adult patients SiobĂĄn B. Keel, et al.

Haematologica 2016; vol. 101 no. 11 - November 2016 http://www.haematologica.org/

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

1351

An operational definition of primary refractory acute myeloid leukemia allowing early identification of patients who may benefit from allogeneic stem cell transplantation Paul Ferguson, et al.

1359

Effect of age and body weight on toxicity and survival in pediatric acute myeloid leukemia: results from NOPHO-AML 2004 Ditte J. A. Løhmann, et al.

Acute Lymphoblastic Leukemia

1368

Optimal interleukin-7 receptor-mediated signaling, cell cycle progression and viability of T-cell acute lymphoblastic leukemia cells rely on casein kinase 2 activity Alice Melão, et al.

Non-Hodgkin Lymphoma

1380

Alterations of microRNA and microRNA-regulated messenger RNA expression in germinal center B-cell lymphomas determined by integrative sequencing analysis Kebria Hezaveh, et al.

Plasma Cell Disorders

1390

Upfront autologous stem cell transplantation for newly diagnosed elderly multiple myeloma patients: a prospective multicenter study Laurent Garderet, et al.

Cell Therapy and Immunotherapy

1398

Autotransplant with and without induction chemotherapy in older multiple myeloma patients: long-term outcome of a randomized trial Christian Straka, et al.

1407

Splenic pooling and loss of VCAM-1 causes an engraftment defect in patients with myelofibrosis after allogeneic hematopoietic stem cell transplantation Christina Hart, et al.

1417

A prospective randomized trial comparing cyclosporine/methotrexate and tacrolimus/sirolimus as graft-versus-host disease prophylaxis after allogeneic hematopoietic stem cell transplantation Johan Törlén, et al.

1426

The prognostic value of serum C-reactive protein, ferritin, and albumin prior to allogeneic transplantation for acute myeloid leukemia and myelodysplastic syndromes Andrew S. Artz, et al.

Immunology & Infiammation

1434

A score of low-grade inflammation and risk of mortality: prospective findings from the Moli-sani study Marialaura Bonaccio, et al.

Letters to the Editor Letters are available online only at www.haematologica.org/content/101/11.toc

e431

Senicapoc: a potent candidate for the treatment of a subset of hereditary xerocytosis caused by mutations in the Gardos channel Raphael Rapetti-Mauss, et al. http://www.haematologica.org/content/101/11/e431

e436

Mutating heme oxygenase-1 into a peroxidase causes a defect in bilirubin synthesis associated with microcytic anemia and severe hyperinflammation Johann Greil, et al. http://www.haematologica.org/content/101/11/e436

Haematologica 2016; vol. 101 no. 11 - November 2016 http://www.haematologica.org/

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

e440

Interleukin-6 receptor-alpha signaling drives anti-RBC alloantibody production and T-follicular helper cell differentiation in a murine model of red blood cell alloimmunization Abhinav Arneja, et al. http://www.haematologica.org/content/101/11/e440

e445

BNIP3L in myelodysplastic syndromes and acute myeloid leukemia: impact on disease outcome and cellular response to decitabine Mariana Lazarini, et al. http://www.haematologica.org/content/101/11/e445

e449

Erlotinib synergizes with the poly(ADP-ribose) glycohydrolase inhibitor ethacridine in acute myeloid leukemia cells Lianne E. Rotin, et al. http://www.haematologica.org/content/101/11/e449

e454

MicroRNA expression-based outcome prediction in acute myeloid leukemia: novel insights through cross-platform integrative analyses Velizar Shivarov, et al. http://www.haematologica.org/content/101/11/e454

e457

Mutational correlates of response to hypomethylating agent therapy in acute myeloid leukemia Catherine C. Coombs, et al. http://www.haematologica.org/content/101/11/e457

e461

Innate lymphoid cells are expanded and functionally altered in chronic lymphocytic leukemia Iris de Weerdt, et al. http://www.haematologica.org/content/101/11/e461

e465

Cytomegalovirus viremia, disease, and impact on relapse in T-cell replete peripheral blood haploidentical hematopoietic cell transplantation with post-transplant cyclophosphamide Scott R. Goldsmith, et al. http://www.haematologica.org/content/101/11/e465

Haematologica 2016; vol. 101 no. 11 - November 2016 http://www.haematologica.org/

EDITORIALS Genetic panels in young patients with bone marrow failure: are they clinically relevant? Amy E. DeZern1 and Robert A. Brodsky2 1

Division of Hematologic Malignancies; The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins; and 2Division of Hematology, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, MD, USA E-mail: adezern1@jhmi.edu doi:10.3324/haematol.2016.152389

one marrow failure can be acquired or inherited. Acquired forms are usually immune-mediated; inherited forms may be due to DNA repair defects (Fanconi anemia), ribosomopathies (Shwachman-Diamond and Diamond-Blackfan anemia), telomere defects (dyskeratosis congenita), and a smattering of other germline mutations that lead to cytopenias and predispose to myelodysplastic syndromes (MDS). Acquired severe aplastic anemia (SAA) is treated with allogeneic hematopoietic stem cell transplantation (HSCT) or immunosuppressive therapy (IST) in patients without a suitable donor. Inherited bone marrow failure syndromes are treated with HSCT if they have severe marrow failure, as IST is ineffective. Importantly, the HSCT conditioning regimen is not the same for all of these disorders. Patients with Fanconi anemia must receive a less intensive HSCT conditioning regimen. Inherited forms of bone marrow failure almost always present in the first decade of life and often, but not always, have either physical stigmata or a family history that is suggestive of the diagnosis. However, in many cases the overlap in clinical presentation and bone marrow features of inherited and acquired bone marrow failure syndromes and pediatric MDS can present a diagnostic challenge. A reliable method to distinguish between inherited and acquired bone marrow failure syndromes would accelerate the diagnosis and could influence the choice of therapy and/or the choice of donor for HSCT. In this issue of Haematologica, Keel et al.1 applied a multiplex targeted capture assay to investigate 63 genes in 208 children and young adults with AA (aplastic anemia) or MDS. These genes had all previously been described in patients with inherited bone marrow failure syndromes. All patients underwent HSCT between 1990-2012. The reported mutation frequencies in each cohort were commensurate with previous reports with similar sequencing methodology.2-6 A thoughtful and extensive characterization of the cohort retrospectively reviewed the patient’s chart history and physical exam characteristics as well as clinical outcomes.1 The AA cohort had 53 pediatric (age ≤18 years) and 45 young adult (age >18 years but ≤40) patients. All of them met the criteria for severe disease at the time of HSCT. Physical anomalies were present in 11% of the AA patients and 40% had a family history suggestive of inherited conditions. The authors found known mutations in 5 of 98 AA patients. The mutations identified for the 5 patients in the AA group included DKC1 (n=2), MPL (n=2), and TP53 (n=1). These all represent known constitutional mutations which have previously been described to result in the clinical syndromes of dyskeratosis congenita,7 congenital amegakaryocytic thrombocytopenia,8 and Li Fraumeni,9 respectively. The MDS cohort had 46 pediatric (age ≤18 years) and 64 young adult (age >18 years but ≤46) patients. Physical anomhaematologica | 2016; 101(11)

alies were present in 24% of patients, with family histories relevant in 52% of the patients. The authors found mutations in 15 of 110 MDS patients. The mutations identified were constitutional in 14 out of these 15 patients. They included compound heterozygous mutations in FANCA, MPL, RTEL1, and SBDS with heterozygous mutations identified in GATA2, RUNX1, TERT, TINF2, and TP53. These have been reported to cause Fanconi anemia,10 congenital amegakaryocytic thrombocytopenia,8 dyskeratosis congenita,7 and Shwachman-Diamond syndrome.11 Additionally mutations in GATA2 and RUNX1 are known to have an inherited predisposition to leukemia and MDS.12,13 The fifteenth patient with mutated bone marrow DNA carried a heterozygous mutation in RUNX1. The authors note a comparison with bowel and skin DNA (wild-type RUNX1) in the patient that ultimately identified this as a somatic mutation.14 The strengths of this work by Keel and colleagues relate to the large number of samples in their biorepository and the meticulous work in performing the genetic analysis. In terms of clonal hematopoiesis, the authors used a panel which focused on genes that are previously described in inherited as well as acquired AA and MDS. Newer techniques and sequencing data could possibly have identified even more genes associated with these diseases of marrow failure. The relevance of a panel such as this to the marrow failure population at large is that it can identify markers (even somatic) of clonal hematopoiesis. In AA, in particular, this clonal hematopoiesis has been closely linked to the evolution of late clonal disorders, including MDS, leukemia, and PNH, even after successful treatment with IST. The detection and close monitoring of somatic mutations may help with predictions of outcomes and earlier diagnosis of clonal evolution, all leading to better management of patients with AA. The authors of the study under discussion attempted to retrospectively ascertain a “pre-test” probability of inherited syndromes through a review of the chart history and physical anomalies, but the amount of incomplete clinical data may limit the wider applicability. Also, since the biorepository used patients who received an allogenic HSCT, it likely excluded patients with moderate aplastic anemia. The authors concluded that the history and physical examination did not identify a subset of patients with underlying mutations. However, in this small cohort, conclusions were not made based on age; four of the 5 AA patients with germline mutations were less than 10 years of age, and the 33-year-old patient had a known family history consistent with dyskeratosis congenita. Thus, given the added cost of these gene panels and the low yield in older patients with no family history or physical stigmata, it may be more cost effective to restrict their use to patients under 18 years of age. Further stratification may be possible after screening for a PNH clone. It has been previously described that the presence of a PNH 1275

Editorials

clone essentially rules out inherited conditions, as this is a marker of acquired disease.15 Accordingly, in patients found to have a PNH clone by flow cytometry, none had a germline mutation. Thus, it remains unclear as to whether gene panels will be useful, especially in patients beyond their second decade of life, and particularly if they have a PNH clone. Another potential benefit of targeted gene panels may relate to HSCT donor selection. As the use of alternative donor sources (matched unrelated donors, haploidentical donors, and cord blood) increases, we need assurances that we are not transplanting defective stem cells. One could argue to always use unrelated donors. However, there is ample evidence that time to treatment matters in severe pancytopenia. Thus, the use of a related donor without increased susceptibility to marrow failure would decrease the time to HSCT without having to search for an unrelated donor. In conclusion, the study by Keel and colleagues is an important first step in helping to define the incidence and clinical importance of germline mutations in young patients with severe bone marrow failure. Future prospective studies and improved technology are needed before a more widespread application of targeted gene panels and/or genome sequencing can be recommended in routine clinical practice.

References 1. Keel SB, Scott A, Sanchez-Bonilla M, et al. Genetic features of myelodysplastic syndrome and aplastic anemia in pediatric and young adult patients. Haematologica. 2016 Jul 14. [Epub ahead of print] 2. Zhang J, Walsh MF, Wu G, et al. Germline mutations in predisposi-

tion genes in pediatric cancer. N Engl J Med. 2015;373(24):2336-2346. 3. Ghemlas I, Li H, Zlateska B, et al. Improving diagnostic precision, care and syndrome definitions using comprehensive next-generation sequencing for the inherited bone marrow failure syndromes. J Med Genet. 2015;52(9):575-584. 4. Tsangaris E, Klaassen R, Fernandez CV, et al. Genetic analysis of inherited bone marrow failure syndromes from one prospective, comprehensive and population-based cohort and identification of novel mutations. J Med Genet. 2011;48(9):618-628. 5. Teo JT, Klaassen R, Fernandez CV, et al. Clinical and genetic analysis of unclassifiable inherited bone marrow failure syndromes. Pediatrics. 2008;122(1):e139-148. 6. Zhang MY, Keel SB, Walsh T, et al. Genomic analysis of bone marrow failure and myelodysplastic syndromes reveals phenotypic and diagnostic complexity. Haematologica. 2015;100(1):42-48. 7. Vulliamy TJ, Marrone A, Knight SW, Walne A, Mason PJ, Dokal I. Mutations in dyskeratosis congenita: their impact on telomere length and the diversity of clinical presentation. Blood. 2006;107(7):26802685. 8. Savoia A, Dufour C, Locatelli F, et al. Congenital amegakaryocytic thrombocytopenia: clinical and biological consequences of five novel mutations. Haematologica. 2007;92(9):1186-1193. 9. Felix CA, Hosler MR, Provisor D, et al. The p53 gene in pediatric therapy-related leukemia and myelodysplasia. Blood. 1996;87(10):4376-4381. 10. D'Andrea AD, Grompe M. The Fanconi anaemia/BRCA pathway. Nat Rev Cancer. 2003;3(1):23-34. 11. Goobie S, Popovic M, Morrison J, et al. Shwachman-Diamond syndrome with exocrine pancreatic dysfunction and bone marrow failure maps to the centromeric region of chromosome 7. Am J Hum Genet. 2001;68(4):1048-1054. 12. Hsu AP, Johnson KD, Falcone EL, et al. GATA2 haploinsufficiency caused by mutations in a conserved intronic element leads to MonoMAC syndrome. Blood. 2013;121(19):3830-3837, S1-7. 13. Song WJ, Sullivan MG, Legare RD, et al. Haploinsufficiency of CBFA2 causes familial thrombocytopenia with propensity to develop acute myelogenous leukaemia. Nat Genet. 1999;23(2):166-175. 14. 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. 15. DeZern AE, Symons HJ, Resar LS, Borowitz MJ, Armanios MY, Brodsky RA. Detection of paroxysmal nocturnal hemoglobinuria clones to exclude inherited bone marrow failure syndromes. Eur J Haematol. 2014;92(6):467-470.

Autotransplants in older multiple myeloma patients: hype or hope in the era of novel agents? Monika Engelhardt,1 Gabriele Ihorst,2 Jo Caers,3 Andreas Günther,4 and Ralph Wäsch1 1

Department of Medicine I, Hematology, Oncology & Stem Cell Transplantation, Medical Center - University of Freiburg, Faculty of Medicine, Germany; 2Clinical Trials Unit, Medical Center - University of Freiburg, Faculty of Medicine, Germany; 3Department of Hematology, University Hospital of Liège, Belgium; and 4Division of Stem Cell Transplantation and Immunotherapy, Medical Department 2, University of Kiel, Germany E-mail: monika.engelhardt@uniklinik-freiburg.de doi:10.3324/haematol.2016.154807

ultiple myeloma (MM) is a malignant disease characterized by the proliferation of clonal plasma cells (PCs) in the bone marrow (BM), and typically accompanied by the secretion of monoclonal immunoglobulins that are detectable in the serum and/or urine. Increased understanding of the genetic alterations, the interactions between malignant PCs and the BM niche and their role in disease progression and the acquisition of therapy resistance, has helped in the development of novel agents, used in combination with cytostatic therapy, including autologous stem cell transplantation (ASCT). The most common indication for ASCT in Europe and the

1276

United States is MM, nevertheless elderly patients are often excluded from ASCTs, due to the patients’ and/or physicians’ choices, subjectivity towards its effectiveness in older cohorts, large prospective studies mostly lacking in elderly cohorts, the effectiveness and broad availability of novel agents and the fear of transplant-related toxicity.1,2 The median age of MM patients at diagnosis is approximately 70 years, with 60% aged 65 or older and ~30% being older than 75 years. The transplant age cutoff has been proposed to be <70 years. In clinical trials for ASCT, the age cutoff is even lower, and commonly 65 years, even if the feasibility of ASCT is established as being up to the haematologica | 2016; 101(11)

Editorials

age of 70-75 years in fit patients.3–5 This age cutoff is unfortunate, since many elderly patients are excluded from ASCT, albeit this population is largely increasing: the percentage of Europeans aged >65 years is projected to amplify from 85 million in 2008 to 151 million in 2060, urging us to designate therapy protocols for elderly cohorts.5 ASCT in younger patients, ≤65 years of age, has shown superiority compared to novel agent-based standard treatment: in 2014, the GIMEMA study group reported improved time to next treatment, progression-free survival (PFS) and overall survival (OS) of tandem melphalan 200mg/m2 (MEL200) with ASCT vs. 6 cycles of melphalanprednisone (MP) with lenalidomide (MPR). PFS was improved by 20 months (median PFS 43 vs. 22.4 months, HR 0.44, 95% CI 0.32-0.61, P<0.001) and the 4-year OS rate was 82% vs. 65% (HR 0.55, 95% CI 0.32-0.93, P=0.02), respectively.6 A similar randomized, multicenter, phase 3 trial with ASCT vs. cyclophosphamide-dexamethasone-lenalidomide chemotherapy alone confirmed the benefit of ASCT.7 Both studies tested immunomodulatory drugs (IMiDs), rather than bortezomib-based induction. As a result, preliminary results of the IFM/DFCI 2009 trial were of particular interest as they verified higher complete responses (58% vs. 46%, P<0.01), lower minimal residual disease persistence and a higher 3-year PFS rate (61% vs. 48%, HR 1.5, 95% CI 12-1.9, P<0.0002) with bortezomib, lenalidomide, dexamethasone (VRD) plus ASCT vs. VRD alone.8 Since novel agent treatment is available today, and has impressively demonstrated its superiority to MP alone, leading to FDA and EMA approval of MP-thalidomide (MPT), bortezomib-MP (VMP), MPR and lenalidomidedexamethasone (Rd) in non-transplant eligible MM patients, the question of standard vs. novel agent treatment has been answered in favor of the latter in elderly patients.1,9 Whether ASCT adds to induction in 60-70 year old patients has been marginally addressed, despite the fact that various groups have verified that ASCT is feasible and that due to novel agents (and possibly also transplants), the prognosis has improved: the Mayo Clinic grouped 1038 patients into two 5-year periods by diagnosis; the median OS for patients in the 2001-2005 cohort vs. the 2006-2010 cohort was 4.6 vs. 6.1 years, respectively (P=0.002). The improvement was primarily seen among patients >65 years, where the 6-year OS strikingly improved from 31% to 56% (P<0.001). Only 10% of patients died during the first year compared with 16% in the earlier cohort (P<0.01). This improved outcome was closely linked to the use of novel agents.10 To thoroughly test novel agent combinations compared to standard treatment and ASCT in elderly patients, the IFM 99-06 study randomized standard MP vs. MPT vs. ASCT, whereby MEL100 conditioning was applied. This trial demonstrated that both MPT and ASCT with ‘lowdose conditioning’ were superior to MP alone. However, this study was hampered by the fact that the chosen dose of MEL100 made the protocol more applicable, but also reduced its efficacy.11 Randomized trials using higher MEL140 or MEL200 conditioning have rarely been performed in elderly patients, although a pragmatic age limit of 70 has been suggested, above which a full dose of MEL200 may generally be inappropriate.3 haematologica | 2016; 101(11)

The randomized multicenter study by the German Multiple Myeloma Study Group (DSMM II) in 434 patients aged 60-70 years is therefore a long awaited trial endeavor, that, before the era of novel agents, tested non-induction with short-term dexamethasone alone vs. 4 cycles of conventional anthracycline dexamethasone induction (mostly VAD) with tandem MEL140 conditioning.12 The treatment duration was short with a median of 7.7 months with induction and 4.6 months without it. The median PFS on the intention-to-treat basis with induction vs. without was 21.4 months vs. 20 months (HR 1.04; 95% CI 0.84-1.28; P=0.36), respectively. Importantly, for patients ≥65 years of age, the outcome was not inferior to those <65 years of age. As expected, patients with low-risk cytogenetics (defined as the absence of del17p13, t(4;14) and 1q21 gains) showed a favorable OS compared to those with high-risk cytogenetics. Of note, MEL140 was associated with a tolerable safety profile and treatment-related deaths were low (1%). Remarkable features of the study were that it represented the largest prospective multicenter tandem ASCT trial in elderly patients, that MEL140 could promptly be repeated 2 months after the first ASCT, that tandem ASCT was well tolerated, with deaths occurring early with induction10 rather than with ASCT itself (6% vs. 1%, respectively), and that even without novel, and at that time unavailable induction and maintenance treatment, long-term survival was achieved. An interesting subgroup of 27 patients (6.4%) were survivors in first remission at 5 years; these were characterized by the presence of low-risk cytogenetics (100%), double transplant (85%) and ISS stage I/II (70%).12 Despite the fact that cross-comparison of other trials and representative historical data sets is problematic, median OS in the IFM 99-06 study with MP, MPT and MEL100 was 33.2, 51.6 and 38.3 months, respectively,11 and in the Medical Research Council (MRC) study with MP, cyclophosphamide-thalidomide-dexamethasone (CTD) vs. ASCT in patients >64 years 30.6, 33.2 and 53 months, respectively,13 thus the median OS in the DSMM II trial (median follow-up: 5.2 years) of 53.4 months with induction and 55.9 months without induction is encouraging, the more so since no modern induction or maintenance treatment were available and therefore not used in this trial.12 The paper by Straka et al.12 encounters today's challenge, however, due to its enrollment from 2001 to 2006, the long follow-up until its publication and the unprecedented MM success, that non-induction or VAD induction is no longer employed, rather, highly effective induction, consolidation and maintenance approaches are employed as pre- and post-transplant strategies.1,6,7,14,15 Thus, the MEL140 tandem ASCT back-bone of the DSMM II trial seems the most relevant today. This well-tolerated treatment element has indeed been transferred to the followup DSMM study testing Rd with or without tandem MEL140, followed by lenalidomide maintenance in newly diagnosed 60-75 year old symptomatic MM patients.16 Additional questions that the DSMM II trial could not answer were: which patients with what assessment tools are best assigned to ASCT, whether MEL140 vs. 200 should be used, and which induction and maintenance strategy is best in elderly patients? A much smaller French multicenter trial in ≥65 year old patients used bortezomib1277

Editorials

based induction, MEL140 in 18 (36%) and MEL200 in 32 (64%) patients and consolidation with either Rd or bortezomib-based treatment, confirming the safety and efficacy of ASCT as first-line treatment in elderly MM patients.17 Although this study was not sufficiently powered to pick up differences between the two melphalan schedules, and the median follow-up, at 21 months, was shorter, the estimated PFS and OS rates at 2 years were encouraging with 76% and 88%, respectively, suggesting that MEL200 may induce superior PFS and OS rates in elderly patients. Nevertheless, since this was a non-randomized study and patients were selected (e.g., those with MEL200 were fitter and not comparable to all newly diagnosed elderly MM patients), objective, prospective and proficiently performed fitness tools might be of benefit before intensive treatment is induced, the more so, since patients and physicians fitness ratings are not as objective as defined tests and scores.1,18,19 However, geriatric tests have been criticized as being time-consuming, and few data as yet show a correlation between geriatric assessment and clinical outcome.5 We and others have, however, shown that one can get a straightforward score and homepage help to swiftly assess MM patients within 1-2 minutes,18–20 which can be used before intensive therapeutic interventions, such as ASCT. This seems important, since the population of elderly patients is heterogeneous and older patients are likely to have frailties complicating their management. For the elderly with MM, novel drugs, ASCT and advances in supportive care have increased response rates and OS in the past several years.1,5,9,10 Present clinical research focuses on the balance between treatment efficacy and quality of life, the optimum sequencing of treatment, and how to induce long-term remission. Given the results of the DSMM II trial, ASCTs should be considered in elderly patients, if these are appropriately assessed and deemed fit for the procedure. Modern, well tolerated induction and maintenance approaches, e.g., with IMiDs and/or proteasome inhibitors, have been shown to improve PFS and OS in MM,6,7,14,15,21 and therefore are currently used. Moreover, immunotherapy to stimulate antitumor immunity after ASCT is of particular interest, since T cell exhaustion has been identified as a distinguishing feature of relapse after ASCT.5 The pipeline of promising new treatments raise hopes for continuous improvements, the Straka12 and Garderet17 trials demonstrating another essential treatment element, how this can be achieved in elderly patients. Funding This work is supported by the Deutsche Krebshilfe [grants 1095969 and 111424 to ME and RW]

References 1. Engelhardt M, Terpos E, Kleber M, et al. European Myeloma Network recommendations on the evaluation and treatment of newly diagnosed patients with multiple myeloma. Haematologica. 2014;99(2):232-242.

1278

2. Caers J, Fernández de Larrea C, Leleu X, et al. The Changing Landscape of Smoldering Multiple Myeloma: A European Perspective. The Oncologist. 2016;21(3):333-342. 3. Morgan GJ. Transplants for the elderly in myeloma. Blood. 2013;122(8):1332-1334. 4. Auner HW, Szydlo R, Hoek J, et al. Trends in autologous hematopoietic cell transplantation for multiple myeloma in Europe: increased use and improved outcomes in elderly patients in recent years. Bone Marrow Transplant. 2015;50(2):209-215. 5. Moreau P, Attal M, Facon T. Frontline therapy of multiple myeloma. Blood. 2015;125(20):3076-3084. 6. Palumbo A, Cavallo F, Gay F, et al. Autologous transplantation and maintenance therapy in multiple myeloma. N Engl J Med. 371:895905, 2014 7. Gay F, Oliva S, Petrucci MT, et al. Chemotherapy plus lenalidomide versus autologous transplantation, followed by lenalidomide plus prednisone versus lenalidomide maintenance, in patients with multiple myeloma: a randomised, multicentre, phase 3 trial. Lancet Oncol. 2015;16(16):1617-1629. 8. Attal M, Lauwers-Cances V, Hulin C, et al. Autologous Transplantation for Multiple Myeloma in the Era of New Drugs: A Phase III Study of the Intergroupe Francophone Du Myelome (IFM/DFCI 2009 Trial). Blood. 2015;126(23):391. 9. Ludwig H, Sonneveld P, Davies F, et al. European perspective on multiple myeloma treatment strategies in 2014. The Oncologist. 2014;19(8):829–844. 10. 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. 11. Facon T, Mary JY, Hulin C, et al. Melphalan and prednisone plus thalidomide versus melphalan and prednisone alone or reducedintensity autologous stem cell transplantation in elderly patients with multiple myeloma (IFM 99-06): a randomised trial. Lancet. 2007;370(9594):1209-1218. 12. Straka C, Liebisch P, Salwender H, et al. Autotransplant with and without induction chemotherapy in older multiple myeloma patients: Long-term outcome of a randomized trial. Haematologica. 2016;101(11):1398-1406. 13. Morgan GJ, Davies FE, Gregory WM, et al. Cyclophosphamide, thalidomide, and dexamethasone (CTD) as initial therapy for patients with multiple myeloma unsuitable for autologous transplantation. Blood. 2011;118(5):1231-1238. 14. Mellqvist U-H, Gimsing P, Hjertner O, et al. Bortezomib consolidation after autologous stem cell transplantation in multiple myeloma: a Nordic Myeloma Study Group randomized phase 3 trial. Blood. 2013;121(23):4647-4654. 15. Cavo M, Pantani L, Petrucci MT, et al. Bortezomib-thalidomide-dexamethasone is superior to thalidomide-dexamethasone as consolidation therapy after autologous hematopoietic stem cell transplantation in patients with newly diagnosed multiple myeloma. Blood. 2012;120(1):9-19. 16. Straka C, Schaefer-Eckart K, Bassermann F, et al. Lenalidomide with Low-Dose Dexamethasone (Rd) Continuously Versus Rd Induction, Tandem MEL140 with Autologous Transplantation and Lenalidomide Maintenance: Planned Interim Analysis of a Prospective Randomized Trial in Patients 60-75 Years of Age with Multiple Myeloma. Blood. 2014;124(21):3969-3969. 17. Garderet L, Caillot D, Stoppa AM, et al. Upfront autologous stem cell transplantation for newly diagnosed elderly multiple myeloma patients: a prospective multicenter study. Haematologica. 2016;101 (11):1390-1397. 18. Kleber M, Ihorst G, Gross B, et al. Validation of the Freiburg Comorbidity Index in 466 multiple myeloma patients and combination with the international staging system are highly predictive for outcome. Clin Lymphoma Myeloma Leuk. 2013;13(5):541-551. 19. Engelhardt M, Dold SM, Ihorst G, et al. Geriatric assessment in multiple myeloma patients: validation of the International Myeloma Working Group (IMWG) score and comparison with other common comorbitity scores. Haematologica. 2016;101(9):1110-1119. 20. Engelhardt M, Dold SM, Ihorst G, et al: R-MCI webpage [Internet], 2015.Available from: http://www.myelomacomorbidityindex.org 21. Gay F, Magarotto V, Crippa C, et al. Bortezomib induction, reducedintensity transplantation, and lenalidomide consolidation-maintenance for myeloma: updated results. Blood. 2013;122(8):1376-1383.

haematologica | 2016; 101(11)

Editorials

FDG-PET as a biomarker for early response in diffuse large B-cell lymphoma as well as in Hodgkin lymphoma? Ready for implementation in clinical practice? Josée M. Zijlstra,1 Coreline N. Burggraaff,1 Marie José Kersten,2 and Sally F. Barrington3 1

Department of Hematology, VU University medical center, Amsterdam, the Netherlands; 2Department of Hematology, Amsterdam Medical Center, the Netherlands; and 3PET Imaging Centre at St Thomas' Hospital, Division of Imaging Sciences and Biomedical Engineering, King’s College London, UK; on behalf of the EHA Scientific Working Group on Lymphoma E-mail: j.zijlstra@vumc.nl doi:10.3324/haematol.2016.142752

A short history Major changes have taken place in the staging and response assessment of malignant lymphoma in the last two decades. With the introduction of fluorodeoxyglucose-positron emission tomography (FDG-PET) and positron emission tomography-computed tomography (PET-CT), the criteria for staging and monitoring response have changed dramatically. In the revised Cheson criteria published in 2007,1 staging with FDGPET was still optional, and end-of treatment assessment using FDG-PET and CT was obligatory for Hodgkin lymphoma (HL) and diffuse large B-cell lymphoma (DLBCL). In the Lugano criteria published in 2014,2 PET-CT is recommended for staging as well as response assessment following therapy, as it is the most accurate imaging modality. However, one of the characteristics of (molecular) metabolic imaging is to be able to assess metabolic changes early in treatment. The question arises whether ‘interim’ FDG-PET-CT (iPET) can be used as a biomarker to differentiate good and poor responders during treatment, in order to modify therapy and to improve outcome. Recent clinical trials have addressed these questions, and we discuss the results and the implications for clinical practice.

Assessment of interim-PET scans International guidelines recommend the use of a 5point scale [also called the Deauville score (DS)] for grading FDG-uptake in lymphoma, compared to physiological uptake in the mediastinum and liver, for response assessment in daily practice and clinical trials.2-4 No FDG uptake is graded as DS 1; uptake less than or equal in intensity to the mediastinum as DS 2; lesions with FDG uptake between mediastinum and liver are assessed as DS 3; uptake more intense than liver is scored as DS 4; and markedly increased uptake or new lymphoma-related lesions as DS 5 (Figure 1). This categorization has a high interobserver agreement in HL and DLBCL.5,6 However, FDG-PET is also a quantitative imaging technique, allowing semi-quantitative imaging interpretation, using standardized uptake values (SUV). Reporting change of FDG uptake (usually expressed as a relative change) can also be used for interim response assessment. The reliability of the results depends on having comparable procedures for patient preparation and injection, and scanning and image reconstruction protocols, as well as comparable data analysis. Quality control and quality assurance procedures are also required to maintain the accuracy and precision of quantification. haematologica | 2016; 101(11)

Recently, the European Association of Nuclear Medicine (EANM) guidelines for FDG-PET in tumor imaging for trials and clinical practice have been up-dated,7 and an accreditation system is available (EARL; http://earl.eanm.org). Within clinical studies, these changes in SUV are being compared with visual assessment. Besides SUV, metabolically active tumor volume defined with FDG-PET is being investigated.

Interim-PET in Hodgkin lymphoma Hodgkin lymphoma is a lymphoma entity with cure rates of up to 90%. iPET predicts response early during treatment and PET-guided therapy is a new strategy in development for HL. The goal of current and recently completed clinical trials is to achieve optimal efficacy in terms of progression-free survival (PFS) and overall survival (OS), and to reduce long-term adverse effects. The first reports using iPET to de-escalate therapy in responding individuals with early-stage disease have been published. The UK RAPID study8 and the EORTC H10 study9 have randomized patients with complete metabolic response (CMR) on iPET after 2-4 cycles of doxorubicin, bleomycin, vinblastine, dacarbazine (ABVD) treatment to receive radiotherapy (RT) or no further treatment (NFT). Both were non-inferiority studies, with a slightly different design. Involved field was used in RAPID and involvednode RT in H10. RAPID investigators accepted that by abandoning RT some loss of disease control was inevitable, whereas H10 investigators designed their trial to demonstrate that patients could be spared RT without any compromise in disease control. Both studies demonstrated a modest PFS advantage for patients receiving RT (Table 1). In the RAPID trial, the 3-year PFS was 97.1% using RT versus 90.8% for NFT in a per-protocol analysis (HR 2.36; 1.13, 4.95). There was no significant difference in 3-year OS: 97.1% (RT) versus 99.0% (NFT). In the H10 study, 1year PFS was 100% (favorable disease) and 97.3% (unfavorable disease) using RT versus 94.9% (favorable) and 94.7% (unfavorable) for NFT. The H10 study was halted early for patients with CMR as it was felt unlikely to demonstrate non-inferiority for the NFT option with a 10% decrease in 5-year PFS where the threshold for non-inferiority was set at a hazard ratio of respectively 3.2 and 2.1 for the favorable and unfavorable subgroups. Nonetheless, patients had excellent outcomes in both trials whether or not they received RT. However, follow up in both trials is still short, and (late) adverse effects of radiotherapy may become apparent over time.10 Results from the HD16 and HD17 tri1279

Editorials Table 1. Studies with i-PET adapted therapy in Hodgkin lymphoma and diffuse large B-cell lymphoma.

Author/study Year

Design

Type +stage

Number i-PET after

Pos criteria

i-PET negative therapy

i-PET positive therapy

Median FUP

Outcome i-PET -/+

IF RT or NFT

1x ABVD + RT

60 mo

3-yr PFS: IF RT: i-t-t: 94.6% p-p a: 97.1% vs. NFT: 90.8%. 3-yr OS: IF RT: 97.1% vs. NFT: 99.0%

Favorable: 2x ABVD or 1x ABVD+INRT Unfavorable: 4x ABVD or 2xABVD+INRT

Favorable: 2x BEACOPPesc+ INRT Unfavorable: 2x BEACOPPesc + INRT

1.1 yr

1-yr PFS fav. IN-RT: 100% NFT 94.9%. 1-yr PFS unfav. IN-RT: 97.3% NFT 94.7%

HL Radford/ RAPID8

2015

RCT

st IA/IIA nonbulky HL

571

3x ABVD

DS 3/4/5

Raemaekers/ EORTC H109

2014

RCT

st I/II supradiaphragmatic HL

1137

2x ABVD

IHP

Press/US Intergroup S081611

2016

phase II

st III/IV HL

336

2x ABVD

DS 4/5

4x ABVD

6x BEACOPP-esc 39.7 mo

2-yr PFS: 82%/64% sign 2-yr OS: 98%

Johnson/ RATHL12

2015

RCT

st II-IV HL

1137

2x ABVD

DS 4/5

4x ABVD or 4x AVD

BEACOPP-14 or BEACOPP-esc

3-yr PFS: ABVD: 85.5%; AVD: 84.5% / i-PET pos:68% 3-yr OS: ABVD:97.0%; AVD: 97.5% / i-PET pos: 86%

32 mo

Straus/ 2015 CALGB Alliance 5060420

phase II

non-bulky st I/II HL

164

2x ABVD

DS 4/5

2x ABVD

2x BEACOPPesc+ IF RT

2 yr

3-yr PFS: 92%/66% sign

Ganesan21

2015

phase II

st IIB/III/IV HL

2x ABVD

DS 4/5

2x ABVD

4x BEACOPP-esc 24.7 mo

2-yr EFS: 82%/50% sign

Hertzberg22

2015

phase II

poor risk DLBCL

151

4x R-CHOP14

IHP

2x R-CHOP +2R

3x R-ICE + ZBEAM ASCT

35 mo

2-yr PFS: 74% /67% NS 2-yr OS: 88%/78% NS

Swinnen/ E340423

2015

phase II

DLBCL st II(bulky)/ III/IV

3x R-CHOP

‘ECOG criteria'

2x R-CHOP

4th R-CHOP +4x R-ICE

4.6 yr

2-yr PFS: 76% /42% NS 3-yr OS: 93%/69% NS

Stewart24

2014

phase II

adv st DLBCL

2x R-CHOP21 >Liver at >1 site

4x R-CHOP

1x R-DICEP + R-BEAM ASCT

41 mo

3-yr PFS: 65.2%/52.7% NS 3-yr OS: 68.4%/70.5% NS

Pardal25

2014

phase II

DLBCL/ gr 3B FL

3x R-MegaCHOP

IHP

3x R-MegaCHOP

2x R-IFE + BEAM ASCT

42.8 mo

3-yr PFS: 81%/57% sign 3-yr OS: 89%/73% NS

Dührsen/ PETAL19

2014

RCT

aggressive NHL (~80% DLBCL)

853

2x R-CHOP

<66% DSUV reduction

4x R-CHOP or 4x R-CHOP+2R

6x R-CHOP or 6x 'Burkitt protocol'

33 mo

2-yr TTTF: 79% i-PET+/47% i-PET- sign.

Sehn26

2014

phase II

adv stage DLBCL/PMBCL

150

4x R-CHOP21

IHP

2x R-CHOP21

4x R-ICE (+RT 45 mo if end of treatment PET pos)

4-yr PFS: 91%/59% sign 4-yr OS: 96%/73% sign

Casasnovas16

2011

phase II

DLBCL/PMBCL

102

2x R-CHOP14 or 2x R-ACVBP

IHP

R-CHOP14 or MTX+ R-ifosVP-16 +AraC

MTXiv + Z-BEAM ASCT

19 mo

PET 2: 2-yr PFS 73%/77% NS 2-yr OS 93%/ 84% NS PET 4: 2-yr PFS 81%/73% NS 2-yr OS 94%/83% NS

Moskowitz14

2010 prospective

adv stage DLBCL

3x ICE

biopsy neg: 3x ICE; biopsy pos:2x ICE+ 1x R-ICE+ASCT

44 mo

PFS NS OS NS

Kasamon27

2009

aggressive B-cell lymphoma

(R-)CHOP14 or 21

2x (R-)ESHAP or 33.6 mo 2x R-ICE

DLBCL

phase II

4x R-CHOP14 >local bg

2 or 3X (R-)CHOP

> bg

2-yr EFS 89%/75% 3-yr EFS: 82%/65%

HL: Hodgkin lymphoma; DLBCL: diffuse large B-cell lymphoma; PMBCL: primary mediastinal large B-cell lymphoma; NHL: non-Hodgkin lymphoma; iPET: interim positron emission tomography; RCT: randomized clinical trial; phase II: prospective phase II study; st: stage; adv: advanced; gr: grade; ABVD: doxorubicin, bleomycin, vinblastine, dacarbazine; (R-)CHOP: (rituximab), cyclophosphamide, doxorubicin, vincristine, prednisone; R-ACVBP: rituximab, doxorubicin, vindesine, bleomycin, prednisone; DS: Deauville score; IHP: International Harmonization Project; SUV: standardized uptake value; bg: background; rand: randomization; IF RT: involved field radiotherapy; NFT: no further treatment; INRT: involved node radiotherapy; AVD: doxorubicin, vinblastine and dacarbazine; 2R: 2 cycles rituximab; MTX: methotrexate; R-ifos-VP-16: rituximab, ifosfamide, vindesine; AraC: cytosine arabinoside; RT: radiotherapy; (R-)ICE: (rituximab), ifosfamide, carboplatin, etoposide; BEACOPP: bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, prednisolone; esc: escalated; BEAM: carmustine, etoposide, cytarabine, melphalan; ASCT: autologous stem cell transplantation; RDICEP: rituximab, dose-intensive cyclophosphamide, etoposide, cisplatin; R-IFE: rituximab, ifosfamide, etoposide; MTXiv: intravenous methotrexate; Z-BEAM: ibritumomab tiuxetan, carmustine, etoposide, cytarabine, melphalan; R-ESHAP: rituximab, etoposide, cisplatin, high-dose cytarabine, methylprednisone; FUP: follow up; mo: months; yr: years; PFS: progression-free survival; i-t-t: intention-to-treat; p-p A: per-protocol analysis; fav.: favorable; unfav.: unfavorable; OS: overall survival; EFS: event-free survival; TTTF: time to treatment failure; NS: not significant; Sign: statistically significant; ECOG: Eastern Cooperative Oncology Group.

1280

haematologica | 2016; 101(11)

Editorials

als of the German Hodgkin Study Group are currently awaited. Both trials are comparing standard combined modality treatment with a PET-directed regimen, omitting radiotherapy for patients with complete metabolic response after chemotherapy (www.ghsg.org). So de-escalation has become a real option in clinical practice, but requires detailed discussions between patients, hematologists and radiation oncologists. Balancing the risks and benefits of chemotherapy alone versus combined modality treatment depends on patient age, fitness, disease distribution and, most importantly, the individual assessment of that risk in the decision-making process. The recently published US Intergroup Trial of responseadapted therapy for stage III-IV Hodgkin lymphoma used early interim PET after 2 cycles of ABVD to escalate therapy for patients with Deauville score 4 or 5 to BEACOPP escalated. The authors concluded that response-adapted therapy based on iPET imaging seemed promising with a 2-year PFS of 64% for PET2-positive patients compared to historical series with 2-year PFS of 15%-30% for PET-positive patients treated with ABVD.11 Unpublished data presented in early and advanced disease from the EORTC H10 and the recently published UK Response Adapted Therapy in Advanced Hodgkin Lymphoma (RATHL) studies12 also suggest that escalation from ABVD to BEACOPP may be beneficial in patients with an inadequate response on iPET after 2 cycles. In RATHL, patients randomized to receive AVD rather than

ABVD on the basis of CMR on iPET had less pulmonary toxicity but no significant difference in 3-year PFS/OS. Published data are awaited for the EORTC H10 trial but in the meantime, at least in centers that participated in RATHL, this strategy is being offered to patients in clinical practice. The H10 and RAPID trials used the mediastinal blood pool (equivalent to DS 2) as the reference region for CMR; the RATHL study used the liver (DS 3). To avoid undertreatment, it may be desirable to use the mediastinal blood pool in trials testing de-escalation. The RATHL study, which tested both treatment escalation and de-escalation, used DS 3 as a cut off for CMR. The liver is a more reliable threshold for reporting iPET with respect to inter-reporter agreement and there was good agreement amongst reporters in local PET centers with expert central reviewers in RATHL.4 This supports the use of DS 3 for assessment of CMR in patients undergoing standard treatment but, in the authorsâ&#x20AC;&#x2122; opinion, in early stage disease for deescalation it is still prudent to use DS2. It is imperative that those reporting PET results and clinicians understand how the DS should be used for response-adaptation in clinical practice. Nowadays, many imaging specialists are educated in using DS not only for clinical trials, but also for clinical practice.

Interim-PET in diffuse large B-cell lymphoma R-CHOP is the standard therapy in DLBCL and will cure approximately 60% of patients. Standard treatment for

Figure 1. Coronal slices from 5 patients are shown at baseline and response. The level of uptake at residual sites, where present (arrowed) is graded according to the 5-point Deauville score.

haematologica | 2016; 101(11)

1281

Editorials

the significant proportion of patients up to the age of 70 years with relapsed or refractory disease is platinumbased immunochemotherapy followed by high-dose chemotherapy and autologous stem cell transplantation (ASCT). However, the results of second-line immunochemotherapy are disappointing, especially for patients who relapse within one year of completing R-CHOP treatment. Early identification of non-responders is of the utmost importance to maximize the chances of successful secondline therapy and to decrease side-effects associated with ineffective first-line therapy. To distinguish responders from non-responders, observational studies have indicated that iPET may be an effective predictive biomarker of outcome in DLBCL, but there are inconsistencies.13,14 It is unclear to what extent these are due to differences in the timing of PET during therapy, the choice of therapy and/or different PET reporting criteria. The current recommendation is to use DS, but earlier studies used International Harmonization Project criteria which separated PET into ‘positive’ and ‘negative’ by comparing FDG uptake with the intensity of the blood pool or nearby normal structures, if less than 2 cm, to offset partial volume effects.15 Standardized uptake value based methods have also been used to assess response in DLBCL. To date, most studies have applied the change in FDG uptake in the pixel with the highest uptake (SUVmax) before and during/after treatment (DSUV).6 Casasnovas et al. advocate DSUV as the most accurate criterion for response assessment. For lymphomas, in which cure is feasible and a rapid drop in SUV is common, cut offs for a clinically relevant interim assessment of response have been reported to range from 66% to 91%.16 Finally, metabolic tumor volume at baseline, perhaps combined with iPET response, has recently been reported as demonstrating predictive value.17 Currently, an international consortium called PETRA (PET-Re-Analyses) is pooling clinical studies in DLBCL to perform an individual patient data meta-analysis and compare different methods in assessing interim-PET.18 Hopefully, this will reveal the optimal time point and best visual or semi-quantitative PET-metrics to use for interim assessment. Another important issue is whether early identification of patients who are likely to be refractory to R-CHOP will result in better outcomes if these patients can be salvaged with high-dose chemotherapy or novel non-chemotherapeutic agents. Progress in targeted therapies in DLBCL might shift treatment paradigms from broad-spectrum poly-chemotherapy towards more targeted therapies based on genetic heterogeneity and complexity. These new drugs are currently being tested within phase I-II trials and results are awaited. Predicting response or resistance to a specific therapy will not only expedite the introduction of the most effective therapy to the patient but will also most likely be necessary to reduce the overall costs. Nowadays, international guidelines do not recommend changing standard treatment on iPET unless there is clear 1282

evidence of progression. Nonetheless, if mid-treatment imaging is performed, PET is better than CT at predicting prognosis and can be useful to exclude the possibility of progression. Preliminary published data and data presented only in abstract form suggest that, for patients with inadequate response on iPET, current chemotherapybased escalation strategies may not overcome treatment resistance19,23-24 (Table 1). For these patients, a more effective initial therapy regimen is needed.

Conclusions FDG-PET is a reliable biomarker for assessing early response in HL. The high negative predictive value of CMR after 2-3 cycles of ABVD has been the basis for recent trials exploring de-escalation of therapy in earlystage disease. The high positive predictive value in advanced disease has also been the focus of clinical trials, with promising data presented for patients escalated from ABVD to BEACOPP if they do not achieve a CMR after 2 cycles. In HL, PET-adapted therapy based on early response is rapidly becoming a clinical reality. In DLBCL, the ability to escalate treatment early for patients unlikely to respond to first-line immunochemotherapy is highly desirable, as these patients do not have good salvage options. Obtaining a CMR on interim PET has a high negative predictive value, but partial metabolic response is also often associated with good outcomes. Modifying treatment for patients who do not achieve an early CMR in DLBCL is likely to lead to overtreatment of a significant proportion of patients, with associated costs and patient anxiety.28 Early data suggest that patients with early failure also show treatment resistance with currently available salvage therapies, and novel, more targeted treatment strategies are clearly needed. Acknowledgment The authors wish to acknowledge Prof. dr. GJ Ossenkoppele and Prof. dr. OS Hoekstra for critically reviewing the manuscript.

References 1. Cheson BD, Pfistner B, Juweid ME, et al. Revised response criteria for malignant lymphoma. J Clin Oncol. 2007;25(5):579-586. 2. Cheson BD, Fisher RI, Barrington SF, et al. Recommendations for Initial Evaluation, Staging, and Response Assessment of Hodgkin and NonHodgkin Lymphoma: The Lugano Classification. J Clin Oncol. 2014;32(27):3059-3067. 3. Barrington SF, Mikhaeel NG, Kostakoglu L, et al. Role of imaging in the staging and response assessment of lymphoma: consensus of the international conference on malignant lymphomas imaging working group. J Clin Oncol. 2014;32(27):3048-3058. 4. Barrington SF, Kirkwood AA, Franceschetto A, et al. PET-CT for staging & early response: results from ‘Response Adapted Therapy in Advanced Hodgkin Lymphoma’ (RATHL) (CRUK/07/033). Blood 2016;127(12):1531-1538. 5. Biggi A, Gallamini A, Chauvri S, et al. International validation study for interim PET in ABVD-treated, advanced-stage Hodgkin lymphoma: interpretation criteria and concordance rate among reviewers. J Nucl Med. 2013;54(5):683-690. 6. Itti E, Meignan M, Berriolo-Riedinger A, et al. An international confirmatory study of the prognostic value of early PET/CT in diffuse large B-cell lymphoma: comparison between Deauville criteria and DSUVmax. Eur J Nucl Med Mol Imaging. 2013;40(9):1312-1320. 7. Boellaard R, Delgado-Bolton R, Oyen WJG, et al. FDG PET/CT: EANM procedure guidelines for tumour imaging: version 2.0. Eur J Nucl Med

haematologica | 2016; 101(11)

Editorials

Mol Imaging. 2015;42(2):328-354. 8. Radford J, Illidge T, Counsell N, et al. Results of a trial of PET-directed therapy for early-stage Hodgkin's lymphoma. N Engl J Med. 2015;372(17):1598-1607. 9. Raemaekers JM, André MP, Federico M, et al. Omitting radiotherapy in early positron emission tomography-negative stage I/II Hodgkin lymphoma is associated with an increased risk of early relapse: Clinical results of the preplanned interim analysis of the randomized EORTC/LYSA/FIL H10 trial. J Clin Oncol. 2014;32(12):1188-1194. 10. Meyer RM, Gospodarowicz MK, Connors JM, et al. ABVD alone versus radiation-based therapy in limited-stage Hodgkin’s lymphoma. N Engl J Med. 2012;366(5):399-408. 11. Press OW, Li H, Schöder H, et al. US Intergroup Trial of ResponseAdapted Therapy for Stage III to IV Hodgkin Lymphoma Using Early Interim Fluorodeoxyglucose-Positron Emission Tomography Imaging: Southwest Oncology Group S0816. J Clin Oncol. 2016;34(17):2020-2027. 12. Johnson PW, Federico M, Kirkwood A, Fossa A, et al. Adapted treatment guided by interim PET-CT scan in advanced Hodgkin’s lymphoma. N Engl J Med. 2016;374(25):2419-2429. 13. Terasawa T, Lau J, Bardet S, et al. Fluorine-18-fluorodeoxyglucose positron emission tomography for interim response assessment of advanced-stage Hodgkin’s lymphoma and diffuse large B-cell lymphoma: a systematic review. J Clin Oncol. 2009;27(11):1906-1914. 14. Moskowitz CH, Schöder H, Teruya-Feldstein J, et al. Risk-adapted DoseDense Immunochemotherapy Determined by Interim FDG-PET in Advanced-Stage Diffuse Large B-Cell Lymphoma. J Clin Oncol. 2010;28(11):1896-1903. 15. Juweid ME, Stroobants S, Hoekstra OS, et al. Use of positron emission tomography for response assessment of lymphoma: consensus of the Imaging Subcommittee of International Harmonization Project in Lymphoma. J Clin Oncol. 2007;25(5):571-578. 16. Casasnovas RO, Meignan M, Berriolo-Riedinger A, et al. SUVmax reduction improves early prognosis value of interim positron emission tomography scan in diffuse large B-cell lymphoma. Blood 2011;118(1):37-43. 17. Mikhaeel NG, Smith D, Dunn JT, et al. Combination of baseline metabolic tumour volume and early response on PET/CT improves progression-free survival prediction in DLBCL. Eur J Nucl Med Mol Imaging. 2016;43(7):1209-1219. 18. Zijlstra JM, Hoekstra OS, de Vet HCW. Validation of interim PET as a biomarker of response in NHL- a study on PET timing, therapies, response criteria, type of NHL and cost-effectiveness. Menton 2014.

haematologica | 2016; 101(11)

19.

20.

21. 22.

23.

24.

25. 26. 27.

28.

Available from: http://www.lymphomapet.com/files/Poster% 20Session%202014.pdf Dührsen U, Hüttmann A, Müller S, et al. Positron Emission Tomography (PET) Guided Therapy of Aggressive Lymphomas – a Randomized Controlled Trial Comparing Different Treatment Approaches Based on Interim PET Results (PETAL Trial). Blood 2014;124(21):(Abstract 391). Straus DJ, Pitcher B, Kostakoglu L , et al. Initial Results of US Intergroup Trial of Response-Adapted Chemotherapy or Chemotherapy/Radiation Therapy Based on PET for Non-Bulky Stage I and II Hodgkin Lymphoma (HL) (CALGB/Alliance 50604). Blood. 2015;126(23):(Abstract 578). Ganesan P, Rajendranath R, Kannan K, et al. Phase II study of interim PET-CT-guided response-adapted therapy in advanced Hodgkin's lymphoma. Ann Oncol. 2015;26(6):1170-1174. Hertzberg MS, Gandhi MK, Butcher B, et al. Early Treatment Intensification with R-ICE Chemotherapy Followed By Autologous Stem Cell Transplantation (ASCT) Using Zevalin-BEAM for Patients with Poor Risk Diffuse Large B-Cell Lymphoma (DLBCL) As Identified By Interim PET/CT Scan Performed after Four Cycles of R-CHOP-14: A Multicenter Phase II Study of the Australasian Leukaemia Lymphoma Study Group (ALLG.) Blood. 2015;126(23):(Abstract 815). Swinnen LJ, Li H, Quon A, et al. Response-adapted therapy for aggressive non-Hodgkin's lymphomas based on early [18F] FDG-PET scanning: ECOG-ACRIN Cancer Research Group study (E3404). Br J Haematol. 2015;170(1):56-65. Stewart DA, Kloiber R, Owen C, et al. Results of a prospective phase II trial evaluating interim positron emission tomography-guided high dose therapy for poor prognosis diffuse large B-cell lymphoma. Leuk Lymph. 2014;55(9):2064-2070. Pardal E, Coronado M, Martin A, et al. Intensification treatment based on early FDG-PET in patients with high-risk diffuse large B-cell lymphoma: a phase II GELTAMO trial. Br J Haematol. 2014;167(3):327-336. Sehn LH, Hardy ELG, Gill KK, et al. Phase 2 Trial of Interim PET ScanTailored Therapy in Patients with Advanced Stage Diffuse Large B-Cell Lymphoma (DLBCL) in British Columbia (BC). Blood 2014;124:392. Kasamon YL, Wahl RL, Ziessmann HA, et al. Phase II study of risk adapted therapy of newly diagnosed, aggressive Non-Hodgkin Lymphoma based on midtreatment FDG-PET scanning. Biol Blood Marrow Transplant. 2009;15(2):242-248. Barrington SF, Mikhaeel NG. PET-scans for staging and restaging in diffuse large B-cell and follicular lymphoma. Curr Hematol Rep. 2016;11(3):185-195.

1283

REVIEW ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

New insights on hereditary erythrocyte membrane defects Immacolata Andolfo,1,2 Roberta Russo,1,2 Antonella Gambale,1,2 and Achille Iolascon1,2

Dipartimento di Medicina Molecolare e Biotecnologie Mediche, UniversitĂ degli Studi di Napoli Federico II; and 2CEINGE Biotecnologie Avanzate, Napoli, Italy

Haematologica 2016 Volume 101(11):1284-1294

Correspondence: achille.iolascon@unina.it

Received: April 22, 2016. Accepted: June 16, 2016. Pre-published: October 18, 2016. doi:10.3324/haematol.2016.142463

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1284

ÂŠ2016 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights are reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

1284

ABSTRACT

fter the first proposed model of the red blood cell membrane skeleton 36 years ago, several additional proteins have been discovered during the intervening years, and their relationship with the pathogenesis of the related disorders have been somewhat defined. The knowledge of erythrocyte membrane structure is important because it represents the model for spectrin-based membrane skeletons in all cells and because defects in its structure underlie multiple hemolytic anemias. This review summarizes the main features of erythrocyte membrane disorders, dividing them into structural and altered permeability defects, focusing particularly on the most recent advances. New proteins involved in alterations of the red blood cell membrane permeability were recently described. The mechanoreceptor PIEZO1 is the largest ion channel identified to date, the fundamental regulator of erythrocyte volume homeostasis. Missense, gain-of-function mutations in the PIEZO1 gene have been identified in several families as causative of dehydrated hereditary stomatocytosis or xerocytosis. Similarly, the KCNN4 gene, codifying the so called Gardos channel, has been recently identified as a second causative gene of hereditary xerocytosis. Finally, ABCB6 missense mutations were identified in different pedigrees of familial pseudohyperkalemia. New genomic technologies have improved the quality and reduced the time of diagnosis of these diseases. Moreover, they are essential for the identification of the new causative genes. However, many questions remain to solve, and are currently objects of intensive studies.

Introduction Red blood cell (RBC) membrane disorders are inherited conditions due to mutations in genes encoding for membrane or cytoskeletal proteins as well as for transmembrane transporters or channels, resulting in decreased red cell deformability and permeability, a reduced half-life and premature removal of the erythrocytes from the bloodstream. Extensive studies on the RBC membrane have allowed the comprehension of both structure and function of this subcellular compartment. Thus, the molecular bases of the overwhelming majority of cases of hemolytic anemia due to RBC membrane defects have been currently defined. They are counted as a subtype of hereditary hemolytic anemias that embrace a highly heterogeneous group of chronic disorders with a highly variable clinical picture. In this review we summarize the biological, clinical and molecular aspects of red cell membrane defects, allowing for a better basis for diagnosis and treatment.

Red blood cell membrane: genesis, structure and function During its long life span of 120 days, the RBC is forced to cross the pores of splenic sinusoids thousands of times. This cell has an ongoing relationship with the spleen that contributes to remodeling during the first week of its life, participating in the passage from reticulocyte to erythrocyte. Moreover the spleen plays a prihaematologica | 2016; 101(11)

Update of the red cell membrane disorders

mary role in the removal of aged RBCs. In order to perform these journeys RBCs must possess and maintain a significant deformability. The main author of this property is certainly the membrane, that ensures both mechanical stability and deformability. After the first proposed model of the RBC membrane skeleton 36 years ago,1 containing the core elements of the modern model, many additional proteins have been discovered during the intervening decades, and their structures and interactions have been defined. RBC membrane structure has been extensively covered by excellent reviews.2,3 Herein we summarize the main concepts. RBC membrane is composed by a fluid double layer of lipids in which approximately 20 major proteins and at least 850 minor ones are embedded.4 The membrane is attached to an intracellular cytoskeleton by protein-protein and lipid-protein interactions that confer the erythrocyte shape, stability and deformability. The transmembrane proteins have mainly a transporter function. However, several of these also have a structural function, usually performed by an intracytoplasmic domain interacting with cytoskeletal proteins. The lipid bilayer acts as a barrier for the retention of cations and anions within the red cells, while it allows water molecules to pass through freely. Human erythrocytes have high intracellular K+ and low intracellular Na+ contents when compared with the corresponding ion concentrations in the plasma. The maintenance of this cation gradient between the cell and its environment involves a passive outward movement of K+, which is pumped back by the action of an ATP-dependent Na+/K+ pump in exchange for Na+ ions. This protein belongs to a class of transmembrane proteins with a transport function (Figure 1). The third and more important component of the RBC membrane is the cytoskeleton, a protein network that

laminates the inner surface of the membrane. Spectrin aand β-chains, proteins 4.1, or 4.1R, and actin are the main components of this skeleton, maintaining the biconcave shape of the RBC. These components are connected to each other in two protein complexes; ankyrin and protein 4.1 complex. The former is composed by band 3 tetramers, Rh, RhAG, CD47, glycophorin A and protein 4.2. Whereas the protein 4.1 complex is composed by band 3 dimers binding adducins a- and β-, glycophorin C, GLUT1 and stomatin (Figure 1). The ends of spectrin tetramers converge toward a protein 4.1 complex (junctional complex). Electron microscopy (EM) shows that this latter links the tail of six spectrin tetramers, forming a pseudo-hexagonal arrangement.5 Spectrin tetramers include anion transporters (band 3 or chloride/bicarbonate exchange). The capability of these transporters to form aggregates could define the half-life of RBCs, causing antibody binding and removal by the spleen. Defects that interrupt this vertical structure (spectrin-actin interaction) underlie the biochemical and molecular basis of hereditary spherocytosis (HS), whereas defects in horizontal interactions (skeletal attachment to membrane proteins) cause hereditary elliptocytosis (HE). Membrane protein synthesis is an important part of the differentiation process of erythroid cells in bone marrow and it starts very early. Cell culture studies established that this production is asynchronous (spectrin production starts before the synthesis of other cytoskeletal components) and is quantitatively exuberant (the production of a-spectrin exceeds that of β-spectrin three or four times).6 This pattern of production seems to play an important role in the genetics of both HS and HE: as a matter of fact only homozygous or double heterozygous defects of a-spectrin could cause HS; whereas the presence of hypomorphic alleles (such as a-LELY, Low expression Lyon) is complete-

Figure 1. Simplified cross-section of the erythrocyte membrane. The red blood cell membrane is composed of integral membrane proteins incorporated into a phospholipid bilayer. The network of cytoskeletal proteins is anchored to the membrane via several transmembrane proteins with a transport function: band 3, anion transporter; GLUT1, glucose and L-dehydroascorbic acid transporter; RhAG, gas transporter, in particular CO2; various cation pumps and transporters including, Na+-K+-ATPase, Ca++-ATPase, Na+-K+-2Cl− and Na+-Cl−, Na+-K+, K+-Cl− co-transporters and Gardos channel. The most recently described proteins PIEZO1, KCNN4 and ABCB6, involved in the modulation of RBC membrane permeability, and their putative interactions are also shown. The relative positions of the proteins to each other within the various complexes are mostly unknown. The shapes of the major proteins are mostly imaginary. GPA, glycophorin A; Rh, Rhesus polypeptide; B-4.1, protein band 4.1; B-4.2, protein band 4.2; GPC, glycophorin C; RhAG, Rh-associated glycoprotein; RBC: red blood cells. *Proteins that are known to be affected by pathogenic mutations so far.

haematologica | 2016; 101(11)

1285

I. Andolfo et al.

ly asymptomatic. However, due to its limiting amount (with respect to a-spectrin), the deficiency of β-spectrin causes HS in the heterozygous state as well. Band 3 and ankyrin synthesis are the latest to occur and they seem to play a critical role in assembly. Protein 4.1 and ankyrin are the last cytoskeletal protein components to continue to be synthesized and assembled. This is at least partly due to the fact that ankyrin and protein 4.1 mRNA persist late into erythropoiesis when the levels of the majority of cytoplasmic RNAs, including those for band 3 and spectrins, have declined precipitously.7

Classification, diagnostic criteria and epidemiology of erythrocyte membrane defect-related anemias From the genetic standpoint, 15 different types of anemias due to RBC membrane defects are currently included in the Online Mendelian Inheritance in Man (OMIM) compendium of human genes and genetic phenotypes (Table 1). Of note, the gene mutations identified so far refer only to a restricted number of patients; indeed, the molecular defect is still unknown for several patients. We can divide RBC membrane disorders into two main sub-

groups: (i) structural defects, and (ii) altered permeability of the RBC membrane. The first subgroup comprises: HS, HE, hereditary pyropoikilocytosis (HPP), and Southeast Asian ovalocytosis (SAO); the second subgroup contains: dehydrated hereditary stomatocytosis (DHS), overhydrated hereditary stomatocytosis (OHS), familial pseudohyperkalemia (FP), and cryohydrocytosis (CHC).

Hereditary anemias due to RBC structural defects Hereditary spherocytosis HS is the most common non-immune hemolytic anemia with a prevalence of 1:2000-5000 in the Caucasian population.8 This value is probably higher due to under-diagnosed mild/moderate forms. HS refers to a group of heterogeneous inherited anemias showing a broad spectrum of clinical severity, ranging from asymptomatic to severe transfusion-dependent forms, even within the same family. The intra-familial heterogeneity can be ascribed to the co-inheritance of genetic variants involved in erythrocyte defects themselves or in other disorders, such as enzymopathies, thalassemias and Gilbert syndrome.9 However, HS clinical findings are summarized by hemolytic anemia, jaundice and splenomegaly. Reticulocytosis (6-10% to 35% in severe cases), increased

Table 1. Classification of erythrocyte membrane disorders by OMIM database.

Phenotype MIM number

Gene location

Protein name§

Hereditary spherocytosis type 1

182900

Ankyrin-1

HS2

Hereditary spherocytosis type 2

616649

Spectrin β chain, erythrocytic

HS3

Hereditary spherocytosis type 3

270970

Spectrin a chain, erythrocytic 1

HS4

Hereditary spherocytosis type 4

612653

Band 3 anion transport protein

HS5

Hereditary spherocytosis type 5

612690

Erythrocyte membrane protein band 4.2

HE1

Hereditary elliptocytosis 1

611804

Protein band 4.1

HE2

Hereditary elliptocytosis 2

130600

Spectrin a chain, erythrocytic 1

HE3

Hereditary elliptocytosis 3

Spectrin β chain, erythrocytic

HPP

Hereditary Pyropoikilocytosis

266140

Spectrin a chain, erythrocytic 1

SAO

Ovalocytosis Southeast Asian type

166900

Band 3 anion transport protein

OHS

Overhydrated hereditary stomatocytosis

185000

Ammonium transporter Rh type A

DHS1

194380

DHS2

Dehydrated hereditary stomatocytosis with or without pseudohyperkalemia and/or perinatal edema Dehydrated hereditary stomatocytosis 2

Familial pseudohyperkalemia

609153

CHC

Cryohydrocytosis

185020

ANK1 8p11.21 SPTB 14q23.3 SPTA1 1q23.1 SLC4A1 17q21.31 EPB42 15q15.2 EPB41 1p35.3 SPTA1 1q23.1 SPTB 14q23.3 SPTA1 1q23.1 SLC4A1 17q21.31 RHAG 6p12.3 PIEZO1 16q24.3 KCNN4 19q13.31 ABCB6 2q35-q36 SLC4A1 17q21.31

Disease symbol

Phenotype

HS1

616689

Inheritance

Piezo-type mechanosensitive ion channel AD component 1 Intermediate conductance AD calcium-activated potassium channel protein 4 ATP-binding cassette sub-family B AD member 6 Band 3 anion transport protein AD

Protein name reported in Uniprot database. AD: Autosomal dominant; AR: Autosomal recessive; ATP: adenosine triphosphate; Rh: Rhesus; OMIM: Online Mendelian Inheritance in Man; MIM: Mendelian Inheritance in Man.

1286

haematologica | 2016; 101(11)

Update of the red cell membrane disorders

mean corpuscular Hb concentration (MCHC > 34.5g/dL), increased RBC distribution width (RDW >14), and normal or slightly decreased MCV are the main laboratory findings. Anemia in most patients is mild (Hb >11 g/dL) or moderate (Hb 8-11 g/dL), due to poorly compensated hemolysis.10 Symptomless or mildly anemic patients are often diagnosed after hemolytic or aplastic crises, while in mildly affected women the condition often becomes evident during pregnancy, but transfusions are required only rarely. A small percentage of patients present a severe form (Hb 6-8 g/dL), which needs regular blood transfusions. One of the most common complications of chronic hemolytic anemia is cholelithiasis, which is more frequent in patients who co-inherited Gilbert Syndrome.11 Of note, the co-inheritance of HS and Gilbert disease can be misdiagnosed as Crigler-Najjar syndrome type II. The third component of the HS triad is splenomegaly, observed in almost all adult patients. The spleen enlargement is mild or moderate, rarely massive: only one patient with spontaneous rupture has been described,12 while in few patients splenic infarction has been observed.13,14 Extramedullary erythropoiesis and iron overload can also

be observed. Hemosiderosis is more relevant in transfusion-dependent patients or in those who have co-inherited mutations in the causative genes of hereditary hemochromatosis.15 The diagnosis of HS is based on clinical features, positive familial history and the observation of a peripheral blood (PB) smear, in which a variable percentage of spherocytes, related to the degree of anemia, mushroom red cells, poikilocytosis, acanthocytes and ovalostomatocytes can be found (Figure 2).16 The diagnostic guidelines of HS from the British Committee for Standards in Haematology do not recommend any additional tests for patients with classical clinical features and laboratory data.17 Whenever necessary, indirect tests can also be performed. Among these, the eosin-5â&#x20AC;˛-maleimide (EMA) binding test shows high sensitivity (92-93%) and specificity (nearly 99%), although a positive test can also be obtained in patients affected by related conditions, such as congenital dyserythropoietic anemia type II (CDA II).16-18 Additional tests, such as the osmotic fragility (OF) test, acidified glycerol lysis test (AGLT) and the pink test, exhibit lower sensitivity compared to the EMA test (68%, 61% and 91%,

Figure 2. Flow diagram for the differential diagnosis of hemolytic anemias due to RBC membrane defects. The flow diagram shows the main steps for guiding the clinical suspicion toward the diagnosis of different subtypes of hereditary erythrocyte membrane disorders. First-, second-, and third-line investigations are also shown. The cut-off for the EMA binding test is still debated: currently, a test with a reduction of EMA binding > 21%, in comparison with controls, is defined positive, whereas a test with a reduction of EMA binding < 16% is considered negative. Values between 16-21% are not conclusive, although a cut-off of 11% has been proposed. Hb: hemoglobin; MCH: mean corpuscular hemoglobin; MCV: mean cellular volume; MCHC; mean corpuscular hemoglobin concentration; CBC: complete blood count; RBC: red blood cells; OHS: overhydrated hereditary stomatocytosis; DHS: dehydrated hereditary stomatocytosis; AD: autosomal dominant; AR: autosomal recessive; EMA: eosin-5-maleimide; SDS: Sodium dodecyl sulfate; NGS: next generation sequencing; RHAG: rhesus blood groupassociated glycoprotein. PB: peripheral blood.

haematologica | 2016; 101(11)

1287

I. Andolfo et al.

respectively). Nevertheless, the combination of the EMA and pink tests or those of the EMA and AGLT tests improves the sensitivity to 99% and 100%, respectively.19 Ektacytometry is a highly sensitive test of membrane deformability.10 In HS the characteristic features of ektacytometry are: decreased DI max, in conjunction with a shift of the Omin point to the right (reduced surface to volume ratio), and a shift of the O' or hyper point to the left (increased dehydration of the red cells) (Figure 3).16 As a third-line of investigation, the analysis of major erythrocyte membrane proteins via SDS-PAGE still represents invaluable support for the identification of different subsets of HS patients; however, several subjects remain unclassified by this technique. As discussed later in this review, the current availability of advanced genomic surveys, such as next generation sequencing (NGS) technologies, allows one to overcome the limitations of previous analytic methods. However, the biochemical analyses may be of great use in the interpretation of NGS data in order to assess the pathogenicity of identified genetic variants. In HS, the phenotype variability is linked to different molecular defects. The increased membrane fragility is caused by heterogeneous molecular defects due to deficiency and/or dysfunction in erythrocyte membrane pro-

teins, ankyrin (ANK1), a- and Î˛-spectrin (SPTA and STPB), band 3 (SLC4A1), and protein 4.2 (EPB42). Approximately 75% of HS cases exhibit an autosomal dominant (AD) pattern of inheritance, associated with mutations in ANK1, SPTB, EPB42 and SLC4A1 genes. In the remaining 25% of patients autosomal recessive (AR) and de novo mutations were observed (Table 1). In rare cases, HS can be associated to psychomotor developmental delay and autism in contiguous gene syndromes due to large genomic deletions, including ANK120,21 or SPTB genes.22,23 Finally, prenatal hydrops fetalis has been rarely observed in patients with mutations in SLC4A124 and SPTA-SPTB genes.25,26

Hereditary elliptocytosis and pyropoikilocytosis HE is characterized by the presence of elliptical-shaped erythrocytes (elliptocytes) on the PB smear associated to variable clinical manifestations. The worldwide incidence of HE is 1:2000-4000 individuals, but it results higher in some African regions (1:100). The majority of patients present no anemia or hemolysis, and diagnosis is made incidentally, after worsening of anemia due to infections or after diagnosis in symptomatic relatives. Severe anemia was observed only in rare cases. A good indicator for the severity of the disease is the percentage of spectrin dimers. A subtype of HE is HPP, a rare severe hemolytic anemia characterized by poikilocytosis and fragmented erythro-

Figure 3. Examples of ektacytometric curves of different hereditary erythrocyte membrane disorders. The ektacytometer is a laser diffraction viscometer that measures RBCs (red blood cells) deformability at constant shear stress as a continuous function of suspending Osmolarity. Three principal features of the osmotic gradient ektacytometry profiles are: the Omin point (red asterisk), which corresponds to the osmolarity at which 50% of the red cells are lysed in the classical osmotic fragility test and represents the surface area to volume ratio; the maximal deformability index (DImax, black asterisk) value, which represents the maximal cellular deformability of the red cell population; and the O' or hyper point (blue asterisk), which corresponds to the osmolarity at DImax/2, which reflects the hydration status of the red cells. Ektacytometric analysis of (A) HS, (B) OHS, (C) DHS1 and (D) DHS2 patients are shown. The dotted line is those relative to the control subjects. HS: hereditary spherocytosis; OHS: overhydrated hereditary stomatocytosis; DHS1: dehydrated hereditary stomatocytosis 1; DHS2: dehydrated hereditary stomatocytosis 2.

1288

haematologica | 2016; 101(11)

Update of the red cell membrane disorders

cytes, resulting in low MCV (50-60 fL) and microspherocytes.16 HPP patients show marked splenomegaly, and splenectomy is therefore usually recommended. There is a strong association between HE and HPP. The main defect in HE erythrocytes is mechanical weakness or fragility of the erythrocyte membrane skeleton due to defective horizontal connections of cytoskeletal proteins, such as spectrin dimer-dimer interactions and spectrin-actin-protein 4.1 at the junctional complex. For the main part, HE is inherited as AD disease, with rare cases of de novo mutations, whereas HPP patients exhibit an AR inheritance. HE can be due to mutations in EPB41, SPTA1 and SPTB genes that lead to serious damage in the association of spectrin dimers/tetramers.27 Also, HE shows high inter- and intrafamilial phenotypic variability, due to the modifier alleles. One example is the a-LELY in the SPTA1 gene, a hypomorphic haplotype composed of two variants, the missense Leu1857Val and the splicing variant in intron 45. This hypomorphic haplotype alone causes minimum damage in both heterozygous and homozygous states since the spectrin a chains are produced in excess (3- to 4fold compared to Î˛-chains); otherwise, when it is associated with a HE mutation in SPTA1, the resulting phenotype is severe, i.e., HPP (Table 1).27

Southeast Asian ovalocytosis SAO is a very common condition in the aboriginal peoples from Papua New Guinea, Indonesia, Malaysia, the Philippines and southern Thailand, in areas where malaria is endemic, with prevalence varying between 5% and 25%. Indeed, this condition offers protection against all forms of malaria.27 Despite the reduced in vitro deformability of SAO erythrocytes, patients are asymptomatic and the diagnosis is made accidentally as a result of a PB smear examination, showing the characteristic rounded elliptocytes (ovalocytes). However, in newborns it may manifest as hemolytic anemia and require phototherapy. SAO is an AD condition caused by the deletion of 27 nucleotides in the SLC4A1 gene, leading to the loss of the amino acids 400-408 of protein band 3.2 The deletion is in linkage disequilibrium with the Memphis polymorphism (p.Lys56Glu) in SLC4A1 (Table 1). SAO erythrocytes show a slight loss of monovalent cations when exposed to low temperatures, with a reduction of anions flux. Thus, the condition may be classified as a genetic disease affecting the permeability of the RBC membrane. Despite the frequency of heterozygotes, only one case homozygous for the 27 nucleotide deletion has been described so far. This patient showed severe phenotype with intrauterine transfusions, transfusion dependent anemia and distal renal tubular acidosis due to the loss of band 3, which is also expressed in the kidneys.28

Other conditions Erythrocyte abnormalities can also be observed in other hereditary and acquired conditions. For example, the autoimmune hemolytic anemias are characterized by shortened RBC survival due to the presence of auto-antibodies directed toward red cells, with a positive Coombs test. RBCs are typically coated with auto-antibodies and trapped by macrophages in the cords of the spleen. The interaction of trapped RBCs with splenic macrophages may result in phagocytosis of the entire cell or partial phagocytosis with the formation of spherocytes, present in the blood film.29 Micro- and macrospherocytes, associhaematologica | 2016; 101(11)

ated with increased osmotic fragility, were also seen in patients affected by chronic hepatitis C virus treated with protease inhibitors (telaprevir and boceprevir). In these patients oxidative stress, induced by drugs, damages membrane-cytoskeletal stability, reducing a- and Î˛-spectrins.30 Alterations in the RBCs membrane are also present in neuroacanthocytosis, a heterogeneous group of diseases that include chorea-acanthocytosis, McLeod and Huntington's disease-like syndromes. These conditions are characterized by alterations of post-translational modifications, mostly phosphorylation, of erythrocyte membrane proteins and significant neurological symptoms.31

Anemias due to altered permeability of RBC membrane Dehydrated hereditary stomatocytosis or xerocytosis HST includes both DHS and OHS, which show alteration of the RBC membrane permeability to monovalent cations Na+ and K+, with a consequent alteration of the intracellular cationic content and alterations of cell volume.32 DHS is the most highly represented among HST, with an incidence of approximately 1:50000 births. It is 10-20 times less frequent than HS, with which it may be, however, confused. Of note, based on our experience, the incidence of this anemia could be underestimated because it is often undiagnosed. The phenotype ranges from asymptomatic to severe forms, with massive hemolysis. Generally, DHS patients show hemolytic well-compensated anemia, with a high reticulocyte count, a tendency to macrocytosis and mild jaundice. The main characteristic of RBCs is cell dehydration caused by the loss of the cation content, with a subsequent increase of MCHC (>36 g/dL). At blood smear the stomatocytes, erythrocytes with a characteristic central mouth-shaped spot, are quite rare, which often makes diagnosis difficult. In addition, it may be difficult when the clinical picture is associated with pseudohyperkalemia and/or perinatal edema, in the so-called pleiotropic syndrome form.33,34 For these reasons, the condition may be overlooked for years or decades before reaching a conclusive diagnosis. Osmotic gradient ektacytometry is a useful tool to diagnose this condition; it shows a leftward shift of the minimum in the deformability index (Omin) at low osmolarities, as well as a decrease in DImax (Figure 3). DHS patients also exhibit a tendency toward having iron overload, regardless of the transfusion regimen or splenectomy.35 The study of the iron metabolism in this condition is an open and interesting field of investigation, enabling the discovery of new drugs to treat the iron overload. DHS is inherited as an AD trait. The candidate gene locus was first localized at 16q23-24.36,37 Several years later, PIEZO1 was identified as the causative gene of both isolated and syndromic forms of DHS1 by exome sequencing (Table 1).38,39 PIEZO1 encodes a mechanoreceptor, an ion channel activated by pressure. This protein has been identified in the RBC membrane, and in mice it has been shown to form a tetramer of about 1.2 million daltons; it is therefore the largest ion channel identified to date, and moreover it regulates mechanotransductive release of ATP from human RBCs.40-43 The identified mutations are missense and mainly located in the highly conserved C-terminus of the protein, recently described to form the pore of the channel.44 Several electrophysiology studies demonstrated that the mutations cause a gain-of-function phenotype with delayed inactivation of the channel,38,39,45 sug1289

I. Andolfo et al.

gesting increased cation permeability leads to DHS erythrocyte dehydration. PIEZO1 is currently the subject of intense research and has been shown to be involved in several physiological and pathophysiological processes. The study of this mechanoreceptor will shed light on the hydration pathways in healthy and diseased RBCs. Recently, a novel gene, KCNN4, has been identified as causative of a second form of DHS, named DHS2, in six different families (Table 1).46-48 The KCNN4 gene encodes the Gardos channel, a widely expressed Ca2+-dependent K+ channel of intermediate conductance that mediates the major K+ conductance of erythrocytes.46-48 Mutated KCNN4 channels showed a higher current compared to WT resulting from changes in the open probability, in the trafficking, and in the unitary conductance of the channel. This is suggestive of the pathogenic mechanism associated with several mutations affecting PIEZO1. In addition, this observation could suggest that PIEZO1 and the Gardos channel might act in the same stretch-induced cation pathway involved in cell volume homeostasis. Unlike DHS1, patients affected by DHS2 show a normal pattern of ektacytometry analysis (Figure 3),46,47 whereas they exhibit iron overload similar to that in DHS1 patients. HST can also be associated with band 3 mutations, characterized by the conversion of band 3 from an anion exchanger to a cation transporter.49

Overhydrated hereditary stomatocytosis OHS is a very rare subtype among HST, with 20 cases reported overall worldwide. Contrary to DHS, RBCs are hydrated due to an increase, from 20 to 40 times, in the loss of cations.32 OHS is associated with more severe phenotypes compared to DHS. In addition to reticulocytosis, it is characterized by a sharp increase in MCV (>110 fL) and decreased MCHC (24-30 g/dL). The number of stomatocytes is usually much higher than that observed in DHS. The causative gene of this condition is RHAG, encoding the Rh-associated glycoprotein (RhAG) which acts as an ammonia channel (Table 1).50 Stomatin has been found at low or absent levels in OHS patients, but no mutations have been found in the encoding gene so far.34 Moreover, a complex syndrome named stomatin-deficient cryohydrocytosis has been described. It is characterized by mental retardation, seizures, cataracts and massive hepatomegaly. RBCs showed dramatic resumption of the leak in vitro when stored at low temperatures and in the absence of stomatin.51,52 Recently, this syndrome was associated with mutations in SLC2A1 that cause both loss of glucose transport and a cation leak.53

Familial pseudohyperkalemia and cryohydrocytosis FP and CHC are additional forms of stomatocytosis. FP is not associated with hemolytic anemia and stomatocytes are rarely observable on PB smear. Conversely, CHC patients show hemolytic anemia of variable degrees.54 RBCs from FP patients exhibit a loss of K+ at low temperatures (<37°C, mostly 8-10°C), but not at 37°C. In CHC the main feature is the temperature dependence of the loss of cations: instead of being around 8-10°C, the minimum is approximately 23°C. The gene responsible for FP was mapped at 2q35-q36,55 and later identified in the ABCB6 gene,56 encoding the homonymous protein, ABCB6. It belongs to the family of ABC transporters with the binding cassette for ATP, one of the most abundant families of 1290

integral membrane proteins. ABCB6 was previously identified as a porphyrin transporter, thus we now find that it is currently highly debated because several other studies identified its expression in the plasma membrane of RBCs and in the endo-lysomal compartment, excluding the mitochondrial uptake of porphyrins.57 Moreover, in erythrocyte membranes it bears the Langereis (Lan) blood group antigen system.58,59 ABCB6 expression increases during erythroid differentiation of CD34+ cells.56 The ABCB6 missense mutations in FP does not alter mRNA or protein levels, or subcellular localization in mature erythrocytes or erythroid precursor cells, but are predicted to have a pathogenic consequence on protein function. Recently, ionic flux assays on the mutations found in FP patients have demonstrated that the mutations are gain-of-function, causing an abnormal loss of potassium from cells at low temperatures.60 These changes could lead either to a cation leak through the normal substrate translocation pathway of ABCB6, or to the generation of a novel constitutive or cyclic leak pathway through the protein. Recently, the ABCB6 variants R723Q, V454A, and R276W were found in FP patients who are regular blood donors.60,61 The blood of these patients exhibited increased potassium leakage upon storage at temperatures below 37°C. Of note, all these variants are annotated in public databases, suggesting that FP is common in the general population. Particularly, the variant R276W has been found in one of 327 random blood donors (0.3%). The storage of blood of these subjects leads to significantly increased K+ levels, with serious clinical implications for neonates and infants receiving large-volume transfusions of whole blood. This interesting finding encouraged further study on the implications for neonates and infants receiving transfusion of whole blood from undiagnosed FP subjects.60 CHC is due to mutations in the SLC4A1 gene; these are gain-of-function mutations, since they are able to transform the band 3 anion exchanger to a cation transporter (Table 1).

Differential diagnosis Hemolytic anemias caused by RBC membrane defects can often be misdiagnosed with other hemolytic anemias. In particular, HS can be confused with autoimmune hemolytic anemia that shows spherocytes on the PB smear. Thus, it is crucial to perform additional diagnostic tests, such as a Coombs' test (the direct antiglobulin test) to distinguish between these conditions. Other conditions, including liver disease, thermal injury, micro- and macroangiopathic hemolytic anemias, clostridial sepsis, transfusion reaction with hemolysis, severe hypophosphatemia, ABO incompatibility, and poisoning with certain snake, spider, and hymenoptera venoms can be associated with the presence of spherocytes on the PB smear. It is critical to evaluate the disorder in the proper clinical context and to evaluate the family history and transmission pattern.10 Additionally, membrane defects can be misdiagnosed with enzymatic defects, particularly with glucose-6-phosphate dehydrogenase (G6PD) and pyruvate kinase (PK) deficiencies. In cases where red cell enzyme deficiency is suspected, the diagnosis should be confirmed by performing the most common red cell enzyme assays and by analyzing the inheritance of the disease that is X-linked for haematologica | 2016; 101(11)

Update of the red cell membrane disorders

G6PD deficiency and autosomal recessive for PK defect. CDA II also shares some characteristics with HS; in fact, CDA II patients are often erroneously diagnosed as HS.62 It is critical to evaluate the reticulocyte count, which is elevated in HS while it is inadequate for the degree of anemia in CDA II, because the marrow stress is higher in respect to HS for the same Hb level. This observation is confirmed by the increased sTfR levels observed in CDA II patients compared to those with HS. RDW is also characteristically increased in CDA II, while the Hb distribution width (HDW) is increased in HS, resulting in an RDW/HDW ratio that is significantly greater in CDA II than HS.63 Additionally, a new clinical index, named BM responsiveness index (BMRI), has been developed to discriminate a well-compensated hemolytic anemia from an ineffective erythropoiesis one. BMRI is calculated as [(patient's absolute reticulocyte count) Ă&#x2014; (patient's Hb/normal Hb)] and showed high specificity and sensitivity.64 This index represents the degree of production of new RBCs in a condition of anemia. Of note, it is crucial to distinguish between these anemias because HS patients take advantage of splenectomy; conversely, this intervention lacks substantial improvement in CDA II patients. Furthermore, both OHS and DHS can be erroneously diagnosed with CDA type I (CDA I). Both CDA I and HST present macrocytosis associated with hemolytic signs. Likewise in this case, the accurate evaluation of the reticulocyte count can be enlightening, since it is inadequate for the degree of anemia in CDA I. Whenever applicable, a bone marrow aspirate can provide a clue, since it shows hypercellularity and erythroid hyperplasia with the pathognomonic morphological features of CDA I, i.e, the presence of thin chromatin bridges between the nuclei pairs of erythroblasts.62

Therapy and management A correct diagnosis of RBC membrane disorders should be obtained before starting treatment. The first-line treatment is often based only on supportive care. For example, HS neonatal patients often need phototherapy or exsanguinous-transfusions, in case of significant jaundice.16 In the children of affected parents, bilirubin values should be strictly monitored. Moreover, some newborns affected by erythrocyte membrane defects may necessitate transfusions: in these patients recombinant erythropoietin (rEPO) can reduce transfusion requirements.65 Similarly, transfusions may be administered as needed in adulthood, mostly during aplastic or hemolytic crises. In patients with increased folate demands, such as children, pregnant women or in patients with evidence of folate deficiency, supplements of this vitamin is suggested. Splenectomy could be suggested in several cases of HS and HE/HPP, resulting in an increased life span of RBCs. Currently, total splenectomy is recommended in adult patients with severe forms of HS and HPP; otherwise, in patients with mild or moderate anemia, splenectomy is not recommended thus far. However, splenectomy increases the risk of both infections (mostly from encapsulated organisms) and thrombosis.66 Vaccinations against encapsulated bacteria, such as Neisseria meningitidis, streptococcus pneumonia and haemophilus influenza,16 should be administered before surgery, even if thus far there is no standard protocol regarding antibiotic prophylaxis after haematologica | 2016; 101(11)

splenectomy. Whenever possible, splenectomy in children should be delayed until they reach 6 years of age, otherwise partial splenectomy (residual splenic mass 20-30%) can be performed.17 Laparoscopic splenectomy is preferred for shorter postoperative discomfort and hospital stay, even if it involves a longer operative time and needs more surgical experience.67 However, laparoscopy is not applicable for spleens which are too large. In the presence of cholelithiasis, contemporary cholecystectomy can be performed. In patients with moderate HS, with poor resistance to fatigue or growth retardation, the role of splenectomy is still debated. Recently Pincez and colleagues have proposed subtotal splenectomy (residual splenic mass 1525%) for severe and moderate HS.68 They propose this approach in order to reduce long-term complications: after a long follow-up, subtotal splenectomy led to a good hematologic response in many patients, but 5 out of 79 patients remained transfusion dependent, 3 were considered as asplenic and 1 showed no significant increase of Hb. In our opinion, in severe HS cases total splenectomy would be preferred, while the role of subtotal splenectomy in symptomatic non-transfusion dependent patients needs to be evaluated case by case. In a few HPP patients, subtotal splenectomy (resection of 80-90% of spleen volume) has been performed, with no advantage as compared to total splenectomy.69 Conversely, splenectomy is contraindicated in DHS due to the increased risk of thromboembolic complications,70,71 probably related to the augmented number of undestroyed stomatocytes in the bloodstream. Thrombotic complications (arterial and venous events, including portal vein thrombosis and pulmonary hypertension) have been described at a high rate following splenectomy.70,71 In literature, data on splenectomy of OHS patients are lacking. Splenectomy should be avoided in this condition because the percentage of stomatocytes is even higher compared to DHS.

Animal models Several animal models have been developed for the study of RBC defects (Table 2). Concerning the mouse models, although the main stream of human hemolytic anemias caused by RBC membrane defects are inherited as AD, in mice these conditions are usually inherited as AR (Table 2). Recently, within hereditary stomatocytosis, mouse models of PIEZO1 were developed. Mice deficient in Piezo1 die in utero at mid-gestation due to defective vasculogenesis.72 Thus, another model of Piezo1 was developed by specific deletion in the hematopoietic system (Vav1-P1cKO mice).72 Hematological analysis of blood from Vav1-P1cKO mice revealed elevated MCV, MCH and reduced MCHC. RBCs exhibited increased osmotic fragility, demonstrating that Piezo1-deficient RBCs were overhydrated. At the moment a knock-in mouse model carrying missense mutations found in DHS1 patients doesn't exist; its creation will further elucidate the role of PIEZO1 in the hydration pathways of RBCs. Beyond mouse models, which in several cases don't recapitulate the main characteristics of human hematological disease, a powerful model for the study of erythropoiesis is the zebrafish, Danio rerio, because of its small size, its ability to generate a large number of embryos, and 1291

I. Andolfo et al. Table 2. Animal models for the study of RBC membrane defects.

Organism Mutant

Altered gene

Type of alteration

Mus musculus ja/ja sph/sph Nan

Sptb Spta Klf1

Deficiency of Î˛-spectrin Deficiency of a-spectrin Missense E339D

ENU-generated

Ank1

Nonsense E924X

wan

Slc4a1

Premature stop codon

sphDem/sphDem

Spta

Vav1-P1cKO

Piezo1

Inframe deletion of 46 amino acids that alters spectrin dimer/tetramer stability Gene deletion in the hematopoietic system

Danio rerio merlot Morpholino ZFN knockout

Epb41 Piezo1 Piezo1

Mutation

Phenotype

References

Severe hemolytic anemia, reticulocytosis Neonatal anemia; in adult mice, hemolytic anemia with decreased RBCs, hematocrit, Hb, and elevated zinc protoporphyrin levels Heterozygous mice: low MCV, elevated RBC counts, reticulocytosis, reduced EMA intensity, and increased osmotic fragility Homozygous mice: severe anemia with marked anisocytosis and spherocytosis on the PB smear Increased MCV, decreased MCHC, marked reticulocytosis (50%). PB smears: elliptocytes, spherocytes and occasional poikilocytes, as seen in severe human HE Elevated MCV, MCH and reduced MCHC. RBCs exhibit increased osmotic fragility

Spiculated RBC membranes; hemolytic anemia, cardiomegaly, splenomegaly Knockdown of gene expression Severe anemia with swollen, fragile and spherocytic RBCs Frameshift in exon 8 No anemia or dysmorphic erythrocyte morphology

84 85 86

88 89

90 74 75

RBC: red blood cells; Hb: hemoglobin; MCV: mean cellular volume; EMA: eosin-5-maleimide; PB: peripheral blood; MCHC: mean corpuscular hemoglobin concentration; HE: hereditary elliptocytosis; MCH: mean corpuscular hemoglobin.

its transparency that facilitates the visualization of erythroid cell migration.73 Notably, the high conservation of hematopoietic genes among vertebrates and the ability to successfully transplant hematopoietic cells have enabled the establishment of models of human anemic diseases in zebrafish (Table 2). Recently, zebrafish models have also been created for PIEZO1. Morpholino-knockdown of Piezo1 expression in the Danio rerio was reported to result in severe anemia.74 However, the phenotype observed in the morpholino-knockdown model was not present in an independent zebrafish model carrying a predicted truncated form of PIEZO1 (Table 2).75 The debate on the phenotype observed in the two different models is still open.76,77 It is notable that patients with homozygous loss-of-function mutations in human PIEZO1 show lymphatic dysplasia and an asymptomatic, fully compensated, very mild hemolytic state of incomplete penetrance.78,79 In conclusion, both mouse and zebrafish models appear not to better recapitulate the human pathogenesis, but they are useful to study the function of newly identified proteins such as PIEZO1.

Conclusions and perspectives Hereditary anemias due to RBC membrane defects represent a heterogeneous group of hereditary defects with very overlapping phenotypes. Indeed, the clinical definition of patients is often difficult. For some conditions, the great phenotypic variability is partially explained by the high genetic heterogeneity; otherwise, it is sometimes complicated to distinguish one form from the others since the signs can be veiled in symptom-free carriers or in mildly affected patients. Moreover, some subtypes of RBC membrane disorders can be easily confused with 1292

other clinically-related hereditary hemolytic conditions, as classically reported for differential diagnosis of HS and CDA II.64 Thus, when there is a suspected hereditary erythrocyte membrane disorder, after the exclusion of other common diseases, it is essential to perform a depth analysis of PB smear and pedigree transmission of the disease. Biochemical tests can be useful, especially in HS, but they do not have high sensitivity. The combination of an EMA test with ektacytometry is of great help for the majority of these conditions, but ektacytometry analysis is of limited availability. Thus, the genetic analysis becomes crucial, mainly in cases with an ambiguous phenotype. The absence of clear genotype/phenotype correlations is often problematic for both genetic counseling and suitable treatments. It is proper in this context that new genomic technologies are utilized. In the last few years, remarkable progress has been made in discovering new disease genes involved in these disorders by means of unbiased genomic approaches, such as whole exome sequencing.39,46,47,56 However, the increasing genetic heterogeneity underlines the problem of a very complex differential diagnosis. In the next generation sequencing (NGS) era, the genetic testing is going to move from few candidate genes to wider panels of genes, namely targeted (t)-NGS. Recent studies have already demonstrated the usefulness of t-NGS as a comprehensive and invaluable diagnostic tool by means of achieving a correct diagnosis and proceeding with careful management of these patients.80,81 Indeed, t-NGS approaches will be increasingly useful to accelerate the analysis, reduce costs and provide a clear diagnosis. One of the most important aspects of the use of t-NGS gene panels in clinical practice is their ability to be easily upgradable in view of novel discoveries. Despite their wide use in clinical practice, the major drawback of current NGS applications is represented by the data prohaematologica | 2016; 101(11)

Update of the red cell membrane disorders

cessing steps, mainly by the difficulty in determining the pathogenicity of the numerous identified variants. One of the ways to overcome this limitation is the simultaneous evaluation of all family members, allowing one to establish the inheritance pattern of the identified variants and thus to understand its pathogenetic role, although functional characterizations are often necessary. Finally, a future scenario in this field will be the implementation of novel therapeutic molecules. This is the case of Gardos channel blockers, such as senicapoc (ICA17043), and PIEZO1 channel inhibitors for the treatment of primary and secondary disorders of erythrocyte hydra-

References 1. Lux SE. Dissecting the red cell membrane skeleton. Nature. 1979;281(5731):426-429. 2. Mohandas N, Gallagher PG. Red cell membrane: past, present, and future. Blood. 2008;112(10):3939-3948. 3. Lux SEt. Anatomy of the red cell membrane skeleton: unanswered questions. Blood. 2016;127(2):187-199. 4. Pesciotta EN, Sriswasdi S, Tang HY, Mason PJ, Bessler M, Speicher DW. A label-free proteome analysis strategy for identifying quantitative changes in erythrocyte membranes induced by red cell disorders. J Proteomics. 2012;76 Spec No.:194-202. 5. Liu SC, Derick LH, Palek J. Visualization of the hexagonal lattice in the erythrocyte membrane skeleton. J Cell Biol. 1987;104 (3):527-536. 6. Chen K, Liu J, Heck S, Chasis JA, An X, Mohandas N. Resolving the distinct stages in erythroid differentiation based on dynamic changes in membrane protein expression during erythropoiesis. Proc Natl Acad Sci U S A. 2009;106(41):17413-17418. 7. Lazarides E, Woods C. Biogenesis of the red blood cell membrane-skeleton and the control of erythroid morphogenesis. Annu Rev Cell Biol. 1989;5:427-452. 8. Bogardus HH, Maksimova YD, Forget BG, Gallagher PG. A de novo band 3 mutation in hereditary spherocytosis. Pediatr Blood Cancer. 2012;58(6):1004. 9. Jamwal M, Aggarwal A, Kumar V, et al. Disease-modifying influences of coexistent G6PD-deficiency, Gilbert syndrome and deletional alpha thalassemia in hereditary spherocytosis: A report of three cases. Clin Chim Acta. 2016;458:51-54. 10. Gallagher PG. Abnormalities of the erythrocyte membrane.pediatr Clin North Am. 2013;60(6):1349-1362. 11. Lee JH, Moon KR. Coexistence of gilbert syndrome and hereditary spherocytosis in a child presenting with extreme jaundice. Pediatr Gastroenterol Hepatol Nutr. 2014;17(4):266-269. 12. Berne JD, Asensio JA, Falabella A, Gomez H. Traumatic rupture of the spleen in a patient with hereditary spherocytosis. J Trauma. 1997;42(2):323-326. 13. Jones L, Refai Z, Linney M. An adolescent with hereditary spherocytosis who presented with splenic infarction. BMJ Case Rep. 2015;2015. 14. Suzuki Y, Shichishima T, Mukae M, et al. Splenic infarction after Epstein-Barr virus infection in a patient with hereditary spherocytosis. Int J Hematol. 2007;85(5):380-383.

haematologica | 2016; 101(11)

tion. Indeed, a clinical trial with senicapoc has already been established for the treatment of disorders of secondary erythrocyte hydration, such as sickle cell disease, demonstrating increased Hb and reduced markers of hemolysis in treated patients.82,83 Acknowledgments This manuscript was partially supported by grants from the Italian Ministero dellâ&#x20AC;&#x2122;UniversitĂ e della Ricerca, by grants MUR-PS 35-126/Ind, by grants from Regione Campania (DGRC2362/07), by EU Contract LSHM-CT-2006-037296, by PRIN to AI (20128PNX83), by SIR to RR (RBSI144KXC).

15. Perrotta S, Gallagher PG, Mohandas N. Hereditary spherocytosis. Lancet. 2008;372(9647):1411-1426. 16. Da Costa L, Galimand J, Fenneteau O, Mohandas N. Hereditary spherocytosis, elliptocytosis, and other red cell membrane disorders. Blood reviews. 2013;27(4):167-178. 17. Bolton-Maggs PH, Langer JC, Iolascon A, Tittensor P, King MJ, General Haematology Task Force of the British Committee for Standards in H. Guidelines for the diagnosis and management of hereditary spherocytosis--2011 update. Br J Haematol. 2012;156 (1):37-49. 18. King MJ, Telfer P, MacKinnon H, et al. Using the eosin-5-maleimide binding test in the differential diagnosis of hereditary spherocytosis and hereditary pyropoikilocytosis. Cytometry B Clin Cytom. 2008;74(4):244250. 19. Bianchi P, Fermo E, Vercellati C, et al. Diagnostic power of laboratory tests for hereditary spherocytosis: a comparison study in 150 patients grouped according to molecular and clinical characteristics. Haematologica. 2012;97(4):516-523. 20. Okamoto N, Wada Y, Nakamura Y, et al. Hereditary spherocytic anemia with deletion of the short arm of chromosome 8. Am J Med Genet. 1995;58(3):225-229. 21. Miya K, Shimojima K, Sugawara M, et al. A de novo interstitial deletion of 8p11.2 including ANK1 identified in a patient with spherocytosis, psychomotor developmental delay, and distinctive facial features. Gene. 2012;506(1):146-149. 22. Griswold AJ, Ma D, Sacharow SJ, et al. A de novo 1.5 Mb microdeletion on chromosome 14q23.2-23.3 in a patient with autism and spherocytosis. Autism Res. 2011;4(3):221227. 23. Lybaek H, Oyen N, Fauske L, Houge G. A 2.1 Mb deletion adjacent but distal to a 14q21q23 paracentric inversion in a family with spherocytosis and severe learning difficulties. Clin Genet. 2008;74(6):553-559. 24. Ribeiro ML, Alloisio N, Almeida H, et al. Severe hereditary spherocytosis and distal renal tubular acidosis associated with the total absence of band 3. Blood. 2000;96(4):1602-1604. 25. Whitfield CF, Follweiler JB, LoprestiMorrow L, Miller BA. Deficiency of alphaspectrin synthesis in burst-forming unitserythroid in lethal hereditary spherocytosis. Blood. 1991;78(11):3043-3051. 26. Gallagher PG, Weed SA, Tse WT, et al. Recurrent fatal hydrops fetalis associated with a nucleotide substitution in the erythrocyte beta-spectrin gene. J Clin Invest. 1995;95(3):1174-1182.

27. Iolascon A, Perrotta S, Stewart GW. Red blood cell membrane defects. Rev Clin Exp Hematol. 2003;7(1):22-56. 28. Picard V, Proust A, Eveillard M, et al. Homozygous Southeast Asian ovalocytosis is a severe dyserythropoietic anemia associated with distal renal tubular acidosis. Blood. 2014;123(12):1963-1965. 29. Packman CH. The Clinical Pictures of Autoimmune Hemolytic Anemia. Transfus Med Hemother. 2015;42(5):317-324. 30. Lupo F, Russo R, Iolascon A, et al. Protease inhibitors-based therapy induces acquired spherocytic-like anaemia and ineffective erythropoiesis in chronic hepatitis C virus patients. Liver Int. 2016;36(1):49-58. 31. De Franceschi L, Bosman GJ, Mohandas N. Abnormal red cell features associated with hereditary neurodegenerative disorders: the neuroacanthocytosis syndromes. Curr Opin Hematol. 2014;21(3):201-209. 32. Delaunay J. The molecular basis of hereditary red cell membrane disorders. Blood Rev. 2007;21(1):1-20. 33. Grootenboer S, Schischmanoff PO, Cynober T, et al. A genetic syndrome associating dehydrated hereditary stomatocytosis, pseudohyperkalaemia and perinatal oedema. Br J Haematol. 1998;103(2):383-386. 34. Delaunay J. The hereditary stomatocytoses: genetic disorders of the red cell membrane permeability to monovalent cations. Semin Hematol. 2004;41(2):165-172. 35. Syfuss PY, Ciupea A, Brahimi S, et al. Mild dehydrated hereditary stomatocytosis revealed by marked hepatosiderosis. Clin Lab Haematol. 2006;28(4):270-274. 36. Carella M, Stewart G, Ajetunmobi JF, et al. Genomewide search for dehydrated hereditary stomatocytosis (hereditary xerocytosis): mapping of locus to chromosome 16 (16q23qter). Am J Hum Genet. 1998;63(3):810-816. 37. Grootenboer S, Schischmanoff PO, Laurendeau I, et al. Pleiotropic syndrome of dehydrated hereditary stomatocytosis, pseudohyperkalemia, and perinatal edema maps to 16q23-q24. Blood. 2000;96(7):25992605. 38. Zarychanski R, Schulz VP, Houston BL, et al. Mutations in the mechanotransduction protein PIEZO1 are associated with hereditary xerocytosis. Blood. 2012;120(9):1908-1915. 39. Andolfo I, Alper SL, De Franceschi L, et al. Multiple clinical forms of dehydrated hereditary stomatocytosis arise from mutations in PIEZO1. Blood. 2013;121(19):3925-3935, S112. 40. Coste B, Mathur J, Schmidt M, et al. Piezo1 and Piezo2 are essential components of distinct mechanically activated cation channels. Science. 2010;330(6000):55-60.

1293

I. Andolfo et al. 41. Coste B, Xiao B, Santos JS, et al. Piezo proteins are pore-forming subunits of mechanically activated channels. Nature. 2012;483(7388):176-181. 42. Ge J, Li W, Zhao Q, et al. Architecture of the mammalian mechanosensitive Piezo1 channel. Nature. 2015;527(7576):64-69. 43. Cinar E, Zhou S, DeCourcey J, Wang Y, Waugh RE, Wan J. Piezo1 regulates mechanotransductive release of ATP from human RBCs. Proc Natl Acad Sci USA. 2015; 112(38):11783-11788. 44. Coste B, Murthy SE, Mathur J, et al. Piezo1 ion channel pore properties are dictated by C-terminal region. Nat Commun. 2015; 6:7223. 45. Bae C, Gnanasambandam R, Nicolai C, Sachs F, Gottlieb PA. Xerocytosis is caused by mutations that alter the kinetics of the mechanosensitive channel PIEZO1. Proc Natl Acad Sci USA. 2013;110(12):E1162-1168. 46. Rapetti-Mauss R, Lacoste C, Picard V, et al. A mutation in the Gardos channel is associated with hereditary xerocytosis. Blood. 2015;126(11):1273-1280. 47. Andolfo I, Russo R, Manna F, et al. Novel Gardos channel mutations linked to dehydrated hereditary stomatocytosis (xerocytosis). Am J Hematol. 2015;90(10):921-926. 48. Glogowska E, Lezon-Geyda K, Maksimova Y, Schulz VP, Gallagher PG. Mutations in the Gardos channel (KCNN4) are associated with hereditary xerocytosis. Blood. 2015;126(11):1281-1284. 49. Iolascon A, De Falco L, Borgese F, et al. A novel erythroid anion exchange variant (Gly796Arg) of hereditary stomatocytosis associated with dyserythropoiesis. Haematologica. 2009;94(8):1049-1059. 50. Genetet S, Ripoche P, Picot J, et al. Human RhAG ammonia channel is impaired by the Phe65Ser mutation in overhydrated stomatocytic red cells. Am J Physiol Cell Physiol. 2012;302(2):C419-428. 51. Fricke B, Argent AC, Chetty MC, et al. The "stomatin" gene and protein in overhydrated hereditary stomatocytosis. Blood. 2003;102 (6):2268-2277. 52. Fricke B, Jarvis HG, Reid CD, et al. Four new cases of stomatin-deficient hereditary stomatocytosis syndrome: association of the stomatin-deficient cryohydrocytosis variant with neurological dysfunction. Br J Haematol. 2004;125(6):796-803. 53. Flatt JF, Guizouarn H, Burton NM, et al. Stomatin-deficient cryohydrocytosis results from mutations in SLC2A1: a novel form of GLUT1 deficiency syndrome. Blood. 2011;118(19):5267-5277. 54. Stewart GW, Corrall RJ, Fyffe JA, Stockdill G, Strong JA. Familial pseudohyperkalaemia. A new syndrome. Lancet. 1979;2(8135):175-177. 55. Carella M, d'Adamo AP, GrootenboerMignot S, et al. A second locus mapping to 2q35-36 for familial pseudohyperkalaemia. Eur J Hum Genet. 2004;12(12):1073-1076. 56. Andolfo I, Alper SL, Delaunay J, et al. Missense mutations in the ABCB6 transporter cause dominant familial pseudohyperkalemia. Am J Hematol. 2013;88(1): 66-72. 57. Krishnamurthy PC, Du G, Fukuda Y, et al. Identification of a mammalian mitochondrial porphyrin transporter. Nature. 2006;443(7111):586-589. 58. Helias V, Saison C, Ballif BA, et al. ABCB6 is dispensable for erythropoiesis and specifies the new blood group system Langereis.

1294

Nature genetics. 2012;44(2):170-173. 59. Kiss K, Brozik A, Kucsma N, et al. Shifting the paradigm: the putative mitochondrial protein ABCB6 resides in the lysosomes of cells and in the plasma membrane of erythrocytes. PloS one. 2012;7(5):e37378. 60. Andolfo I, Russo R, Manna F, et al. Functional characterization of novel ABCB6 mutations and their clinical implications in familial pseudohyperkalemia. Haematologica. 2016;101(8):909-917. 61. Bawazir WM, Flatt JF, Wallis JP, et al. Familial pseudohyperkalemia in blood donors: a novel mutation with implications for transfusion practice. Transfusion. 2014;54(12): 3043-3050. 62. Gambale A, Iolascon A, Andolfo I, Russo R. Diagnosis and management of congenital dyserythropoietic anemias. Expert Rev Hematol. 2016;9(3):283-296. 63. Danise P, Amendola G, Nobili B, et al. Flowcytometric analysis of erythrocytes and reticulocytes in congenital dyserythropoietic anaemia type II (CDA II): value in differential diagnosis with hereditary spherocytosis. Clin Lab Haematol. 2001;23(1):7-13. 64. Russo R, Gambale A, Langella C, Andolfo I, Unal S, Iolascon A. Retrospective cohort study of 205 cases with congenital dyserythropoietic anemia type II: definition of clinical and molecular spectrum and identification of new diagnostic scores. Am J Hematol. 2014;89(10):E169-175. 65. Tchernia G, Delhommeau F, Perrotta S, et al. Recombinant erythropoietin therapy as an alternative to blood transfusions in infants with hereditary spherocytosis. Hematol J. 2000;1(3):146-152. 66. O'Neal HR, Jr., Niven AS, Karam GH. Critical Illness in Patients with Asplenia. Chest. 2016 Apr 8. Epub ahead of print. 67. Rogulski R, Adamowicz-Salach A, Matysiak M, et al. Laparoscopic splenectomy for hereditary spherocytosis-preliminary report. Eur J Haematol. 2016;96(6):637-642. 68. Pincez T, Guitton C, Gauthier F, et al. Longterm follow-up of subtotal splenectomy for hereditary spherocytosis: a single-center study. Blood. 2016;127(12):1616-1618. 69. Medejel N, Garcon L, Guitton C, Cynober T, Bader-Meunier B. Effect of subtotal splenectomy for management of hereditary pyropoikilocytosis. Br J Haematol. 2008;142(2):315-317. 70. Stewart GW, Amess JA, Eber SW, et al. Thrombo-embolic disease after splenectomy for hereditary stomatocytosis. Br J Haematol. 1996;93(2):303-310. 71. Jais X, Till SJ, Cynober T, et al. An extreme consequence of splenectomy in dehydrated hereditary stomatocytosis: gradual thrombo-embolic pulmonary hypertension and lung-heart transplantation. Hemoglobin. 2003;27(3):139-147. 72. Cahalan SM, Lukacs V, Ranade SS, Chien S, Bandell M, Patapoutian A. Piezo1 links mechanical forces to red blood cell volume. eLife. 2015;22;4. 73. Kulkeaw K, Sugiyama D. Zebrafish erythropoiesis and the utility of fish as models of anemia. Stem Cell Res Ther. 2012;3(6):55. 74. Faucherre A, Kissa K, Nargeot J, Mangoni ME, Jopling C. Piezo1 plays a role in erythrocyte volume homeostasis. Haematologica. 2014;99(1):70-75. 75. Shmukler BE, Huston NC, Thon JN, et al. Homozygous knockout of the piezo1 gene in the zebrafish is not associated with anemia. Haematologica. 2015;100(12):e483-485.

76. Shmukler BE, Lawson ND, Paw BH, Alper SL. Authors' response to "Comment on: 'Homozygous knockout of the piezo1 gene in the zebrafish is not associated with anemia'". Haematologica. 2016;101(1):e39. 77. Faucherre A, Kissa K, Nargeot J, Mangoni ME, Jopling C. Comment on: 'Homozygous knockout of the piezo1 gene in the zebrafish is not associated with anemia'. Haematologica. 2016;101(1):e38. 78. Lukacs V, Mathur J, Mao R, et al. Impaired PIEZO1 function in patients with a novel autosomal recessive congenital lymphatic dysplasia. Nat Commun. 2015;6:8329. 79. Fotiou E, Martin-Almedina S, Simpson MA, et al. Novel mutations in PIEZO1 cause an autosomal recessive generalized lymphatic dysplasia with non-immune hydrops fetalis. Nature Commun. 2015;6:8085. 80. Sangle N, Christensen RD, Salama ME, Prchal J, Yaish H, Agarwal AM. The Clinical Utility Of Next-Generation Sequencing In The Diagnosis Of Hereditary Hemolytic Anemias. Blood. 2013;122(21):3421. 81. Russo R AI, Manna F, Gambale A, Pignataro P, De Rosa G, Iolascon A. RedPlex: a targeted next generation sequencing-based diagnosis for patients with hereditary hemolytic anemias. Haematologica. 2016;101(s1):1. 82. Nagalla S, Ballas SK. Drugs for preventing red blood cell dehydration in people with sickle cell disease. Cochrane Database Syst Rev. 2016;3:CD003426. 83. Ataga KI, Reid M, Ballas SK, et al. Improvements in haemolysis and indicators of erythrocyte survival do not correlate with acute vaso-occlusive crises in patients with sickle cell disease: a phase III randomized, placebo-controlled, double-blind study of the Gardos channel blocker senicapoc (ICA17043). Br J Haematol. 2011;153(1):92-104. 84. Bloom ML, Kaysser TM, Birkenmeier CS, Barker JE. The murine mutation jaundiced is caused by replacement of an arginine with a stop codon in the mRNA encoding the ninth repeat of beta-spectrin. Proc Natl Acad Sci U S A. 1994;91(21):10099-10103. 85. Wandersee NJ, Birkenmeier CS, Gifford EJ, Mohandas N, Barker JE. Murine recessive hereditary spherocytosis, sph/sph, is caused by a mutation in the erythroid alpha-spectrin gene. Hematol J. 2000;1(4):235-242. 86. Heruth DP, Hawkins T, Logsdon DP, et al. Mutation in erythroid specific transcription factor KLF1 causes Hereditary Spherocytosis in the Nan hemolytic anemia mouse model. Genomics. 2010;96(5):303-307. 87. Hughes MR, Anderson N, Maltby S, et al. A novel ENU-generated truncation mutation lacking the spectrin-binding and C-terminal regulatory domains of Ank1 models severe hemolytic hereditary spherocytosis. Exp Hematol. 2011;39(3):305-320. 88. Peters LL, Swearingen RA, Andersen SG, et al. Identification of quantitative trait loci that modify the severity of hereditary spherocytosis in wan, a new mouse model of band-3 deficiency. Blood. 2004;103(8):3233-3240. 89. Wandersee NJ, Roesch AN, Hamblen NR, et al. Defective spectrin integrity and neonatal thrombosis in the first mouse model for severe hereditary elliptocytosis. Blood. 2001;97(2):543-550. 90. Shafizadeh E, Paw BH, Foott H, et al. Characterization of zebrafish merlot/chablis as non-mammalian vertebrate models for severe congenital anemia due to protein 4.1 deficiency. Development. 2002;129(18): 4359-4370.

haematologica | 2016; 101(11)

REVIEW ARTICLE

Osteonecrosis in children with acute lymphoblastic leukemia

Marina Kunstreich,1 Sebastian Kummer,2 Hans-Juergen Laws,1 Arndt Borkhardt,1 and Michaela Kuhlen1

University of Duesseldorf, Medical Faculty, Department of Pediatric Oncology, Hematology, and Clinical Immunology, Center for Child and Adolescent Health; and University of Duesseldorf, Medical Faculty, Department of General Pediatrics, Neonatology and Pediatric Cardiology, Center for Child and Adolescent Health, Germany

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

1 2

Haematologica 2016 Volume 101(11):1295-1305

ABSTRACT

he morbidity and toxicity associated with current intensive treatment protocols for acute lymphoblastic leukemia in childhood become even more important as the vast majority of children can be cured and become long-term survivors. Osteonecrosis is one of the most common therapy-related and debilitating side effects of antileukemic treatment and can adversely affect long-term quality of life. Incidence and risk factors vary substantially between study groups and therapeutic regimens. We therefore analyzed 22 clinical trials of childhood acute lymphoblastic leukemia in terms of osteonecrosis incidence and risk factors. Adolescent age is the most significant risk factor, with patients >10 years old at the highest risk. Uncritical modification or even significant reduction of glucocorticoid dosage cannot be recommended at this stage. A novel and innovative approach to reduce osteonecrosisassociated morbidity might be systematic early screening for osteonecrosis by serial magnetic resonance images. However, discriminating patients at risk of functional impairment and debilitating progressive joint disease from asymptomatic patients still remains challenging.

Correspondence: michaela.kuhlen@med.uni-duesseldorf.de

Background Survival of children with acute lymphoblastic leukemia (ALL) has dramatically improved over the last decades due to the progressive intensification of multi-agent chemotherapy. Currently, more than 90% of children and adolescents can be cured and become long-term survivors.1,2 Thus, the long-term adverse effects of treatment become increasingly important. Osteonecrosis is one of the most common and debilitating therapy-related side effects of anti-leukemic treatment and can adversely affect long-term quality of life.3 Incidence (1.6â&#x20AC;&#x201C;17.6%) and risk factors for the development of osteonecrosis have been investigated in many studies, but results vary substantially between study groups and therapeutic regimens.4-9 Adolescence is the most consistently identified and most significant risk factor, with patients >10 years old at the highest risk.7-11 As this dominates all other therapy-related and patient-specific risk factors, it suggests that the underlying pathophysiology for the development of osteonecrosis likely has to be attributed to agespecific factors ultimately affecting bone morphology, metabolism, and/or nourishment. This may be due, at least in part, to increased end-organ susceptibility caused by a markedly increased growth rate and specific hormonal changes in this period of life.12

Current concepts of osteonecrosis pathogenesis The early events leading to osteonecrosis are poorly understood. Multiple factors for the development of osteonecrosis are discussed, which probably act synergistically in the context of anti-leukemic treatment. All contributing mechanisms finally lead to an imbalance between the actual and the required bone perfusion, which haematologica | 2016; 101(11)

Received: May 4, 2016. Accepted: June 23, 2016. Pre-published: October 14, 2016. doi:10.3324/haematol.2016.147595

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1295

1295

M. Kunstreich et al.

may be related to intravascular clotting/embolism (intraluminal obliteration), increased marrow pressure (extraluminal obliteration), and direct blood vessel injury. In addition, the direct toxic effects of chemotherapy on bone marrow and bone cells may disturb bone integrity and contribute to osteonecrosis .13 Although the underlying disease and the exposure to damaging agents, such as glucocorticoids (GCs), are of a systemic nature, osteonecrosis predominantly develops in vulnerable areas such as long bone epiphysis and metaphysis (Table 1).

increases and can be passed through to the epiphyseal part of the bone.

Direct blood vessel injury Disruption of the vascular supply to the bone is a preceding event to glucocorticoid-induced osteonecrosis in a murine model.20 This is mainly mediated by damaging effects on the endothelial and smooth muscle cells of nutrient arteries and venous vessels, which promote further vascular stasis, ischemia, and arteriopathy.21

Altered integrity of bone structure Disrupted blood supply to the bone Bone is a highly perfused tissue. The blood supply to the endosteal cavity is delivered by the nutrient artery, which enters through the diaphysis and branches into marrow sinusoids, and finally ramifies into small vessels in the cortex. The epiphyseal and metaphyseal vascular zones of prepubertal children are separated by the growth plate, which receives its blood supply only from dia- and epiphyseal vessels and anastomoses in the perichondrium, respectively (Figure 1). Neural, humoral, and hormonal factors contribute to the regulation of vascular resistance, and, thus, influence the blood supply to the bone.

Longitudinal bone growth occurs by endochondral bone formation, particularly in the growth plates at the proximal and distal ends of long bones, whereas bone growth in width occurs by bone modeling. During remodeling, the bone tissue is continuously turned over. Both osteoclasts and osteoblasts are fundamentally involved in this process and influence bone development during childhood and adolescence.22,23 During the pubertal growth spurt, particularly, the bone length increases. Furthermore, sexual hormones impact bone (re)modeling and, thus, affect bone strength and mass.22

Direct cell toxicity Intraluminal obliteration Liver-to-bone marrow lipid emboli trigger thrombotic and/or embolic ischemia, resulting in cell damage and subsequent bone marrow edema (BME). This leads to ischemic necrosis of metabolically active/vulnerable regions such as the epiphyses.14,15 By triggering intravascular coagulation in the intraosseous microcirculation (capillaries and venous sinusoids), increased prothrombotic factors (e.g., thrombin, cholesterol) contribute to the development of osteonecrosis.16

Extraluminal obliteration Intramedullary lipocyte proliferation (compromising the sinusoidal circulation) and osteocyte lipid hypertrophy (e.g., related to GCs or dyslipidemia), proliferation of histiocytes in storage disorders (e.g., Gaucher disease), or bleeding within the bone marrow cause increased intramedullary pressure. Because of the inelasticity of the bone, intraosseous compartment syndrome develops, further reducing intramedullary blood flow and predisposing for hemostasis in the intraosseous blood vessels.17-19 The epiphyseal plate of the immature bone during childhood growth provides elasticity to compensate for the increasing intraosseous pressure, while with epiphyseal closure during adolescence, the intramedullary pressure Table 1. Distribution pattern of osteonecrosis in children and adolescents with ALL (acute lymphoblastic leukemia) according to published data.

Joints affected

Shoulder Elbow Hip Knee Ankle Multiple joints

13-24 3-15 35-67 45-88 13-44 29-90

1296

References 7, 39, 43, 46, 62 7, 39, 43, 46, 62 3, 7, 8, 10, 33, 34, 39, 43, 46, 62 3, 7, 8, 10, 33, 39, 43, 46, 62 7, 8, 10, 34, 39, 43, 46, 62 3, 7, 8, 10, 32, 34, 39, 41, 42, 46, 62

GCs are reported to induce gradual lipid accumulation within osteocytes, osteocyte death, increased osteocyte apoptosis, suppression of osteoblastic differentiation of marrow stem cells, decreased cell division of osteoblasts near osteonecrosis lesions, and increased mesenchymal stem cell differentiation into lipocytes at the expense of osteogenesis.14,20,24-29

Defective bone repair During revascularization following ischemia, changes occur in the hematopoietic marrow, fatty marrow, and vascular structures. The surrounding bony architecture within the area of infarction becomes weakened by resorption of subchondral dead bone along the reactive interface. The repair process at least temporarily compromises bone mass integrity. Continued cellular stress, mechanical load/weight-bearing stress fractures, collapse of the chondral bony support system, cartilage disintegration, and deformity of articular surfaces may ultimately lead to progressive joint collapse and degenerative joint disease.30,31

Osteonecrosis in the context of anti-leukemic treatment Osteonecrosis has only recently been recognized as one of the most significant toxicities of anti-leukemic treatment (see Tables 2 and 3). This is in stark contrast to historical experience in which osteonecrosis was considered a rare complication of ALL therapy. In 2000, Mattano et al. reported on a large retrospective multi-center survey on symptomatic osteonecrosis in children with high-risk ALL treated according to the CCG-1882 protocol between 1989 and 1995.7 With a cumulative osteonecrosis incidence of 9.3% and orthopedic interventions in 24% of the affected children, this report highlighted, for the first time, osteonecrosis as a serious problem of modern chemotherapy. A trend to better outcome after occurhaematologica | 2016; 101(11)

Osteonecrosis in children with ALL

rence of osteonecrosis further emphasized the challenge of treating these children with therapy that maximizes cure rates but is associated with unanticipated and – to a certain extent – unacceptable toxicity. Notably, this was chronologically associated with the introduction of dexamethasone for delayed intensification with improved survival rates, particularly in the most affected group of adolescents. A retrospective study on two consecutive DFCI trials reported a slightly lower osteonecrosis incidence (7%) but an even higher rate (30%) of orthopedic interventions,32 the former speculatively owed to the fact that dexamethasone was only given in DFCI trial 91-01. A magnetic resonance imaging (MRI) screening based prospective study determined a significantly higher osteonecrosis incidence (15.5%) without the impact of steroid dose or dexamethasone administration (see Table 4).33 This was even exceeded by a prospective study analyzing the Nordic ALL protocols, which reported an osteonecrosis incidence as high as 24%, identified by MRI screening at the end of treatment.8 As the earlier reports were based on retrospective data collection, the true

osteonecrosis incidence was most likely underestimated. However, 6 of the 17 affected patients reported by Ribeiro et al.33 and 16 of the 23 patients reported by Niinimäki et al.8 were only detected by MRI, and the patients remained asymptomatic until the end of the study. Two independent retrospective reports on BerlinFrankfurt-Muenster (BFM)-based trials with quite similar therapy (AIEOP-ALL 95,34 ALL-BFM 9510) from the late 90s determined a much lower, but almost identical, osteonecrosis incidence of 1.6–1.8%. However, in patients aged ≥10 years, the osteonecrosis incidence was reported to be 8.9% and even higher in those ≥15 years (16.7%).10 Thus, when comparing these studies, one has to keep in mind that appropriate age groups must be compared, and that there might be a significant difference in the patients’ age distribution in each study, which certainly influences the overall incidence of osteonecrosis. In line with this, Mattano et al.7 only evaluated high-risk patients, but young patients usually make up only about one third of the high-risk group. However, the osteonecrosis incidence in these retrospective studies was

Figure 1. Schematic illustration of osteonecrosis pathogenesis. (A) Blood supply of bone is delivered by the nutrient arteries, which enter in the dia- and epiphysis and branch into marrow sinusoids. The growth plate receives its blood supply from dia- and epiphyseal vessels and anastomoses in the perichondrium, respectively. (B-C) Bone growth occurs by endochondral bone formation and bone modeling. The bone tissue is continuously turning over, with osteoclasts and osteoblasts being fundamentally involved in this process. In the context of osteonecrosis development, osteoblast differentiation from mesenchymal progenitor cells is disturbed by gradual lipid accumulation within osteoblasts and osteocytes and increased cell death, both mainly induced by GCs (glucocorticosteroids), and results in defective bone repair. (D-E) Bone perfusion is disturbed by intraluminal obliteration induced by lipid emboli and intravascular coagulation and extraluminal obliteration induced by intramedullary lipocyte proliferation and lipid hypertrophy. Intraosseous compartment syndrome may develop, increasingly reducing intramedullary blood flow and predisposing for coagulation in the intraosseous blood vessels. (F-G) Direct blood vessel injury and disruption of the vascular supply to the bone is mainly mediated by damaging effects on the endothelial and smooth muscle cells of nutrient arteries and venous vessels. HSC: hematopoietic stem cell; MP: mesenchymal progenitor cell.

haematologica | 2016; 101(11)

1297

M. Kunstreich et al.

probably underestimated as this toxicity was unanticipated, and therefore not listed as a reportable event on the case report forms. Furthermore, the treating physicians in those days were not aware of this toxicity, and a standardized diagnostic approach was lacking. However, if one assumes that osteonecrosis was underreported in these trials, and exposure to dexamethasone increases the risk of osteonecrosis, particularly during delayed intensification, one would expect a much higher incidence of osteonecrosis in the subsequent trial ALL-BFM 2000. Notwithstanding that the overall incidence of osteonecrosis was substantially higher (4.7%)35 and exceeded that reported in the trial NOPHO ALL-2008 (3.1%),36 and

EORTC-CLG 58951 (2.5%),37 it still remained lower than that of CCG,38,39 DFCI,40,41 DCOG,42,43 and UK44,45 trials. Even when comparing only the group of older patients, the incidence (in prospective studies on symptomatic osteonecrosis) is much higher but still varies substantially between the trials (CCG-196139 9.9% 10-15 years, 20% 16-21 years; UKALL 200345 16% 10-15 years, 15% >16 years; NOPHO ALL200836 11% 10-14 years, 6.5% 15-17 years; ALL-BFM 200035 14.5% 10-15 years; DFCI-ALL 000141 14% 10-18 years). However, a factor which remained consistent throughout all the studies was that older children and adolescents are at a much higher risk of developing osteonecrosis.

Table 2. Overview of retrospective studies reporting incidences and risk factors for symptomatic osteonecrosis in children and adolescents with ALL.

Protocol & Study cohort recruiting period Data source & no. of participating centers

No of patients (with ON/ALL)

Incidence

CCG-1882 05/89-06/95

High-risk ALL 111/1,409 9.3% CI mostly confirmed • 1-9 y with WBC ≥50x109/L 27 pts with by radiographic imaging • ≥10 y orthop. interv. Survey / 56 centers

DFCI 87-01 DFCI 91-01 11/87-12/95 AIEOP-ALL 95 05/95-12/99

ALL • 0-18 y Records / Single center Non B-ALL • <18 y Data recall / Multicenter

13/176 7% CI confirmed 4 pts with by radiographic imaging orthop. interv. 15/1421 1.6% CI

ALL-BFM 95 01/96-06/00

ALL • 0-18 y Questionnaire / Multicenter

31/1951 13 pts with orthop. interv.

1.8% CI

UKALL97 01/97-12/07

ALL Records / Single center

18/186

9.7% CI

CoALL-07-03 09/03-12/09

ALL 22/124 25% CI confirmed by MRI • 1-18 y 8 pts with Records / Single center orthop. interv. ALL 153/3.126 5% • 1-24 y 30 pts with Records & questionnaire / orthop. interv. Multicenter ALL, LBL 18/251 7% CI confirmed by MRI • Children & adolescents Records / Single center ALL 65/730 8.9% • 1-18 y Cohort study

UKALL 2003 10/03-06/11

ANZCHOG 8 2002-11 DFCI 05-001 2005-11

Risk factors

Reference

• 0.9% <10 y vs. 13.5% Mattano et al.7 10-15 y vs. 18% 16-20 y (S) • 12.2% females vs. 7.7% males (S) • 10-15 y: 19.2% females vs. 9.8% males (S) • 16-20 y: 13.2% females vs. 20.7% males • ≥10 y: 16.7% whites vs. 3.3% blacks (S) • Slight trend to better outcome after occurrence of ON • 4% <9 y vs. 21% 9-18 y (S) Strauss et al.32 • 9% DEX vs. 6% PRED (NS) • Sex, risk group, WBC (NS) • 0.3% 0-5 y vs. 0.7% 6-9 y Arico et al.34 vs. 7.4% 10-17 y (S) • 2.5% female vs. 0.7% male (S) • 2.4% SR vs. 1.0% IR vs. 5.8% HR (S) • Highest risk: Females aged 10-17 y (S) • 0.2% <10 y vs. 8.9% ≥10 y (S) Burger et al.10 • 1.3% <15 y vs. 16.7% ≥15 y (S) • 0.2% SR vs. 2.7% MR (S) • 2.7% MR vs. 3.5% HR (NS) • 2.4% female vs. 1.4% male (NS) • Age >9 y (S) Elmantaser et al.44 • 9% female vs. 10% male (NS) • 11% DEX vs. 3.5% PRED • 13.4% <10 y vs. 52.3% ≥10 y (S) Kuhlen et al.46 • 16.1% females vs. 36.2% males (NS) • 8.3% LR vs. 39.7% HR (S) • 0.7% <10 y vs. 16% 10-15 y Amin et al. (abstr.)45 vs. 15% >16 y (S) • Ethnic group, sex, number of delayed intensifications (NS) • 29% >10 y (S) Padhye et al.62 • 3.4% SR vs. 7.5% MR vs. 13.8% VHR (NS) • 5.2% males vs. 11.2% females (NS) • 3.3% in Hispanic vs. 10.3% Kahn et al.(abstr.)40 in non-Hispanic (S)

Year

2000

2001

2003

2005

2010

2014

2015

2016

2015

PRED: prednisolone; DEX: dexamethasone; CI: cumulative incidence; S: significant: NS: not significant; y: years; Ind: Induction; Intensif: intensification; Maint: maintenance; Cont: continuation; Reind: reinduction; Postrem: postremission; Reintensif: reintensification; Cons: consolidation. ALL : acute lymphoblastic leukemia; ON: osteonecrosis; WBC : white blood count; MRI : magnetic resonance imaging; B-ALL: B cell acute lymphoblastic leukemia; LBL: lymphoblastic lymphoma; pts: patients; orthop. interv.: orthopedic interventions; HR: high-risk; SR: standard-risk; MR: medium risk; IR: intermediate-risk; VHR: very high-risk; LR: low-risk.

1298

haematologica | 2016; 101(11)

Osteonecrosis in children with ALL

It may be speculated that these differences in osteonecrosis incidence may be due to reporting bias, incompleteness of data, and different methods of analysis, and might further be substantially influenced by treatment related or non-treatment related risk factors.

Risk factors As osteonecrosis seems to be a particularly predominant problem in children and adolescents diagnosed with acute lymphoblastic leukemia, leukemia itself might contribute to the development of osteonecrosis. Lymphoblasts are known to have bone-resorbing effects. However, neither areas of leukemic infiltration of bone26 nor white blood count at diagnosis32,33,46 and immunophenotype (T- versus B-cell leukemia)38,46 are associated with osteonecrosis risk.

Treatment related risk factors Glucocorticoids GCs are major contributors to the development of osteonecrosis, with the cumulative dose of received GCs correlating with the risk of osteonecrosis (see Table 5).3,47 In study AALL0232,38 an excess risk of osteonecrosis was found in older patients with dexamethasone at 10 mg/m2/d x 14 days (24%) versus prednisone at 60 mg/m2/d x 28 days (16%). Most studies3,5,32,37,48 report no obviously increased risk of osteonecrosis with the administration of dexamethasone compared to prednisone, even in the risk group of older patients (for example in ALL-BFM 2000,35 the osteonecrosis incidence in patients treated with dexamethasone was 14% versus 19% with prednisone). To make the different trials immediately comparable, many authors calculated the equipotent anti-inflammatory doses

Table 3. Overview of prospective studies reporting incidences and risk factors for symptomatic osteonecrosis in children and adolescents with ALL.

Protocol & recruiting period

Study cohort No. of participating centers

ALL97 ALL97/99 04/97-06/02

ALL • 1-18 y Multicenter ALL Multicenter

DCOG-ALL9 01/97-11/04 CCG-1961 09/96-05/02

High-risk ALL • 1-21 y • WBC ≥50x109/L • ≥10 y Multicenter

DFCI-ALL 00-01 09/00-12/04

ALL • 1-18 y Multicenter ALL • <18 y Multicenter ALL • 4-18 y

EORTC-CLG 58951 12/98-08/08 DCOG-ALL9 01/97-11/04

COG-AALL043 01/07-07/14 ALL-BFM 2000 07/00-07/06

NOPHO ALL2008 06/09-ongoing

No of patients (with ON/ALL)

Incidence

15/1603

1% NCI grade 3 / 4

38/694 7 pts with orthop. interv. 143/2056 62 pts with orthop. interv.

6.1% CI confirmed by MRI 7.7% CI confirmed by MRI

23/408

49/1.947

2.5%

30/466

6.4% confirmed by MRI

T-ALL • 1-30 y Multicenter ALL • 1-18 y Multicenter

69/1,155

8% CI imaging confirmed

84/1,737

4.7% CI

ALL • 1-17 y Multicenter

29/934

3.1%

Risk factors

Reference

Year

Mitchell et al.48

2005

• Age (S) te Winkel et al.43 • Female sex (S) • NHR vs. HR (NS) • 1.0% 1-9 y vs. 9.9% 10-15 y Mattano et al.39 vs. 20.0% 16-21 y (S) • 15.7% female vs. 9.3% male aged 10-21 y (S) • 8.7% alternate-week vs. 17.0% continuous DEX aged ≥10 y during delayed-intensification (S) • >10 y: 5-EFS 85.8% with vs. 68.2% without ON, in males and females (S) • 3.5% <10 y vs. 14% 10-18 y (S) Vrooman et al.41 • 5% PRED vs. 23% DEX aged 10-18 y (S)

2011

Domenech et al.37

2014

• Older age, female sex (S) • DEX vs. PRED (NS)

• 2.5% DEX vs. 2.6% PRED (NS)

2012

2013

2015 • lower mean bone mineral density den Hoed et al.42 (BMD) of the lumbar spine (LS) and total body (TB) at cessation of treatment (S) • steeper BMDLS and BMDTB decline in pts with ON during follow-up (S) 2.6% 1-9 y vs. 14.6% ≥10 y CI (S) Mattano et al. (abstr.)38 2014 • • 5% females vs. 6% males • 0.5% 1-<6 y vs. 1.3% 6-<10 y Möricke et al.35 vs. 14.5% 10-<15 y vs. 22.7% 15-<18 y (S) • 1-<10 y: 0.8% DEX vs. 0.6% PRED; 10-<18 y: 13.8% DEX vs. 19.2% PRED (NS) • 1.5% 1-9 y vs. 11.0% 10-14 y vs. Toft et al.36 6.5% 15-17 y (S)

2016

PRED: prednisolone; DEX: dexamethasone; CI: cumulative incidence; S: significant, NS: not significant; y: years; Ind: Induction; Intensif: intensification; Maint: maintenance; Cont: continuation; Reind: reinduction; Postrem: postremission; Reintensif: reintensification; Cons: consolidation; ALL: acute lymphoblastic leukemia; ON: osteonecrosis; T-ALL: T cell acute lymphoblastic leukemia; WBC: white blood count; pts: patients; orthop. interv.: orthopedic interventions; MRI: magnetic resonance imaging; NCI: National Cancer Institute; NHR: non high-risk; HR: highrisk; EFS: event-free survival.

haematologica | 2016; 101(11)

1299

M. Kunstreich et al.

of dexamethasone and compared cumulative prednisoneequivalent doses of GCs, showing no significant correlation with the occurrence of osteonecrosis.10,33,46 However, as dexamethasone is known to be more toxic to the skeletal system than prednisone, and low dexamethasone clearance was linked to severe osteonecrosis,6 this approach might conceal differences. On the contrary, the toxic effects of dexamethasone during delayed intensification may be additive or synergistic with those of GCs administered during the induction phase. Reducing the duration of exposure to dexamethasone seems to reduce the risk for symptomatic osteonecrosis and outweighs the cumulative dose as a risk factor for the development of treatment-related osteonecrosis.8,39,49 GCs might affect antithrombin and protein S levels, with the latter further worsened by the additional administration of asparaginase, thus leading to hypercoagulability.50

Nonglucocorticoid drugs Given the varying frequencies of osteonecrosis in different ALL treatment regimens, nonglucocorticoid drugs such as asparaginase (ASP) and methotrexate (MTX) may additionally modify the risk of osteonecrosis.51 ASP treatment leads to increased plasma concentrations of dexamethasone,6,51,52 whereas ASP allergy is associated with decreased systemic exposure to ASP and with decreased risk of osteonecrosis.53 These effects might further be influenced by the different preparations of ASP used and, to some extent, explain the above-mentioned conflicting results regarding the risk of osteonecrosis in older patients, with the administration of dexamethasone compared to prednisone in trial ALL-BFM 2000 (native asparaginase)35 and AALL0232 (pegylated asparaginase).38 High-dose MTX can damage the growth plate and primary bone, and the long-term use of MTX can reduce primary bone formation, likely due to decreased osteoblast

function as well as increased osteoclast formation and Methylenetetrahydrofolate reductase function.54,55 (MTHFR) polymorphisms can lead to mild to moderate increases in plasma homocysteine levels with homocysteinemia, leading to an increased risk of venous thrombosis.56,57 Alkylating agents may harm gonadal function and lead to primary hypogonadism, which compromises bone mineralization if not adequately treated.58 Ifosfamide can induce renal tubulopathy/Fanconi syndrome, and may subsequently manifest as hypophosphatemic rickets, compromising bone structure.59 Due to hypercoagulability, vascular endothelial damage, and disruption of bone formation, purine antimetabolites can impair proliferation of chondrocytes.60

Other treatment related factors Compared to chemotherapy alone, patients undergoing hematopoietic stem cell transplantation are at an increased risk of developing osteonecrosis.3 Furthermore, total body irradiation (TBI) and chronic graft-versus-host disease correlate with the incidence of osteonecrosis.61

Non-treatment related factors Osteonecrosis occurs more frequently in white patients than in black patients and in non-Hispanics than in Hispanics.7,9,40 Girls between the ages of 10 and 14 years old are especially affected by osteonecrosis, whereas boys are at the highest risk above the age of 15 years.7,46 There is no clear consensus on a risk difference between males and females. Even in groups that used essentially the same treatment regime, there are disparate results in this regard.7,10,32,34,45,46,62 Inconsistent results have also been reported for the influence of obesity and BMI as risk factors.8,63,64

Table 4. Overview of MRI screening studies for osteonecrosis in children and adolescents with ALL.

Protocol & recruiting period

Study cohort No of patients No. of (with ON/ALL) participating centers

Total Therapy XIIIA, NHL XIII 12/91-08/94

ALL, advanced-stage NHL • <18 y Single center ALL • 1-16 y 2 centers 3 pts with orthop. interv.

17/116 incl. Classified acc. 6 asympt. to Ficat 1 pt with (earliest MRI 1 year orthop. interv. after ALL diagnosis) 23/97 incl. 7 sympt. At the end of therapy

ALL Single center

69/364 exclud. 190 asympt.

Nordic ALL protocols 07/92-12/05

St. Jude total XV 06/00-07

Assessment

Incidence Risk factors

Reference

Year

Ribeiro et al.33

2001

15.5%

• Age >10 y (S) • Sex, WBC, BMI, MTX dose, steroid dose, DEX (NS)

24%

• 6% SR vs. 30% IR vs. 35% HR Niinimäki et al.8 2007 • High BMI, female sex, older age, higher cumulative DEX dose (S) • 7% ≤2 weeks vs. 36% >3 weeks DEX during delayed-intensification (S) • no difference in prednisone equivalents • Age >10 y, SR/HR treatment arm (S) Kawedia et al.6 2011 • Older age, lower albumin, higher lipid levels, poor DEX clearance (S)

after reind. 71.8% CI I & II and at any ON completion of therapy 17.6% CI sympt ON

DEX: dexamethasone; S: significant; NS: not significant; y: years; Reind: reinduction; Pt: patient; incl: inclusive; ALL: acute lymphoblastic leukemia; ON: osteonecrosis; NHL: non Hodgkin lymphoma; MRI: magnetic resonance imaging; WBC: white blood count; BMI : body mass index; MTX: methotrexate; SR : standard-risk; IR: intermediate-risk; HR: high-risk; asympt: asymptomatic; orthop. interv.: orthopedic intervention(s);sympt : symptomatic; acc: according.

1300

haematologica | 2016; 101(11)

Osteonecrosis in children with ALL

Genetic risk factors Various genetic risk factors for the development of osteonecrosis in children with ALL and in steroid-induced osteonecrosis have been identified in numerous studies using candidate gene approaches and large genome-wide

association studies (GWAS).6,9,65-68 Polymorphisms in the plasminogen activator inhibitor-1 (PAI-1) gene were initially reported to be associated with an increased risk of osteonecrosis,4,66 but this finding could not be confirmed by subsequent GWAS studies.68

Table 5. Overview of cumulative corticosteroid doses in pediatric ALL (acute lymphoblastic leukemia) studies.

Protocol Retrospective studies

Cumulative steroid dose (assigned to treatment phase)

CCG-1882

Ind. (PRED, 28 d plus taper): 1,815 mg/m2 Regimens A+B: Delayed intensif. (DEX, 21 d plus taper): 235 mg/m2; Maint. (PRED, 5 d cycles): Males 7,000 mg/m2, females 4,400 mg/m2 Regimen C: Delayed intensif. (DEX, 21 d plus taper): 470 mg/m2; Maint. (PRED, 5 d cycles): Males 6,200 mg/m2, females 3,600 mg/m2 DFCI 87-01 Ind. (PRED, 21 d): 840 mg/m2; Intensif./cont. (PRED, 5 d cycles): SR 6,760 mg/m2, HR 20,400 mg/m2 DFCI 91-01 Ind. (PRED, 21 d): 1,120 mg/m2; Intensif./cont. (DEX, 5 d cycles): SR 1,020 mg/m2, HR 3,060 mg/m2 Ind. (PRED, 28 d): 1,680 mg/m2; Reind. (DEX, 21 d): 210 mg/m2; Cont. (PRED, 5 d once): 200 mg/m2 HR only Ind. (PRED, 28 d plus taper): 1,837 mg/m2, 1,417 mg/m2 in HR only; Reintensif. (DEX, 22 d plus taper): 236 mg/m2; Int. interim cons. (DEX, 5 d each cycle): 600 mg/m2 MR only PRED 7,728/11,019/7,117/9,938 mg/m2; DEX 1,230/1,652/1,067/1,490 mg/m2

DFCI 87-01 DFCI 91-01 AIEOP-ALL 95 ALL-BFM 95 UKALL97 UKALL97/01 CoALL-07-03 UKALL 2003 ANZCHOG 8 DFCI 05-001

LR red, LR stand, HR red, HR stand: Ind. (PRED, 28 d): 1,680 mg/m2; Reind. (DEX) LR red (7 d): 70 mg/m2; LR stand (14 d). 140 mg/m2; HR red (2x7 d): 140 mg/m2; HR stand (2x14 d): 280 mg/m2 Ind. (DEX, 28 d): 168 mg/m² plus taper; IM 1 (2x5 d): 60 mg/m²; DI 1 (2x7 d): 140 mg/m²; IM 2 (2x5 d): 60 mg/m²; DI 2 (2x7 d): 140 mg/m²; Maint. (3x5 d): 90 mg/m² per cycle, girls 7-8 cycles, boys 11-12 cycles Ind. (PRED, 28 d plus taper); Reind. (DEX, 21 d plus taper) Cumulative steroid exposure 3,143 mg/m² prednisolone equivalents Ind. (PRED, 28 d): 1,120 mg/m2 plus prophase; Cons IC VHR only. (DEX, 5 d): 90 mg/m2; CNS (DEX, 5 d): SR 30 mg/m², HR/VHR 90 mg/m²; Cons II (DEX, 5 d/cycle): SR 30 mg/m², HR/VHR 90 mg/m² approx. 9 cycles; Cont. (DEX, 5 d/cycle): 30 mg/m2

Prospective studies ALL97 ALL97/99

DCOG-ALL9 CCG-1961 DFCI-ALL 00-01 EORTC-CLG 58951 COG-AALL043 ALL-BFM 2000 NOPHO ALL2008

ALL97: Ind. (28 d): PRED 1,120 mg/m² vs. DEX 182 mg/m²; 1. Intensif. (PRED, 7 d): 280 mg/m²; CNS–dir. treat. (PRED, 5 d every 4 weeks): 600 mg/m²; 2. Intensif. (PRED, 7 d): 280 mg/m²; Interim CT (PRED, 5 d every 4 weeks): 600 mg/m²; 3. Intensif. (DEX, 10 d plus taper): 100 mg/m²; CT (PRED, 5 d every 4 weeks): 3,200 mg/m² ALL97/99: Ind. (28 d plus taper): PRED 1,160 mg/m² vs. DEX 188,5 mg/m²; Interim maint. 1 (2x5 d): PRED 400 mg/m² vs. DEX 65 mg/m²; Delayed intensif. 1 (2x7 d): DEX 140 mg/m²; Interim maint. 2 (2x5 d): PRED 400 mg/m² vs. DEX 65 mg/m²; Delayed intensif. 2 (2x7 d): DEX 140 mg/m²; Cont. (5x5 d): PRED 600 mg/m² vs. DEX 97,5 mg/m² Ind. (6 weeks) & repetitive pulses during maintenance; NHR 1,370 mg/m2 DEX; HR 1,244 mg/m2 DEX Ind. (PRED, 28 d plus taper): 1,815 mg/m2; Delayed intensif. A (DEX, 21 d); Delayed intensif. B (DEX, 2x7 d) Ind. (PRED, 28 d): 1,120 mg/m2 plus taper; Intensif. (10x5 d per cycle): DEX 300 vs. PRED 2,000 mg/m2; Cont. (5 d per cycle): DEX 30 vs. PRED 200 mg/m2 approx. 23 cycles Ind. R1 (28 d plus taper): PRED 1,680 mg/m² vs. DEX 168 mg/m²; Reind. (DEX, 21 d plus taper): 126 mg/m²; Maint. (6x7 d): PRED 2,520 mg/m² vs. DEX 252 mg/m²; VHR only: Cons. (DEX 3x5 d): 150 mg/m²; R-Blocks (DEX 3x5 d): 150 mg/m² Ind. (PRED, 28 d): 1,680 mg/m2; Delayed intensif. (DEX, 21 d): 1-9 y 210 mg/m2; Maint. (DEX, 5 d): 30 mg/m2 every 4 weeks after 9/2008: Delayed intensif. (DEX, 2x7 d): 140 mg/m2; Maint. (PRED, 5 d): 200 mg/m2 every 4 weeks; Maint. 1 year longer in males Ind. (28 d): DEX 280 mg/m2 vs. PRED 1,680 mg/m2 plus pre-phase and taper; Reind./Prot. II (DEX, 21 d plus taper): 210 mg/m2; Reind./Prot. III (DEX, 14 d plus taper): 140 mg/m2; HR-Blocks (DEX, 3x5 d): 300 mg/m2 Ind. preB-ALL and WBC <100x109/L (PRED, 28 d): 1,680 mg/m²; T-ALL a/o WBC ≥100x109/L (DEX, 21 d): 210 mg/m²; HR Block B1 (DEX. 5 d): 100 mg/m²

MRI screening studies Total Therapy XIIIA, NHL XIII Nordic ALL protocols

St. Jude total XV

Ind. (PRED, 28 d plus taper): 1,120 mg/m2; Cont. HR (PRED, 7 d cycles): 280 mg/m2 every 4 weeks; Reind. HR (PRED, 28 d plus taper): 1,120 mg/m2; Postrem. LR (PRED, 7 d cycles): 280 mg/m2 every 4 weeks SR 86 PRED 4,800 mg/m2; SR 92 PRED 4,740 mg/m2; SR 00 PRED 2,400 mg/m2, DEX (5 d) 150/390 mg/m2 IR 86 PRED 1,980 mg/m2, delayed intensif. (28 d plus taper): DEX 320 mg/m2; IR 92 PRED 4,260 mg/m2, delayed intensif. (21 d plus taper) DEX 250 mg/m2; IR 00 PRED 2,400 mg/m2, intensif. (14 d) DEX 264/504 mg/m2 HR 92 PRED 2,800/3,200 mg/m2, delayed intensif. (21 d plus taper) DEX 240 mg/m2; HR 00 PRED 2,400 mg/m2, intensif. (14 d) DEX 430 mg/m2 Cont. (DEX, 3x5 d): LR 120 mg/m2, SR/HR 180 mg/m2 Reind. I & II (DEX, 2x4x8 d): 448 mg/m2

PRED: prednisolone, DEX: dexamethasone; Ind: Induction; Intensif : intensification; Maint: maintenance; Cont: continuation; Reind: reinduction; Postrem: postremission; Reintensif: reintensification; Cons: consolidation. SR: standard-risk; HR: high-risk; MR: medium-risk; LR: low-risk;VHR: very high-risk; IC VHR: consolidation 1C very high-risk; NHR: non high-risk; CT: continuing therapy; CNS : central nervous system; d: day(s); preB-ALL: precursor B acute lymphoblastic leukemia; WBC: white blood count; T-ALL: T cell acute lymphoblastic leukemia; IR: intermediate-risk.

haematologica | 2016; 101(11)

1301

M. Kunstreich et al.

Likewise, findings about polymorphisms involved in lipid homeostasis (acid phosphatase locus 1, ACP1),6 antifolate pharmacodynamics (thymidylate synthetase, TYMS), and steroid hormone response (vitamin D receptor, VDR), have been reported to be associated with osteonecrosis,9 but were not reproducible in GWAS studies.68 According to recent GWAS studies, the glutamate receptor pathway seems to be of crucial importance for the pathogenesis of osteonecrosis in patients with prolonged exposure to corticosteroids. Mechanical load opens mechanosensitive calcium channels in osteocytes, leading to exocytosis of glutamate, which activates osteoblast receptors and impairs endothelial barrier function.67-70 In addition, SNPs in adipogenesis pathways and in enhancers active in mesenchymal stem cells are significantly associated with osteonecrosis development.67 Bone morphogenetic protein (BMP) is toxic to vascular smooth muscle and is released in response to bone damage and mechanical stress. To summarize, osteonecrosis risk is influenced by germline polymorphisms in genes linked to pharmacodynamics of chemotherapy, bone metabolism, adipogenesis, glutamate signaling pathway, and mesenchymal stem cell differentiation. However, given the lack of a single consistent genetic factor being undoubtedly identified, predictive diagnostic testing that helps to evaluate the risk of osteonecrosis development is not established. Even in the context of genetic variants that increase the risk of osteonecrosis, the occurrence of osteonecrosis remains highly dependent on the patient’s age and the specific therapeutic regimen, and, conversely, genetic risk factors significantly depend on the patient’s age.

Adolescence Age is the most consistently identified and most significant risk factor, with patients ≥10 years old at the highest risk across treatment regimens and study groups (Table 1).7-11,43,62 In contrast, the incidence is lower in adults undergoing ALL therapy.36 Thus, the pathogenesis of osteonecrosis is likely strongly associated with factors being most prominent in adolescent age, thereby causing the highest vulnerability for osteonecrosis in this age. There are several adolescent physiological processes that differ fundamentally from younger children and older individuals. These can mainly be attributed to hormonal changes that might lead to increased osteonecrosis susceptibility via interaction with different mechanisms, such as increased local metabolic/perfusion requirements, skeletal maturation (e.g., growth plate structure and development), the coagulation system, or osseous blood vessel supply. All these processes are induced by the beginning and maturation of sexual hormone production and a physiological peak of growth hormone production during puberty. Increasing sexual hormone and especially estrogen concentrations during puberty have procoagulatory effects, thus adolescence is associated with the highest risk of the development of venous thromboembolism.71 Additional risk factors, such as thrombophilia or hypofibrinolysis, can further increase the risk of thrombosis.25,72 In experimental settings, testosterone increases nitrogen monoxide release of endothelial cells, and inhibits platelet aggregation.73 Estrogen further promotes intracortical bone 1302

remodeling. Bone material is added to the endosteal surface, increasing cortical density and bone mass during puberty.22 These estrogenic effects result in a peak in bone mass gain, and the procoagulatory effects of estrogens predispose adolescents to an imbalance between osseous metabolic/blood supply demands and actual osseous blood supply. This effect might further explain the trend to witness more osteonecrosis in females compared to males. The growth hormone/IGF1 axis is physiologically stimulated during puberty, 1.5- to 3-fold compared to pre- and postpubertal individuals.74 The IGF1 level peaks in females at an average of 14 years, and in males at 15 years of age.75 Peak growth velocity/pubertal growth spurt can be expected at 12 years of age in females, and at 14 years of age in males.76 This leads to excessive metabolic activity in growth plates and bones, such as increased oxygen consumption with increased hypoxic effects in growth plates.77-79 This helps to explain why areas of bone with late epiphyseal closure and extensive contribution to pubertal length growth, such as long limb bones, are predominantly affected by osteonecrosis.7 Pubertal epiphyseal maturation and ossification progressively reduce mechanically compliant areas in bone architecture, which might then lose their ability to compensate for increased bone marrow pressure. Concentrations of pro- and anticoagulant factors change crucially during growth.80,81 Major turning points occur after the first six months of life, and between adolescence (11–16 years) and adulthood.82 Coagulant factors (II, V, VII, IX, X, XI, XII, bleeding time), anticoagulant factors (a2M, HCII, Protein C), and the fibrinolytic system (plasminogen, TPA, PAI) are substantially modulated during adolescence and differ significantly from adult levels.80 Both elevated estrogen and testosterone levels further increase the impact of underlying thrombophilia (Factor V Leiden, MTHFR polymorphisms, prothrombinemia, Protein C deficiency, Protein S deficiency hyperhomocysteinemia), and hypofibrinolysis (PAI polymorphisms, increased plasminogen activator inhibitor activity).4,25,66,83 In addition, lifestyle factors, such as smoking, substance abuse, obesity, and use of oral hormonal contraceptives gain importance during adolescence and further contribute to venous thromboembolism (VTE) risk.71 For example, contraceptives with high estrogen content influence the protein C pathway, with subsequently increased activated protein C (APC) resistance84 and platelet aggregation.85 This likely increases the risk of intraluminal obliteration and ischemia in the rapidly growing bone.

Prevention and screening Long continuous exposure during delayed intensification plays a pivotal role in the development of osteonecrosis. Therefore, this was modified in two trials, either by replacing continuous with alternate-week dexamethasone or by entirely reducing the duration of administration.8,39 osteonecrosis incidence in patients treated according to the altered dexamethasone schedule was significantly reduced. However, high-risk ALL patients with osteonecrosis had a 17.6% better event-free survival than patients without osteonecrosis.39 Hence, modifying treatment must be carefully monitored in future prospective trials. haematologica | 2016; 101(11)

Osteonecrosis in children with ALL

A different approach to reduce osteonecrosis-associated debilitating long-term effects is early screening for osteonecrosis by MRI to prevent functional impairment. This has been carried out by Ribeiro et al.33 one year after diagnosis of ALL, by Niinimäki et al.86 at the end of therapy, by Kaste et al.87 at 6.5 and 9 months from diagnosis and at completion of therapy, and by Kawedia et al.6 after reinduction I and II and at completion of therapy. The number of patients diagnosed with radiographic osteonecrosis was high (15.5%,33 24%86 and 71.8%,6 respectively) with a substantial proportion of patients remaining asymptomatic until the end of the study (35%, 70%, and 73%, respectively). Kaste et al.87 further distinguished between limited and extensive (involving more than 30% of the head surface) femoral head osteonecrosis, the latter being a crucial predictor of joint infraction. As radiological classification of osteonecrosis was not uniform, comparability of these results is limited. Furthermore, a recent study by Niinimäki et al. identified critical deficiencies in all available radiological osteonecrosis classification systems and recommended a new, joint-specific classification system.88 With a cumulative incidence of 71.8% of any osteonecrosis, the study by Kawedia et al.6 highlights the need for further research, with particular regard to followup, as the course of osteonecrosis may be transient and reversible and some changes may resolve without symptoms. Even when patients present with joint pain and radiographic changes, the clinical course remains unpredictable. Thus, identifying patients at risk of functional impairment and debilitating progressive joint disease still remains challenging. Precise prospective evaluation of side effects and toxicity in children undergoing treatment for ALL in childhood is therefore an important aspect of modern therapy to reduce compromising outcome after successful treatment. Hence, we initiated the multi-center OPAL trial

References 1. Schrappe M, Moricke A, Reiter A, et al. Key treatment questions in childhood acute lymphoblastic leukemia: results in 5 consecutive trials performed by the ALL-BFM study group from 1981 to 2000. Klin Padiatr. 2013;225 Suppl 1:S62-72. 2. Pui CH, Campana D, Pei D, et al. JCO 2015: Childhood Acute Lymphoblastic Leukemia: Progress Through CollaborationTreating. N Engl J Med. 2009;360(26):2730-2741. 3. Girard P, Auquier P, Barlogis V, et al. Symptomatic osteonecrosis in childhood leukemia survivors: prevalence, risk factors and impact on quality of life in adulthood. Haematologica. 2013;98(7):1089-1097. 4. Bond J, Adams S, Richards S, Vora A, Mitchell C, Goulden N. Polymorphism in the PAI-1 (SERPINE1) gene and the risk of osteonecrosis in children with acute lymphoblastic leukemia. Blood. 2011;118(9): 2632-2633. 5. Kadan-Lottick NS, Dinu I, WasilewskiMasker K, et al. Osteonecrosis in adult survivors of childhood cancer: a report from the childhood cancer survivor study. J Clin Oncol. 2008;26(18):3038-3045. 6. Kawedia JD, Kaste SC, Pei D, et al. Pharmacokinetic, pharmacodynamic, and

haematologica | 2016; 101(11)

10.

11.

(Osteonecrosis in Pediatric patients with Acute lymphoblastic Leukemia and lymphoblastic lymphoma [LBL]), which is still ongoing. In this trial, we prospectively evaluate children aged ≥10 years diagnosed with ALL or LBL, who are treated according to the AIEOP-BFM 2009 and the CoALL-08-09 protocol with a combination of MRI screening and symptom-oriented anamnesis and functional examination at defined time points during ALL treatment. The trial aims to define the proportion of children who can be diagnosed with early asymptomatic osteonecrosis by MRI and subsequently develop symptomatic osteonecrosis, to identify critical time points of osteonecrosis development during ALL treatment, and to describe the natural course of asymptomatic osteonecrosis lesions only identified by MRI. These data are still lacking and are mandatory for the subsequent evaluation of interventions aimed at preventing osteonecrosis progression and functional impairment. These aspects strongly underline the need for intensive future research in the field of pediatric osteonecrosis.

Conclusions Osteonecrosis is the most common therapy-related side effect in children with acute lymphoblastic leukemia. Better understanding of the associated therapy-related and non-therapy-related risk factors is needed to improve prediction, management, and, preferably, prevention of this sequelae. Acknowledgments This work was supported by the German Childhood Cancer Foundation (DKS 2011.11). The authors would like to thank Dr. Jessica I. Höll for her critical reading of the manuscript and editorial assistance.

pharmacogenetic determinants of osteonecrosis in children with acute lymphoblastic leukemia. Blood. 2011;117(8): 2340-2347; quiz 2556. Mattano LA, Jr., Sather HN, Trigg ME, Nachman JB. Osteonecrosis as a complication of treating acute lymphoblastic leukemia in children: a report from the Children's Cancer Group. J Clin Oncol. 2000;18(18):3262-3272. Niinimaki RA, Harila-Saari AH, Jartti AE, et al. High body mass index increases the risk for osteonecrosis in children with acute lymphoblastic leukemia. J Clin Oncol. 2007;25 (12):1498-1504. Relling MV, Yang W, Das S, et al. Pharmacogenetic risk factors for osteonecrosis of the hip among children with leukemia. J Clin Oncol. 2004;22(19):3930-3936. Burger B, Beier R, Zimmermann M, Beck JD, Reiter A, Schrappe M. Osteonecrosis: a treatment related toxicity in childhood acute lymphoblastic leukemia (ALL)--experiences from trial ALL-BFM 95. Pediatr Blood Cancer. 2005;44(3):220-225. Lackner H, Benesch M, Moser A, et al. Aseptic osteonecrosis in children and adolescents treated for hemato-oncologic diseases: a 13-year longitudinal observational study. J Pediatr Hematol Oncol. 2005;27 (5):259-263.

12. Bukowinski AJ, Burns KC, Parsons K, Perentesis JP, O'Brien MM. Toxicity of Cancer Therapy in Adolescents and Young Adults (AYAs). Semin Oncol Nurs. 2015;31(3):216-226. 13. Lafforgue P. Pathophysiology and natural history of avascular necrosis of bone. Joint Bone Spine. 2006;73(5):500-507. 14. Jones JP, Jr. Fat embolism, intravascular coagulation, and osteonecrosis. Clin Orthop Relat Res. 1993;(292):294-308. 15. Laroche M. Intraosseous circulation from physiology to disease. Joint Bone Spine. 2002;69(3):262-269. 16. Jones JP, Jr. Coagulopathies and osteonecrosis. Acta Orthop Belg. 1999;65 Suppl 1:5-8. 17. Kaste SC, Karimova EJ, Neel MD. Osteonecrosis in children after therapy for malignancy. AJR Am J Roentgenol. 2011;196 (5):1011-1018. 18. Miyanishi K, Yamamoto T, Irisa T, et al. Bone marrow fat cell enlargement and a rise in intraosseous pressure in steroid-treated rabbits with osteonecrosis. Bone. 2002;30 (1):185-190. 19. Motomura G, Yamamoto T, Miyanishi K, Yamashita A, Sueishi K, Iwamoto Y. Bone marrow fat-cell enlargement in early steroidinduced osteonecrosis--a histomorphometric study of autopsy cases. Pathol Res Pract. 2005;200(11-12):807-811.

1303

M. Kunstreich et al. 20. Janke LJ, Liu C, Vogel P, et al. Primary epiphyseal arteriopathy in a mouse model of steroid-induced osteonecrosis. Am J Pathol. 2013;183(1):19-25. 21. Assouline-Dayan Y, Chang C, Greenspan A, Shoenfeld Y, Gershwin ME. Pathogenesis and natural history of osteonecrosis. Semin Arthritis Rheum. 2002;32(2):94-124. 22. Schoenau E. Bone mass increase in puberty: what makes it happen? Horm Res. 2006;65 Suppl 2:2-10. 23. Schoenau E, Saggese G, Peter F, et al. From bone biology to bone analysis. Horm Res. 2004;61(6):257-269. 24. Weinstein RS, Nicholas RW, Manolagas SC. Apoptosis of osteocytes in glucocorticoidinduced osteonecrosis of the hip. J Clin Endocrinol Metab. 2000;85(8):2907-2912. 25. Glueck CJ, Freiberg RA, Fontaine RN, Tracy T, Wang P. Hypofibrinolysis, thrombophilia, osteonecrosis. Clin Orthop Relat Res. 2001;(386):19-33. 26. Ojala AE, Paakko E, Lanning FP, Lanning M. Osteonecrosis during the treatment of childhood acute lymphoblastic leukemia: a prospective MRI study. Med Pediatr Oncol. 1999;32(1):11-17. 27. Wang GJ, Sweet DE, Reger SI, Thompson RC. Fat-cell changes as a mechanism of avascular necrosis of the femoral head in cortisone-treated rabbits. J Bone Joint Surg Am. 1977;59(6):729-735. 28. Kawai K, Tamaki A, Hirohata K. Steroidinduced accumulation of lipid in the osteocytes of the rabbit femoral head. A histochemical and electron microscopic study. J Bone Joint Surg Am. 1985;67(5):755-763. 29. Yin L, Li YB, Wang YS. Dexamethasoneinduced adipogenesis in primary marrow stromal cell cultures: mechanism of steroidinduced osteonecrosis. Chin Med J (Engl). 2006;119(7):581-588. 30. Mankin HJ. Nontraumatic necrosis of bone (osteonecrosis). N Engl J Med. 1992;326(22): 1473-1479. 31. Sala A, Mattano LA, Jr., Barr RD. Osteonecrosis in children and adolescents with cancer - an adverse effect of systemic therapy. Eur J Cancer. 2007;43(4):683-689. 32. Strauss AJ, Su JT, Dalton VM, Gelber RD, Sallan SE, Silverman LB. Bony morbidity in children treated for acute lymphoblastic leukemia. J Clin Oncol. 2001;19(12):30663072. 33. Ribeiro RC, Fletcher BD, Kennedy W, et al. Magnetic resonance imaging detection of avascular necrosis of the bone in children receiving intensive prednisone therapy for acute lymphoblastic leukemia or nonHodgkin lymphoma. Leukemia. 2001;15(6): 891-897. 34. Arico M, Boccalatte MF, Silvestri D, et al. Osteonecrosis: An emerging complication of intensive chemotherapy for childhood acute lymphoblastic leukemia. Haematologica. 2003;88(7):747-753. 35. Möricke 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. 36. Toft N, Birgens H, Abrahamsson J, et al. Toxicity profile and treatment delays in NOPHO ALL2008-comparing adults and children with Philadelphia chromosomenegative acute lymphoblastic leukemia. Eur J Haematol. 2016;96(2):160-169. 37. Domenech C, Suciu S, De Moerloose B, et al. Dexamethasone (6 mg/m2/day) and prednisolone (60 mg/m2/day) were equally effective as induction therapy for childhood acute lymphoblastic leukemia in the

1304

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49. 50.

EORTC CLG 58951 randomized trial. Haematologica. 2014;99(7):1220-1227. Mattano LA, Devidas M, Chen S, et al. Effect of High-Dose Methotrexate (HDMTX) Vs Capizzi Methotrexate/Pegaspargase (CMTX/ASNase) on Osteonecrosis (ON) Incidence in Children and Young Adults with T-Acute Lymphoblastic Leukemia (TALL): Results of Children’s Oncology Group (COG) Study AALL0434. Blood. 2014;124 (21):3649-3649. Mattano LA, Jr., Devidas M, Nachman JB, et al. Effect of alternate-week versus continuous dexamethasone scheduling on the risk of osteonecrosis in paediatric patients with acute lymphoblastic leukaemia: results from the CCG-1961 randomised cohort trial. Lancet Oncol. 2012;13(9):906-915. Kahn J, Barrera S, Davila R, et al. Higher Incidence of Treatment-Related Toxicities in Non-Hispanic Patients Undergoing Therapy for Newly Diagnosed Pediatric Acute Lymphoblastic Leukemia on Dana-Farber Cancer Institute ALL Consortium Protocol 05-001. Blood. 2015;126(23):248-248. Vrooman LM, Stevenson KE, Supko JG, et al. Postinduction dexamethasone and individualized dosing of Escherichia Coli Lasparaginase each improve outcome of children and adolescents with newly diagnosed acute lymphoblastic leukemia: results from a randomized study--DanaFarber Cancer Institute ALL Consortium Protocol 00-01. J Clin Oncol. 2013;31(9):1202-1210. den Hoed MA, Pluijm SM, Te Winkel ML, et al. Aggravated bone density decline following symptomatic osteonecrosis in children with acute lymphoblastic leukemia. Haematologica. 2015;100(12):1564-1570. te Winkel ML, Pieters R, Hop WC, et al. Prospective study on incidence, risk factors, and long-term outcome of osteonecrosis in pediatric acute lymphoblastic leukemia. J Clin Oncol. 2011;29(31):4143-4150. Elmantaser M, Stewart G, Young D, Duncan R, Gibson B, Ahmed SF. Skeletal morbidity in children receiving chemotherapy for acute lymphoblastic leukaemia. Arch Dis Child. 2010;95(10):805-809. Amin NL, Kinsey S, Feltbower R, et al. Analysis of Long-Term Outcomes, Management and Prevalence of Osteonecrosis in UKALL 2003: 3.5% of Adolescents and Young Adults over 10 Years of Age with Acute Lymphoblastic Leukaemia Required Hip Replacement. Blood. 2015;126(23):2083-2083. Kuhlen M, Moldovan A, Krull K, Meisel R, Borkhardt A. Osteonecrosis in paediatric patients with acute lymphoblastic leukaemia treated on Co-ALL-07-03 trial: a single centre analysis. Klin Padiatr. 2014;226(3):154-160. McAvoy S, Baker KS, Mulrooney D, et al. Corticosteroid dose as a risk factor for avascular necrosis of the bone after hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2010;16(9):1231-1236. Mitchell CD, Richards SM, Kinsey SE, et al. Benefit of dexamethasone compared with prednisolone for childhood acute lymphoblastic leukaemia: results of the UK Medical Research Council ALL97 randomized trial. Br J Haematol. 2005;129(6):734-745. van den Heuvel-Eibrink MM, Pieters R. Steroids and risk of osteonecrosis in ALL: take a break. Lancet Oncol. 2012;13(9):855-857. te Winkel ML, Appel IM, Pieters R, van den Heuvel-Eibrink MM. Impaired dexamethasone-related increase of anticoagulants is

51.

52.

53.

54.

55. 56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

associated with the development of osteonecrosis in childhood acute lymphoblastic leukemia. Haematologica. 2008;93(10):1570-1574. Yang L, Panetta JC, Cai X, et al. Asparaginase may influence dexamethasone pharmacokinetics in acute lymphoblastic leukemia. J Clin Oncol. 2008;26(12):1932-1939. Hanada T, Horigome Y, Inudoh M, Takita H. Osteonecrosis of vertebrae in a child with acute lymphocytic leukaemia during Lasparaginase therapy. Eur J Pediatr. 1989;149(3):162-163. Liu C, Kawedia JD, Cheng C, et al. Clinical utility and implications of asparaginase antibodies in acute lymphoblastic leukemia. Leukemia. 2012;26(11):2303-2309. Fan C, Cool JC, Scherer MA, et al. Damaging effects of chronic low-dose methotrexate usage on primary bone formation in young rats and potential protective effects of folinic acid supplementary treatment. Bone. 2009;44(1):61-70. Schwartz AM, Leonidas JC. Methotrexate osteopathy. Skeletal Radiol. 1984;11(1):1316. Refsum H, Wesenberg F, Ueland PM. Plasma homocysteine in children with acute lymphoblastic leukemia: changes during a chemotherapeutic regimen including methotrexate. Cancer Res. 1991;51(3):828835. Whayne TF, Jr. Methylenetetrahydrofolate reductase C677T polymorphism, venous thrombosis, cardiovascular risk, and other effects. Angiology. 2015;66(5):401-404. Müller HL, Langer T, Schnabel D. Growth and bone metabolism after oncological disease in childhood and adolescence. Monatsschr Kinderheilkd. 2015;163(2):135141. Loebstein R, Atanackovic G, Bishai R, et al. Risk factors for long-term outcome of ifosfamide-induced nephrotoxicity in children. J Clin Pharmacol. 1999;39(5):454-461. Robson H, Anderson E, Eden OB, Isaksson O, Shalet S. Chemotherapeutic agents used in the treatment of childhood malignancies have direct effects on growth plate chondrocyte proliferation. J Endocrinol. 1998;157 (2):225-235. Faraci M, Calevo MG, Lanino E, et al. Osteonecrosis after allogeneic stem cell transplantation in childhood. A case-control study in Italy. Haematologica. 2006;91 (8):1096-1099. Padhye B, Dalla-Pozza L, Little D, Munns C. Incidence and outcome of osteonecrosis in children and adolescents after intensive therapy for acute lymphoblastic leukemia (ALL). Cancer Med. 2016;5(5):960-967. Kaste SC, Shidler TJ, Tong X, et al. Bone mineral density and osteonecrosis in survivors of childhood allogeneic bone marrow transplantation. Bone Marrow Transplant. 2004;33(4):435-441. Sharma S, Yang S, Rochester R, et al. Prevalence of osteonecrosis and associated risk factors in children before allogeneic BMT. Bone Marrow Transplant. 2011;46 (6):813-819. Asano T, Takahashi KA, Fujioka M, et al. Genetic analysis of steroid-induced osteonecrosis of the femoral head. J Orthop Sci. 2003;8(3):329-333. French D, Hamilton LH, Mattano LA, Jr., et al. A PAI-1 (SERPINE1) polymorphism predicts osteonecrosis in children with acute lymphoblastic leukemia: a report from the Children's Oncology Group. Blood. 2008;111(9):4496-4499. Karol SE, Mattano LA, Jr., Yang W, et al.

haematologica | 2016; 101(11)

Osteonecrosis in children with ALL

68.

69.

70. 71.

72.

73.

74.

Genetic risk factors for the development of osteonecrosis in children under age 10 treated for acute lymphoblastic leukemia. Blood. 2016;127(5):558-564. Karol SE, Yang W, Van Driest SL, et al. Genetics of glucocorticoid-associated osteonecrosis in children with acute lymphoblastic leukemia. Blood. 2015;126(15): 1770-1776. Sharp CD, Houghton J, Elrod JW, et al. Nmethyl-D-aspartate receptor activation in human cerebral endothelium promotes intracellular oxidant stress. Am J Physiol Heart Circ Physiol. 2005;288(4):H1893-1899. Brakspear KS, Mason DJ. Glutamate signaling in bone. Front Endocrinol (Lausanne). 2012;3:97. Nowak-Gottl U, Kenet G. Challenging aspects of managing hemostasis in adolescents. Acta Haematol. 2014;132(3-4):326330. Glueck CJ, Friedman J, Hafeez A, Hassan A, Wang P. Testosterone, thrombophilia, thrombosis. Blood Coagul Fibrinolysis. 2014;25(7):683-687. Campelo AE, Cutini PH, Massheimer VL. Testosterone modulates platelet aggregation and endothelial cell growth through nitric oxide pathway. J Endocrinol. 2012;213(1): 77-87. Rose SR, Municchi G, Barnes KM, et al. Spontaneous growth hormone secretion

haematologica | 2016; 101(11)

75.

76.

77. 78.

79.

80.

81.

increases during puberty in normal girls and boys. J Clin Endocrinol Metab. 1991;73(2): 428-435. Brabant G, von zur Muhlen A, Wuster C, et al. Serum insulin-like growth factor I reference values for an automated chemiluminescence immunoassay system: results from a multicenter study. Horm Res. 2003;60(2):5360. Prader A, Largo RH, Molinari L, Issler C. Physical growth of Swiss children from birth to 20 years of age. First Zurich longitudinal study of growth and development. Helv Paediatr Acta Suppl. 1989;52:1-125. Shim KS. Pubertal growth and epiphyseal fusion. Ann Pediatr Endocrinol Metab. 2015;20(1):8-12. Stewart AJ, Houston B, Farquharson C. Elevated expression of hypoxia inducible factor-2alpha in terminally differentiating growth plate chondrocytes. J Cell Physiol. 2006;206(2):435-440. Emons J, Chagin AS, Hultenby K, et al. Epiphyseal Fusion in the Human Growth Plate Does not Involve Classical Apoptosis. Pediatr Res. 2009;66(6):654-659. Andrew M, Vegh P, Johnston M, Bowker J, Ofosu F, Mitchell L. Maturation of the hemostatic system during childhood. Blood. 1992;80(8):1998-2005. Monagle P, Barnes C, Ignjatovic V, et al. Developmental haemostasis. Impact for

82.

83. 84.

85.

86.

87.

88.

clinical haemostasis laboratories. Thromb Haemost. 2006;95(2):362-372. Jaffray J, Young G. Developmental hemostasis: clinical implications from the fetus to the adolescent. Pediatr Clin North Am. 2013;60(6):1407-1417. Sprengers ED, Kluft C. Plasminogen activator inhibitors. Blood. 1987;69(2):381-387. Tans G, Bouma BN, Buller HR, Rosing J. Changes of hemostatic variables during oral contraceptive use. Semin Vasc Med. 2003;3 (1):61-68. Jayachandran M, Miller VM. Ovariectomy upregulates expression of estrogen receptors, NOS, and HSPs in porcine platelets. Am J Physiol Heart Circ Physiol. 2002;283(1):H220-226. Niinimaki RA, Harila-Saari AH, Jartti AE, et al. Osteonecrosis in children treated for lymphoma or solid tumors. J Pediatr Hematol Oncol. 2008;30(11):798-802. Kaste SC, Pei D, Cheng C, et al. Utility of early screening magnetic resonance imaging for extensive hip osteonecrosis in pediatric patients treated with glucocorticoids. J Clin Oncol. 2015;33(6):610-615. Niinimaki T, Niinimaki J, Halonen J, Hanninen P, Harila-Saari A, Niinimaki R. The classification of osteonecrosis in patients with cancer: validation of a new radiological classification system. Clin Radiol. 2015;70(12):1439-1444.

1305

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Red Cell Biology & its Disorders

Ferrata Storti Foundation

Gene panel sequencing improves the diagnostic work-up of patients with idiopathic erythrocytosis and identifies new mutations Carme Camps,1,2 Nayia Petousi,3 Celeste Bento,4 Holger Cario,5 Richard R. Copley,1,2 Mary Frances McMullin,6 Richard van Wijk,7 WGS500 Consortium,8 Peter J. Ratcliffe,3 Peter A. Robbins,9 and Jenny C. Taylor1,2

National Institute for Health Research (NIHR) Comprehensive Biomedical Research Centre, Oxford, UK; 2Wellcome Trust Centre for Human Genetics, University of Oxford, UK; 3 Nuffield Department of Medicine, University of Oxford, UK; 4Hematology Department, Centro Hospitalar e Universitário de Coimbra, Portugal; 5Department of Pediatrics and Adolescent Medicine, University Medical Center, Ulm, Germany; 6Centre for Cancer Research and Cell Biology, Queen’s University, Belfast, UK; 7Department of Clinical Chemistry and Hematology, University Medical Center Utrecht, the Netherlands; 8A list of members and affiliations is provided in the Online Supplementary Information; and 9 Department of Physiology, Anatomy and Genetics, University of Oxford, UK 1

Haematologica 2016 Volume 101(11):1306-1318

*CC and NP contributed equally to this work **PJR, PAR and JCT jointly supervised this work

ABSTRACT

Correspondence: nayia.petousi@ndm.ox.ac.uk

Received: February 9, 2016. Accepted: July 26, 2016. Pre-published: September 20, 2016. doi:10.3324/haematol.2016.144063

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1306

©2016 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to the Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. Permission in writing from the publisher is required for any other use.

1306

rythrocytosis is a rare disorder characterized by increased red cell mass and elevated hemoglobin concentration and hematocrit. Several genetic variants have been identified as causes for erythrocytosis in genes belonging to different pathways including oxygen sensing, erythropoiesis and oxygen transport. However, despite clinical investigation and screening for these mutations, the cause of disease cannot be found in a considerable number of patients, who are classified as having idiopathic erythrocytosis. In this study, we developed a targeted next-generation sequencing panel encompassing the exonic regions of 21 genes from relevant pathways (~79 Kb) and sequenced 125 patients with idiopathic erythrocytosis. The panel effectively screened 97% of coding regions of these genes, with an average coverage of 450X. It identified 51 different rare variants, all leading to alterations of protein sequence, with 57 out of 125 cases (45.6%) having at least one of these variants. Ten of these were known erythrocytosis-causing variants, which had been missed following existing diagnostic algorithms. Twenty-two were novel variants in erythrocytosis-associated genes (EGLN1, EPAS1, VHL, BPGM, JAK2, SH2B3) and in novel genes included in the panel (e.g. EPO, EGLN2, HIF3A, OS9), some with a high likelihood of functionality, for which future segregation, functional and replication studies will be useful to provide further evidence for causality. The rest were classified as polymorphisms. Overall, these results demonstrate the benefits of using a gene panel rather than existing methods in which focused genetic screening is performed depending on biochemical measurements: the gene panel improves diagnostic accuracy and provides the opportunity for discovery of novel variants.

Introduction Erythrocytosis is a clinical condition characterized by increased red cell mass and typically elevated hemoglobin concentration and hematocrit.1 It can be congenital (e.g. genetic) or acquired and classified as primary or secondary1 (Figure 1A). Several causal genetic mutations have been identified. Heterozygous mutations in the erythropoietin receptor (EPOR) gene cause primary congenital haematologica | 2016; 101(11)

Gene panel for idiopathic erythrocytosis

erythrocytosis,2,3 while JAK2 mutations are predominantly associated with primary acquired erythrocytosis i.e. polycythemia vera.4-6 Homozygous germline mutations in VHL e.g. Chuvash polycythemia and heterozygous germline mutations in EGLN1 (PHD2) and EPAS1 (HIF2A) have been found in patients with secondary congenital erythrocytosis.2,7 Regarding EPAS1, somatic gainof-function mutations have been detected in pheochromocytomas and paragangliomas in patients with congenital erythrocytosis, attributed to tissue mosaicism.8 Some patients, particularly those with polycythemia vera and some forms of genetic erythrocytosis, have increased incidences of both arterial and venous thromboembolic events.9 Other congenital lesions include high oxygenaffinity hemoglobinopathies or 2,3-bisphosphoglycerate deficiency,10-12 caused by mutations in globin genes (HBA1, HBA2, HBB) or the BPGM gene, respectively. These genes belong to key pathways involved in the pathogenesis of erythrocytosis e.g. the oxygen-sensing (hypoxia-inducible factor, HIF) pathway, erythropoiesis and oxygen transport (Figure 1B). Briefly, HIF are transcription factors composed of two subunits: HIFa, which is oxygen-sensitive, and HIFÎ˛. There are three HIFa isoforms, but HIF2a (EPAS1) is erythropoietin's (EPO) main transcriptional regulator.13,14 In normoxia, HIFa is hydroxylated by oxygen-dependent prolyl hydroxylases (encoded by EGLN1, EGLN2 and EGLN3), binds to VHL and becomes ubiquitinated and degraded. In hypoxia, hydroxylation diminishes and HIFa stabilizes and initiates the transcription of target genes, including EPO.15 Erythropoietin binds to the EPOR of erythroid progenitor cells in the bone marrow, stimulating proliferation and differentiation into red blood cells, through a JAK2mediated signaling cascade. In red blood cells, BPGM promotes the release of oxygen to local tissues by producing 2,3-bisphosphoglycerate, which decreases the affinity of hemoglobin to oxygen. Even if fully investigated (including screening for known mutations), a considerable proportion of patients (~70%) remain without an identified cause of their erythrocytosis and are described as having idiopathic erythrocytosis.3,9 About two thirds of these patients have inappropriately normal or elevated erythropoietin levels suggesting a defect in oxygen-sensing or oxygen delivery pathways. Most patients have early-onset disease and/or often a family history, suggesting a high probability of genetic etiology. Logically, further investigation of these patients should begin by fully sequencing genes in which genetic variants are already known to cause erythrocytosis as opposed to simply screening for particular known variants. As many of these are in the HIF pathway, sequencing other key genes in this pathway (in which variants have not yet been observed) and also other erythropoiesis-related genes, is likely to be fruitful in the effort to resolve functional variants. Using traditional DNA sequencing methods, e.g. Sanger sequencing, to comprehensively sequence a large number of genes in a substantial number of patients with a relatively rare disease is time-consuming, labor-intensive and impractical. Conversely, high-throughput technology e.g. whole-genome sequencing (WGS), has its own drawbacks with generation of huge volumes of data, high cost and complex bioinformatic analysis. A way forward is the development of disease-relevant, targeted, next-generation sequencing gene panels. haematologica | 2016; 101(11)

We developed a next-generation sequencing erythrocytosis gene panel, using an ultra-high multiplex polymerase chain reaction method (AmpliSeq, Thermo Fisher), which allows rapid high-throughput sequencing of the full length of multiple genes in multiple samples. We defined a custom-made panel of 21 candidate genes from key pathways involved in the pathogenesis of erythrocytosis, and used it to sequence 125 patients with idiopathic erythrocytosis. We also included novel candidate genes suggested by an initial WGS study, the WGS500 project,16 in which 500 samples across a diverse spectrum of clinical disorders were sequenced, including some cases of idiopathic erythrocytosis strongly suspected of having a genetic cause. The aims of the study were: (i) to create a targeted sequencing panel, as a research tool, for the genetic investigation of erythrocytosis; (ii) to evaluate the panelâ&#x20AC;&#x2122;s diagnostic utility in a cohort of patients with idiopathic erythrocytosis; (iii) to search for novel variants in erythrocytosis-associated genes; and (iv) to include new candidate genes identified in WGS500 to determine whether they are mutated in additional patients.

Methods Patients DNA samples extracted from the blood of patients with idiopathic erythrocytosis were acquired from four separate idiopathic erythrocytosis databases (UK, Portugal, Germany and The Netherlands). Participants gave informed consent and appropriate ethical approval was gained. The inclusion criteria were: (i) confirmed absolute erythrocytosis with a red cell mass >125% predicted, and hemoglobin >180 g/L and hematocrit >0.52% in adult males or hemoglobin >160 g/L and hematocrit >0.48% in adult females, or hemoglobin and hematocrit levels above the 99th centile of age-appropriate reference values in children; (ii) registered as idiopathic (unidentified cause of illness), following appropriate investigation at each Center (Online Supplementary Figure S1); and (iii) early-onset disease, or cases with long-standing idiopathic erythrocytosis. Details are given in the Online Supplementary Information. Ten samples were whole-genome sequenced as part of the WGS500 project, whereas we used our erythrocytosis gene panel to sequence 125 samples from patients with idiopathic erythrocytosis as well as ten positive controls.

Whole-genome sequencing Samples were sequenced at a 30X depth with Illumina HiSeq2000. Details are provided in the Online Supplementary Information.

Ion Torrent sequencing and analysis A customized panel, encompassing the coding and untranslated regions of the candidate genes (Table 1), was created using the Ion AmpliSeq Designer (Thermo Fisher), whereby 635 primer pairs generating amplicons of ~200 bp were designed. This panel covered 90.3% of the target region (78.96 Kb), with 97.4% average coverage of the coding regions. The primers, synthesized in two multiplex pools, were used with the Ion Ampliseq Library kit 2.0 and Ion Xpress barcode adapters (Thermo Fisher) to create libraries. Library quality and concentration were assessed using a 2100 Bioanalyzer (Agilent Technologies). Pools of eight libraries were used for template preparation, loaded into an Ion 316 chip and sequenced on an Ion PGM instrument (500 flows). 1307

C. Camps et al.

The Torrent Suite Software (Thermo Fisher) was used for quality control and alignment of the sequencing data to the human genome (Hg19). Variants were called with the Ion Reporter Software v4.2 (Thermo Fisher), using the germline workflow for single samples and the default parameters, and annotated with ANNOVAR.17 Only variants fulfilling all of the following conditions were selected for further analysis: confidence ≥40, read depth ≥20, frequency in 1000 Genomes (1000G) ≤3% and frequency in NHLBI ESP exomes (6500si) ≤3%. Provean and the SIFT and PolyPhen2-HDIV scores and cut-offs from the ANNOVAR LJB23 database were used to

assess causality of non-synonymous variants. Synonymous variants were investigated for possible splicing effects using Human Splicing Finder, NetGene2 and FSPLICE. Further details are given in the Online Supplementary Information.

Sanger validation All relevant variants identified by Ion Torrent sequencing were confirmed by Sanger sequencing. For protocol and primer details see the Online Supplementary Information and Online Supplementary Table S1.

1308

Figure 1. Classification and pathogenesis of erythrocytosis. (A) Causes of erythrocytosis. Erythrocytosis can be congenital or acquired. It is classified as primary, when there is an intrinsic defect in erythropoietic cells and erythropoietin (Epo) levels are low, or secondary, when the increased red cell production is externally driven through increased EPO production and EPO levels are high or inappropriately normal. Note: in this article, the term erythrocytosis rather than polycythemia is used consistently throughout (B) Pathways involved in the pathogenesis of erythrocytosis. (i) Hypoxia inducible factor (HIF) oxygen sensing pathway in renal EPO-producing cells. HIF are dimeric transcription factors composed of one a- and one β- subunit. In normoxia, HIFa subunits are hydroxylated by oxygendependent prolyl-hydroxylases (PHD) and asparaginyl hydroxylase (HIF1AN). The hydroxylated prolines (P) are recognized by VHL, which mediates the ubiquitination and proteasomal degradation of HIFa. The hydroxylated asparagine (N) compromises the interaction of HIFa with cofactors necessary for transcriptional activity (p300/CBP). In hypoxia, PHD and HIF1AN are less active, HIFa subunits stabilize and translocate into the nucleus where they interact with the HIFβ subunit and cofactors and initiate transcription of target genes, including EPO (ii) Erythropoiesis in the bone marrow. This is triggered by the binding of EPO to the EPO receptor (EPOR) located on the surface of erythroid progenitor cells and subsequent activation of the JAK2-signaling cascade. The process is inhibited by the interaction of SH2B3 and JAK2. (iii) Hemoglobin (Hb) synthesis and oxygen transport. BPGM produces 2,3BPG, which promotes the release of oxygen to local tissues by decreasing the affinity of deoxygenated Hb to oxygen. Alterations in the Hb chains (Hb-a and Hb-β) or BPGM could shift the Hb-oxygen dissociation curve and alter oxygen levels, which directly influence EPO production. (PV, polycythemia vera; ECYT 1-4, erythrocytosis type 1-4; Hb, hemoglobin; O2, oxygen; 2,3-BPG, 2,3-bisphosphoglycerate; RBC, red blood cells; EPO, erythropoietin; PHDs, prolyl hydroxylases). PHDs: PHD1 (EGLN2), PHD2 (EGLN1) and PHD3 (EGLN3).

haematologica | 2016; 101(11)

Gene panel for idiopathic erythrocytosis

Table 1. Genes included in the custom-made erythrocytosis gene panel.

Candidate gene

Position

N. of exons

Transcript ID

Pathway

Candidacy

Inheritance

VHL

Chr3:10183319-10195354

NM_000551

Oxygen-sensing

EPAS1

Chr2:46524541-46613842

NM_001430

Oxygen-sensing

EGLN1

Chr1:231499497-231560790

NM_022051

Oxygen-sensing

HIF1A

Chr14:62162119-62214977

NM_001530

Oxygen-sensing

Known erythrocytosis-causing variants Known erythrocytosis-causing variants Known erythrocytosis-causing variants Key gene of the HIF pathway

HIF3A

Chr19:46800303-46846690

NM_022462

Oxygen-sensing

Key gene of the HIF pathway

EGLN2

Chr19:41305048-41314346

NM_053046

Oxygen-sensing

Key gene of the HIF pathway

EGLN3

Chr14:34393421- 34420284

NM_022073

Oxygen-sensing

Key gene of the HIF pathway

HIF1AN

Chr10:102295641-102313681

NM_017902

Oxygen-sensing

Key gene of the HIF pathway

EPO

Chr7:100318423-100321323

NM_000799

Erythropoiesis/ oxygen-sensing

1. Key gene in erythropoiesis

EPOR

Chr19:11487881-11495018

NM_000121

Erythropoiesis

JAK2

Chr9:4985245-5128183

NM_004972

Erythropoiesis

SH2B3

Chr12:111843752-111889427

NM_005475

Erythropoiesis

BPGM

Chr7:134331531-134364567

NM_001724

Oxygen transport

HBB

Chr11: 5246696-5248301

NM_000518

HBA1

Chr16:226650-227521

NM_000558

HBA2

Chr16:222846-223709

NM_000517

KDM6A

ChrX:44732421-44971857

NM_021140

GFI1B

Chr9:135854098-135867084

NM_004188

Oxygen transport/ hemoglobin synthesis Oxygen transport/ hemoglobin synthesis Oxygen transport/ hemoglobin synthesis Oxygen-regulated demethylase Erythropoiesis

BHLHE41

Chr12:26272959-26278003

NM_030762

Factor associated with HIF

Identified in WGS500

OS9

Chr12:58087738-58115340

NM_001261421

Factor associating with HIF

Related to the HIF pathway

ZNF197

Chr3: 44666511-44689963

NM_006991

Factor associating with HIF

Related to the HIF pathway

Recessive / compound heterozygous (based on reported cases) Dominant (based on reported cases) Dominant (due to haploinsufficiency, based on reported cases) Unknown (no cases reported). Likely dominant by function similarity to EPAS1 Unknown (no cases reported). Likely dominant by function similarity to EPAS1 Unknown (no cases reported). Likely dominant by function similarity to EGLN1 Unknown (no cases reported). Likely dominant by function similarity to EGLN1 Unknown (no cases reported). No function similarity to any known associated gene Dominant (based on WGS500 variant), but cannot discard other patterns of inheritance Dominant (based on reported cases) Somatic (based on reported cases) Dominant (somatic or germline, based on reported cases) Dominant / compound heterozygous / recessive (based on reported cases) Dominant (based on reported cases) Dominant (based on reported cases) Dominant (based on reported cases) Recessive X-linked inheritance (based on WGS500 variant) Recessive (based on WGS500 variant), but cannot discard other patterns of inheritance Recessive (based on WGS500 variant), but cannot discard other patterns of inheritance Unknown (no cases reported). No function similarity to any known associated gene Unknown (no cases reported). No function similarity to any known associated gene

2. Identified in WGS500 Known erythrocytosis-causing variants Known erythrocytosis-causing variants Known erythrocytosis-causing variants 1. Known erythrocytosis-causing variants 2. Identified in WGS Known erythrocytosis-causing variants Key gene in oxygen transport Key gene in oxygen transport Identified in WGS500 Identified in WGS500

haematologica | 2016; 101(11)

1309

C. Camps et al.

Results Novel candidate genes and variants were identified by whole-genome sequencing The whole genomes of a small number of idiopathic erythrocytosis cases strongly suspected of having a genetic cause were sequenced as part of the WGS500 project. Candidate variants were found in novel genes, not previously associated with erythrocytosis: EPO, GFI1B, KDM6A and BHLHE41. Details of the rationale and criteria used to select these genes as candidates are given in the Online Supplementary Information and Online Supplementary Table S2. On this basis, these genes were included in the next-generation sequencing gene panel along with other erythrocytosis candidate genes (Table 1).

The erythrocytosis gene panel has high performance in sequencing and variant detection Overall, 135 samples were sequenced on the Ion Torrent using the gene panel (125 undiagnosed patients, 10 positive controls). On average, 89% of mapped reads were on target regions, which indicates a successful custom panel according to the manufacturer’s guidance. The average coverage depth of the amplicons generated was 450X (Figure 2A). Most samples (133 out of 135) had over 92% of amplicons with coverage above 20X (Figure 2B). Only two samples presented substantial failure across the panel (Figure 2B), which was related to DNA quality. Only 17

amplicons (2.6%) had an average coverage below 20X across samples, indicating a general poor amplification of these regions within the highly-multiplexed reactions (Online Supplementary Table S3). Ten of these (1.6% of all amplicons) had complete failure (coverage <20X in all samples), probably due to sequence context issues. The sequencing was, therefore, generally successful across samples, with a high percentage of the target sequence included at a good depth for germline variant calling. We compiled a list of all known erythrocytosis-associated variants from the literature,2,3 including the variants identified in the WGS study, and cross-referenced their genomic coordinates with those of the generated amplicons. With the exception of two missense variants in VHL, all the other variants were within amplicons that performed well. The two VHL missense variants – c.235C>T and c.311G>T – fall within an amplicon in exon 1 that showed complete failure and would not, therefore, be detected. Importantly, our panel reliably detected ten known variants – in different genes and hence in different amplicons – in the positive control samples, in which mutations had previously been identified either through WGS or Sanger sequencing (Online Supplementary Table S4).

Fifty-one exonic variants were identified across 57 patients by the erythrocytosis gene panel and validated by Sanger sequencing We identified 98 different variants across the coding

Figure 2. Coverage of the amplicons generated by the erythrocytosis gene panel across 135 samples. (A) Each boxplot represents the distribution of the number of reads obtained for all the amplicons generated by the panel within each sample. The horizontal line across the plot shows the average coverage (450X). (B) Each dot represents the percentage of amplicons with coverage over 20X within each sample.

1310

haematologica | 2016; 101(11)

Gene panel for idiopathic erythrocytosis

regions of the genes examined, of which 19 were insertions or deletions (INDEL), 49 non-synonymous single nucleotide variations (SNV) and 30 synonymous SNV (Figure 3). None of the synonymous SNV is predicted to alter splicing according to Human Splicing Finder, NetGene2 and FSPLICE. We, therefore, focused on variants resulting in protein sequence alterations: following Sanger sequencing, 17 out of the 19 INDEL appeared to be false positives but two were confirmed. All 49 non-synonymous SNV were confirmed, although for one SNV there was a single base discrepancy: Ion Torrent detected a triple base change (CAA>ATT) in exon 12 of JAK2 (chr9:5070025-5070027) but only a double change (AA>TT, chr9:5070026-5070027) was confirmed by Sanger sequencing. As a result, a total of 51 variants (49 SNV, 2 INDEL) were detected (Online Supplementary Table S5). Therefore, 57 out of 125 cases had at least one exonic variant (45.6%); of those, 38 patients had only one exonic variant detected (30.4%), while 19 had more than one (15.2%). To investigate whether the variants discovered are unique to erythrocytosis patients (and therefore more likely to be disease-causing), we used in silico data from the 1000G project as a control. For this, we examined the variant calls from the 1000G project after integrating both exome and low coverage data across 1041 individuals and extracted the SNV identified within the coordinates of the amplicons generated by our gene panel. We found that of the 49 non-synonymous SNV discovered, 30 were uniquely found in our erythrocytosis cohort and not in the 1000G in silico control cohort, whereas the other 19 were also found in the control cohort at similar or higher frequencies (Fisher exact test and Benjamini and Hochberg false dis-

covery correction18) (Figure 3). Those 19 SNV (Online Supplementary Table S6) are thus unlikely to be diseasecausing mutations and most likely represent polymorphisms. Out of the 30 uniquely identified variants in our cohort of patients, ten had been previously reported in the literature as causing erythrocytosis and hence are classified here as disease-causing variants (Table 2). The remaining 20 had no previous clinical associations. No exonic variants were identified in EGLN3, HIF1AN (FIH), HBA1, HBA2, GFI1B or ZNF197.

Novel genes and variants identified by the erythrocytosis gene panel The 22 novel variants (20 SNV and 2 INDEL) identified (Table 3) are extremely rare: nine were absent from both the dbSNP142 and Exome Aggregation Consortium (ExAC) databases, the latter containing data from 60,706 unrelated individuals; eight were reported only in ExAC at extremely low allele frequencies (â&#x2030;¤0.0007), and only five were reported in both databases at very low allele frequencies (â&#x2030;¤0.005). Fourteen of these novel or very rare variants were found in known erythrocytosis-associated genes, such as VHL, EPAS1, JAK2, SH2B3 (LNK), EGLN1 and BPGM. Some of these variants have a high likelihood of causality based on the location and predicted effect of the protein coding change as well as on genetic evidence for causality, and are of particular physiological interest. For example, EPAS1 p.Y532H, a novel exon 12 mutation, is located one position downstream of residue 531, which is the prolyl hydroxylation site on HIF2a on the C-terminal oxygendependent degradation domain (ODD). Furthermore, it is

Figure 3. Overview of the exonic variants detected with Ion Torrent sequencing among 125 patients with erythrocytosis, their validation and further classification.

haematologica | 2016; 101(11)

1311

C. Camps et al.

part of a six-residue domain which is highly conserved both across all HIFa isoforms and across species and which interacts with the VHL complex.19 Thus, this mutation likely interferes with hydroxylation of HIF2a by prolyl hydroxylases and binding to the VHL complex, leading to upregulation of erythropoietin. It was found in two

related patients, father and son, both of whom had idiopathic erythrocytosis with raised erythropoietin levels, and was, therefore, inherited in an autosomal dominant manner. EGLN1 p.L279P affects a conserved residue, previously reported as altered (p.L279Tfs43, a frameshift variant) in a patient with erythrocytosis.20 Structurally, this

Table 2. Variants detected by the erythrocytosis gene panel, known to cause erythrocytosis.

Genomic location

Gene

chr2:46607420 G>A

EPAS1

c.G1609A p.G537R

Het

chr3:10191578 C>G

VHL

c.C571G p.H191D

Hom

chr3:10191605 C>T

VHL

c.C598T p.R200W

Het*

chr9:5070026 AA>TT

JAK2

c.1615_ 1616invAA p.K539L

Het

chr11:5246832 T>G

HBB

c.A440C p.H147P

Het

chr11:5246840 G>C

HBB

c.C432G p.H144Q

Het

chr11:5246944 C>T chr11:5247816 C>G chr12:111856571 G>C chr12:111885310 G>A

HBB HBB SH2B3

SH2B3

cDNA/protein Genotype N. change of cases

c.G328A p.V110M c.G306C p.E102D c.G622C p.E208Q c.G1198A p.E400K

Het

Patient information

Type of erythrocytosis

Mechanism of action

Previous publication

Female; age at diagnosis, 13 y; Hb, 196 g/L; Hct, 59.1%; Epo, 7.5 mIU/mL; chronic headache; pulmonary hypertension Male; age at diagnosis, 12 y; Hb, 154 g/L; Hct, 59%; Epo, 23 mIU/mL Patient 1: Male; age at diagnosis, 47 y; Hb, 182 g/L; Hct, 54%; Epo, 8 mIU/mL; no family history Patient 2: Female; age at diagnosis, 48 y; Hb, 199 g/L; Hct, 67%; Epo, 22 mIU/mL; no family history Patient 3: Male; age at diagnosis, 1 y; Hb, 179 g/L; Hct, 54%; Epo, 60 mIU/mL; brother of patient 4 Patient 4: Male; age at diagnosis, 2 y; Hb, 146 g/L; Hct, 44.4%; Epo, high; brother of patient 3 Male, age at diagnosis, 35 y; Hb, 155 g/L; Hct, 52%; RBC, 6.37x1012/L; WBC and platelets, normal range; Epo, 5-46 mIU/mL (after venesection); first presentation with large stroke; splenomegaly; BM biopsy, erythropoietic hyperplasia; JAK2 V617F negative Female; age at diagnosis, 27 y; Hb, 173 g/L; Hct, 52.4%; Epo, 24 mIU/mL; family history, two brothers, mother, grandfather and great grand-mother affected by erythrocytosis (maternal line) Male; age at diagnosis, 34 y; Hb, 200 g/L; Hct, 58%; Epo, 7.5-27 mIU/mL; asymptomatic Male; age at diagnosis, 49 y; Hb, 181 g/L; Hct, 53%; Epo, 25 mIU/mL Male; age at diagnosis, 25 y; Hb, 204 g/L; Hct, 58%; Epo, 5.5 mIU/mL Male; age at diagnosis, 14 y; Hb, 210 g/L; Hct, not available; Epo, 20.2 mIU/mL (after venesection) Male; age at diagnosis, 40 y; Hb, 189 g/L; Hct, 52.7%; RBC, 5.78x1012/L; Epo, 2.5 mIU/mL

Secondary

Gain of function of HIF2A

Percy et al. 200831, Gale et al. 200832

Loss of function Tomasic et al. 201333 (enhances HIF regulated gene expression) Secondary Loss of Ang et al.200241 function (decreased HIF binding & hydroxylation, enhances HIF-regulated gene expression)

Secondary

Primary

Gain of function of JAK2 (K539L)

Scott et al. 20075

Secondary

High oxygen affinity Hb (Hb York)

Misgeld et al. 200137

Secondary

High oxygen affinity Hb (Hb Little Rock)

Bromberg et al.197336,

Secondary Secondary Primary

Primary

High oxygen affinity Hb (Hb San Diego) High oxygen affinity Hb (Hb Potomac) Enhances JAK2 signaling Interacts with JAK2 signaling

Wajcman et al. 199638 Wajcman et al. 199638, Gonzalez et al. 200935 Charache et al. 197834 Spolverini et al. 201339

McMullin et al. 201140, Spolverini et al. 201339

Official gene symbols according to the HUGO Gene Nomenclature Committee are given here. Other gene symbols used frequently in the literature are: HIF2A (EPAS1), LNK (SH2B3). Chr : chromosome; Het: heterozygous; Hom: homozygous; y: years; Hb: hemoglobin; Hct: hematocrit; WBC: white blood cell count; RBC: red blood cell count; BM: bone marrow. Typical normal ranges: Hb, 130-180 g/L (adult males) and 115-155 g/L (adult females); Hct, 45-52% (adult males) and 37-48% (adult females); RBC, 4.7-6.1 x1012/L (adult males) and 4.2-5.4 x1012/L (adult females); Epo, 3.3-15.8 mIU/mL (adult males and females, although this range can vary between laboratories).*This variant causes Chuvash polycythemia in the homozygous state. In one of the patients, this variant was discovered in this study, whereas in the other three it had been detected in previous genetic tests.

1312

haematologica | 2016; 101(11)

Gene panel for idiopathic erythrocytosis

Table 3. Novel variants detected by the erythrocytosis gene panel.

Genomic location

Gene

cDNA/ protein change

Genotype N. of cases

Patient information

chr1:231556799 A>G

EGLN1

c.T836C p.L279P

Het

chr2:46607405 T>C

EPAS1

c.T1594C p.Y532H

Het

chr3:10183685 G>T

VHL

c.G154T p.E52X

Het

chr7:100319185 TC>T

EPO

c.19delC p.P7fs

Het

chr2:46574031 AAGG>A

EPAS1

c.47delAGG p.del17E

Het

Male; age at diagnosis, 47 y; Hb, 186 g/L; Hct, 58.6%; Epo, normal range; headaches and dizziness

chr9:5050747 A>T

JAK2

c.A530T p.E177V

Het

chr12:111856181 G>A

SH2B3

c.G232A p.E78K

Het

chr12:111884812 G>A

SH2B3

c.G901A p.E301K

Het

Male; age at diagnosis, 52 y; Hb, 198 g/L; Hct, 61.2%; RBC, 6.42x1012/L; Epo, 12.2 mIU/mL; headaches and dizziness; family history

Type of SIFT/ Allele frequency erythrocytosis Polyphen/ dbSNP142 Provean ExAc

DNA studies in family members

Evidence of causality

Putative secondary

D/D/D

Not found Not found

Not available

Predicted structural/ functional effects

Patient 1: Putative Male; age at diagnosis, 12 y; secondary Hb, 190 g/L; Hct, 54%; Epo, not available; clinically well; no pulmonary hypertension; family history, father with congenital erythrocytosis (Patient 2) Patient 2: Male; age at diagnosis, 42 y

D/D/D

Not found Not found

Variant present in both affected father and son

Predicted structural/ functional effects & segregation

Putative secondary

T/NA/NA

Not found 4.16E-05

Not available

Predicted structural/ functional effects

Female; age at diagnosis, 3 y; Putative Hb, 194 g/L; Hct, 58%; secondary Epo, 4.1 mIU/mL; asymptomatic; family history, affected father and paternal grandmother with high Hb and Hct

NA/NA/NA

Not found Not found

Variant present in father, who has high Hb and Hct

Segregation

Putative secondary

NA/NA/D

Not found Not found

Not available

1. Deleterious by at least two prediction tools 2. Most in known erythrocytosiscausing genes 3. Most not found in large population databases

Male; age at diagnosis, 14 y; Hb, 158 g/L; Hct, 52%; Epo, 7.6 mIU/mL; normal liver and spleen; normal cardiopulmonary function; no family history

Putative primary

D/D/D

Not found Not found

Not available

Female; age at diagnosis, not reported (current age 85 y); Hb, 196 g/L; Hct, 57%; Epo, 12.2 mIU/mL; normal liver and spleen size; no family history

Putative primary

D/P/D

Not found 7.40E-04

Not available

D/D/Neutral

3.20E-05 3.30E-05

Not available

Male; age at diagnosis, 28 y; Hb, 184 g/L; Hct, 54.6%; Epo, not available; headaches and dizziness; family history, affected brother

Male; age at diagnosis, 46 y; Putative Hb, 187 g/L; Hct, 52%; primary RBC, 6.16x1012/L; WBC and platelets, normal range; Epo, 10 mIU/mL; fatigue; mild splenomegaly; BM biopsy, increased erythropoiesis and slight dyserythropoiesis, normal megakaryopoiesis, no myeloproliferative neoplasia; no family history

Continued on the next page

haematologica | 2016; 101(11)

1313

C. Camps et al. continued from the previous page

Genomic location

Gene

cDNA/ protein change

Genotype

N. of cases

chr12:111885466 SH2B3 C>T

c.C1243T p.R415C

Hom

chr19:41313427 G>T

EGLN2

c.G1139T p.R380L

Het

chr2:46611651 T>C

EPAS1

c.T2465C p.M822T

Het

chr3:10183605 C>T

VHL

c.C74T p.P25L

Het

chr7:134346563 C>A

BPGM

c.C304A p.Q102K

Het

chr9:5022168 G>A

JAK2

c.G181A p.E61K

Het

chr9:5054775 G>C

JAK2

c.G827C p.G276A

Het

chr12:58109559 G>A

OS9

c.G497A p.G166D

Het

chr19:46811511 A>C

HIF3A

c.A190C p.I64L

Het

chr19:46823777 C>A

HIF3A

c.C896A p.A299D

Het

chr9:5072561 G>A

JAK2

c.G1711A p.G571S*

Het

chr12:26276001 BHLHE41 c.T447G A>C p.F149L

Het

Patient information

Male; age at diagnosis, 18 y; Hb, 188 g/L; Hct, 57%; Epo, not available; splenomegaly; no family history Male; age at diagnosis, 16 y; Hb, 183 g/L; Hct, 53.7%; Epo, 2.8 mIU/mL; family history Female; age at diagnosis, 9 y; Hb, 162 g/L; Hct, 48%; Epo, 5.8 mIU/mL; no family history Patient 1: Male; age at diagnosis, 21 y; Gitelman syndrome (with positive SLC12A3 mutation); Patient 2: Male; age at diagnosis, 15 y; Hb, 190 g/L; Hct, 55%; Epo, not available; mild headaches; family history, father also affected Male; age at diagnosis, 52 y; Hb, 186 g/L; Hct, 52.5%; Epo, normal range; myocardial infarction; no family history Male; age at diagnosis, 41 y; Hb, 172 g/L; Hct, 53%; WBC and platelets, normal range; Epo, 10-27 mIU/mL (while venesected); no family history Female; age at diagnosis, 33 y; Hb, 172 g/L; Hct, 53.3%; Epo, 6 mIU/mL; headaches; dizziness; family history, affected brother Male; age at diagnosis, 37 y; Hb, 188 g/L; Hct, 51.8%; Epo, 8.24 mIU/mL Female; age at diagnosis, 19 y; Hb, 183 g/L; Hct, 58%; Epo, not available Male; age at diagnosis, 53 y; Hb, 188 g/L; Hct, 58%; Epo, 23 mIU/mL; coronary heart disease Male; age at diagnosis, 4 y; Hb, 180 g/L; Hct, not available; Epo, 12.4 mIU/mL Male; age at diagnosis, 23 y; Hb, 183 g/L; Hct, 57%; Epo, not available

Type of SIFT/ Allele frequency erythrocytosis Polyphen/ dbSNP142 Provean ExAc

DNA studies in family members

Putative primary

D/D/D

1.00E-03 4.19E-05

Not available

Putative secondary

D/D/D

Not found Not found

Not available

Putative secondary

D/B/Neutral Not found 8.24E-06

Not available

Putative secondary

D/B/Neutral

Not available

Putative secondary

D/B/Neutral Not found Not found

Not available

Putative primary

T/B/Neutral Not found Not found

Not available

Putative primary

T/B/Neutral Not found 8.29E-06

Not available

Unknown

Not found 3.42E-05

Not available

Putative secondary

D/B/Neutral Not found 1.65E-05

Not available

Putative secondary

T/B/Neutral

0.00066 7.93E-04

Not available

Unknown

T/D/D

7.4 E-04 4.81E-04

Variant present in non-affected parent

Unknown

T/D/D

4.00E-04 5.17E-03

T/B/Neutral Not found 1.54E-04

Evidence of causality

Extremely rare variants

Unlikely diseasecausing in erythrocytosis

Variant present in non-affected paternal aunt; absent in non-affected mother and sibling Continued on the next page

1314

haematologica | 2016; 101(11)

Gene panel for idiopathic erythrocytosis continued from the previous page

Genomic location

Gene

cDNA/ protein change

Genotype

N. of cases

Patient information

chr7:100320336 A>G

EPO

c.A296G p.E99G**

Het

chr7:100320290 G>C

EPO

c.G250C p.G84R***

Het

Male; age at diagnosis, 13 y; Hb, 149 g/L; Hct, 49.6%; Epo, 7.5 mIU/mL; subsequent red cell mass measurement negative for absolute erythrocytosis despite high Hb Patient 1: Male; age at diagnosis, 12 y; Hb, 190 g/L; Hct, 54%; Epo, not available; clinically well; no pulmonary hypertension; family history, father with congenital erythrocytosis (Patient 2) Patient 2: Male; age at diagnosis, 42 y

Type of SIFT/ Allele frequency DNA erythrocytosis Polyphen/ dbSNP142 studies Provean ExAc in family members Not absolute erythrocytosis

D/D/D

Not found

Not available

Putative secondary

T/D/Neutral

Not found 8.04E-05

Variant present in both affected father and son

Evidence of causality

Official gene symbols according to the HUGO Gene Nomenclature Committee are given here. Other gene symbols used frequently in the literature are: HIF2A (EPAS1), PHD2 (EGLN1), PHD1 (EGLN2), LNK (SH2B3), DEC2 (BHLHE41). Chr: chromosome; Het: heterozygous; Hom: homozygous; y: years; Hb: hemoglobin; Hct: hematocrit; WBC: white blood cell count; RBC: red blood cell count; BM: bone marrow; D: deleterious (applicable to SIFT and Provean predictions) and probably damaging (applicable to Polyphen2 HDIV predictions); T: tolerated by SIFT; P: possibly damaging by Polyphen2 HDIV; B: benign by Polyphen2 HDIV and NA, non-applicable. Typical normal ranges: Hb, 130-180 g/L (adult males) and 115-155 g/L (adult females); Hct, 45-52% (adult males) and 37-48% (adult females); RBC, 4.7-6.1x1012/L (adult males) and 4.2-5.4 x1012/L (adult females); Epo: 3.3-15. 8 mIU/mL (adult males and females, although this range can vary between laboratories). ExAc: Exome Aggregation Consortium, Cambridge, MA, USA (http://exac.broadinstitute.org). All variants except the ones included in the “unlikely disease-causing in erythrocytosis” section have been submitted to a dedicated database (www.erythrocytosis.org). * JAK2 p.G571S has been reported previously in myeloproliferative disorders47 but it is thought to be a silent non-functioning polymorphism; ** predicted deleterious but although patient has high Hb and Hct was subsequently found to have normal red cell mass. *** Likely to be non-pathogenic: predicted benign/tolerated and present in patients with an identified variant in EPAS1 with very high likelihood of causality.

residue is located on helix 3, which interacts with both Nterminal and C-terminal ODD hydroxylation domains on HIFa;21 a proline substitution may affect protein stability and diminish ODD binding, reducing HIFa hydroxylation. The VHL p.E52X variant introduces a stop codon, predicting translation termination of the long VHL isoform (p30) while allowing translation only of the alternative form of VHL (p19) from a translation site at M54. To date, only a few variants upstream of the VHL internal start codon 54 have been described and have been associated with either pheochromocytomas (codon 25 and 38) or with von Hippel-Landau (VHL) disease (p.E46X and p.E52K).22-24 The role of the heterozygous VHL p.E52X in producing erythrocytosis in the patient in our study is not clear and the patient will be advised to undergo investigations for the presence of VHL disease; there is evidence that erythrocytosis is seen in about 5-20% of patients with VHL disease.25 For the remaining variants, most were classified as deleterious by either SIFT, PolyPhen2 or Provean (Table 3), with a high degree of agreement between tools, so further investigations are needed to elucidate their functional impact. Eight variants were identified in novel genes included in the panel because of their association with the oxygensensing pathway but in which no previous erythrocytosisassociated mutation has been reported, such as EGLN2, HIF3A and OS9 (Table 3). In addition, novel variants were also found in EPO and BHLHE41, two genes without previous genetic association with erythrocytosis which were revealed by WGS500. For EPO, the most striking variant found is a frameshift, p.P7fs, detected in a heterozygous state in one patient. Although at present it is difficult to link an apparently inactivating mutation to the generation of erythrocytosis, the variant has since been confirmed in haematologica | 2016; 101(11)

a heterozygous state in the patient’s father who also has high hematocrit and hemoglobin levels. Two other EPO SNV were detected in other patients but these are most likely very rare polymorphisms (Table 3). Regarding BHLHE41, the novel missense variant identified (p.F149L) is classified as benign by Provean, PolyPhen2 and SIFT and is thus unlikely to be pathogenic, a notion supported by segregation analysis in the patient’s family (Table 3).

Discussion The technical progress in next-generation sequencing, together with the increasing understanding of the biological pathways underlying the pathogenesis of erythrocytosis, provide new opportunities to advance the genetic investigation of patients with erythrocytosis. Our approach allowed the creation of a next-generation sequencing targeted gene panel with the capacity to process a large group of samples and simultaneously examine a large number of genes across several biological pathways in a systematic and efficient manner. Our panel exhibited high performance and reliability. It produced high quality sequencing data with good target coverage. It accurately detected variants in ten positive controls. It was excellent at reliably calling SNV, with all SNV identified subsequently validated in all samples by Sanger sequencing. Nevertheless, a few limitations are recognized and should be taken into account when considering its future applications. For example, a few amplicons – including a region on VHL exon 1 – showed complete failure across samples and thus potential variants within them would not be detected. Furthermore, there were some false positive INDEL, as previously reported by other Ion 1315

C. Camps et al.

Torrent sequencing users.26-28 These could be addressed by re-designing primers covering that particular VHL genomic region, optimizing the variant calling bioinformatics workflow and employing recently proposed strategies to increase the accuracy of INDEL detection.26,28 Another limitation of the panel – related to the nature of its technology – is that it can only identify SNV and short INDEL but not other structural variants such as large INDEL or copy number variations. Also, variant detection in genes with high sequence similarity such as HBA1 and HBA2 is challenging and caution is needed for variant calling. Currently, the clinical consensus for investigating erythrocytosis involves: establishing the diagnosis of absolute erythrocytosis, excluding systemic causes (e.g. hypoxic lung diseases or tumors) and then proceeding to focused genetic testing based on algorithms that attempt to predict the type of mutation that might be present. The procedures employed at different Centers vary (Online Supplementary Figure S1), but as a general rule if the patient’s erythropoietin level is low, variants in genes involved in erythropoiesis (EPOR, JAK2) are screened for. If the patient’s erythropoietin level is high or normal, the P50 (partial pressure of oxygen at which 50% of hemoglobin is saturated with oxygen) is calculated and if low, hemoglobin electrophoresis is performed and/or variants in oxygen-delivery pathways (globin genes, BPGM) are screened for; if P50 is normal or not available, variants in the oxygen-sensing HIF pathway (VHL, EPAS1, EGLN1) are screened for.2,29,30 Using our gene panel we were able to provide definitive genetic diagnoses in nine patients whose mutations had been previously missed. For example, a variant in EPAS1, p.G537R – a well-described gain-of-function mutation found in erythrocytosis patients31,32 – was detected. This was previously missed because the patient was not screened for EPAS1 variants, owing to the fact that the erythropoietin level was not high enough (and investigations were thus directed to a different branch of the diagnostic algorithm). Similarly, we identified a homozygous VHL variant (p.H191D) known to cause erythrocytosis.33 Interestingly, we found four variants in the HBB gene, all relating to high-affinity hemoglobinopathies associated with erythrocytosis: HBB p.H147P (Hb York), HBB p.H144Q (Hb Little Rock), HBB p.V110M (Hb San Diego) and HBB p.E102D (Hb Potomac).34-38 These were missed previously, either because conventional screening with hemoglobin electrophoresis can miss hemoglobinopathies38 or because of difficulties in obtaining optimal fresh venous blood samples for P50 measurements in all patients. In addition, we identified a heterozygous variant in JAK2 (p.K539L) and two in SH2B3 (p.E208Q and p.E400K), all known to associate with erythrocytosis.5,39,40 The patient with variant JAK2 p.K439L, originally classified as having idiopathic erythrocytosis as the conventional criteria for polycythemia vera, including JAK2 p.V617F screening, were not met, should now be considered as having polycythemia vera with a JAK2 exon 12 mutation. As highlighted in previous studies,5,6 the clinical picture of this subtype of polycythemia vera is indistinguishable from that of idiopathic erythrocytosis. This emphasizes that JAK2 exon 12 mutations should be actively screened for in patients with idiopathic erythrocytosis. Furthermore, the finding of SH2B3 variants highlights that this gene should also be surveyed, which is currently not done routinely. The erythrocytosis 1316

gene panel can successfully do both. Thus, we demonstrated that the panel allows reliable detection of known erythrocytosis-causing mutations, avoiding pitfalls that may occur when following existing algorithms. In this study, four out of the 125 patients were heterozygous for VHL p.R200W. In the homozygous state, this variant causes Chuvash polycythemia.41,42 Congenital erythrocytosis also occurs in patients who are compound heterozygotes,43-45 but heterozygous carriers are usually unaffected. Nevertheless, VHL p.R200W heterozygous mutations feature significantly more frequently in erythrocytosis databases3 than in general populations,46 suggesting a causal role for this mutation. For one of the four patients here, the variant was newly identified. For the other three, previous genetic tests had also identified it. Thus, within this study we aimed to detect additional genetic changes that might explain the patients’ clinical phenotype. We did not detect any other variants within VHL, except for two single nucleotide polymorphisms in the 3’ untranslated region with high minor allelic frequencies (≥0.35 in dbSNP142). Alternatively, the co-occurrence of this heterozygous variant with another heterozygous variant in a separate gene of the same biological pathway could act in synergy to produce disease. We did not obtain conclusive evidence for this in our four patients: two did not have an additional variant and in the other two, the VHL p.R200W co-occurred with heterozygous missense variants classified as polymorphisms (Online Supplementary Table S6), i.e. with EGLN1 p.A157Q and EGLN2 p.T405M in one patient and with EPOR p.G46E and EGLN2 p.S58L in the other. As this research panel provides full-gene sequencing instead of specific mutation screening, it allowed the detection of 22 novel variants. For some of these, there is a strong likelihood of causality, based on the location of the mutated residues on functional or regulatory domains and the expected disturbance they would cause to protein structure and function (as explained in the Results section for EGLN1 p.L279P, EPAS1 p.Y532H and VHL p.E52X), and based on genetic evidence of familial segregation (e.g. EPAS1 p.Y532H and EPO p.P7fs, which are dominantly inherited). For other variants, mostly found in known erythrocytosis-associated genes, there is strong consensus in the in silico prediction of deleterious effect, whereas for some there is less evidence of functional candidacy (Table 3). While the functional significance of newly identified variants cannot currently be confirmed – and indeed clinical causation cannot be concluded – future functional studies and screening of larger cohorts of erythrocytosis patients are needed to replicate the findings and to provide further evidence of causality. In this study we explored some genes, not previously associated with erythrocytosis, because of their involvement in the HIF pathway or their discovery through WGS500 as potential candidates. Candidate variants were found in EGLN2 and HIF3A but not in key HIF pathway genes such as EGLN3, HIF1AN and importantly, HIF1A. This is consistent with existing literature in which variation in EPAS1, but not HIF1A, is associated with erythrocytosis. The precise WGS-identified variants in EPO, GFI1B, KDM6A and BHLHE41 were not found in this cohort of 125 cases, suggesting that larger cohorts of patients need to be sequenced before the significance of variation in these genes can be properly interpreted. However, in the case of EPO, other variants were identihaematologica | 2016; 101(11)

Gene panel for idiopathic erythrocytosis

Figure 4. Proposed use of a gene panel in the investigation of erythrocytosis. A gene panel would make genetic testing more efficient and streamlined. It enables the simultaneous survey of the full length of 21 candidate genes, in a systematic and unbiased manner, allowing the detection of known causal variants as well as novel variants in known and novel genes.

fied suggesting that EPO should be actively surveyed as an erythrocytosis-associated candidate gene. Accrued use of the panel in further patients will provide insight into which novel genes play a role in erythrocytosis and will allow refinement of any future diagnostic panels. One limitation of our study is the lack of DNA from a source other than blood to determine germline or somatic status. This would only be a concern for JAK2 and SH2B3, in which somatic mutations are associated with polycythemia vera and myeloproliferative diseases. When variants in JAK2 and SH2B3 are found by the panel, further studies in skin/nail DNA are probably warranted. For all other genes, variants detected in blood with the panel are most likely germline. While somatic mutations in EPAS1 can be found in tumors of patients with erythrocytosis,8 these would not be detectable in blood with our methodology. Thus, despite the few technical limitations described, the erythrocytosis gene panel is useful in the genetic investigation of patients with erythrocytosis from a research perspective. Furthermore, following appropriate optimization and refinement, gene panel sequencing has the potential to improve the diagnostic work-up of erythrocytosis patients in clinical practice. A point to note is that the gene panel in our study was applied to a highly-selected group of patients who had undergone significant clinical and genetic “filtering” (Online Supplementary Figure S1) before inclusion in the study. Despite this, candidate variants – known causal and novel – were detected in 29% of patients. Thus, we propose that gene panel sequencing should be applied haematologica | 2016; 101(11)

directly to “erythrocytosis cases where a genetic cause is suspected”, i.e. after clinical exclusion of acquired systemic causes and at the point where genetic testing is considered (Figure 4). This would undoubtedly increase the diagnostic yield and, because genetic testing would be conducted in an unbiased manner, it would improve diagnostic accuracy by decreasing the number of missed diagnoses. In conclusion, we hope to demonstrate the immediate utility of a targeted gene panel in the investigation of erythrocytosis at a time when next-generation sequencing is revolutionizing clinical medicine. Acknowledgments The authors would like to thank the patients and their families who consented to this study, Melissa M. Pentony for the support provided with the management of Ion Torrent data and the Core and administration services at the Wellcome Trust Centre for Human Genetics, which are funded by the Wellcome Trust Core Award [090532/Z/09/Z]. This work was supported by the National Institute for Health Research (NIHR) Biomedical Research Centre Oxford with funding from the Department of Health’s NIHR Biomedical Research Centre’s funding scheme. The WGS500 study was funded by the Wellcome Trust Core Award (090532/Z/09/Z) and a Medical Research Council Hub grant (G0900747 91070) to Peter Donnelly (director of the Wellcome Trust Centre of Human Genetics), the NIHR Biomedical Research Centre Oxford, the UK Department of Health’s NIHR Biomedical Research Centres funding scheme and Illumina. NP is funded via a NIHR Clinical Lectureship. PJR is a member of the Ludwig Institute for Cancer Research. 1317

C. Camps et al.

References 1. McMullin MF. The classification and diagnosis of erythrocytosis. Int J Lab Hematol. 2008;30(6):447-459. 2. Hussein K, Percy M, McMullin MF. Clinical utility gene card for: familial erythrocytosis. Eur J Hum Genet. 2012;20(5): 3. Bento C, Percy MJ, Gardie B, et al. Genetic basis of congenital erythrocytosis: mutation update and online databases. Hum Mutat. 2014;35(1):15-26. 4. Percy MJ, Jones FG, Green AR, Reilly JT, McMullin MF. The incidence of the JAK2 V617F mutation in patients with idiopathic erythrocytosis. Haematologica. 2006;91(3): 413-414. 5. Scott LM, Tong W, Levine RL, et al. JAK2 exon 12 mutations in polycythemia vera and idiopathic erythrocytosis. N Engl J Med. 2007;356(5):459-468. 6. Percy MJ, Scott LM, Erber WN, et al. The frequency of JAK2 exon 12 mutations in idiopathic erythrocytosis patients with low serum erythropoietin levels. Haematologica. 2007;92(12):1607-1614. 7. Lee FS, Percy MJ, McMullin MF. Oxygen sensing: recent insights from idiopathic erythrocytosis. Cell Cycle. 2006;5(9):941-945. 8. Zhuang Z, Yang C, Lorenzo F, et al. Somatic HIF2A gain-of-function mutations in paraganglioma with polycythemia. N Engl J Med. 2012;367(10):922-930. 9. McMullin MF. Idiopathic erythrocytosis: a disappearing entity. Hematology Am Soc Hematol Educ Program. 2009;629-635. 10. Percy MJ, Butt NN, Crotty GM, et al. Identification of high oxygen affinity hemoglobin variants in the investigation of patients with erythrocytosis. Haematologica. 2009;94(9):1321-1322. 11. Hoyer JD, Allen SL, Beutler E, Kubik K, West C, Fairbanks VF. Erythrocytosis due to bisphosphoglycerate mutase deficiency with concurrent glucose-6-phosphate dehydrogenase (G-6-PD) deficiency. Am J Hematol. 2004;75(4):205-208. 12. Petousi N, Copley RR, Lappin TR, et al. Erythrocytosis associated with a novel missense mutation in the BPGM gene. Haematologica. 2014;99(10):e201-204. 13. Gruber M, Hu CJ, Johnson RS, Brown EJ, Keith B, Simon MC. Acute postnatal ablation of Hif-2alpha results in anemia. Proc Natl Acad Sci USA. 2007;104(7):2301-2306. 14. Rankin EB, Biju MP, Liu Q, et al. Hypoxiainducible factor-2 (HIF-2) regulates hepatic erythropoietin in vivo. J Clin Invest. 2007;117(4):1068-1077. 15. Webb JD, Coleman ML, Pugh CW. Hypoxia, hypoxia-inducible factors (HIF), HIF hydroxylases and oxygen sensing. Cell Mol Life Sci. 2009;66(22):3539-3554. 16. Taylor JC, Martin HC, Lise S, et al. Factors influencing success of clinical genome sequencing across a broad spectrum of disorders. Nat Genet. 2015;47(7):717-726. 17. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data.

1318

Nucleic Acids Res. 2010;38(16):e164. 18. Benjamini Y, Hochberg Y. Controlling the false discovery rate - a practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol. 1995;57(1):289-300. 19. Min JH, Yang H, Ivan M, Gertler F, Kaelin WG Jr., Pavletich NP. Structure of an HIF1alpha -pVHL complex: hydroxyproline recognition in signaling. Science. 2002;296(5574):1886-1889. 20. Jang JH, Seo JY, Jang J, et al. Hereditary gene mutations in Korean patients with isolated erythrocytosis. Ann Hematol. 2014;93(6): 931-935. 21. Chowdhury R, McDonough MA, Mecinovic J, et al. Structural basis for binding of hypoxia-inducible factor to the oxygen-sensing prolyl hydroxylases. Structure. 2009;17(7):981-989. 22. van der Harst E, de Krijger RR, Dinjens WN, et al. Germline mutations in the vhl gene in patients presenting with phaeochromocytomas. Int J Cancer. 1998;77 (3):337-340. 23. Olschwang S, Richard S, Boisson C, et al. Germline mutation profile of the VHL gene in von Hippel-Lindau disease and in sporadic hemangioblastoma. Hum Mutat. 1998;12(6):424-430. 24. Dollfus H, Massin P, Taupin P, et al. Retinal hemangioblastoma in von Hippel-Lindau disease: a clinical and molecular study. Invest Ophthalmol Vis Sci. 2002;43(9): 3067-3074. 25. Friedrich CA. Genotype-phenotype correlation in von Hippel-Lindau syndrome. Hum Mol Genet. 2001;10(7):763-767. 26. Costa JL, Sousa S, Justino A, et al. Nonoptical massive parallel DNA sequencing of BRCA1 and BRCA2 genes in a diagnostic setting. Hum Mutat. 2013;34(4):629-635. 27. Junemann S, Sedlazeck FJ, Prior K, et al. Updating benchtop sequencing performance comparison. Nat Biotechnol. 2013;31(4):294-296. 28. Yeo ZX, Chan M, Yap YS, Ang P, Rozen S, Lee AS. Improving indel detection specificity of the Ion Torrent PGM benchtop sequencer. PLoS One. 2012;7(9):e45798. 29. Cario H, McMullin MF, Bento C, et al. Erythrocytosis in children and adolescentsclassification, characterization, and consensus recommendations for the diagnostic approach. Pediatr Blood Cancer. 2013;60 (11):1734-1738. 30. Bento C, Almeida H, Maia TM, et al. Molecular study of congenital erythrocytosis in 70 unrelated patients revealed a potential causal mutation in less than half of the cases (Where is/are the missing gene(s)?). Eur J Haematol. 2013;91(4):361368. 31. Percy MJ, Furlow PW, Lucas GS, et al. A gain-of-function mutation in the HIF2A gene in familial erythrocytosis. N Engl J Med. 2008;358(2):162-168. 32. Gale DP, Harten SK, Reid CD, Tuddenham EG, Maxwell PH. Autosomal dominant erythrocytosis and pulmonary arterial hypertension associated with an activating HIF2 alpha mutation. Blood. 2008;112 (3):919-921.

33. Tomasic NL, Piterkova L, Huff C, et al. The phenotype of polycythemia due to Croatian homozygous VHL (571C>G:H191D) mutation is different from that of Chuvash polycythemia (VHL 598C>T:R200W). Haematologica. 2013;98 (4):560-567. 34. Charache S, Jacobson R, Brimhall B, et al. Hb Potomac (101 Glu replaced by Asp): speculations on placental oxygen transport in carriers of high-affinity hemoglobins. Blood. 1978;51(2):331-338. 35. Gonzalez Fernandez FA, Villegas A, Ropero P, et al. Haemoglobinopathies with high oxygen affinity. Experience of Erythropathology Cooperative Spanish Group. Ann Hematol. 2009;88(3):235-238. 36. Bromberg PA, Alben JO, Bare GH, et al. High oxygen affinity variant of haemoglobin Little Rock with unique properties. Nat New Biol. 1973;243(127):177-179. 37. Misgeld E, Gattermann N, Wehmeier A, Weiland C, Peters U, Kohne E. Hemoglobinopathy York [beta146 (HC3) His==>Pro]: first report of a family history. Ann Hematol. 2001;80(6):365-367. 38. Wajcman H, Galacteros F. Abnormal hemoglobins with high oxygen affinity and erythrocytosis. Hematol Cell Ther. 1996;38(4): 305-312. 39. Spolverini A, Pieri L, Guglielmelli P, et al. Infrequent occurrence of mutations in the PH domain of LNK in patients with JAK2 mutation-negative 'idiopathic' erythrocytosis. Haematologica. 2013;98(9):e101-102. 40. McMullin MF, Wu C, Percy MJ, Tong W. A nonsynonymous LNK polymorphism associated with idiopathic erythrocytosis. Am J Hematol. 2011;86(11):962-964. 41. Ang SO, Chen H, Hirota K, et al. Disruption of oxygen homeostasis underlies congenital Chuvash polycythemia. Nat Genet. 2002;32(4):614-621. 42. Smith TG, Brooks JT, Balanos GM, et al. Mutation of von Hippel-Lindau tumour suppressor and human cardiopulmonary physiology. PLoS Med. 2006;3(7):e290. 43. Bento MC, Chang KT, Guan Y, et al. Congenital polycythemia with homozygous and heterozygous mutations of von HippelLindau gene: five new Caucasian patients. Haematologica. 2005;90(1):128-129. 44. Cario H, Schwarz K, Jorch N, et al. Mutations in the von Hippel-Lindau (VHL) tumor suppressor gene and VHL-haplotype analysis in patients with presumable congenital erythrocytosis. Haematologica. 2005;90(1):19-24. 45. Percy MJ, McMullin MF, Jowitt SN, et al. Chuvash-type congenital polycythemia in 4 families of Asian and Western European ancestry. Blood. 2003;102(3):1097-1099. 46. Liu E, Percy MJ, Amos CI, et al. The worldwide distribution of the VHL 598C>T mutation indicates a single founding event. Blood. 2004;103(5):1937-1940. 47. Panovska-Stavridis I, Eftimov A, PivkovaVeljanovska A, Ivanovski M, Cevreska L, Dimovski AJ. Familiar JAK2 G571S variant not linked with essential trombocythemia. Blood. 2014:124(21);558.

haematologica | 2016; 101(11)

ARTICLE

Blood Transfusion

Human neutrophil peptides and complement factor Bb in pathogenesis of acquired thrombotic thrombocytopenic purpura Wenjing Cao, Huy P. Pham, Lance A. Williams, Jenny McDaniel, Rance C. Siniard, Robin G. Lorenz, Marisa B. Marques, and X. Long Zheng

Division of Laboratory Medicine, Department of Pathology, University of Alabama at Birmingham, AL, USA

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2016 Volume 101(11):1319-1326

ABSTRACT

cquired thrombotic thrombocytopenic purpura is primarily caused by the deficiency of plasma ADAMTS13 activity resulting from autoantibodies against ADAMTS13. However, ADAMTS13 deficiency alone is often not sufficient to cause acute thrombotic thrombocytopenic purpura. Infections or systemic inflammation may precede acute bursts of the disease, but the underlying mechanisms are not fully understood. Herein, 52 patients with acquired autoimmune thrombotic thrombocytopenic purpura and 30 blood donor controls were recruited for the study. The plasma levels of human neutrophil peptides 1-3 and complement activation fragments (i.e. Bb, iC3b, C4d, and sC5b-9) were determined by enzyme-linked immunosorbent assays. Univariate analyses were performed to determine the correlation between each biomarker and clinical outcomes. We found that the plasma levels of human neutrophil peptides 1-3 and Bb in patients with acute thrombotic thrombocytopenic purpura were significantly higher than those in the control (P<0.0001). The plasma levels of HNP1-3 correlated with the levels of plasma complement fragment Bb (rho=0.48, P=0.0004) and serum lactate dehydrogenase (rho=0.28, P=0.04); in addition, the plasma levels of Bb correlated with iC3b (rho=0.55, P<0.0001), sC5b-9 (rho=0.63, P<0.0001), serum creatinine (rho=0.42, p=0.0011), and lactate dehydrogenase (rho=0.40, P=0.0034), respectively. Moreover, the plasma levels of iC3b and sC5b-9 were correlated (rho=0.72, P<0.0001), despite no statistically significant difference of the two markers between thrombotic thrombocytopenic purpura patients and the control. We conclude that innate immunity, i.e. neutrophil and complement activation via the alternative pathway, may play a role in the pathogenesis of acute autoimmune thrombotic thrombocytopenic purpura, and a therapy targeted at these pathways may be considered in a subset of these patients.

Introduction Thrombotic thrombocytopenic purpura (TTP) is characterized by the formation of disseminated microvascular thrombosis in small arterioles and capillaries.1 TTP patients manifest with severe thrombocytopenia (usually less than 20,000/ml of platelet counts), microangiopathic hemolytic anemia with elevated levels of serum lactate dehydrogenase (LDH) and schistocytes on a peripheral blood smear, and signs and symptoms of end-organ dysfunction, including renal failure and/or myocardial or cerebral infarctions.2 Severe deficiency of plasma ADAMTS13 activity (usually <10%), resulting from ADAMTS13 mutations or autoantibodies against ADAMTS13, appears to be the key pathogenic factor of TTP.3,4 ADAMTS13 is a plasma metalloprotease that cleaves von Willebrand factor (VWF) at the Tyr1605-Met1606 bond, thereby regulating hemostasis and preventing thrombosis after vascular injury.4 In patients with hereditary TTP, the lower the haematologica | 2016; 101(11)

Correspondence: xzheng@uabmc.edu

Received: May 7, 2016. Accepted: July 29, 2016. Pre-published: August 4, 2016. doi:10.3324/haematol.2016.149021

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1319

ÂŠ2016 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. A permission in writing by the publisher is required for any other use.

1319

W. Cao et al.

plasma ADAMTS13 activity is, the earlier the initial TTP episode occurs,5 suggesting that any residual ADAMTS13 activity may be protective. Treatment with plasma infusion, aimed at increasing plasma ADAMTS13 activity to greater than 5%, is clinically effective in hereditary TTP.6 In patients with acquired TTP, the immunoglobulin (IgG) type of autoantibodies, which bind primarily to the spacer domain of ADAMTS13,7 result in competitive inhibition of plasma ADAMTS13 activity. Therapeutic plasma exchange, often used in combination with immunosuppression, including corticosteroids, vincristine, cyclophosphamide, and rituximab, etc., remains the treatment of choice for acquired TTP with inhibitors. Clinical observations have also demonstrated that most hereditary TTP patients with plasma ADAMTS13 activity of less than 10% remain asymptomatic for many years before experiencing their first episode.8,9 Patients with acquired autoimmune TTP may achieve clinical remission after therapeutic plasma exchange and other adjunctive therapies, despite ongoing severe deficiency of plasma ADAMTS13 activity and the presence of inhibitors.10 These findings indicate that a triggering event may be necessary to provoke the initial onset of TTP and subsequent recurrent or relapsing episodes. For example, central catheter infection, systemic inflammation, certain medications, and pregnancy are known to be the potential inciting factors for TTP. However, the mechanisms underlying such a triggering event remain poorly understood. Based on our recently published study, wherein we demonstrated that HNP1-3, a group of 29-30 amino acid anti-microbial peptides, potentially released from activated human neutrophils, is a potent inhibitor of ADAMTS13-medidated VWF proteolysis.11 We hypothesize that the locally released HNP1-3 may play a role in the pathogenesis of acute TTP, particularly in those with severely low circulating ADAMTS13 activity. Moreover, several recent studies have indicated that complement activation may be another inciting factor that affects the onset, clinical presentation, and outcome of thrombotic microangiopathy (TMA), including TTP. For instance, serum from patients with TMA caused C3 and membrane attack complex (MAC) deposition on human microvascular endothelial cells (HMEC)-1 and its cytotoxic effect was abolished by complement inhibition.12 Additionally, plasma levels of C3a and C5a were significantly elevated in patients during acute TTP as compared with those in remission.13 Most importantly, complement factor H mutations were identified in 5 out of 6 patients with ticlopidine (anti-platelet drug)-associated TTP with severe deficiency of plasma ADAMTS13 activity.14 Together, these preliminary findings suggest that complement activation via the alternative pathway and severe ADAMTS13 deficiency may play a synergistic role in the pathogenesis of TTP. However, the relationship between the measurement of inflammatory and complement activation markers at the onset of acute TTP and clinical consequences has not been investigated in a large cohort of acquired autoimmune TTP patients.

Table 1. Demographic and clinical data of 52 patients with acquired TTP.

Sex Female Male Race White African American Hispanic CNS signs and symptoms No Yes NA Disease status Initial Rel./Refrac. Episodes One >1 Renal function Cr<1.2 Crâ&#x2030;Ľ1.2 Comorbidity No Yes NA Blood culture Negative Positive NA Corticosteroids No Yes NA Rituximab Yes No NA Outcome Remission Death

Number of cases

Percentage

27 25

51.9 48.1

13 38 1

25.0 73.1 1.9

24 26 2

46.2 50.0 3.8

40 12

76.9 23.1

28 24

53.8 46.2

33 19

63.5 36.5

12 39 1

23.1 75.0 1.9

20 6 26

38.5 11.5 50.0

7 41 4

13.5 78.8 7.7

30 18 4

57.7 34.6 7.7

43 9

82.7 13.3

n/a: not applicable or information was not available; CNS: central nervous system; Rel/Refrac.: relapse/refractory; CR: complete remission; Cr: creatinine.

was diagnosed in patients with thrombocytopenia (platelet count <150x103/mL) and microangiopathic hemolytic anemia (indicated by reduced hematocrit, increased serum LDH, and the presence of the fragmentation of red blood cells, i.e. schistocytes on the peripheral blood smear), with or without signs and/or symptoms of major organ damage and in the absence of an alternative diagnosis to explain the TMA.15 All patients underwent therapeutic plasma exchange. Blood samples anticoagulated with sodium citrate (0.32%) were obtained prior to the initiation of therapeutic plasma exchange. Control samples were obtained from healthy blood donors whose medical history details were not collected or available for the purpose of this study. Blood was centrifuged at 3,000 rpm for 15 min, and plasma was aspirated from the top and stored in aliquots at -80 ÂşC until assays were performed. All samples were frozen and thawed only once prior to the study.

Assays for plasma ADAMTS13 activity and inhibitors Methods Patients The Institutional Review Board (IRB) of the University of Alabama at Birmingham (UAB), USA, approved the study. TTP 1320

A commercial FRETS-VWF73 assay in a reference lab (Blood Center of Wisconsin, Milwaukee, WI, USA)16 and a homegrown recombinant FRETS-VWF73 (rF-VWF73) assay determined plasma ADAMTS13 activity as described previously.17 ADAMTS13 inhibitor titers were similarly determined by measuring the residual enzyme activity in normal human plasma (NHP) after being haematologica | 2016; 101(11)

The role of innate immunity in TTP Table 2. Laboratory results of TTP patients and healthy controls.

TTP patients (n=52) Median (range) Mean ± SEM

Test Results Platelet counts (x103/ L) Hct (%) WBC (x103/ L) Neutrophil (%) LDH (U/L) Creatinine (mg/dL) ADAMTS13 activity (%) HNPs1-3 (ng/mL) Bb (g/mL) iC3b (g/mL) sC5b-9 (g/mL) C4d (g/mL)

14.3 (5.4-108) 25 (14-40) 11.8 (4.5-24.5) 70 (43-90) 878.5 (226-7,332) 1.3 (0.1-8.5) 1.2 (0-8.1)*** 33.1 (12.8-239.4)*** 2.6 (0.9-9.6)*** 13.4 (3.8-58.6) 1.3 (0.5-7.4) 2.6 (1.1-14.0)

Healthy Controls (n=30) Median (range) Mean ± SEM

19.1 ± 2.6 24.2 ± 0.7 12.4 ± 0.7 69.9 ± 1.5 1136.6 ± 150.3 1.6 ± 0.2 2.0 ± 0.3*** 47.1 ± 5.6*** 3.1 ± 0.3*** 15.6 ± 1.4 1.7 ± 0.1 3.3 ± 0.2

117.7 (84.6-228.1) 2.4 (1.8-14.5) 1.3 (0.8-2.5) 12.5 (6.1-22.2) 1.6 (0.3-3.0) 2.3 (1.4-4.2)

122.5 ± 5,8 3.5 ± 0.6 1.4 ± 0.1 12.1 ± 0.6 1.7 ± 0.1 2.4 ± 0.2

***P values<0.0001, the reference ranges for platelet count 150-400x103/L, hematocrit (Hct) 33-45%, white blood cells (WBC) 4-11x103/mL, neutrophil 35-73%, lactate dehydrogenase (LDH) 120-240 U/L, creatinine 0.4-1.2 mg/dL; SEM: standard error of the mean; TTP: thrombotic thrombocytopenic purpura; HNPs: human neutrophil peptides.

mixed (50:50) and incubated for 30 min with the patient’s plasma (various dilutions) at 37 ºC, using the prior mentioned commercial FRETS-VWF73 assay in the reference lab.

P<0.0001

Assays for plasma complement activation markers Enzyme-linked immunosorbent assay (ELISA) kits for complement activation markers, including iC3b, sC5b-9, Bb and C4d, were obtained from a commercial source (MicroVue, San Diego, CA, USA), and the assays were performed on the diluted plasma samples according to the manufacturers’ recommendations.

Statistical analysis

A Nonparametric Mann-Whitney test was used for comparison between the two groups. A nonparametric Spearman rank correlation (rho) was determined between the parameters. The P values <0.05 and <0.01 were considered statistically significant and highly significant, respectively.

P<0.0001

Results Characteristics of TTP patients A total of 52 patients with acquired TTP were recruited into the study. The detailed demographic information, clinical data, and laboratory results are shown in the Online Supplementary Table S1. Of these, the female to male ratio was 27/25 with 73% being African-American, 25% Caucasian and 2% Hispanic. Comorbidities, including obesity, diabetes mellitus, hypertension, rheumatoid arthritis, systemic lupus erythematosus, etc., were present in 75% of patients. The majority (77%) of patients presented with an initial episode of TTP, and the remainder (23%) presented with a relapse or exacerbation. Nearly 54% of patients had only one episode and 46% had ≥2 episodes. The blood culture performed was positive in 6/26 (23%) of cases (Table 1). Other laboratory tests on admission showed a median platelet count of 14.3 (4.7108)x103/mL, a median hematocrit of 25% (14-40%), a median white blood cell count of 11.8 (4.5-24.5) x103/mL, a median serum LDH of 878.6 (226-7,332) U/dL, and a median serum creatinine of 1.3 (0.1-8.5) mg/dL. The median plasma ADAMTS13 activity was 1.2% (0-8%), which was significantly lower than the control levels of 118% (85-226%) (Online Supplementary Table S1, Table 2, and Figure 1A). The mixing study identified inhibitors (>0.4 haematologica | 2016; 101(11)

C P<0.0001

Figure 1. Plasma ADAMTS13 activity and HNP1-3 in patients with TTP and controls. Plasma ADAMTS13 activity (A), anti-ADAMTS13 IgG (B), and HNP1-3 (C) in TTP patients and controls were determined by the cleavage of rF-VWF73 and commercial ELISAs according to manufacturer’s recommendation. A nonparametric Mann-Whitney U test was performed. The P values <0.0001 are highly statistically significant. The horizontal lines within the dots represent the geometric mean ± 95% confidential interval; TTP: thrombotic thrombocytopenic purpura; HNP: human neutrophil peptide; ELISA: enzyme-linked immunosorbent assay. IgG; immunoglobin G.

1321

W. Cao et al. B

A P<0.0001

P=0.31

C P=0.05

U/ml) in 90% of TTP patients with a median titer of 2 (0.4 to >8 U/ml) (Online Supplementary Table S1 and Figure 1B). An immunoassay confirmed the presence of antiADAMTS13 IgG in all TTP patients including 6 patients with <0.4 U/ml by the mixing assay (Online Supplementary Table S1). These results support the fact that all patients with ADAMTS13 activity <10% in the study cohort had an acquired autoimmune TTP. Therapeutic plasma exchange, corticosteroids, and rituximab were offered on admission to 100%, 79%, and 58% of patients, respectively. Follow-up results indicated that the mortality rate was 19.9%, with all but one patient dying within 30 days of admission (Online Supplementary Table S1 and Table 1), consistent with results reported by other groups.18,19

Plasma HNP1-3 in TTP patients and controls Neutrophil activation is common in patients with acute TTP.20,21 To determine if HNP1-3 are released from activated neutrophils, we measured their plasma concentrations in patients with acute TTP by an immunoassay. As shown in Table 2 and Figure 1C, the median plasma level of HNP1-3 in acute TTP patients was 33.1 (12.8-239.4) ng/ml, significantly higher than the median level of 2.4 (1.8-14.5) in the blood donor controls (P<0.0001). Plasma HNP1-3 was independent of total white blood cell counts (rho=0.10, P=0.47), the percentage of neutrophils (rho=0.13, P=0.36), and absolute neutrophil counts (rho=0.15, P=0.28) (Online Supplementary Table S1). These results suggest that neutrophil activation rather than neutrophil counts contribute to the increased levels of plasma HNP in patients with acute TTP. 1322

P=0.15

Figure 2. Plasma levels of complement fragments Bb, C4d, iC3b, and sC5b-9 in TTP patients and controls. Commercial ELISAs were used to determine plasma levels of Bb (A), C4d (B), iC3b (C), and sC5b-9 complexes (D) according to the manufacturer's protocol. The horizontal lines within the dots represent the geometric means Âą 95% confidential intervals. A Mann-Whitney U test was performed and the P value (<0.0001) in panel A is statistically highly significant. The P values in all other panels are greater than 0.05, not statistically significant. TTP: thrombotic thrombocytopenic purpura; ELISA: enzyme-linked immunosorbent assay.

Plasma complement activation fragments in TTP and controls Complement activation is the primary cause of atypical hemolytic uremic syndrome (aHUS),22 but it may also be a risk factor for the development of acquired TTP. Plasma complement fragments generated as a consequence of activation via the classical pathway (C4d), the alternative pathway (Bb), both the pathways of (iC3b), and the formations of the terminal complexes (sC5b-9)22 were determined by immunoassays in 52 TTP plasma samples. The results showed that the median level of plasma Bb in TTP patients was 2.6 (0.9-9.6) mg/mL, significantly higher than the median level of 1.3 (0.8-2.5) mg/mL in the blood donor controls (P<0.0001) (Table 2 and Figure 2A). A small fraction of TTP patients exhibited high levels of plasma C4d (Figure 2B), iC3b (Figure 2C), and sC5b-9 (Figure 2D), but there was no statistically significant difference of these fragments in TTP samples as a group when compared with those in the controls (Table 2 and Figure 2). These results indicate that complement activation via the alternative pathway is present in a subset of patients with acute autoimmune TTP.

Correlations between HNP1-3 or Bb and other key clinical and laboratory parameters To determine the relationship between inflammation, complement activation and end-organ damage during acute episodes, Spearman rank correlation coefficients (rho) were determined between HNP1-3 or Bb and various other biomarkers and clinical parameters. Our results showed that in TTP patients their plasma levels of HNP1-3 correlated with plasma levels of Bb (rho=0.48, P<0.001) (Figure 3A) and serum LDH (rho=0.28, P<0.05) haematologica | 2016; 101(11)

The role of innate immunity in TTP

A rho=0.48 P<0.001

rho=0.28 P<0.05

C rho=0.10 P>0.05

(Figure 3B), but not with the inhibitor titer (P>0.05) (Figure 3C) and serum creatinine (P>0.05) (Figure 3D). As expected, plasma Bb in these patients significantly correlated with iC3b (rho=0.55, P<0.0001) (Figure 4A), sC5b-9 (r=0.64, P<0.0001) (Figure 4B), serum LDH (rho=0.40, P=0.003) (Figure 4C), and serum creatinine (rho=0.42, P=0.0011) (Figure 4D), respectively. In addition, plasma levels of iC3b in TTP patients highly correlated with plasma sC5b-9 (rho=0.72, P<0.0001) (Figure 4E), but not with serum creatinine (Figure 4F) and LDH (Figure 4G). No statistically significant correlation was observed between plasma levels of sC5b-9 and serum creatinine (Figure 4H). None of these biomarkers measured in this cohort of acute TTP samples predicted the relapse and mortality rate (data not shown). Nevertheless, these results suggest that innate immunity, including neutrophil activation and complement activation via an alternative pathway, may participate in the onset and progress of the disease in a subset of TTP patients.

Discussion The present study demonstrates the significant increase in plasma HNP1-3 and Bb in patients with acute autoimmune TTP when compared with the control; the increased plasma levels of HNP1-3 correlate with Bb, which, in turn, correlates with iC3b and sC5b-9, as well as serum LDH and/or creatinine. This is, to our knowledge, by far the largest and the most comprehensive analysis of the relationship between the innate immunity (i.e. neutrophil activation and complement activation) and the disease onset, progression, and long-term outcome in patients with TTP. haematologica | 2016; 101(11)

rho=0.18 P>0.05

Figure 3. Correlations between HNP1-3 and the other key laboratory parameters. Nonparametric Spearman rank correlation coefficient (rho) tests were performed to determine the correlations between plasma levels of HNP1-3 and Bb (A), serum LDH (B), anti-ADAMTS13 inhibitor titer (C), and serum creatinine (D), respectively, in patients with acquired TTP. The P values in panel A (P<0.001) and panel B (P<0.05) are considered to be statistically highly significant and significant, respectively. The P values in panels C and D (â&#x2030;Ľ 0.05) are not statistically significant. HNP: human neutrophil peptide; LDH: lactate dehydrogenase; rho: Spearman rank correlation coefficient.

How could HNP1-3 contribute to the onset of TTP? Nearly all acquired TTP with severe deficiency of plasma ADAMTS13 activity (< 10% of normal) is caused by autoantibodies against ADAMTS13.23,24 The autoantibodies bind to the spacer domain of ADAMTS13, which may physically block its substrate recognition.7 The inability to cleave newly released ultralarge (UL) VWF results in the accumulation of ULVWF polymers on endothelial cells25 and in the circulation.26 Consequently, the ULVWF multimers serve as templates for rapidly recruiting platelets,27,28 neutrophils,28,29 and complement components30 from the circulation to the sites of vascular injury. Platelets play critical roles in both hemostasis and inflammation. Studies have shown that platelet surface glycoprotein (GP) Ib-VI31 and p-selectin32 or ADP,33 released from activated platelets, may provide receptors or signal for recruiting neutrophils and monocytes. The accumulation and activation/degranulation of neutrophils may result in the massive release of granular contents, including neutrophil extracellular traps (NETs)20 and HNP1-334 at the sites of vascular injury. HNP1-3 is highly cationic and has hydrophobic peptides, which can bind to various plasma proteins and cellular components. Only a small fraction of HNP1-3 may be able to escape from the sites of their release and circulate in plasma. Therefore, 7-10 fold increases of plasma HNP1-3 in our cohort of TTP patients may reflect the massive local release of HNP1-3 in situ from activated and accumulated neutrophils. HNP1-3 exhibits a broad-spectrum of bactericidal activity contributing to human innate immunity. The positive charge of amino acid side chains is responsible for the initial interaction with negatively charged bacterial membranes, resulting in cytotoxicity. HNP1-3 may also have a variety of other biological functions including activation of 1323

W. Cao et al. platelets,35 stimulation of inflammatory responses and cytokine release,36 and inhibition of fibrinolysis.37 In addition, HNP1-3 has been shown to cause endothelial dysfunction and increased endothelial permeability.38 More recently, we demonstrated that HNP1-3 inhibits proteolytic cleavage of VWF by ADAMTS13.11 Dramatic increases

A rho=0.55 P<0.0001

in plasma HNP1-3 are also reported in patients with myocardial infarction,39,40 systemic lupus erythematosus,41 and septic meningitis,42 suggesting that HNP may also play a role in the pathogenesis of other inflammatory and thrombotic disorders. In TTP patients, we speculate that the massive release of HNP1-3 at the sites of vascular

D rho=0.42 P=0.0011

rho=0.40 P<0.01

rho=0.63 P<0.0001

rho=0.72 P<0.0001

rho=0.03 P=0.83

H rho=0.19 P=0.18

rho=0.26 P=0.04

Figure 4. Correlations between complement fragments and the severity of end-organ damage. Nonparametric Spearman rank correlation coefficients (rho) were determined between plasma Bb and iC3b (A), between Bb and sC5b-9 (B), between Bb and serum LDH (C), and between Bb and creatinine (D). Also, the Spearman rank correlation coefficients between iC3b and sC5b-9 (E), between iC3b and serum creatinine (F), between iC3b and serum LDH (G), and between sC5b-9 and creatinine (H) were determined. The P values < 0.05 and < 0.01 are considered to be statistically significant and highly significant, respectively. LDH: lactate dehydrogenase; rho; Spearman rank correlation coefficient; Cr: creatinine.

1324

haematologica | 2016; 101(11)

The role of innate immunity in TTP

injury may be extremely detrimental when the circulating ADAMTS13 is limited. The finding that plasma HNP1-3 correlates with patient’s serum LDH, indicative of tissue ischemia in TTP,43 supports the potential involvement of HNP in the pathogenesis of TTP. How may HNP and complement activation be related? HNP1 has been shown to inhibit the classical and lectin pathways of complement activation.44 However, its relationship with the alternative pathway is not known. The levels of plasma HNP1-3 appear to correlate with plasma Bb, suggesting that the HNP1-3 may activate complement or vice versa. While our results did not find a statistically significant increase in iC3b and sC5b-9 complexes in TTP patients when compared with healthy controls, other studies have demonstrated the increased levels of C3a and sC5b-9 during acute TTP.45 Moreover, the increased levels of Bb, C3a, and C5a, and sC5b-9 appear to correlate with a worse outcome in TTP.46 We did not find an association between any of these markers we measured and the relapse and mortality rate in our cohort of patients. The reasons for the discrepant results of plasma C5b-9 compared to the literature may be multifactorial, including sample collection and storage (freeze and thaw), assay methodology (dilution factor), and patient population (African vs. European ancestry). However, we do demonstrate the significant association between plasma levels of Bb and the evidence of organ tissue ischemia (i.e. serum LDH and creatinine), although the

References 1. Hosler GA, Cusumano AM, Hutchins GM. Thrombotic thrombocytopenic purpura and hemolytic uremic syndrome are distinct pathologic entities. A review of 56 autopsy cases. Arch Pathol Lab Med. 2003;127(7):834-839. 2. Bell WR, Braine HG, Ness PM, Kickler TS. Improved survival in thrombotic thrombocytopenic purpura-hemolytic uremic syndrome. Clinical experience in 108 patients. N Engl J Med. 1991;325(6):398-403. 3. Tsai HM. Thrombotic thrombocytopenic purpura: a thrombotic disorder caused by ADAMTS13 deficiency. Hematol Oncol Clin North Am. 2007;21(4):609-632. 4. Zheng XL. ADAMTS13 and von Willebrand Factor in Thrombotic Thrombocytopenic Purpura Annu Rev Med. 2015;66:211-225. 5. Lotta LA, Wu HM, Mackie IJ, et al. Residual plasmatic activity of ADAMTS13 is correlated with phenotype severity in congenital thrombotic thrombocytopenic purpura. Blood. 2012;120(2):440-448. 6. Barbot J, Costa E, Guerra M, et al. Ten years of prophylactic treatment with freshfrozen plasma in a child with chronic relapsing thrombotic thrombocytopenic purpura as a result of a congenital deficiency of von Willebrand factor-cleaving protease. Br J Haematol. 2001;113(3):649-651. 7. Casina V, Hu WB, Mao JH, et al. High Resolution Epitope Mapping by HX MS Reveals the Pathogenic Mechanism and a Possible Therapy for Autoimmune TTP

haematologica | 2016; 101(11)

levels of plasma complement activation components are variable in TTP patients. Such a variation may be partially attributed to the comorbidity, but perhaps to other unknown genetic predispositions in a subset of patients. More interestingly, our unpublished preliminary results demonstrated that an administration of anti-complement factor H IgG (mAb7.1) in Adamts13–/– and wild-type mice, which provokes complement activation via the alternative pathway, resulted in more severe thrombocytopenia in Adamts13–/– mice than in wild-type mice, but a similar degree of renal insufficiency in an otherwise TMA-resistant mouse strain (C57BL/6),1 suggesting a potential causative role of complement activation rather than merely a biomarker of acute TTP. In conclusion, we found significantly increased levels of plasma HNP1-3 and Bb in patients with acute autoimmune TTP; plasma levels of HNP1-3 and/or Bb correlated with the downstream complement activation markers and organ tissue ischemia in TTP patients. Our findings may provide the molecular basis for additional targeted therapies, including the blockage of neutrophil activation and degranulation with colchicine and complement activation with eculizumab, in a subset of patients with acquired autoimmune TTP. Funding This work is partially supported by Answering T.T.P. foundation, Institutional Adam’s resident research grant, and NIHR01HL115187-01A1.

Syndrome. PNAS. 2015;112(31):9620-9625. 8. Mise K, Ubara Y, Matsumoto M, et al. Long term follow up of congenital thrombotic thrombocytopenic purpura (UpshawSchulman syndrome) on hemodialysis for 19 years: a case report. BMC Nephrol. 2013;14:156. 9. Fujimura Y, Matsumoto M, Isonishi A, et al. Natural history of Upshaw-Schulman syndrome based on ADAMTS13 gene analysis in Japan. J Thromb Haemost. 2011;9 Suppl 1:283-301. 10. Zheng XL, Kaufman RM, Goodnough LT, Sadler JE. Effect of plasma exchange on plasma ADAMTS13 metalloprotease activity, inhibitor level, and clinical outcome in patients with idiopathic and nonidiopathic thrombotic thrombocytopenic purpura. Blood. 2004;103(11):4043-4049. 11. Pillai VG, Bao J, Zander CB, et al. Human neutrophil peptides inhibit cleavage of von Willebrand factor by ADAMTS13: a potential link of inflammation to TTP. Blood. 2016;128(1):110-119. 12. Ruiz-Torres MP, Casiraghi F, Galbusera M, et al. Complement activation: the missing link between ADAMTS-13 deficiency and microvascular thrombosis of thrombotic microangiopathies. Thromb Haemost. 2005;93(3):443-452. 13. Westwood JP, Langley K, Heelas E, Machin SJ, Scully M. Complement and cytokine response in acute Thrombotic Thrombocytopenic Purpura. Br J Haematol. 2014;164(6):858-866. 14. Chapin J, Eyler S, Smith R, Tsai HM, Laurence J. Complement factor H mutations are present in ADAMTS13-deficient, ticlopidine-associated thrombotic microan-

giopathies. Blood. 2013;121(19):4012-4013. 15. George JN. How I treat patients with thrombotic thrombocytopenic purpura: 2010. Blood. 2010;116(20):4060-4069. 16. Kokame K, Nobe Y, Kokubo Y, Okayama A, Miyata T. FRETS-VWF73, a first fluorogenic substrate for ADAMTS13 assay. Br J Haematol. 2005;129(1):93-100. 17. Raife TJ, Cao W, Atkinson BS, et al. Leukocyte proteases cleave von Willebrand factor at or near the ADAMTS13 cleavage site. Blood. 2009;114(8):1666-1674. 18. Deford CC, Reese JA, Schwartz LH, et al. Multiple major morbidities and increased mortality during long-term follow-up after recovery from thrombotic thrombocytopenic purpura. Blood. 2013;122(12):20232029; quiz 2142. 19. Korkmaz S, Keklik M, Sivgin S, et al. Therapeutic plasma exchange in patients with thrombotic thrombocytopenic purpura: a retrospective multicenter study. Transfus Apher Sci. 2013;48(3):353-358. 20. Fuchs TA, Kremer Hovinga JA, Schatzberg D, Wagner DD, Lammle B. Circulating DNA and myeloperoxidase indicate disease activity in patients with thrombotic microangiopathies. Blood. 2012; 120(6):1157-1164. 21. Mikes B, Sinkovits G, Farkas P, et al. Elevated plasma neutrophil elastase concentration is associated with disease activity in patients with thrombotic thrombocytopenic purpura. Thromb Res. 2014;133(4):616-621. 22. Noris M, Mescia F, Remuzzi G. STEC-HUS, atypical HUS and TTP are all diseases of complement activation. Nat Rev Nephrol. 2012;8(11):622-633.

1325

W. Cao et al. 23. Zheng X, Pallera AM, Goodnough LT, Sadler JE, Blinder MA. Remission of chronic thrombotic thrombocytopenic purpura after treatment with cyclophosphamide and rituximab. Ann Intern Med. 2003;138 (2):105-108. 24. Li D, Xiao J, Paessler M, Zheng XL. Novel recombinant glycosylphosphatidylinositol (GPI)-anchored ADAMTS13 and variants for assessment of anti-ADAMTS13 autoantibodies in patients with thrombotic thrombocytopenic purpura. Thromb Haemost. 2011;106(5):947-958. 25. Dong JF, Moake JL, Nolasco L, et al. ADAMTS-13 rapidly cleaves newly secreted ultralarge von Willebrand factor multimers on the endothelial surface under flowing conditions. Blood. 2002;100(12):40334039. 26. Moake JL, Rudy CK, Troll JH, et al. Unusually large plasma factor VIII:von Willebrand factor multimers in chronic relapsing thrombotic thrombocytopenic purpura. N Engl J Med. 1982;307(23):14321435. 27. Bernardo A, Ball C, Nolasco L, et al. Platelets adhered to endothelial cell-bound ultra-large von Willebrand factor strings support leukocyte tethering and rolling under high shear stress. J Thromb Haemost. 2005;3(3):562-570. 28. Chauhan AK, Motto DG, Lamb CB, et al. Systemic antithrombotic effects of ADAMTS13. J Exp Med. 2006;203(3):767776. 29. Gandhi C, Khan MM, Lentz SR, Chauhan AK. ADAMTS13 reduces vascular inflammation and the development of early atherosclerosis in mice. Blood. 2012;119 (10):2385-2391. 30. Turner NA, Moake J. Assembly and activa-

1326

31.

32.

33.

34. 35.

36.

37.

38.

tion of alternative complement components on endothelial cell-anchored ultralarge von Willebrand factor links complement and hemostasis-thrombosis. PloS one. 2013;8(3):e59372. Corken A, Russell S, Dent J, Post SR, Ware J. Platelet glycoprotein Ib-IX as a regulator of systemic inflammation. Arterioscler Thromb Vasc Biol. 2014;34(5):996-1001. Downing LJ, Wakefield TW, Strieter RM, et al. Anti-P-selectin antibody decreases inflammation and thrombus formation in venous thrombosis. J Vasc Surg. 1997;25(5):816-827. Song ZF, Chen DY, Du B, Ji XP. Poly (ADPribose) polymerase inhibitor reduces heart ischaemia/reperfusion injury via inflammation and Akt signalling in rats. Chin Med J (Engl). 2013;126(10):1913-1917. Zhao L, Lu W. Defensins in innate immunity. Curr Opin Hematol. 2014;21(1):37-42. Horn M, Bertling A, Brodde MF, et al. Human neutrophil alpha-defensins induce formation of fibrinogen and thrombospondin-1 amyloid-like structures and activate platelets via glycoprotein IIb/IIIa. J Thromb Haemost. 2012;10(4):647-661. Chaly YV, Paleolog EM, Kolesnikova TS, et al. Neutrophil alpha-defensin human neutrophil peptide modulates cytokine production in human monocytes and adhesion molecule expression in endothelial cells. Eur Cytokine Netw. 2000;11(2):257-266. Higazi AA, Ganz T, Kariko K, Cines DB. Defensin modulates tissue-type plasminogen activator and plasminogen binding to fibrin and endothelial cells. J Biol Chem. 1996;271(30):17650-17655. Kougias P, Chai H, Lin PH, et al. Neutrophil antimicrobial peptide alpha-defensin causes endothelial dysfunction in porcine coro-

39.

40.

41.

42.

43.

44.

45.

46.

nary arteries. J Vasc Surg. 2006;43(2):357363. Quinn K, Henriques M, Parker T, Slutsky AS, Zhang H. Human neutrophil peptides: a novel potential mediator of inflammatory cardiovascular diseases. Am J Physiol Heart Circ Physiol. 2008;295(5):H1817-1824. Zhao H, Yan H, Yamashita S, et al. Acute ST-segment elevation myocardial infarction is associated with decreased human antimicrobial peptide LL-37 and increased human neutrophil peptide-1 to 3 in plasma. J Atheroscler Thromb. 2012;19(4):357-368. Sthoeger ZM, Bezalel S, Chapnik N, Asher I, Froy O. High alpha-defensin levels in patients with systemic lupus erythematosus. Immunology. 2009;127(1):116-122. Panyutich AV, Panyutich EA, Krapivin VA, Baturevich EA, Ganz T. Plasma defensin concentrations are elevated in patients with septicemia or bacterial meningitis. J Lab Clin Med. 1993;122(2):202-207. Cohen JA, Brecher ME, Bandarenko N. Cellular source of serum lactate dehydrogenase elevation in patients with thrombotic thrombocytopenic purpura. J Clin Apher. 1998;13(1):16-19. Groeneveld TW, Ramwadhdoebe TH, Trouw LA, et al. Human neutrophil peptide-1 inhibits both the classical and the lectin pathway of complement activation. Mol Immunol. 2007;44(14):3608-3614. Reti M, Farkas P, Csuka D, et al. Complement activation in thrombotic thrombocytopenic purpura. J Thromb Haemost. 2012;10(5):791-798. Wu TC, Yang S, Haven S, et al. Complement activation and mortality during an acute episode of thrombotic thrombocytopenic purpura. J Thromb Haemost. 2013;11(10):1925-1927.

haematologica | 2016; 101(11)

ARTICLE

Platelet Biology & its Disorders

Comparison of two dosing schedules for subcutaneous injections of low-dose anti-CD20 veltuzumab in relapsed immune thrombocytopenia

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Howard A. Liebman,1 Mansoor N. Saleh,2 James B. Bussel,3 O. George Negrea,4 Heather Horne,5 William A. Wegener,5 and David M. Goldenberg5

Internal Medicine, Jane Anne Nohl Division of Hematology, Keck School of Medicine of University of Southern California, Los Angeles, CA; 2Georgia Cancer Specialists, Atlanta, GA; 3 Platelet Disorders Center, Division of Pediatric Hematology-Oncology, New York Presbyterian Hospital, NY; 4Low Country Cancer Care Assoc., Savannah, GA; and 5 Immunomedics, Inc., Morris Plains, NJ, USA 1

Haematologica 2016 Volume 101(11):1327-1332

ABSTRACT

e compared two dosing schedules for subcutaneous injections of a low-dose humanized anti-CD20 antibody, veltuzumab, in immune thrombocytopenia. Fifty adults with primary immune thrombocytopenia, in whom one or more lines of standard therapy had failed and who had a platelet count <30x109/L but no major bleeding, initially received escalating 80, 160, or 320 mg doses of subcutaneous veltuzumab administered twice, 2 weeks apart; the last group received once-weekly doses of 320 mg for 4 weeks. In all dose groups, injection reactions were transient and mild to moderate; there were no other safety issues. Forty-seven response-evaluable patients had 23 (49%) objective responses (platelet counts ≥30x109/L and ≥2 x baseline) including 15 (32%) complete responses (platelets ≥100x109/L). Responses (including complete responses) and bleeding reduction occurred in all dose groups and were not dose-dependent. In contrast, response duration increased progressively with total dose, reaching a median of 2.7 years with the four once-weekly 320-mg doses. Among nine responders retreated at relapse, three at higher dose levels responded again, including one patient who was retreated four times. In all dose groups, B-cell depletion occurred after the first dose until recovery starting 12 to 16 weeks after treatment. Veltuzumab serum levels increased with dose group according to total dose administered, but terminal half-life and clearance were comparable. Human anti-veltuzumab antibody titers developed without apparent dose dependence in nine patients, of whom six responded including five who had complete responses. Subcutaneous veltuzumab was convenient, well-tolerated, and active, without causing significant safety concerns. Platelet responses and bleeding reduction occurred in all dose groups, and response durability appeared to improve with higher doses. Clinicaltrials.gov identifier: NCT00547066

Correspondence: liebman@usc.edu or dmg.gscancer@att.net

Received: March 29, 2016. Accepted: August 4, 2016. Pre-published: August 11, 2016. doi:10.3324/haematol.2016.146738

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1327

Introduction Veltuzumab is a second-generation, humanized, anti-CD20 antibody with structural and functional differences from rituximab.1 Clinical studies in B-cell malignancies found that relatively low doses of this antibody were effective when administered by intravenous infusion2 or subcutaneous (SC) injection.3,4 Case reports in systemic lupus erythematosus5 and pemphigus vulgaris6 indicated that veltuzumab might also be effective in autoimmune disease. We therefore undertook a clinical study in immune thrombocytopenia (ITP), previously reporting that low doses of intravenous or SC veltuzumab could improve platelet counts when administered twice 2 weeks apart,7 haematologica | 2016; 101(11)

1327

H.A. Liebman et al.

a dosing schedule used with rituximab in rheumatoid arthritis and other autoimmune diseases. With veltuzumab formulated at 80 mg/mL, patients received their SC injections at one of three planned increasing dose levels of 80, 160 or 320 mg (delivered by 1 mL, 2 mL, or two 2-mL injections given separately several minutes apart, respectively). No dose-limiting toxicity was encountered during the initial dose escalation and this portion of the study was completed by adding several other patients to provide additional experience. We subsequently amended this study to also evaluate SC veltuzumab administered once-weekly for 4 consecutive weeks, a dosing schedule often used with rituximab in hematologic malignancies and in ITP as well, and for this part of the study only the highest dose level of 320 mg was evaluated. With platelet follow-up data now mature, we report the final results for all ITP patients treated with SC veltuzumab, comparing platelet count improvements, bleeding reduction, and other metrics across both dosing schedules.

Methods This was an open-label, multi-center, phase I study of patients with relapsed ITP who received either 80, 160, or 320 mg doses of SC veltuzumab administered twice, 2 weeks apart, or else onceweekly 320 mg doses for 4 consecutive weeks. Eligible patients were ≥18 years old with primary ITP according to American Society of Hematology guidelines8 in whom one or more standard ITP therapies had failed and who had platelet counts <30x109/L on two occasions at least 1 week apart. Patients could be recruited irrespectively of whether they had or had not undergone splenectomy and in all stages of their disease (newly-diagnosed, <3 months; persistent, 3-12 months; or chronic, >1 year).9 Using the ITP Bleeding Scale (IBLS) with bleeding at any anatomic site graded as 0 (none), 1 (mild), or marked (2),10 patients with marked bleeding were excluded, as were patients with other significant cytopenias (patients with Evans syndrome, etc.). Patients had to be off ITP medications, except for prednisone ≤20 mg/day and danazol, which were allowed if the patients were continuing on stable doses. Patients who had previously been treated with rituximab could enter the trial only if they had achieved at least a partial response for ≥6 months and were either 1 year beyond rituximab therapy or had evidence of B-cell recovery. The treatment response to SC veltuzumab was based upon a patient’s best platelet count in the absence of major bleeding or rescue interventions. An objective response was defined as platelet counts ≥30x109/L twice, 1 week apart, and at least doubled from baseline. Objective responses were categorized as a complete response if the platelet counts were ≥100x109/L twice, 1 week apart, and otherwise as a partial response.9 Time to response was measured from the first injection to onset of the objective response. Responders were followed up to 5 years, with time to relapse measured from first injection to first occurrence of platelet counts <30x109/L on at least two separate occasions at least 1 day apart. Adverse events and safety with regards to laboratory findings were classified by NCI CTC v3 toxicity grades and bleeding by IBLS grades. Blood B-cell levels (CD19) were used to determine the pharmacodynamics of the drug. For pharmacokinetics and immunogenicity, enzyme-linked immunosorbent assays performed by the sponsor measured veltuzumab serum levels (lower level of quantitation, 0.5 mg/mL) and titers of any human anti-veltuzumab antibody (HAHA) (lower level of quantitation, 50 ng/mL), respectively. Pharmacokinetic parameters following the last injection were determined by WinNonLin 2.1 (Pharsight 1328

Corporation, Mountain View, CA, USA) using a non-compartmental model. Time to relapse was analyzed by Kaplan-Meier methods. Other study results are summarized using descriptive statistics. At each participating institution, the governing ethics committee approved the study, and written informed consent was obtained from all patients.

Results Fifty patients received veltuzumab administered by SC injection. Only two patients had newly-diagnosed ITP (<3 months), while 12 had persistent ITP (3-12 months). Since these 14 patients had previously been treated with only steroids and/or immunoglobulins, they were combined into one group (ITP ≤1 year) for the purposes of comparison with the remaining 36 patients who had chronic disease (ITP >1 year) and had received additional therapies. The demographics and baseline characteristics of the patients are summarized in Table 1. All 50 patients completed their scheduled treatment, receiving veltuzumab twice, 2 weeks apart, at a dose of 80 (n=9), 160 (n=10), or 320 mg (n=15), or once-weekly 320 mg doses for 4 consecutive weeks (n=16). At the investigators’ discretion, 44% of patients (22/50) received acetaminophen alone or with diphenhydramine prior to at least one injection, and two patients also received benzodiazepines for anxiety, but no steroids or other premedication was given.

Platelet response Three patients, while receiving SC injections, initiated high-dose prednisone, cyclosporine, or romiplostim because of their extremely low platelet counts which continued to decrease at the start of treatment. This early use

Table 1. Demographics and baseline characteristics of the patients.

All patients ITP ≤ 1 year ITP > 1 year (n=50) (n=14) (n=36)*

Sex (male/female) Median age (years) Median platelet count at treatment initiation (x109/L) Maximum bleeding score (n) 0 1† 2‡ Currently on steroids (n)§ Prior systemic treatments (n) Steroids IVIG or anti-Rho(D) Azathoprine or danazol TPO-receptor agonists Splenectomy Rituximab Chemotherapy

19/31 54 22

4/10 54 23

15/21 54 22

15 33 2 8

5 7 2 2

10 26 0 6

47 32 15 10 9 7 5

14 8 4 1 0 0 0

33 24 11 9 9 7 5

*ITP duration: 1.2 - 36.6 years. §Eight patients on corticosteroids at study entry discontinued doses prior to receiving veltuzumab (n=1), transiently increased prednisone to 20 mg/day for 5 days for a platelet count of 6x109/L at the start of treatment (n=1), or continued unchanged or tapered (n=6). One patient was on danazol at study entry, none was on azathioprine or mycophenolate. † 33 patients with 44 sites involved: skin (1-5 bruises and/or scattered petechiae, n=26), epistaxis (< 5 min per episode, n=8), oral (gum bleeding < 5 min or 1 blood blister or >5 petechiae, n=3), microscopic hematuria (n=3), spotting outside normal period (n=2), occult gastrointestinal bleeding (n=1), subconjunctival hemorrhage, n=1). ‡2 patients granted waivers for epistaxis >5 min. and >5 bruises with a size >2 cm. not serious enough to preclude participation.

haematologica | 2016; 101(11)

Subcutaneous veltuzumab in ITP

of rescue medications resulted in prolonged platelet improvements but precluded unequivocal assessment of the contribution of veltuzumab to the response. The other 47 patients could be evaluated for platelet response to veltuzumab. These included 23 responders who did not receive any rescue medications or other treatment interventions until relapse (n=18) or who received limited immunoglobulins (n=2), a brief course of steroids (n=2), or platelet transfusions (n=1) during treatment for low platelet levels with improvement continuing substantially beyond any expected transient response to these agents and without further interventions. The remaining 24 patients either had no platelet response after completing the 12-week, post-treatment evaluation period (n=9) or initiated other treatments after having no improvement in platelet counts 1 – 8 weeks following veltuzumab treatment (n=15). Among the 47 patients whose response to veltuzumab could be assessed, there were 49% (23/47) objective responses, including 32% (15/47) complete responses. Response rates (including complete responses) were generally comparable regardless of whether patients had ITP for ≤1 year or longer, and responses were variable across the four dose groups as expected with small numbers of patients with no clear evidence of a dose-response relationship (Table 2). The median time from first veltuzumab dose to response onset was 26 days (range, 7 – 264), with a median time of 38 days (7 – 189) to first occurrence of platelet counts ≥ 100x109/L for patients with complete responses. There was no clear evidence that response onset varied with dose group (data not shown). Of the 23 responders, one patient currently remains in long-term follow-up with a response still ongoing 2.6 years after treatment initiation, 17 have relapsed, and the other five are off study with continuing responses as of last evaluation (lost to follow up after 0.8, 2.7, and 3.2 years; withdrawn after 1.6 years for scheduled surgery; concluded the 5-year study follow-up period). The median time from treatment initiation to relapse for all 23 responders was 1.3 years. As shown in Figure 1, the duration of a response increased with higher veltuzumab doses, with the median time from treatment initiation to relapse reaching 2.7 years in the patients treated with four weekly 320-mg doses. Responders who achieved complete responses or had ITP ≤1 year also showed trends towards having more durable responses (Online Supplementary Figure S1). Nine responders were retreated at the investigators’ discretion after relapsing. Five patients did not respond when

retreated again twice, 2 weeks apart, with the same or higher dose. One patient retreated with four weekly 320-mg doses required rescue medications during retreatment which precluded response assessment. The three other patients responded to retreatment, as follows. One patient who had an initial partial response lasting 3 months responded with a similar duration when retreated with two 320-mg doses 2 weeks apart. Another patient who had an initial complete response lasting 6 months was retreated three times with two 320-mg doses 2 weeks apart and a fourth time with four weekly 320-mg doses, each time responding with a complete response of similar duration. The remaining patient who had an initial complete response lasting 2.7 years was retreated with four weekly 320-mg doses and achieved a response which is currently ongoing 10 months later.

Bleeding reduction At treatment initiation, 68% (34/50) of all patients had one or more sites of bleeding; this percentage progressively decreased to 29% (12/42) of patients assessed at the end of the 12-week, post-treatment evaluation period. Bleeding primarily involved the skin, oral cavity, and epistaxis, with few occurrences at other anatomic sites and no cases of intracranial, intraocular, or pulmonary bleeding. Most bleeding was minor (IBLS grade 1) with marked bleeding (IBLS grade 2) limited to ~10% of patients or less at any evaluation. Bleeding reduction following treatment occurred in all dose groups without evidence of a doseresponse relationship and was primarily limited to patients achieving objective responses (Figure 2).

Immunological changes The first dose of SC veltuzumab effectively depleted peripheral blood B cells in most patients, with median B-cell levels decreasing from 284 cells/mL before treatment to 4 cells/mL by the second dose. Only four patients (all treated

Table 2. Response rates.*

Overall Subgroup: ITP duration ≤1 year >1 year Dose group (total dose) 80 mg x 2 (160 mg) 160 mg x 2 (320 mg) 320 mg x 2 (640 mg) 320 mg x 4 (1280 mg)

Patients, n

Objective response

Complete response

49% (23/47)

32% (15/47)

14 33

57% (8/14) 45% (15/33)

29% (4/14) 33% (11/33)

8 9 15 15

75% (6/8) 33% (3/9) 53% (8/15) 40% (6/15)

50% (4/8) 11% (1/9) 27% (4/15) 40% (6/15)

Treatment response according to International Working Group criteria (Rodeghiero et al., 2009); ITP: immune thrombocytopenia. *Data from 47 patients assessable for treatment response.

haematologica | 2016; 101(11)

Figure 1. Kaplan-Meier estimates of increasing durability of responses with higher veltzumab dose levels. Time to relapse in each responder was measured from the initial treatment dose to first occurrence of a platelet count <30 x 109/L, but was censored at the time of last evaluation (ticks) if discontinued from the study prior to relapse. Results show percentages of responders continuing relapse-free after receiving either 80 (n=6) or 160 (n=3) mg doses twice 2 weeks apart [pooled for clarity], 320 mg (n=8) doses twice weekly 2 weeks apart, or 320 mg doses (n=6) once-weekly for 4 consecutive weeks (pairwise log-rank tests: 320 mg x 4 vs. 320 mg x 2, P= 0.22; 320 mg x 4 vs. 80-160 mg, P=0.17; both 320 mg doses vs. 80-160 mg, P=0.16).

1329

H.A. Liebman et al. twice, 2 weeks apart) did not achieve B-cell levels â&#x2030;¤20 cells/mL by 4 weeks after treatment; however, their B-cell counts decreased 74-94% from baseline and two patients achieved objective responses (both complete responses). Bcell depletion appeared comparable across the four dose groups but recovery towards baseline levels appeared slower among patients treated with the higher doses (Online Supplementary Figure S2). T cells and serum immunoglobulin levels evaluated 4 weeks after treatment showed no consistent pattern of change from baseline, either for patients treated with two doses, 2 weeks apart, or those given four doses of 320 mg at weekly intervals (data not shown).

Pharmacokinetics Veltuzumab serum levels were measured on treatment days and then 1, 2, 3, 4, 8, and 12 weeks after the last dose, with values generally increasing by dose group and reaching a maximum value 1 week after the last treatment dose before subsequently slowly declining over this period (Online Supplementary Figure S3). Evaluation of post-treatment pharmacokinetic parameters showed that the peak value (Cmax) and area-under-the-curve (AUC) increased with dose group, as expected, while the terminal half-life and clearance from the blood differed little across the dose groups (Online Supplementary Table S1).

patient with a cardiac history and resolved spontaneously without treatment or recurrence on subsequent examination. Other than platelet counts, routine hematology and serum chemistry tests showed no consistent pattern of change from baseline (data not shown), and there were no cases of serum sickness, hypogammaglobulinemia, delayed neutropenia, or other side effects that had been previously reported with rituximab.

Immunogenicity Eleven patients had elevated HAHA titers, which subsequently resolved or decreased in patients with available follow-up, without apparent clinical sequelae. Two patients previously treated with rituximab were already HAHA-positive at baseline (titers, 52 and 1006 ng/mL). Both achieved complete responses following two 80-mg doses of SC veltuzumab 2 weeks apart, but one patient who was retreated after relapse did not respond to a second course of treatment. Nine other patients were treated either twice 2 weeks apart with 80 (n=2), 160 (n=3), or 320 mg (n=1) doses or with four weekly 320-mg doses (n=3) and developed elevated titers after receiving veltuzumab (HAHA, 18%). These included eight patients with HAHA detectable after initial treatment (peak titers, 61 â&#x20AC;&#x201C; 2190 ng/mL) and one patient who only had HAHA after being retreated (peak titer, 152 ng/mL). Of the nine patients, six achieved objective responses following initial treatment,

Safety There was only one serious adverse event, grade 3 viral gastroenteritis, which occurred 6 months after treatment and was considered unrelated to the treatment. One patient had increased generalized pain 10 days after the first veltuzumab injection which was considered grade 3 and possibly treatment-related, although the patient had a history of chronic pain with herniated discs and multiple prior surgeries for pain relief. Otherwise, all treatment-related adverse events were grade 1-2 (mild-moderate) and occurred in 39 patients (28 with only grade 1 events), including 30 patients with local injection site reactions (pain, burning or soreness, bruising, rash, erythema), 26 patients with constitutional symptoms (body aches or generalized pain, chills, low-grade fever, headache, fatigue, nausea, rash), and seven patients with other events considered at least possibly related [sore throat, increased thirst and urination, fatigue, headache, edema, abdominal bloating, knee pain, and electrocardiogram abnormalities in one patient (see below)]. Most reactions resolved spontaneously or with acetaminophen on the day of injection, and no patient required steroids. Treatment-related adverse events did not appear to be dose-dependent for patients treated twice, 2 weeks apart, and increases in patients treated with four weekly doses were not unexpected since these patients received double the number of injections (Table 3). Other adverse events included minor infections (predominantly of the upper respiratory tract or sinusitis) in 14 patients, â&#x2030;Ľ1 minor bleeding events (conjunctival x 3; gingival x 2; mucosal purpura, hemoptysis, hematemesis, epistaxis) in seven patients, palpitations at second injection in two patients with a history of hypertension which resolved spontaneously, and brief episodes of tachycardia several weeks after treatment in two patients which were considered unrelated to veltuzumab. Electrocardiography was required in all patients at the time of the last injection; the only finding of clinical significance was asymptomatic atrial fibrillation of uncertain etiology which occurred in one 1330

Figure 2. Percentage of patients with any bleeding at first injection and then at 1, 4, 8 and 12 weeks following treatment with SC veltuzumab. The ITP Bleeding Scale was used to evaluate any bleeding at any site. (A) Results by dose group for patients receiving 80 mg (n=9), 160 mg (n=10), or 320 mg (n=15) doses twice 2 weeks apart or 320 mg doses (n=16) once-weekly for 4 consecutive weeks. (B) Results by platelet response for patients who achieved an objective response (OR, n=23) or were non-responders (NR, n=24). In patients with platelet responses continuing beyond 12 weeks, 84% (16/19) were free from bleeding at 24 weeks and 100% (16/16) at 48 weeks.

haematologica | 2016; 101(11)

Subcutaneous veltuzumab in ITP

Table 3. Treatment-related adverse events.*

Overall Dose group (total dose): 80 mg x 2 (160 mg) 160 mg x 2 (320 mg) 320 mg x 2 (640 mg) 320 mg x 4 (1280 mg)

Patients n

Any adverse event n (%)

Local site reactions n (%)

Constitutional symptoms n (%)

Other events n (%)

39 (78)

30 (60)

25 (50)

7 (14)

9 10 15 16

7 (78) 8 (80) 9 (60) 15 (94)

6 (67) 8 (80) 3 (20) 13 (81)

3 (33) 4 (40) 4 (27) 14 (88)

0 (0) 1 (10) 4 (27) 2 (13)

*Patients with one or more adverse events considered at least possibly treatment-related and categorized as local reactions at injection sites, constitutional symptoms, or other events (see text). Events were limited to grade 1 in 28 patients, grade 2 in ten patients, and grade 3 in one patient with increased generalized pain (see text).

including five with complete responses, but none of three patients retreated after relapse responded to a second course of treatment.

Discussion Since the B-cell antigen burden in patients with ITP is widely believed to be less than that seen in patients with Bcell malignancies, we hypothesized that low doses of veltuzumab would likely be effective in this disease, thus allowing for anti-CD20 therapy by more convenient subcutaneous injections. Following a dosing schedule used with rituximab in other autoimmune diseases such as rheumatoid arthritis, we previously reported that in ITP, two doses of low-dose veltuzumab given either by intravenous infusion or subcutaneous injection 2 weeks apart had activity.7 With this dosing schedule, there was activity at the lowest dose level of 80 mg veltuzumab x 2. Furthermore there was no clear dose dependence or evidence of greater response at the highest level explored with 320-mg doses. We subsequently treated an additional 16 patients with 320-mg doses of SC veltuzumab given once-weekly for 4 consecutive weeks to determine whether the increased dosing frequency, and especially greater cumulative dose delivered with this dosing schedule, which is often used in oncology, would lead to increased activity or any safety issues when used with SC veltuzumab in ITP. With long-term responses and other study data now mature, this report provides the final results from all SC dose groups. Compared to two doses of 80-320 mg SC veltuzumab given 2 weeks apart (total dose, 160-640 mg), there were no increased safety concerns with four weekly 320-mg SC doses (total dose, 1280 mg). The only treatment-related adverse events with SC veltuzumab administered in ITP with either dosing regimen were limited to mild-moderate (predominately grade 1) transient injection reactions, either local reactions at the injection site or constitutional symptoms, most of which resolved spontaneously or with acetaminophen, but did not require steroids. Routine laboratory tests remained unremarkable and no other safety concerns have emerged with SC veltuzumab. Eleven patients had elevated HAHA titers which subsequently resolved or decreased without apparent clinical sequelae. These patients including two who had been previously treated with rituximab and had measurable titers at study entry. Both patients previously treated with rituximab who were already HAHA-positive at study entry achieved complete responses when treated with SC veltuzumab, consistent with case reports of veltuzumab being effective in patients who had become resistant to rituximab.5,6 The other nine patients developed elevated titers haematologica | 2016; 101(11)

after receiving veltuzumab (8 after initial treatment, 1 after retreatment) without any obvious pattern regarding dose group or dosing regimen (HAHA, 18%). While immunogenicity rates with rituximab are not available in ITP, this frequency is comparable to rates reported with rituximab in rheumatoid arthritis (11%)11 and systemic lupus erythematosus (26%).12 Importantly, six of the nine patients achieved objective responses to veltuzumab with initial treatment, including five who had complete responses, consistent with other studies in cancer patients correlating immunogenicity with anti-B-cell antibodies to enhanced outcomes.13,14 However, of four patients with HAHA who received a second course of treatment with veltuzumab, none responded to retreatment, which would lower expectations for continued treatment in HAHA-positive patients. Among evaluable patients, the objective response rate to initial treatment with SC veltuzumab was 49% (complete response rate, 32%) with a median time to relapse of 1.3 years from the first dose. Of nine responders retreated at relapse, three treated at higher levels again responded, including one patient who was retreated four times. This overall result is generally similar to what we reported earlier with intravenous or SC veltuzumab given twice 2 weeks apart.7 Although limited by small numbers, it still appears that complete responses are more durable than partial responses, in agreement with findings reported with rituximab,15 and that responses in adults with long-standing ITP are likely less durable as has also been reported for rituximab in at least one study.16 It is difficult to compare the overall response results with veltuzumab with those reported for rituximab, given the wide range of the latter, which may be due to heterogeneity of the patients studied, ITP severity and duration, and treatment patterns, as well as publication bias, and differing response criteria,17 including recent rituximab studies using treatment failure endpoints other than objective response rates alone.18,19 We evaluated bleeding as an endpoint in this study, finding that treatment with SC veltuzumab was often effective in reducing the mostly minor cases of bleeding that occurred in this population, particularly in patients achieving objective responses and therefore most likely an expected consequence of improved platelet counts. SC veltuzumab was pharmacodynamically active with peripheral blood B cells depleted rapidly in most patients even after one dose of 80 mg veltuzumab, and meaningful platelet responses (including complete responses) were achieved already at the lowest dose group of two 80 mg SC doses given 2 weeks apart. Veltuzumab serum levels increased with each dose group according to the total dose administered, but the terminal half-life and clearance of veltuzumab from the blood after treatment appeared generally comparable across all dose groups. While there was no clear 1331

H.A. Liebman et al.

evidence of a platelet response relationship among the four dose groups, the duration of the platelet responses progressively increased with greater total dose, achieving a median duration of 2.7 years in responders following 4 weekly 320mg doses, including one patient with response still ongoing at last evaluation at 2.6 years. Furthermore, the platelet responses to retreatment occurred in patients treated both initially and at relapse with higher (320 mg) doses. The small numbers of patients make it difficult to draw firm conclusions regarding the most appropriate SC veltuzumab dosing regimen for further consideration despite the possible longer duration of responses at the highest doses. Two other studies with SC veltuzumab also administered 80, 160 or 320 mg SC doses, again finding activity in all dose groups with increased serum levels and no obvious exposure-response relationship, but these were studies in Bcell malignancies and with too few patients to evaluate whether response duration improved with higher doses.3,4 One tentative explanation for the apparent lack of dosedependence of responses might be that even 80-mg doses of SC veltuzumab result in B-cell depletion in peripheral blood irrespective of response. In this model, response would depend on whether B-cell clearance leads to platelet response. Further, even 80-mg doses of veltuzumab may already effectively saturate the mechanisms of action usually considered for immunotherapy (complement dependent cytotoxicity, antibody-dependent cell-mediated cytotoxicity, apoptosis) with little room for improvement at higher doses. It is also possible (at least in chronic lymphocytic

References 1. Goldenberg DM, Morschhauser F, Wegener WA. Veltuzumab (humanized anti-CD20 monoclonal antibody): characterization, current clinical results, and future prospects. Leuk Lymphoma. 2010;51(5):747-755. 2. Morschhauser F, Leonard JP, Fayad L, et al. Humanized anti-CD20 antibody, veltuzumab, in refractory/recurrent non-Hodgkin's lymphoma: phase I/II results. J Clin Oncol. 2009;27(20):3346-3353. 3. Negrea GO, Elstrom R, Allen SL, et al. Subcutaneous injections of low-dose veltuzumab (humanized anti-CD20 antibody) are safe and active in patients with indolent non-Hodgkin's lymphoma. Haematologica. 2011;96(4):567-573. 4. Kalaycio ME, Negrea OG, Allen SL, et al. Subcutaneous injections of low doses of humanized anti-CD20 veltuzumab: a phase I study in chronic lymphocytic leukemia. Leuk Lymphoma. 2016;57(4):803-811. 5. Tahir H, Bhatia A, Wegener WA, Isenberg DA. Humanised anti-CD20 monoclonal antibody in the treatment of severe resistant systemic lupus erythematosus in a patient with antibody against rituximab. Rheumatology. 2005;44(4):561-562. 6. Ellebrecht CT, Choi EJ, Allman DM, et al. Subcutaneous veltuzumab, a humanized anti-CD20 antibody, in the treatment of refractory pemphigus vulgaris. JAMA Dermatol. 2014;150(12):1331-1335. 7. Liebman HA, Saleh MN, Bussel JB, et al. Low-dose anti-CD20 veltuzumab given intravenously or subcutaneously is active in relapsed immune thrombocytopenia: a phase I study. Br J Haematol. 2013;162(5): 693-701. 8. Neunert C, Lim W, Crowther M, Cohen A,

1332

10.

11.

12.

13.

14.

15.

leukemia) that increasing doses might be counterproductive by resulting in removal of CD20 from B-cell surfaces (trogocytosis)20 or by exhausting complement and cell-mediated effector systems needed for effective cytotoxicity.21 Even with rituximab, years after the drugâ&#x20AC;&#x2122;s initial approval, there is still no clear explanation for why some patients respond and others do not, or why increased doses do not improve response rates. In ITP, 100-mg doses of rituximab may be active,22 but other low doses have not been evaluated, and while different rituximab dosing regimens have been compared,23 the 375 or 750 mg/m2 doses are higher than the 80 â&#x20AC;&#x201C; 320 mg doses of veltuzumab in this study. Furthermore for rituximab, the consensus of the limited studies in ITP is that the response rate at the lower doses (100 mg x 4) is the same as that at higher doses, but the duration of the responses is not as long, which is quite similar to the veltuzumab data reported here. In summary, SC veltuzumab was convenient, well-tolerated and did not cause significant safety concerns, with platelet responses and bleeding reduction occurring in all dose groups. Based on a possible longer duration of responses with higher doses, higher SC doses may be warranted and a more concentrated formulation of veltuzumab has recently been developed for this purpose. Acknowledgments The authors thank Lucy Lee, PharmD, Kiril Gordeyev, BS, and Robert M. Sharkey, PhD, for their contributions to the pharmacokinetic and immunoassay analyses.

Solberg L Jr, Crowther MA. The American Society of Hematology 2011 evidence-based practice guideline for immune thrombocytopenia. Blood. 2011;117(16):4190-4207. Rodeghiero F, Stasi R, Gernsheimer T, et al. Standardization of terminology, definitions, and outcome in idiopathic thrombocytopenic purpura (ITP) of adults and children: report from an international working group. Blood. 2009;113(11):2386-2393. Page L, Psaila B, Provan D, et al. The immune thrombocytopenic purpura (ITP) bleeding score: assessment of bleeding in patients with ITP. Br J Haematol. 2007;138(2):245-248. van Vollenhoven RF, Emery P, Bingham CO 3rd, et al. Longterm safety of patients receiving rituximab in rheumatoid arthritis clinical trials. J Rheumatol. 2010;37(3):558-567. Merrill JT, Neuwelt CM, Wallace DJ, et al. Efficacy and safety of rituximab in moderately-to-severely active systemic lupus erythematosus: the randomized, double-blind, phase II/III systemic lupus erythematosus evaluation of rituximab trial. Arthritis Rheum. 2010;62(1):222-233. Miotti S, Negri DR, Valota O, et al. Level of anti-mouse-antibody response induced by bi-specific monoclonal antibody OC/TR in ovarian-carcinoma patients is associated with longer survival. Int J Cancer. 1999;84(1):62-68. Azinovic I, DeNardo GL, Lamborn KR, et al. Survival benefit associated with human antimouse antibody (HAMA) in patients with B-cell malignancies. Cancer Immunol Immunother. 2006;55(12):1451-1458. Patel VL, MahĂŠvas M, Lee SY, et al. Outcomes 5 years after response to rituximab therapy in children and adults with immune thrombocytopenia. Blood. 2012; 119(25):5989-5995.

16. Cooper N, Stasi R, Cunningham-Rundles S, et al. The efficacy and safety of B-cell depletion with anti-CD20 monoclonal antibody in adults with chronic immune thrombocytopenic purpura. Br J Haematol. 2004;125(2): 232-239. 17. Arnold DM, Dentali F, Crowther MA, et al. Systematic review: efficacy and safety of rituximab for adults with idiopathic thrombocytopenic purpura. Ann Intern Med. 2007;46(1):25-33. 18. Arnold DM, Heddle NM, Carruthers J, et al. A pilot randomized trial of adjuvant rituximab or placebo for nonsplenectomized patients with immune thrombocytopenia. Blood. 2012;119(6):1356-1362. 19. Ghanima W, Khelif A, Waage A, et al. Rituximab as second-line treatment for adult immune thrombocytopenia (the RITP trial): a multicentre, randomised, double-blind, placebo-controlled trial. Lancet. 2015;385 (9978):1653-1661. 20. Kennedy AD, Beum PV, Solga MD, et al. Rituximab infusion promotes rapid complement depletion and acute CD20 loss in chronic lymphocytic leukemia. J Immunol. 2004;172 (5):3280-3288. 21. Beurskens FJ, Lindorfer MA, Farooqui M, et al. Exhaustion of cytotoxic effector systems may limit monoclonal antibody-based immunotherapy in cancer patients. J Immunol. 2012;188(7):3532-3541. 22. Zaja F, Vianelli N, Volpetti S, et al. Low-dose rituximab in adult patients with primary immune thrombocytopenia. Eur J Haematol. 2010;85(4):329-334. 23. Zwaginga JJ, van der Holt B, Te Boekhorst PA, et al. Multi-center randomized open label phase II trial on three rituximab dosing schemes in immune thrombocytopenia patients. Haematologica. 2015;100(3):e90e92.

haematologica | 2016; 101(11)

ARTICLE

Platelet Biology & its Disorders

Clinical and pathogenic features of ETV6related thrombocytopenia with predisposition to acute lymphoblastic leukemia

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Federica Melazzini,1 Flavia Palombo,2 Alessandra Balduini,3,4 Daniela De Rocco,5 Caterina Marconi,2 Patrizia Noris,1 Chiara Gnan,5 Tommaso Pippucci,2 Valeria Bozzi,1 Michela Faleschini,5 Serena Barozzi,1 Michael Doubek,6 Christian A. Di Buduo,3 Katerina Stano Kozubik,7 Lenka Radova,7 Giuseppe Loffredo,8 Sarka Pospisilova,7 Caterina Alfano,9 Marco Seri,2 Carlo L. Balduini,1 Alessandro Pecci,1 and Anna Savoia5

Department of Internal Medicine, IRCCS Policlinico San Matteo Foundation and University of Pavia, Italy; 2Department of Medical and Surgical Science, Policlinico Sant’Orsola Malpighi and University of Bologna, Italy; 3Department of Molecular Medicine, University of Pavia, Italy; 4Department of Biomedical Engineering, Tufts University, Medford, MA, USA; 5 Department of Medical, Surgical and Health Sciences, IRCCS Burlo Garofolo and University of Trieste, Italy; 6University Hospital and Masaryk University, Brno, Czech Republic; 7Center of Molecular Medicine, Central European Institute of Technology, Masaryk University, Brno, Czech Republic; 8Department of Oncology, Azienda “SantobonoPausilipon”, Pausilipon Hospital, Napoli, Italy; and 9Maurice Wohl Institute, King's College London, UK 1

Haematologica 2016 Volume 101(11):1333-1342

ABSTRACT

TV6-related thrombocytopenia is an autosomal dominant thrombocytopenia that has been recently identified in a few families and has been suspected to predispose to hematologic malignancies. To gain further information on this disorder, we searched for ETV6 mutations in the 130 families with inherited thrombocytopenia of unknown origin from our cohort of 274 consecutive pedigrees with familial thrombocytopenia. We identified 20 patients with ETV6-related thrombocytopenia from seven pedigrees. They have five different ETV6 variants, including three novel mutations affecting the highly conserved E26 transformationspecific domain. The relative frequency of ETV6-related thrombocytopenia was 2.6% in the whole case series and 4.6% among the families with known forms of inherited thrombocytopenia. The degree of thrombocytopenia and bleeding tendency of the patients with ETV6-related thrombocytopenia were mild, but four subjects developed B-cell acute lymphoblastic leukemia during childhood, resulting in a significantly higher incidence of this condition compared to that in the general population. Clinical and laboratory findings did not identify any particular defects that could lead to the suspicion of this disorder from the routine diagnostic workup. However, at variance with most inherited thrombocytopenias, platelets were not enlarged. In vitro studies revealed that the maturation of the patients’ megakaryocytes was defective and that the patients have impaired proplatelet formation. Moreover, platelets from patients with ETV6-related thrombocytopenia have reduced ability to spread on fibrinogen. Since the dominant thrombocytopenias due to mutations in RUNX1 and ANKRD26 are also characterized by normal platelet size and predispose to hematologic malignancies, we suggest that screening for ETV6, RUNX1 and ANKRD26 mutations should be performed in all subjects with autosomal dominant thrombocytopenia and normal platelet size.

Introduction Until the end of the last century, only a few forms of inherited thrombocytopenia were known, all of which were extremely rare and characterized by a severe bleeding tendency. Since then, knowledge of these thrombocytopenias has improved haematologica | 2016; 101(11)

Correspondence: alessandro.pecci@unipv.it

Received: April 6, 2016. Accepted: June 29, 2016. Pre-published: June 30, 2016. doi:10.3324/haematol.2016.147496

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1333

1333

F. Melazzini et al.

greatly and we presently recognize at least 26 disorders caused by mutations in 30 genes.1,2 This advancement of knowledge revealed that most patients with inherited thrombocytopenias have only mild or moderate thrombocytopenia, with trivial bleeding episodes or no bleeding at all. However, it also became apparent that many patients are exposed to a threat of acquiring additional defects that worsen their quality of life or can even be fatal. Subjects with MYH9-related disease are predisposed to proteinuric nephropathy evolving into end-stage renal failure, those with congenital amegakaryocytic thrombocytopenia always develop bone marrow aplasia, while patients with ANKRD26-related thrombocytopenia (ANKRD26-RT) or familial platelet disorder with predisposition to acute myeloid leukemia (AML) due to RUNX1 mutations (FPD/AML) have increased risk of AML and myelodysplastic syndromes. Thus, bleeding is no longer the unique problem of inherited thrombocytopenia patients. In 2015, four independent studies showed that mutations in the ETV6 gene are responsible for a new form of inherited thrombocytopenia and suggested that ETV6related thrombocytopenia (ETV6-RT) predisposes to acute lymphoblastic leukemia (ALL).3-6 However, only a few families have been reported so far and the clinical and laboratory features of ETV6-RT remain poorly defined. In order to gain further information on this disorder, we screened 130 consecutive unrelated propositi with inherited thrombocytopenia of unknown origin for ETV6 mutations and identified seven affected families. Two of these pedigrees have been briefly reported in a previous paper.4 Here we describe the features of 20 affected subjects, who form the largest cohort of ETV6-RT patients collected so far. As these patients were identified by screening a series of consecutive, unselected probands with familial thrombocytopenia, we could estimate the relative frequency of ETV6-RT among inherited thrombocytopenias and the risk of hematologic malignancies associated with this condition. By reporting the clinical and laboratory features of these patients in detail, we provide indications to raise the level of suspicion of the presence of this disorder from the findings of routine diagnostic workup of probands with inherited thrombocytopenia. Finally, we discuss the pathogenesis of ETV6-RT, having investigated, for the first time, megakaryocytes differentiated from hematopoietic progenitors of patients with ETV6-RT and functionally characterized the patientsâ&#x20AC;&#x2122; platelets.

Society on Thrombosis and Haemostasis bleeding assessment tool.7 The institutional review board of San Matteo Foundation approved the study and all subjects or their legal guardians signed written informed consent in accordance with the Declaration of Helsinki.

Mutation screening and reverse transcriptase polymerase chain reaction analysis Genomic DNA and RNA were extracted from peripheral blood. The ETV6 gene was analyzed using Sanger and whole exome sequencing. Methods of mutation screening and reverse transcriptase polymerase chain reaction analysis are detailed in the Online Supplementary Information.

Bioinformatic tools and analysis of ETV6 structure The bioinformatic tools used to evaluate missense variants together with the methods used to analyze ETV6 structure are reported in the Online Supplementary Information.

Basic blood cell studies Blood cell counts were evaluated by electronic counters. Parameters relative to platelet diameter were measured by software-assisted image analysis on blood smears, as reported elsewhere.8 The following previously defined parameters were computed: mean platelet diameter, platelet diameter distribution width, platelet diameter large cell ratio, and platelet diameter small cell ratio.8 The percentage of large platelets was also estimated empirically, as previously reported8 and detailed in the Online Supplementary Information. Surface expression of platelet glycoproteins (GP) was investigated by flow cytometry as reported, whereas platelet aggregation was evaluated using the densitometric method described by Born.9 The antibodies and platelet agonists used are listed in the Online Supplementary Information.

Platelet activation Platelet activation in response to ADP or TRAP was investigated by flow cytometry as reported previously.10 The protocol is described in detail in the Online Supplementary Information.

Platelet adhesion and spreading Platelet adhesion and spreading on the subendothelium components of the extracellular matrix, type I collagen, von Willebrand factor, or fibrinogen, were investigated as previously described11,12 and as detailed in the Online Supplementary Information.

Investigation of megakaryocytes Methods Patients Between 2003 and 2014, we analyzed at the IRCCS Policlinico San Matteo Foundation of Pavia (Italy) 274 consecutive unrelated probands with familial thrombocytopenia. By applying a welldefined diagnostic algorithm for inherited thrombocytopenias,1 we made a molecular diagnosis in 144 of these families, whereas 130 probands remained without a definite diagnosis as they did not fit the criteria for any known inherited thrombocytopenia. These 130 consecutive propositi with inherited thrombocytopenia of unknown origin have been screened for mutations in ETV6. Whenever ETV6 mutations were identified, the available relatives of probands were also investigated. Bleeding tendency was measured using the International 1334

Megakaryocytes were differentiated in vitro from peripheral blood CD45+ cells as previously reported.13,14 Morphological analysis of megakaryocytes was performed by phase-contrast and fluorescence microscopy, while the percentage of fully differentiated megakaryocytes and megakaryocyte ploidy at the end of the culture were investigated by flow cytometry.14,15 Proplatelet yields were evaluated both in suspension and following adhesion on fibrinogen at the end of the culture, as previously described.13,16 Methods are reported in the Online Supplementary Information.

Statistical analysis Data are presented as means and standard deviations or ranges. Statistical comparisons were performed by the two-tailed Student t test. Incidences of hematologic malignancies (per 100,000 person-years) together with their exact 95% confidence intervals (95% CI) were computed. haematologica | 2016; 101(11)

ETV6-related thrombocytopenia

Results Mutation screening Analysis of the ETV6 gene allowed us to identify five different heterozygous variants in seven unrelated pedigrees. Two variants (c.641C>T/p.P214L and c.1252A>G/ p.R418G+p.N385Vfs*7) have been reported previously in

two families (Figure 1A, families B and G).4 The remaining three novel variants are two missense alterations and one deletion. The two missense variants, c.1105C>T (p.R369W) and c.1138T>A (p.W380R) segregate in the affected family members and are not present in healthy relatives (Figure 1A). They are absent in public genomic databases, such as

Figure 1. Mutations identified in the ETV6 gene and their effect on protein structure. (A) Pedigrees of families enrolled in this study carrying different mutations as indicated (novel mutations in bold). Nucleotide numbering reflects the ETV6 cDNA with +1 corresponding to the A of the ATG translation initiation codon in the reference sequence (RefSeq NM_001987.4). Therefore, the initiation codon is residue 1 in the amino acid sequence. Families B and G have been previously reported (Noetzli et al.4). (B) RT-PCR in affected members (I-2, II-1, and II-2) of family F to determine the consequence of the c.1153-1_1165del mutation on splicing. C+, wildtype control; C-, negative control. The analysis shows two fragments, the wild-type (822 bp) and the exon 7 skipping (721 bp) products. (C) The deletion of the 14 bp (gAACAGAACAAACA) of c.1153-1_1165del is likely due to non-allelic homologous recombination between the two GAACAAACA repeats located at the intron 6 and exon 7 boundary. (D) Domain structure of ETV6 (XP_011518909.1) based on Pfam annotation at http://www.ncbi.nlm.nih.gov/gene/2120 (PNT, pointed N-terminal domain; ETS, C-terminal DNA binding domain), with mutations identified in ETV6, already reported or identified in this study (top). The numbers of families carrying each mutation are in brackets. Mutations leading to skipping of exon 7 are boxed. (E) Structural modeling of the ETS domain with residues R369 (blue) and W380 (green) affected by the p.R369W and p.W380R mutations.

haematologica | 2016; 101(11)

1335

F. Melazzini et al.

dbSNP (www.ncbi.nlm.nih.gov/SNP), 1000 genomes (www.1000genomes.org), and Exome Aggregation Consortium (www.exac.broadinstitute.org). Multiplesequence alignment indicated that they affect highly conserved amino acid residues (data not shown). They are predicted to be deleterious for protein function according to different tools (Online Supplementary Table S1). Moreover, the CADD scores were 26.1 and 23.1 for p.R369W and p.W380R, respectively. Both mutations map in the E26 transformation-specific (ETS) domain which is in the Cterminal half of the protein (residues 337-421). Analysis of the coordinates of the ETS domain of ETV6 (2DAO) (numbered in pdb C8 and L112, so that R369 and W380 correspond to R39 and W50) showed that W380 is well buried in the hydrophobic core and surrounded by a number of hydrophobic residues, such as L341 and M394 (L11 and M64 in the structure) (Figure 1E). W380 is also close to the side chains of H383 and K384 (H53 and K54). Its substitution to an arginine will greatly destabilize the structure by creating both an uncompensated cavity in the hydrophobic core and electrostatic repulsion of nearby positively charged residues. Residue R369 is well exposed

on the protein surface and is predicted to form an electrostatic interaction with the spatially nearby E361 (E31). Its substitution by a tryptophan could destabilize the fold by abolishing this interaction. Alternatively, this residue could be implicated in protein-protein interactions. In this case, its substitution by a much bulkier and uncharged residue could be deleterious. The c.1153-1_1165del deletion variant removes the last "G" nucleotide of intron 6 and the first 13 nucleotides of exon 7. To investigate the effect of this deletion, we carried out reverse transcriptase polymerase chain reaction analysis on the three affected individuals of family F. Sequencing analysis of the altered 721 bp product showed skipping of exon 7 (r.1153_1253del/p.N385Vfs*7; Figure 1B) resulting in truncation of the ETS domain. Since the 721 bp band was fainter than the wild-type product (822 bp), we cannot exclude that the alternatively spliced mRNA was partially degraded. Inspection of the intron 6/exon 7 genomic boundary revealed repeats that are likely to be involved in non-allelic homologous recombination leading to micro-deletions/duplications (Figure 1C). The seven families reported in Figure 1A formed our

Table 1. Main characteristics of the investigated patients.

Family/ Individual

ETV6 mutationb

A/I-1 A/II-1 A/II-2

c.641C>T p.Pro214Leu p.Pro214Leu

Agec, y/ Age at ISTH BAT Platelets, MPV, Gender diagnosisd, y scoree x109/L fLf

MPD, mmg

Hb, g/dL

MCV, fLh

WBC, Neutrophils x109/L x109/L

57/M 20/F 27/F

30 birth birth

3 7 3

115 59 82

8.8 8.6 8.2

2.44 2.24 2.23

14.6 10.4 13.6

99 68 98

7.13 4.98 5.5

4.9 2.3 3.39

B/I-21

43/F

115

2.82

11.1

5.02

1.75

B/II-11 B/II-21 C/I-1 C/II-1 D/I-1 D/II-1 E/I-1 E/I-3

15/M 18/F 48/M 13/M 53/M 7/F 37/F 42/M

birth 2 38 3 47 1 8 5

3 0 3 0 0 0 0 4

66 44 112 87 110 109 105 55†

10.4 10.1 na na 8.4 9.2 8.1 9.1†

2.89 3.26 2.73 2.53 2.42 2.28 na na

14.0 13.1 15.4 14.1 13.7 12.6 14.2 14.9†

91 97 103 86 97 79 97 94†

5.36 4.04 6.3 3.84 5.4 6.82 7.50 8.0†

1.18 1.42 4 1.81 2.84 1.87 5.2 6.1†

45/M 20/M

20 4

0 2

93 60‡

7.9 8.0‡

na 2.73‡

16.9 14.8‡

101 86‡

8.30 5.11‡

4.24 1.97‡

13/M 49/F 12/F 17/F

birth 7 birth birth

4 2 1 2

99 105 57 70

7.4 8.9 8.6 8.7

na 2.55 2.40 2.36

14.0 13.4 14.2 14.4

90 107 97 97

6.15 7.11 6.59 8.24

2.45 4.4 4 5.3

51/F 28/M

20 3

0 2

101 101

7.6 7.8

3.17 2.99

13.6 15.9

97 97

4.71 5.3

2.02 2.39

c.1105C>T p.Arg369Trp p.Arg369Trp

c.1138T>A p.Trp380Arg p.Trp380Arg

E/I-4 E/II-1 E/II-3 F/I-2 F/II-1 F/II-2 G/I-2a G/II-1a

c.1153-1_1165del r.1153_1253del r.1153_1253del p.Asn385Valfs*7 p.Asn385Valfs*7 c.1252A>G p.Arg418Gly p.Arg418Gly + + p.Asn385Valfs*7 p.Asn385Valfs*7

Hematological malignancies

Common ALL at age 7 years B-cell ALL at age 15 years

JAK2V617F+ PV at age 37 years Common ALL at age 7 years

Common ALL at age 3 years

a Previously reported patients. bNucleotide A of the ATG translation initiation start site of the ETV6 cDNA in GenBank sequence NM_001987.4 is indicated as nucleotide +1. Novel germline mutations are in bold. cAge at the last evaluation: the blood parameters and bleeding score reported here were measured at the last evaluation, unless otherwise specified. dAge at diagnosis of thrombocytopenia. eInternational Society on Thrombosis and Haemostasis (ISTH) bleeding assessment tool (BAT) score was calculated as previously reported (Lowe et al.7). fNormal range: 8-13.4 fL. gNormal range: 1.9-3.4 mm. hNormal range: 82-98 fL. †Parameters measured at the last available examination before the development of PV (age 35). ‡Parameters measured before chemotherapy and hematopoietic stem cell transplantation for the development of ALL. Na: not available; ALL: acute lymphoblastic leukemia; PV: polycythemia vera.

1336

haematologica | 2016; 101(11)

ETV6-related thrombocytopenia

cohort of 20 affected individuals who have been studied to characterize the phenotype of ETV6-RT.

Clinical picture A mild bleeding tendency was present in 12 patients, whereas eight subjects did not have any significant bleeding diathesis (Table 1). The more common bleeding symptoms were petechiae, ecchymoses, gum bleeding, epistaxis, and menorrhagia. Thrombocytopenia was discovered in adulthood in five patients, whereas it was identified at birth in six patients because of the family history of low platelet count. One patient (E/I-3) was initially misdiagnosed with immune thrombocytopenia, and received steroids and underwent splenectomy at the age of 9 years without this producing an increase in his platelet count. Ten patients had undergone 17 operations and six had had teeth extracted without excessive bleeding. Four women had given birth to six children, three vaginally and three by Cesarean section. Prophylactic platelet transfusion was deemed necessary to cover one vaginal delivery. Excessive bleeding (800 mL) was reported in another woman (patient F/I-2) who had given birth vaginally. We have no information on her platelet count or function at the time of the delivery; however, it is interesting to note that she had a defective platelet response to low doses of collagen and ADP when she was investigated at our institution (see below). Unilateral polydactyly was observed in one patient, mitral valve prolapse in two subjects, and renal ectopia in one. So, no recurrent extra-hematologic abnormalities have been identified. Four patients from four families developed B-cell ALL during childhood (common ALL in three cases, not better defined in one). Conventional cytogenetic analysis resulted normal in three cases, while patient E/II-1 had hyperdiploid ALL; the search for the ETV6-AML1 transcript was performed in one patient (E/II-1) with normal findings. The incidence of ALL in our case series was 731.3 per 100,000 (95% CI, 274.5-1948.4), while it is 1.4 per 100,000 in the general population according to the National Cancer Institute.17 Three patients obtained remission after conventional chemotherapy, and one after hematopoietic stem cell transplantation from an unrelated donor. Patient E/I-3, who had a history of isolated thrombocytopenia

since childhood (Table 1), at the age of 37 developed an increased hemoglobin level (19.0 g/dL, hematocrit 56%), with mild leukocytosis and thrombocytosis. The JAK2V617F mutation was identified and a diagnosis of polycythemia vera was made. A history of non-hematologic neoplasms was present in three patients. Patient B/II-1 had breast fibroadenoma at 35 years old and meningioma at the age of 42. Patient G/I2 had breast carcinoma at the age of 49, while patient F/II2 developed breast fibroadenoma when she was 14 years old.

Blood cell counts and peripheral blood film examination Table 1 reports the blood cell counts obtained at the last examination for 18 patients, and at the last available examination before the development of polycythemia vera and before hematopoietic stem cell transplantation for patients E/I-3 and E/II-1, respectively. Eleven patients had fewer than 100x109 platelets/L and only one fewer than 50x109/L. For most patients we had platelet counts measured at different ages prior to the evaluation for this study (Online Supplementary Table S2). There was some fluctuations in the patientsâ&#x20AC;&#x2122; platelet counts over time, but none of the patients showed a definite trend toward improvement or worsening of thrombocytopenia during their life. The mean platelet volume was slightly reduced in four cases and normal in the other 14 evaluable patients (Table 1). Peripheral blood film examination in 16 patients showed that mean platelet diameter was similar to that of healthy subjects, confirming that average platelet size is consistently normal in ETV6-RT patients (Table 2). We found very mild but significant increases in platelet diameter distribution width and platelet diameter large cell ratio, which indicate that a mild platelet anisocytosis and a slightly increased proportion of large platelets were frequent features of the investigated patients. In agreement with previous findings,8 empirical measurement of the percentage of platelets larger than half an erythrocyte gave similar results to the assessment of platelet diameter large cell ratio by image analysis (data not shown). Conversely, the increased mean platelet diameter distribution width detected by image analysis did not correspond to increased mean platelet distribution width values

Table 2. Parameters of platelet diameters measured on peripheral blood films in investigated patients.

N.a

MPD, mm mean (SD)

PDDW, mm mean (SD)

PDLCR, % mean (SD)

PDSCR, % mean (SD)

Family A Family B Family C Family D Family E Family F Family G

3 3 2 2 1 3 2

2.30 (0.12) 2.99 (0.24) 2.63 (0.31) 2.35 (0.10) 2.73 2.44 (0.10) 3.08 (0.13)

2.57 (0.06) 2.97 (0.21) 2.75 (0.49) 2.10 (0.14) 2.9 2.57 (0.35) 3.15 (0.49)

8.27 (1.77) 12.3 (4.75) 7 (4.24) 4.5 (4.95) 8.5 8.83 (4.07) 13 (7.07)

5.87 (3.19) 1 (0.86) 1.25 (1.06) 5.5 (3.53) 1 3.16 (1.25) 0.5 (0.71)

Total ETV6-RT patients Healthy subjectsb

16 55

2.63 (0.17) 2.49 (0.32)

2.70 (0.29)* 2.18 (0.58)

9.23 (4.47)* 3.64 (4.93)

2.85 (1.77) 4.35 (5.9)

a Number of investigated subjects. bValues of healthy subjects previously measured in a cohort of 55 healthy volunteers (Noris et al.8). *P<0.01 with respect to healthy subjects. MPD: mean platelet diameter. PDDW: platelet diameter distribution width = difference from the 2.5 to the 97.5 percentile of platelet diameter distribution. PDLCR: platelet diameter large cell ratio = proportion of platelets larger than the 97.5 percentile of MPD of healthy subjects (3.9 Âľm). PDSCR, platelet diameter small cell ratio = proportion of platelets smaller than the 2.5 percentile of the MPD of healthy subjects (1.6 Âľm) (Noris et al.8).

haematologica | 2016; 101(11)

1337

F. Melazzini et al. Table 3. In vitro platelet aggregation and surface expression of major platelet glycoproteins in investigated patients.

Family A B C F G

Platelet aggregation, maximal extent, %a - mean (range) N. of investigated Collagen, ADP, Ristocetin, subjects 4 mg/mL 5 ÎźM 1.5 mg/mL

2 2 2 3 2

80 (71-89) 80 (74-87) 71 (69-73) 54 (50-56) 78 (67-90)

75 (66-84) 81 (77-85) 57 (44-70) 37 (35-39) 82 (78-87)

88 (76-100) 80 (68-93) 67 (57-77) 100 (100-100) 88 (77-100)

Surface expression of platelet glycoproteins, % of controls - mean (range) N. of investigated GPIba (SZ2) GPIX (SZ1) GPIIb (P2) subjects 2 3 2 2 2

131 (130-132) 98.7 (98-99) 147.5 (143-152) 117.5 (110-125) 159.5 (136-165)

130 (116-144) 99 (98-100) 126 (122-130) 125 (121-129) 101 (89-113)

91 (85-97) 98.7 (97-100) 108 (94-122) 127.5 (118-137) 85 (78-92)

Normal ranges: collagen 66-88; ADP 43-76; ristocetin 67-90.

obtained by automated cell counts (data not shown). Mild anemia was observed in one patient with iron deficiency (A/II-1). Mean corpuscular volume was reduced in this subject, increased without any apparent cause in five subjects, and within the normal range in the remaining patients. White blood cell count was normal in all the cases.

In vitro platelet studies Platelet aggregation Among the 11 investigated patients, the three patients from family F had mildly reduced platelet aggregation after stimulation with collagen 4 mg/mL and ADP 5 ÂľM, while individual C/II-1 showed a slightly reduced response to ristocetin 1.5 mg/mL (Table 3). However, all patients had completely normal responses to higher concentrations of these agonists (collagen 20 mg/mL, ADP 20 mM, ristocetin 3 mg/mL, data not shown), indicating that, if present, the aggregation defects were mild.

Platelet flow cytometry As shown in Table 3, flow cytometry performed in 11 patients did not identify any consistent defect of the major glycoproteins of the platelet surface.

Platelet activation Overall, the surface expression of activated GPIIb-IIIa and P-selectin and the reduction of GPIba upon stimulation of platelets with ADP or TRAP, were not significantly different in 11 ETV6-RT patients with respect to those in controls (Online Supplementary Figure S1). A mild reduction of activated GPIIb-IIIa expression after stimulation with TRAP (52% to 65% of the expression in controls) was observed in three patients.

Platelet adhesion and spreading In vitro adhesion of platelets from seven patients to subendothelium components of the extracellular matrix was not different from that of controls. However, the ability of ETV6-RT platelets to spread on fibrinogen was consistently and significantly reduced, while spreading on collagen and von Willebrand factor was normal (Table 4).

In vitro culture of megakaryocytes and assessment of proplatelet formation Megakarocytes from eight patients and eight healthy subjects were cultured in vitro. After 14 days of culture, expression levels of the major megakaryocyte differentia1338

Table 4. In vitro platelet interaction with subendothelium molecules in seven ETV6-RT patients.

Platelet adhesion and spreading, % of controls - mean (SD) N. of adhering % of spread Surface area platelets platelets covered by platelets Fibrinogen Collagen von Willebrand factor

102.9 (27.6) 107.2 (45.1) 80.4 (36.5)

51.5 (33.5)* 103.6 (41.7) 103.8 (41.5)

61 (28.5)* 122.4 (29.3) 120.7 (22.7)

*P<0.01 with respect to controls.

tion surface markers (GPIIIa, GPIIb and GPIba) were similar to those of healthy controls (Figure 2A,B). Conversely, megakaryocyte ploidy was significantly lower in patients than in controls (Figure 2C), and this was paralleled by differences in megakaryocyte diameters (Figure 2D). The analysis of proplatelet formation revealed that, compared to megakaryocytes from controls, megakaryocytes from patients had elongated proplatelet shafts of shorter length and with decreased number of branches. Furthermore, the percentage of proplatelet-forming megakaryocytes was significantly reduced in patients. In contrast, the size of proplatelet tips was similar in patients and in controls (data not shown). Similar results were obtained with megakaryocytes in suspension (Figure 3A,B) and following adhesion on fibrinogen (Figure 3C,D).

Discussion Here we report the molecular and phenotypic characterization of seven families with germline mutations in ETV6. In addition to the variants previously reported,4 we identified three novel alterations, which are likely to be pathogenic. The two novel missense variants (p.R369W and p.W380R) segregated within the families, are absent in public genomic databases, and are expected to be deleterious for protein function according to bioinformatic tools and analysis of protein conformational structure. ETV6 is a modular protein which contains a PNT and an ETS domain sandwiched between regions of potential intrinsically unstructured nature. Both p.R369W and p.W380R affect the ETS domain, a conserved region that interacts directly with DNA consensus sequences. We have shown that the role of W380 is structural, it being surrounded by hydrophobic residues in the domain hydrophobic core. Its substitution by an arginine will haematologica | 2016; 101(11)

ETV6-related thrombocytopenia

Figure 2. Normal differentiation but decreased ploidy of ETV6-RT megakaryocytes. Hematopoietic progenitors from peripheral blood samples of healthy controls (CTRL) and patients (ETV6-RT) were differentiated in vitro into megakaryocytes in the presence of thrombopoietin, interleukin-6 and interleukin-11. (A) Representative immunofluorescence staining of plasma membrane GPIIIa in CTRL and ETV6-RT megakaryocytes (red=GPIIIa; blue=nuclei; scale bar=20 mm). (B) Flow cytometry analysis of GPIIb and GPIbÎą expression revealed comparable percentages of double-stained populations in CTRL and ETV6-RT at the end of the culture. (C) Ploidy of megakaryocytes at the end of the culture was significantly reduced in cells from ETV6-RT patients (*P<0.05). (D) Diameters of megakaryocytes were also significantly lower in ETV6-RT patients (total number of cells analyzed: 1,100, *P<0.01).

therefore severely destabilize the domain structure. Residue R369 is involved in an electrostatic interaction and possibly in protein-protein interactions. It is important to note that the somatic p.R369W has previously been associated with chronic myelomonocytic leukemia, colorectal cancer, and childhood leukemia.5,6 Moreover, Zhang et al. previously reported one ETV6-RT pedigree carrying a different germline missense variant affecting the same residue (p.R369Q). Similarly to our patients with the p.R369W, the subjects with p.R369Q had mild thrombocytopenia and normal platelet morphology.6 Among the eight members of the p.R369Q pedigree, one had chronic myelomonocytic leukemia at the age of 82 and one had colorectal cancer at the age of 43,6 whereas we did not observe neoplasms in our p.R369W patients at a median age at evaluation of 30 years. Finally, both p.R369Q and p.R369W have been associated with genetic predisposition to childhood ALL.3 These observations suggest that arginine 369 is a mutational hot spot. With regards to the c.1153-1_1165del variant, reverse transcriptase polymerase chain reaction analysis demonstrated that it affects the splicing process leading to skipping of exon 7 (p.N385Vfs*7). The same alternative splicing was also caused by the c.1153-5_1153-1del mutation,5 which, as c.1153-1_1165del, is likely to derive from nonallelic homologous recombination between repetitive sequences present at the intron 6/exon 7 boundary. The skipping of exon 7 is also determined by the c.1252A>G substitution.4 Affecting the second to last nucleotide of exon 7, this allele is associated with both correctly (p.R418G) and alternatively (p.N385Vfs*7) spliced mRNA.4 Of the ten different mutant forms identified so far in ETV6-RT families, p.P214L is the only one that does not affect the ETS domain (Figure 1D), but instead alters a less haematologica | 2016; 101(11)

conserved central domain that interacts with several transcription repressors further controlling expression of the target genes. Unlike the other germline mutations, which are mainly private, this substitution was responsible for ETV6-RT in four of the 14 families characterized so far, indicating that it represents another potential mutational hot spot. ETV6 is a transcriptional repressor involved in embryonic development and hematopoietic regulation.18 In particular, animal studies suggested that ETV6 has two independent roles in mouse hematopoiesis: on the one hand it is required for survival of hematopoietic stem cells, on the other it promotes the late phases of megakaryopoiesis. Interest in ETV6 increased greatly at the end of the last century after demonstration that its deregulation due to rearrangements, fusions or deletions is involved in hematologic malignancies.19,20 Moreover, somatic mutations in ETV6 were recently found in a variety of hematologic neoplasm, including AML, T- and B-cell ALL, mixed-phenotype acute leukemia, myelodysplastic syndromes, chronic lymphocytic leukemia and chronic myelogenous leukemia.21 Even more recently, targeted sequencing of ETV6 in 4405 childhood ALL cases identified 31 germline variants potentially related to leukemia in 35 cases.3 Based on this evidence, it is not surprising that the four studies that identified ETV6-RT in 41 subjects from nine families found that 16 patients (39%) had hematologic malignancies, with 12 patients (29%) developing ALL.3-6 Of note, 11 of the 12 subjects with ALL were children. The other blood neoplasms observed in ETV6-RT patients were mixed-phenotype acute leukemia, multiple myeloma, myelodysplastic syndromes and chronic myelomonocytic leukemia. We found that four of 20 consecutive patients with ETV6-RT (20%) developed ALL during childhood, thus 1339

F. Melazzini et al. A

confirming that early leukemic transformation is a major risk in these patients. Moreover, we observed that one patient developed JAK2-positive polycythemia vera at the age of 37, supporting the previous hypothesis that ETV6RT predisposes not only to ALL, but also to other blood neoplasms. The frequency of hematologic malignancies is lower in our study than in the previous ones (25% versus 39%). This is explained by the fact that, in the previous investigations,3-6 the occurrence of hematologic malignancies was one of the criteria for the recruitment of patients, while we examined a series of consecutive, unselected patients with inherited thrombocytopenia of unknown origin. This approach appears more suitable for providing a reliable estimation of the incidence of hematologic neoplasms among ETV6-RT patients. Of course, the analysis of a larger series of patients is needed to confirm our figure. Similarly to this study, we previously searched a large series of unselected patients for ANKRD26 mutations and discovered that ten of 118 (8%) subjects with ANKRD26RT had developed myeloid malignancies.22 Thus, hematologic malignancies seem much more frequent in ETV6-RT than in ANKRD26-RT. The risk of malignancies appears even higher in FPD/AML, since over 40% of such patients had myeloid neoplasms.23 However, as discussed for ETV6-RT, the RUNX1 mutational screening was also generally performed in pedigrees with hematologic malignancies,24 and it is therefore likely that the incidence of transformation has been overestimated. However, each patient with an inherited thrombocytopenia caused by mutations in ETV6, RUNX1 or ANKRD26 has a relevant risk of hematologic malignancies, and recognizing these patients 1340

Figure 3. Aberrant proplatelet formation by ETV6-RT megakaryocytes. (A) Representative light microscopy analysis of proplatelet formation and structure from control (CTRL, i) and patient (ETV6-RT, ii-v) megakaryocytes cultured for 16 h in suspension (scale bar=50 mm). (B) The percentage of proplatelet-forming megakaryocytes was calculated as the number of megakaryocytes displaying at least one filamentous pseudopod with respect to total number of round megakaryocytes per analyzed field (*P<0.01). (C) Representative fluorescence microscopy analysis of proplatelet formation and structure from CTRL (i-ii) and ETV6-RT (iii-vi) megakaryocytes cultured for 16 h with adhesion on fibrinogen. The pictures clearly show defective proplatelet elongation in ETV6-RT (red=Î˛1-tubulin; blue=nuclei; scale bar=30 mm). (D) The percentage of proplatelet-forming megakaryocytes was calculated as the number of Î˛1-tubulin-positive cells displaying at least one pseudopod with respect to total number of round megakaryocytes per analyzed field (*P<0.01).

is important not only to provide effective genetic counseling and appropriate follow-up, but also to give appropriate treatment to patients who develop blood neoplasms and need hematopoietic stem cell transplantation. In fact, as shown in different disorders predisposing to myeloid malignancies,25 the use of an affected family member as the donor would entail the risk of developing malignancies once again. ETV6-RT is a relatively frequent form of inherited thrombocytopenia. In fact, in our series of 274 consecutive propositi, ETV6-RT was identified in seven families and had, therefore, a relative prevalence of 2.6% in the whole case series, and of 4.6% in the series of probands with known inherited thrombocytopenia (7/151). In our cohort, the frequency of ETV6-RT was lower only to that of monoallelic Bernard-Soulier syndrome (12.2% in the whole series), MYH9-related disease (11.4%), ANKRD26RT (9.4%), and biallelic Bernard-Soulier syndrome (5.7%). Since most of our patients with monoallelic BernardSoulier syndrome had the Ala156Val mutation of GPIba (Bolzano mutation), which is exclusive to the Italian population,26 it is expected that the relative frequency of ETV6-RT is even higher in other countries. Our study did not identify any peculiar feature that can be used to raise the suspicion of ETV6-RT from the routine diagnostic workup and the diagnosis does, therefore, remain difficult. A previous investigation, which reported five patients who have been re-evaluated in this study, suggested that red blood cell macrocytosis is a feature of the ETV6-RT phenotype.4 In that investigation, the percentage of patients with increased mean corpuscular volume was 40%,4 whereas it was 25% in the present study. haematologica | 2016; 101(11)

ETV6-related thrombocytopenia

With regards to the five patients reported in both studies, red blood cell macrocytosis was found in two individuals in the previous examination but was not confirmed in the present evaluation. Of note, the absolute mean corpuscular volumes were similar in the two studies (mean 94.5 fL with SD 3.8 versus mean 93.3 fL with SD 8.9) and the discrepancy in the percentage of patients with red cell macrocytosis resulted from the different upper limits of normal range used in the two investigations (95 fL in the previous study and 98 fL in the present one, according to the normal ranges of the different laboratories). On the whole, these findings indicate that red blood cell macrocytosis is present in a minority of patients with ETV6-RT, and suggest that it may be inconstantly found in the same patients over time. Thus, red cell macrocytosis seems to have limited diagnostic value for recognizing this condition. Moreover, we did not identify any distinguishing defect of major platelet glycoproteins or in vitro platelet aggregation and evaluation of peripheral blood films did not reveal any morphological abnormalities, except for mild platelet anisocytosis. However, at variance with most inherited thrombocytopenias, mean platelet diameter and mean platelet volume were consistently normal in ETV6-RT, and it is precisely the normal size of platelets that should raise suspicion of this condition in subjects with an autosomal dominant thrombocytopenia. The other dominant inherited thrombocytopenias with this feature are FDP/AML, ANKRD26-RT, and CYCS-RT. Of note, CYCS-RT is a very rare condition described so far in only two pedigrees,1 whereas the other two disorders are more frequent and, like ETV6-RT, predispose to hematologic malignancies. Thus, we suggest that all subjects with a dominant inherited thrombocytopenia and normal platelet size should be tested for mutations in ETV6, RUNX1, and ANKRD26, in order to identify one of these predisposition syndromes. The psychological impact of receiving a diagnosis of ETV6-RT, as well as of FPD-AML or ANKRD26-RT, should be carefully considered by physicians. We suggest that all patients are correctly informed, before undergoing diagnostic workup for thrombocytopenia of suspected genetic origin, about the possibility of receiving a diagnosis that implicates the risk of malignancies, and have the chance to state in advance whether they want to receive information about the risk of neoplasms, for themselves as well as for their progeny. In this study, the in vitro megakaryopoiesis of ETV6-RT patients was investigated for the first time. We showed that ETV6 pathogenic variants impair megakaryocyte maturation, as demonstrated by the production of smaller megakaryocytes with decreased ploidy. The ability of

References 1. Pecci A. Diagnosis and treatment of inherited thrombocytopenias. Clin Genet. 2016;89(2):141-153. 2. Savoia A. Molecular basis of inherited thrombocytopenias. Clin Genet. 2016:89(2): 154-162. 3. Moriyama T, Metzger ML, Wu G, et al. Germline genetic variation in ETV6 and risk of childhood acute lymphoblastic leukaemia: a systematic genetic study. Lancet Oncol. 2015;16(16):1659-1666. 4. Noetzli L, Lo RW, Lee-Sherick AB, et al.

haematologica | 2016; 101(11)

these immature megakaryocytes to extend fully developed proplatelets was impaired, providing an explanation for the patients’ thrombocytopenia. These findings seem consistent with the results of studies in mice, which suggested a role for ETV6 in terminal megakaryocyte maturation,18 and with the findings obtained with megakaryocyte differentiated from human CD34+ cells transduced with some ETV6 variants.4 We also had the possibility to study platelet function in detail in a substantial number of patients. Although we did not identify any consistent defect of in vitro platelet aggregation, activation or adhesion, we found that the ability of platelets to spread on fibrinogen was reduced in all the investigated patients. As the platelet expression of GPIIb-IIIa was normal, this finding suggests that mutations in the ETV6 transcription factor alter the expression of one or more proteins involved in the GPIIb-IIIa-mediated platelet outside-in signaling after interaction with fibrinogen. Moreover, this defect could contribute to the bleeding diathesis observed in some ETV6-RT individuals. In fact, although the degree of bleeding was always mild, the proportion of patients with spontaneous bleeding (60%) appeared globally high with respect to the very mild degree of thrombocytopenia. In conclusion, our study showed that monoallelic ETV6 mutations cause a relatively frequent form of inherited thrombocytopenia and confirmed that affected subjects have a mild bleeding tendency but propensity to hematologic malignancies, in particular ALL. Since ETV6-RT is one of the few autosomal dominant forms of inherited thrombocytopenia without platelet macrocytosis, screening for ETV6 mutations is recommended in all patients with these characteristics. Acknowledgments The authors would like to thank Dr. Carmine Tinelli for his contribution to the statistical analysis, Prof. Federica Meloni for technical assistance with the flow cytometry analysis, Prof. Joseph Italiano for providing β1-tubulin antibody, and Prof. Enrica Tira for providing purified type I collagen. Funding This study was supported by the ERA-Net for Research Program on Rare Diseases (EUPLANE), Telethon Foundation (grant GGP13082), Cariplo Foundation (2012-0529), Italian Ministry of Health (RF-2010-2309222), the Ministry of Education, Youth and Sports of the Czech Republic under the project CEITEC 2020 (LQ1601), and Czech Ministry of Health (grant AZV 1629447A). MF receives a fellowship from the Associazione Italiana per la Ricerca sul Cancro (n. 18024/16).

Germline mutations in ETV6 are associated with thrombocytopenia, red cell macrocytosis and predisposition to lymphoblastic leukemia. Nat Genet. 2015;47(5):535-538. 5. Topka S, Vijai J, Walsh MF, et al. Germline ETV6 mutations confer susceptibility to acute lymphoblastic leukemia and thrombocytopenia. PLoS Genet. 2015;11(6): e1005262. 6. Zhang MY, Churpek JE, Keel SB, et al. Germline ETV6 mutations in familial thrombocytopenia and hematologic malignancy. Nat Genet. 2015;47(2):180-185. 7. Lowe GC, Lordkipanidzé M, Watson SP. Utility of the ISTH bleeding assessment

tool in predicting platelet defects in participants with suspected inherited platelet function disorders. J Thromb Haemost. 2013;11(9): 1663-1668. 8. Noris P, Biino G, Pecci A, et al. Platelet diameters in inherited thrombocytopenias: analysis of 376 patients with all known disorders. Blood. 2014;124(6):e4-e10. 9. Noris P, Guidetti GF, Conti V, et al. Autosomal dominant thrombocytopenias with reduced expression of glycoprotein Ia. Thromb Haemost. 2006;95(3):483-489. 10. Psaila B, Bussel JB, Linden MD, et al. In vivo effects of eltrombopag on platelet function in immune thrombocytopenia: no evidence

1341

F. Melazzini et al.

11.

12.

13.

14.

15.

1342

of platelet activation. Blood. 2012;119(17): 4066-4072. Pecci A, Bozzi V, Panza E, et al. Mutations responsible for MYH9-related thrombocytopenia impair SDF-1-driven migration of megakaryoblastic cells. Thromb Haemost. 2011;106(4): 693-704. Canobbio I, CatricalĂ S, Di Pasqua LG, et al. Immobilized amyloid A peptides support platelet adhesion and activation. FEBS Lett. 2013; 587(16): 2606-2611. Pecci A, Malara A, Badalucco S, et al. Megakaryocytes of patients with MYH9related thrombocytopenia present an altered proplatelet formation. Thromb Haemost. 2009;102(1):90-96. Bluteau D, Balduini A, Balayn N, et al. Thrombocytopenia-associated mutations in the ANKRD26 regulatory region induce MAPK hyperactivation. J Clin Invest. 2014;124(2):580-591. Balduini A, Di Buduo CA, Malara A, et al. Constitutively released adenosine diphosphate regulates proplatelet formation by human megakaryocytes. Haematologica.

2012;97(11):1657-1665. 16. Di Buduo CA, Moccia F, Battiston M, et al. The importance of calcium in the regulation of megakaryocyte function. Haematologica. 2014;99(4):769-778. 17. National Cancer Institute. Cancer statistics. Available at: www.seer.cancer.gov/statistics/. Accessed December 20, 2015. 18. Hock H, Meade E, Medeiros S, et al. Tel/Etv6 is an essential and selective regulator of adult hematopoietic stem cell survival. Genes Dev. 2004;18(19):2336-2341. 19. Bohlander SK. ETV6: a versatile player in leukemogenesis. Semin Cancer Biol. 2005;15(3):162-174. 20. De Braekeleer E, Douet-Guilbert N, Morel F, Le Bris MJ, Basinko A, De Braekeleer M. ETV6 fusion genes in hematological malignancies: a review. Leuk Res. 2012;36(8): 945-961. 21. Wang Q, Dong S, Yao H, et al. ETV6 mutation in a cohort of 970 patients with hematologic malignancies. Haematologica. 2014;99(10):e176-178. 22. Noris P, Favier R, Alessi MC, et al.

23. 24.

25.

26.

ANKRD26-related thrombocytopenia and myeloid malignancies. Blood. 2013;122(11): 1987-1989. Liew E, Owen C. Familial myelodysplastic syndromes: a review of the literature. Haematologica. 2011;96(10):1536-1542 Owen CJ, Toze CL, Koochin A, et al. Five new pedigrees with inherited RUNX1 mutations causing familial platelet disorder with propensity to myeloid malignancy. Blood. 2008;112(12):4639-4645. Churpek JE, Artz A, Bishop M, Liu H, Godley LA. Correspondence regarding the consensus statement from the Worldwide Network for Blood and Marrow Transplantation Standing Committee on Donor Issues. Biol Blood Marrow Transplant. 2016;22(1):183-184. Noris P, Perrotta S, Bottega R, et al. Clinical and laboratory features of 103 patients from 42 Italian families with inherited thrombocytopenia derived from the monoallelic Ala156Val mutation of GPIb (Bolzano mutation). Haematologica. 2012; 97(1):82-88.

haematologica | 2016; 101(11)

ARTICLE

Bone Marrow Failure

Genetic features of myelodysplastic syndrome and aplastic anemia in pediatric and young adult patients

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Siobán B. Keel,1* Angela Scott,2,3,4* Marilyn Sanchez-Bonilla,5 Phoenix A. Ho,2,3,4 Suleyman Gulsuner,6 Colin C. Pritchard,7 Janis L. Abkowitz,1 Mary-Claire King,6 Tom Walsh,6** and Akiko Shimamura5**

Department of Medicine, Division of Hematology, University of Washington, Seattle, WA; 2Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, WA; 3 Department of Pediatric Hematology/Oncology, Seattle Children’s Hospital, WA; 4 Department of Pediatrics, University of Washington, Seattle, WA; 5Boston Children’s Hospital, Dana Farber Cancer Institute, and Harvard Medical School, MA; 6Department of Medicine and Department of Genome Sciences, University of Washington, Seattle, WA; and 7Department of Laboratory Medicine, University of Washington, Seattle, WA, USA 1

*SBK and ASc contributed equally to this work **TW and ASh are co-senior authors

Haematologica 2016 Volume 101(11):1343-1350

ABSTRACT

he clinical and histopathological distinctions between inherited versus acquired bone marrow failure and myelodysplastic syndromes are challenging. The identification of inherited bone marrow failure/myelodysplastic syndromes is critical to inform appropriate clinical management. To investigate whether a subset of pediatric and young adults undergoing transplant for aplastic anemia or myelodysplastic syndrome have germline mutations in bone marrow failure/myelodysplastic syndrome genes, we performed a targeted genetic screen of samples obtained between 1990-2012 from children and young adults with aplastic anemia or myelodysplastic syndrome transplanted at the Fred Hutchinson Cancer Research Center. Mutations in inherited bone marrow failure/myelodysplastic syndrome genes were found in 5.1% (5/98) of aplastic anemia patients and 13.6% (15/110) of myelodysplastic syndrome patients. While the majority of mutations were constitutional, a RUNX1 mutation present in the peripheral blood at a 51% variant allele fraction was confirmed to be somatically acquired in one myelodysplastic syndrome patient. This highlights the importance of distinguishing germline versus somatic mutations by sequencing DNA from a second tissue or from parents. Pathological mutations were present in DKC1, MPL, and TP53 among the aplastic anemia cohort, and in FANCA, GATA2, MPL, RTEL1, RUNX1, SBDS, TERT, TINF2, and TP53 among the myelodysplastic syndrome cohort. Family history or physical examination failed to reliably predict the presence of germline mutations. This study shows that while any single specific bone marrow failure/myelodysplastic syndrome genetic disorder is rare, screening for these disorders in aggregate identifies a significant subset of patients with inherited bone marrow failure/myelodysplastic syndrome.

Introduction The overlap in clinical presentation and bone marrow features of inherited and acquired bone marrow failure syndromes and pediatric myelodysplastic syndromes (MDS) can pose a significant diagnostic challenge. Idiopathic acquired aplastic anemia (AA) is typically a diagnosis of exclusion after inherited disorders, acquired haematologica | 2016; 101(11)

Correspondence: akiko.shimamura@childrens.harvard.edu or sioban@u.washington.edu

Received: May 16, 2016. Accepted: July 13, 2016. Pre-published: July 14, 2016. doi:10.3324/haematol.2016.149476

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1343

1343

S.B. Keel et al.

MDS, infections, metabolic diseases, or nutritional deficiencies have been ruled out. Childhood MDS is emerging as clinically and biologically distinct from MDS in the elderly.1,2 Importantly, and in contrast to MDS in the elderly, the majority of cases of refractory cytopenia of childhood, the most common subtype of pediatric MDS, are hypocellular with a normal karyotype.3 If MDS-related cytogenetic abnormalities are absent, the distinction between refractory cytopenia of childhood and AA or an inherited bone marrow failure or myelodysplastic predisposition syndrome (inherited BMF/MDS) is challenging. Furthermore, patients with inherited BMF/MDS can initially present with MDS or acute leukemia, obscuring the diagnosis of their underlying disorder. The clinical and histopathological distinctions between germline genetic versus acquired disorders have important therapeutic implications. The major treatments for AA currently consist of either hematopoietic stem cell transplant (HSCT) or immunosuppressive therapy with antithymocyte globulin and cyclosporine. The choice of therapy is largely based on the patient’s age, availability of a matched sibling donor, and co-morbidities. The success of immunosuppressive therapy is limited by lack of response in approximately 30% of patients, relapse in 30% of patients who initially respond, and clonal evolution in 10-20% of patients.4 Progressive improvement in outcomes of matched unrelated donor transplants raises the question of whether this should be considered in lieu of immunosuppression as first-line therapy for a subset of patients. In contrast, patients with inherited bone marrow failure syndromes generally have poor or transient responses to immunosuppressive therapy.5-7 In childhood MDS, the indications for HSCT are currently based on the severity of cytopenias, the degree of marrow dysplasia,8 the peripheral blood and marrow blast percentages, and cytogenetic aberrations. In addition to informing the timing and indication for HSCT, accurate distinction of these entities informs HSCT approaches. Many inherited BMF/MDS are associated with excessive transplant regimen-related toxicities and may require specialized reduced intensity conditioning regimens for optimal outcomes.9,10 Additionally, the careful evaluation of a related stem cell donor is critical in the context of a familial genetic disease. The objectives of this study were to investigate whether pediatric and young adult patients undergoing transplant for AA and MDS harbored pathogenic constitutional mutations in inherited BMF/MDS genes and to examine whether family history or physical findings were associated with mutation status. We employed a multiplexed targeted capture gene panel of known inherited and acquired BMF/MDS genes coupled with next generation sequencing to query 208 pediatric and young adult patients referred for HSCT for AA and MDS for mutations in known inherited BMF/MDS genes.

Methods Subjects The study was conducted in accordance with a protocol approved by the Institutional Review Board of the Fred Hutchinson Cancer Research Center and the Declaration of Helsinki. Genomic DNA was obtained from peripheral blood mononuclear cells from the Fred Hutchinson Cancer Research 1344

Center Transplant Genomics Biorepository. This repository contains tissue samples from Fred Hutchinson Cancer Research Center and Seattle Cancer Care Alliance HSCT patients, donors, and family members. Study inclusion criteria were patients transplanted at the Seattle Cancer Care Alliance for whom a pre-transplant sample was available and who were: ≤40 years old and transplanted for AA or ≤45 years old and transplanted for MDS. Patients <20 years of age with monosomy 7 acute myeloid leukemia were also included. Patients presented between 1990 and 2012. As the majority of adult MDS patients present after the age of 60 years (SEER Cancer Statistics Review 1975-2012), an age cut-off of ≤45 years-old was selected to study the genetic features of MDS presenting at an unusually young age. Clinical data were obtained by retrospective chart review of available pre- and post-transplant medical records. Genomic DNA was isolated from peripheral blood mononuclear cells banked in the Genomics Biorepository and from paraffin-embedded clinical tissue biopsy samples (ReliaPrep™ FFPE gDNA Miniprep System, Promega) with the patients’ informed consent in accordance with a protocol approved by the Fred Hutchinson Cancer Research Center. Additional details of the retrospective chart review are provided in the Online Supplementary Methods section.

Genomics Targeted gene capture and massively parallel sequencing were performed as previously described using a capture assay targeting mutations in known inherited and acquired BMF/MDS genes.11 The genes included on the targeted capture panel are listed in Online Supplementary Table S1. Reads were aligned to the human reference genome (hg19) using the Burrows-Wheeler aligner12 and single nucleotide and small insertion-deletion variants called with GATK, using best practice guidelines, as previously described.13 Alignment to the whole genome facilitated exclusion of variants that fell in pseudogenes. Copy number variants were identified as previously described.14 Mutations were identified by a variant allele fraction consistent with heterozygosity (0.3-0.7). Somatic mutations defined by a variant allele fraction of less than 30% were excluded from the analysis as were recessive mutations inconsistent with a Mendelian pattern of inheritance. Variants were classified as pathogenic by predicted effect on protein function, as previously described.15,16 Pathogenic variants were validated by Sanger sequencing. Compound heterozygous variants were confirmed to be in trans by subcloning and Sanger sequencing. Variants were confirmed as constitutional by sequencing DNA isolated from paraffin-embedded biopsy samples of skin, lip, or gastrointestinal tissue or by chromosome fragility testing of patient-derived lymphoblasts, performed as previously described.17 More detailed descriptions are given in the Online Supplementary Methods section.

Results Characteristics of the patients with aplastic anemia The study group transplanted for AA included 53 pediatric patients (≤18 years old) and 45 young adult patients (>18 years old) (range, 1-40 years old). A family history of a first- or second-degree relative with malignancy or cytopenias was present in 40% (39/98) of patients. Physical anomalies were noted in 11% (11/98) of patients. An HLA-matched sibling transplant was performed for 38 patients, six of whom had received immunosuppressive therapy prior to transplantation. Fifty-four patients received alternative donor HSCT due to refractory or haematologica | 2016; 101(11)

Genetics of MDS and AA in pediatric and young adult patients

relapsed disease following immunosuppressive therapy and the remaining six patients received alternative donor HSCT after other medical therapies had failed (Online Supplementary Table S2).

Characteristics of the patients with myelodysplastic syndrome Among the 135 patients transplanted for MDS, 25 patients with a current or antecedent history of a myeloproliferative disorder were excluded from further analyses, leaving a total of 110 patients for study (Online Supplementary Table S3). The final study group consisted of 46 pediatric patients (â&#x2030;¤18 years old) and 64 young adult patients (>18 years old) (range, 1-46 years old). A family history of a first- or second-degree relative with malignancy or cytopenias was present in 52% (57/110) of patients. Physical anomalies were noted in 24% (26/110) of patients.

Analysis of inherited bone marrow failure or myelodysplastic predisposition syndrome genes Given the diagnostic challenge of discriminating between inherited versus acquired BMF/MDS, we hypothesized that a subset of pediatric and young adult patients who underwent HSCT at our center for AA or MDS had cryptic inherited disorders. This hypothesis was addressed using a targeted capture panel coupled to high-throughput, next-generation sequencing of genes known to contribute to inherited and acquired BMF/MDS (Online Supplementary Table S1).11 For all samples evaluated, the median coverage across the 712 kb targeted region was 444X, with 99.4% of bases having >50X coverage and 99.6% of bases having >10X coverage. This depth of coverage enabled identification of point mutations, small indels, and copy number variants spanning three exons.

Aplastic anemia genetics Pathological constitutional mutations in known inherited BMF/MDS genes were identified in five of 98 (5.1%) pediatric and young adult patients with AA (Table 1 and Online Supplementary Table S4). Hemizygous mutations in DKC1 were identified in patients AA3 (T66A) and AA45

(c.-142 C>G). Compound heterozygous and homozygous mutations in MPL were identified in patients AA25 (R102P and W515X) and AA37 (P394S), respectively, and a heterozygous mutation in TP53 was identified in patient AA79 (R196Q). Constitutional mutations in DKC1, MPL, and TP53 cause X-linked dyskeratosis congenita, autosomal recessive congenital amegakaryocytic thrombocytopenia, and the autosomal dominant familial cancer predisposition syndrome, Li Fraumeni, respectively.7 Patients with germline TP53 mutations have an increased risk of developing acute myeloid leukemia and MDS.18-20 The genetics are described in more detail in the Online Supplementary Results.

Aplastic anemia genotype-phenotype analysis Among the five AA patients in whom constitutional mutations in known inherited BMF/MDS genes were found, two patients were clinically suspected to have a constitutional disorder prior to their initial HSCT. Patient AA3 was diagnosed with dyskeratosis congenita at 10 years old based on the findings of his physical examination (nail dystrophy and a hyperpigmented macule), a family history of four brothers with clinical dyskeratosis congenita and several family members with early cancers. Patient AA25 presented with an intracranial hemorrhage and severe thrombocytopenia at birth and was diagnosed with congenital amegakaryocytic thrombocytopenia which was genetically confirmed. Of the five patients with constitutional mutations, three received an upfront transplant from an HLA-matched sibling donor. One patient (AA37) received immunosuppressive therapy prior to an alternative donor transplant and a second patient, AA25, received granulocyte colony-stimulating factor and erythropoietin prior to undergoing an alternative donor transplant. Patient AA3 died of colon cancer 21 months after transplantation (Table 1). Colon cancer is a known complication of dyskeratosis congenita.21 A negative family history did not reliably distinguish those patients lacking mutations in inherited BMF/MDS genes (Table 2). Absence of physical anomalies also failed to distinguish patients with acquired versus germline genetic forms of AA (Table 2).

Table 1. Clinical and genetic features of AA patients.

Sex

Age (years)

Gene

Mutation

Family history**

Physical anomalies

IST

Sibling donor

AA3

DKC1

T66A

F F

1 7

MPL MPL

R102P/W515X P394S/P394S

Nail dystrophy, hyperpigmented macules -

AA25 AA37

4 brothers with clinical DC; early cancers in family -

AA45 AA79

M M

9 6

DKC1 TP53

c.-142 C>G R196Q

Facial hyperpigmentation -

PostTime from Cause transplant transplant of death complications until death (months) Colon cancer

Failed to engraft initial cord transplant, engrafted 2nd transplant; BOOP + + -

Colon cancer

*Age at transplant; **family history indicates family history of related phenotype or cancer in first- or second-degree relative. AA, idiopathic acquired aplastic anemia; IST: immunosuppressive therapy; DC: dyskeratosis congenita; BOOP: bronchiolitis obliterans organizing pneumonia.Two of the 98 AA patients were included among the pediatric and adult patients with marrow failure or MDS deemed to have idiopathic disease after laboratory and clinically-directed genetic evaluation in the report by Zhang et al.11 (AA92/FH-50 and AA87/FH-13).

haematologica | 2016; 101(11)

1345

S.B. Keel et al. Table 2. Genotype-phenotype analyses in AA patients without or with constitutional mutations.

Phenotypic findings Median age*, years (range) Personal history of cancer prior to HSCT, percent (n) Family history**, percent (n) Physical anomalies, percent (n) Three-year overall survival after HSCT, percent (n) Cancers after HSCT, excluding relapsed disease, percent (n)

Patients without mutations (n=93)

Patients with mutations (n=5)

18.9 (1-40) 4% (4) 41% (38) 10% (9) 76% (71) 10% (9)

12.5 (1-33) 20% (1) 20% (1) 40% (2) 80% (4) 20% (1)

*Age at transplant; **family history indicates family history of related phenotype or cancer in first- or second- degree relative. AA: idiopathic acquired aplastic anemia; HSCT: hematopoietic stem cell transplant.

The presence of a paroxysmal nocturnal hemoglobinuria clone was assessed by flow cytometry using fourcolor or six-color flow cytometry of marrow or peripheral blood, or, in one patient, by the Ham test. A paroxysmal nocturnal hemoglobinuria clone was detected in five of 50 patients tested (10%), none of whom harbored a constitutional mutation in an inherited BMF/MDS gene.

Myelodysplastic syndrome genetics Pathological mutations in known inherited BMF/MDS genes were identified in 15 out of 110 (13.6%) pediatric and young adult patients with MDS (Table 3 and Online Supplementary Table S4). The mutations were constitutional in 14 out of these 15 patients. Compound heterozygous mutations were identified in FANCA (HIP12286: p.Q549X and c.3349-1G>A), MPL (HIP05737: p.L79Efs* and c.393+5G>C), RTEL1 (HIP02696: p.M492I and R1264H; HIP17561: p.T56M and H116R), and SBDS (HIP02099: p.K62X and c.258+2T>C); constitutional mutations in these genes cause autosomal recessive Fanconi anemia, congenital amegakaryocytic thrombocytopenia, dyskeratosis congenita, and Shwachman-Diamond syndrome, respectively. Heterozygous mutations were identified in GATA2 (HIP08919: p.G101Afs*; HIP17707: c.1017+2T>C; HIP18921: p.R330X; HIP18952: p.T354M; HIP20476: p.R361C), RUNX1 (HIP14128: p.R293X), TERT (HIP02687: P704S), TINF2 (HIP05477: R282H), and TP53 (HIP21264: R209Ffs*; HIP01569: G245S); constitutional mutations in these genes cause autosomal dominant inherited predisposition to leukemia and MDS (GATA2, RUNX1), dyskeratosis congenita (TERT, TINF2), and Li Fraumeni syndrome (TP53). Patient HIP14128 is a particularly instructive clinical case. This patient carried a heterozygous nonsense mutation in RUNX1 (c.958C>T [p.R293X]) in DNA isolated from pre-transplant peripheral blood. This mutation was previously reported as a somatic variant in de novo acute myeloid leukemia22 and occurs in a highly conserved amino acid in the transactivation domain of the protein. Importantly, in DNA isolated from the patientâ&#x20AC;&#x2122;s small bowel and skin, the patient was wild-type at this base pair, consistent with a somatically acquired, rather than inherited mutation. This distinction is critical, as germline mutations in RUNX1 cause an autosomal dominant familial platelet disorder with a high risk of developing a myeloid malignancy23 necessitating close clinical monitoring of the patient and appropriate genetic counseling of family members. Distinguishing somatic from germline mutations is critical for clinical management decisions. 1346

Additional details are available in the Online Supplementary Results.

Myelodysplastic syndrome genotype-phenotype analysis Among the 14 MDS patients in whom constitutional mutations in known inherited BMF/MDS genes were found, only three patients were clinically suspected to have a constitutional disorder prior to their initial HSCT. Patient HIP05737 was diagnosed with congenital amegakaryocytic thrombocytopenia during her pre-transplant evaluation and patient HIP02099 was suspected to have Shwachman-Diamond syndrome before transplant based on a history of marrow failure and pancreatic insufficiency presenting in a 3.5-month old boy (without genetic confirmation of the diagnosis). Patient HIP21264 was a member of a known Li-Fraumeni family whose TP53 mutation was formerly defined. Additionally, while patients with constitutional mutations presented for HSCT at a younger age than those lacking mutations (median age 12.1 years old versus 31.2 years old), the absence of a prior personal history of cancer, family history of a first- or second-degree relative with cancer or a related phenotype, or the presence of physical anomalies did not reliably rule out mutations in inherited BMF/MDS genes (Table 4). Only two of the five patients carrying heterozygous pathogenic mutations in GATA2 had a family history suspicious of an inherited disorder (HIP08919 and HIP18921, Table 3). Standard bone marrow transplant conditioning regimens used for MDS can cause excessive and life-threatening toxicities in patients with certain inherited BMF/MDS.7 Some patients, for example those with dyskeratosis congenita,9 may also face a higher risk of graft failure. The clinical course of patient HIP12286, who carried compound heterozygous deleterious mutations in FANCA, illustrates this issue. She initially presented at the age of 20 years (5 years prior to transplantation) with epistaxis and was found to have mild thrombocytopenia and macrocytosis in the context of a hypocellular marrow for her age. Her family history was unremarkable and she lacked physical or structural anomalies suggestive of an inherited BMF/MDS. She developed progressive cytopenias and monosomy 7 and received an unrelated donor bone marrow transplant with busulfan and cyclophosphamide conditioning. She died 23 days after transplantation of severe transplant-related toxicity, including whole body blistering and a desquamative rash requiring surgical debridement, hypoxic respiratory failure with evidence of diffuse alveolar damage, and severe diarrhea with endoscopic findings of necrotic gastric mucosa. Patients with haematologica | 2016; 101(11)

Genetics of MDS and AA in pediatric and young adult patients Table 3. Clinical and genetic features of the MDS patients.

Sex

Age Gene (years)*

Mutation

Family Physical Sibling history ** anomalies donor

HIP12286

FANCA

Q549X; c.3349-1G>A

HIP08919

GATA2

G101Afs*

HIP17707

GATA2 c.1017+2T>C

HIP18921

GATA2

HIP18952 HIP20476 HIP05737

F F F

10 16 11

GATA2 GATA2 MPL

+ +

HIP02696

T354M R361C L79Efs*; c.393+5G>C RTEL1 M492I; R1264H

HIP17561

RTEL1 T56M; H116R

HIP14128 *** M HIP02099 M

40 5

RUNX1 SBDS

R293X K62X; c.258+2T>C

+ +

+ -

HIP02687

TERT

P704S

HIP05477

TINF2

R282H

HIP21264 HIP01569

F M

22 7

TP53 TP53

R209Ffs* G245S

+ +

HIP01569

TP53

G245S

R330X

Post-transplant complications

Graft failure

Time from transplant until death (days)

Cause of death

Extensive acute toxicity: day +4 developed skin bullae requiring debridement, severe mucositis, LFT abnormalities, and renal insufficiency; day +8 developed hypoxia and diffuse alveolar damage; day +12 developed diarrhea with necrotic gastric mucosa -

Transplantrelated toxicity

1085

Persistent AML ~ 2 years after 1st HSCT. Died 75 days after 2nd HSCT with persistent AML of idiopathic pneumonitis on day +4 of reinduction chemotherapy

Day +80 developed sclerodermatous skin changes and pulmonary function tests and CT scan findings suspicious for bronchiolitis obliterans Day +8 developed acute kidney injury and severe liver sinusoidal obstruction syndrome -

2920

Multiorgan system failure

4581

Liver failure presumed alcohol-related

Graft failure complicated by CNS toxicity secondary to cyclosporine (headaches and focal cerebellar changes by MRI); Candida parapsilosis fungemia; gut GvHD; CMV reactivation; Staphylococcus epidermidis and Corynebacterium bacteremia and Clostridium difficile colitis - Post-transplant pulmonary complications and chronic liver GvHD (transaminitis) - Hyperacute GvHD and moderate/severe mucositis

209

2754

Relapsed AML

1092

Hypoxic respiratory failure and chronic GvHD of skin, gut, and lungs receiving immunosuppressive medications

Hyperacute GvHD

306

Hyperacute GvHD

Graft failure X 2; died of multiorgan system failure after three HSCT -

Day +65 after 2nd HSCT, after successful engraftment, of CMV pneumonia requiring intubation and CMV enteritis

Relapsed AML ~9 months after transplant; died of complications of reinduction chemotherapy (capillary leak syndrome) 306 Relapsed AML ~9 months after transplant of complications (capillary leak) from induction chemotherapy

*Age at transplant; **family history indicates family history of related phenotype or cancer in first- or second-degree relative. ***Mutation was not constitutional.AML: acute myeloid leukemia; LFT: liver function test; CT: computed tomography; HSCT: hematopoietic stem cell transplant; CNS: central nervous system; MRI: magnetic resonance imaging; CMV: cytomegalovirus; GvHD: graft-versus-host disease; Two of the 110 MDS patients were included among the pediatric and adult patients with marrow failure or MDS deemed to have idiopathic disease after laboratory and clinically-directed genetic evaluation in the report by Zhang et al.11 (HIP02696/FH-54 and HIP21822/FH-67).

haematologica | 2016; 101(11)

1347

S.B. Keel et al. Table 4. Genotype-phenotype associations in MDS patients without or with constitutional mutations.

Cancers after HSCT, excluding relapsed disease, percent (n)

Patients without mutations (n=96)

Patients with mutations (n=14)

31 (1-46) 17% (16) 49% (47) 24% (23) 60% 38 deaths, 6/38 died >3 years after transplant 8 deaths, 4/8 died >3 years after transplant 3% (3)

12 (2-41) 21% (3) 71% (10) 21% (3) 43% 0 % (0)

*Age at transplant; **family history indicates family history of related phenotype or cancer in first- or second-degree relative. MDS: myelodysplastic syndrome;HSCT: hematopoietic stem cell transplant.

Fanconi anemia are exquisitely sensitive to genotoxic agents (e.g., alkylators and ionizing radiation) and face potentially severe regimen-related toxicity and high mortality rates with standard conditioning regimens containing these agents. Patient HIP02696, who carried compound heterozygous mutations in RTEL1, provides a second example of transplant outcomes in a patient with an unrecognized inherited BMF/MDS at the time of their initial transplant, specifically, Hoyeraal-Hreidarsson syndrome. His transplant courses were marked by recurrent graft failure â&#x20AC;&#x201C; initially after a non-myeloablative HLA-matched unrelated donor transplant utilizing a preparative regimen containing fludarabine and 200 cGy total body irradiation and again after a second non-myeloablative peripheral blood stem cell transplant from the same donor utilizing a preparative regimen containing fludarabine, antithymocyte globulin, and cyclophosphamide. He ultimately engrafted after infusion of bone marrow hematopoietic stem cells without conditioning from the same original donor and died 95 months after his original transplant from multi-organ system failure. Of note, among the six patients with damaging mutations in dyskeratosis congenita genes in this study (Online Supplementary Table S5), pulmonary complications developed in one, about 6 months after transplantation.

Discussion Inherited bone marrow failure syndromes, idiopathic AA, and hypoplastic MDS share significant overlap in clinical presentation and bone marrow findings, which often pose a diagnostic challenge. In the present study we determined whether pediatric and young adults referred to our center for HSCT for either AA or MDS harbor mutations in inherited BMF/MDS genes. We found that 5.1% of AA and 13.6% of MDS patients carried mutations in known inherited BMF/MDS genes. Prior studies have sought to define this incidence by analyzing single genes or small sets of genes. Genetic analyses were confined to single genes or limited gene sets. Yamaguchi et al. screened patients with apparently acquired AA for mutations in TERT, DKC1, NHP2, and NOP10, and reported heterozygous mutations in TERT in ~3.4% (7/205) of patients with AA unresponsive to immunosuppressive therapy.5 Their study included patients aged 2-83 years old with a median age of 34 years 1348

old, which is older than our series of AA patients (median age 18 years old). A second study of TERC mutations in 210 patients with bone marrow failure syndromes, including 150 AA and 55 MDS patients, identified heterozygous mutations in only 3/210 (1.4%) cases.6 Field et al. found only two TERC mutations in 284 blood samples obtained from pediatric patients with AA (n=109), MDS (n=137), or juvenile myelomonocytic leukemia (n=38) under 18 years of age who underwent a stem cell transplant for marrow failure from the National Donor Marrow Program Research Sample Repository.24 The incidence of inherited BMF/MDS appeared to be low based on these studies. A recent German study, limited to GATA2 sequencing, demonstrated a 7% (28/426) incidence of germline GATA2 mutations among children and adolescents with de novo MDS.25 Our study here demonstrated, by expanding the genetic screen to include a large set of inherited BMF/MDS genes, that a significant subset of patients carried germline mutations. These results extend a previous study in which we found that 11% (8/71) of patients, primarily from the pediatric age group, harbored germline BMF/MDS gene mutations.11 While the majority of mutations identified in our previous study affected the GATA2 gene, our current study identified mutations in a wide variety of genes. A subsequent study utilized a targeted next-generation sequencing panel of 72 inherited bone marrow failure genes to screen ten patients with severe AA unresponsive to immunosuppressive therapy.26 Mutations were found in three of ten patients (30%). One patient carried compound heterozygous mutations in RTEL1, one patient had a heterozygous TERT mutation, and a third patient had a heterozygous mutation in microtubule-associated serine/threonine kinase like (MASTL), which had previously been reported in autosomal dominant thrombocytopenia. This group additionally identified mutations involving CXCR4, GATA2, G6PC3, MAST, MYH9, RPL5, RTEL1, TERT, TERC, TINF2, or WAS in 18.1% (15/83) of patients with unclassified inherited bone marrow failure syndromes.26 These studies demonstrate that while mutations in any single gene are rare, mutations in BMF/MDS genes as a group are found in a significant subset of patients. The results of this study underscore the limitations of relying on clinical stigmata or family history to identify patients with an underlying constitutional disorder. Only two of the five AA patients with mutations in an inherited BMF/MDS gene had a physical anomaly and only haematologica | 2016; 101(11)

Genetics of MDS and AA in pediatric and young adult patients

one patient had a family history of cancer or a related phenotype. Among the MDS cohort, only three of the 14 patients with mutations in an inherited BMF/MDS gene had a physical anomaly and while patients with mutations were more likely to have a positive family history of cancer or a related phenotype than those patients without mutations, the absence of a family history did not exclude the possibility of a mutation. Although the retrospective design of this study limits the available physical anomaly and family history data to those captured in the clinical records and further studies are warranted to confirm the generalizability of our findings, our conclusions are consistent with other published work. Our earlier study identified mutations in eight of 71 (11%) pediatric and adult patients (<40 years old at presentation or with a family history of BMF/MDS regardless of age) with idiopathic bone marrow failure, with all eight patients lacking the classical clinical stigmata of these syndromes and only four having a suggestive family history despite extensive longitudinal evaluation.11 Our study both complements and expands on a recent large study of 1,120 pediatric cancer patients which found that 8.5% of patients had germline mutations in inherited cancer predisposition genes and more than half of the children with germline mutations lacked any family history of cancer.27 Our study also highlights the importance of recognizing inherited BMF/MDS to inform appropriate medical care. One of the five AA patients with mutations in inherited BMF/MDS genes received immunosuppressive therapy without response. Patients with inherited BMF/MDS respond poorly or transiently to immunosuppressive therapy and are at increased risk of relapse and clonal evolution with reduced survival.28 Additionally, sibling mutation status would inform transplant donor choice. For relapsed MDS after a sibling donor transplant, it would be informative to determine whether relapsed MDS involved donor or recipient-derived cells. Published reports suggest a high transplant-related mortality and organ toxicity in patients with certain inherited BMF/MDS transplanted using standard preparatory regimens for AA or MDS.7,9,10 The severe and immediate post-transplant complications observed in the Fanconi anemia, subtype A patient in this study (HIP12866) provides an instructive example of this risk. Additionally, dyskeratosis congenita appears to be associated with early and late graft failure and pulmonary toxicity with standard transplant preparatory regimens,9 which was observed in this series (HIP02696 and HIP05477, respectively). Given the increased risks of hematologic and solid tumors associated with inherited BMF/MDS, a clinical surveillance protocol for patients has the potential to improve outcomes in individuals and affected family members of genetically defined probands. These patients may face a particularly high risk of tumors following HSCT. One of the challenges in genetic studies is the ascertainment of pathogenicity for a given variant. As our understanding evolves, reassessment of previously described variants is continually warranted. For example, we observed the TERT c.C1234T (H412Y) variant in a 10year old girl with MDS, a 2-year old boy with AA and a 27-year old man with AA and have also observed this variant in four other patients among the 1042 patients with bone marrow failure or MDS sequenced on our targeted gene sequencing platform. This missense mutation haematologica | 2016; 101(11)

affects an amino acid that is not highly conserved across species and is reported in 121 alleles (among a total of 19080) on the Exome Aggregation Consortium, including two homozygous individuals (ExAC, Cambridge, MA, USA). The lack of amino acid conservation across species and frequent observation in our laboratory and on ExAC suggest that while the variant reduces telomerase activity in vitro,5 it might be clinically benign in humans in the heterozygous state. Du et al. described a family in which the proband had clinical dyskeratosis congenita, very short telomeres, and a homozygous TERT P704S mutation resulting in reduced telomerase activity. The probandâ&#x20AC;&#x2122;s father carried compound heterozygous P704S and H412Y TERT mutations and had very short telomeres. The probandâ&#x20AC;&#x2122;s mother had neither clinical evidence of dyskeratosis congenita nor very short telomeres in peripheral blood mononuclear cells and was heterozygous for the P704S allele. Since both mutations are hypomorphic alleles, impairing but not eliminating telomerase activity,29 these data suggest that hypomorphic TERT mutations might contribute to disease when present together in the compound heterozygous or homozygous state. Further studies are warranted to clarify whether a single heterozygous TERT H412Y variant is pathogenic in humans. Several limitations of this study must be considered. While the rarity of AA and MDS in children and young adults precluded a prospective study, the retrospective study design limits the available clinical history. Further studies with comprehensive clinical data, including a detailed family history and physical examination findings with attention to subtle features associated with bone marrow failure, to accompany the genetic evaluation, are warranted. The frequency of germline mutations in our study group may be underestimated since our variant pipeline filtered out mutations with a variant allele fraction of less than 0.30, which may have eliminated patients with somatic mosaicism.27 In current practice, testing for an underlying genetic disorder is most commonly considered in pediatric and young adult patients presenting with bone marrow failure. Genetic testing is rarely pursued in pediatric and young adult patients presenting with MDS. While screening BMF/MDS patients for mutations in a single gene or small set of genes reveals rare cases of germline mutations, we found that comprehensive screening of BMF/MDS genes in aggregate revealed a significant subset of previously unrecognized patients. As we restricted our AA analysis to those patients with severe AA requiring HSCT, it will be interesting to determine the frequency of constitutional mutations in inherited BMF/MDS genes among patients diagnosed with moderate AA. Since many of these disorders require specialized evaluation of family donors and counseling, reduced intensity conditioning regimens, and surveillance for disease-specific extra-hematopoietic complications, diagnosis prior to transplantation is critical. Our findings suggest that genetic screening to evaluate the inherited BMF/MDS genes collectively should be considered in pediatric and young adult patients presenting with AA or MDS. Acknowledgments The authors would like to thank Sharon Savage, Clinical Genetics Branch, Division of Cancer Epidemiology and 1349

S.B. Keel et al.

Genetics, National Cancer Institute, National Institutes of Health, Bethesda, MD, and Keith Loeb, Department of Pathology, University of Washington, Seattle, WA and J. Arturo Londoño-Vallejo, Institut Curie, Paris, France and Simon Boulton, Francis Crick Institute, UK for helpful discussions.

References 1. Hirabayashi S, Flotho C, Moetter J, et al. Spliceosomal gene aberrations are rare, coexist with oncogenic mutations, and are unlikely to exert a driver effect in childhood MDS and JMML. Blood. 2012;119(11):e9699. 2. Hasle H, Niemeyer CM, Chessells JM, et al. A pediatric approach to the WHO classification of myelodysplastic and myeloproliferative diseases. Leukemia. 2003;17(2):277-282. 3. Niemeyer CM, Baumann I. Classification of childhood aplastic anemia and myelodysplastic syndrome. Hematology Am Soc Hematol Educ Program. 2011;2011:84-89. 4. Young NS, Calado RT, Scheinberg P. Current concepts in the pathophysiology and treatment of aplastic anemia. Blood. 2006;108 (8):2509-2519. 5. Yamaguchi H, Calado RT, Ly H, et al. Mutations in TERT, the gene for telomerase reverse transcriptase, in aplastic anemia. N Engl J Med. 2005;352(14):1413-1424. 6. Yamaguchi H, Baerlocher GM, Lansdorp PM, et al. Mutations of the human telomerase RNA gene (TERC) in aplastic anemia and myelodysplastic syndrome. Blood. 2003;102(3):916-918. 7. Shimamura A, Alter BP. Pathophysiology and management of inherited bone marrow failure syndromes. Blood Rev. 2010;24(3): 101-122. 8. Alter BP, Elghetany MT. The CCC system: is it really the answer to pediatric MDS? J Pediatr Hematol Oncol. 2003;25(5):426-427; author reply 427-428. 9. Gadalla SM, Sales-Bonfim C, Carreras J, et al. Outcomes of allogeneic hematopoietic cell transplantation in patients with dyskeratosis congenita. Biol Blood Marrow Transplant. 2013;19(8):1238-1243. 10. Gluckman E, Devergie A, Schaison G, et al. Bone marrow transplantation in Fanconi anaemia. Br J Haematol. 1980;45(4):557-564. 11. Zhang MY, Keel SB, Walsh T, et al. Genomic

1350

12.

13.

14.

15.

16.

17.

18.

19.

20.

This work was supported by NIH/NIDDK grant R24DK099808 (AS, MCK, and JA), the Aplastic Anemia & Myelodysplastic Syndrome International Foundation, the Ghiglione Aplastic Anemia Fund, and the Julian’s Dinosaur Guild at Seattle Children’s Hospital (AS).

analysis of bone marrow failure and myelodysplastic syndromes reveals phenotypic and diagnostic complexity. Haematologica. 2015;100(1):42-48. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25(14):17541760. Gulsuner S, Walsh T, Watts AC, et al. Spatial and temporal mapping of de novo mutations in schizophrenia to a fetal prefrontal cortical network. Cell. 2013;154(3):518-529. Nord AS, Lee M, King MC, Walsh T. Accurate and exact CNV identification from targeted high-throughput sequence data. BMC Genomics. 2011;12:184. Walsh T, Lee MK, Casadei S, et al. Detection of inherited mutations for breast and ovarian cancer using genomic capture and massively parallel sequencing. Proc Natl Acad Sci USA. 2010;107(28):12629-12633. Walsh T, Casadei S, Lee MK, et al. Mutations in 12 genes for inherited ovarian, fallopian tube, and peritoneal carcinoma identified by massively parallel sequencing. Proc Natl Acad Sci USA. 2011;108(44): 18032-18037. Shimamura A, Montes de Oca R, Svenson JL, et al. A novel diagnostic screen for defects in the Fanconi anemia pathway. Blood. 2002;100(13):4649-4654. Kuribayashi K, Matsunaga T, Sakai T, et al. A patient with TP53 germline mutation developed Bowen's disease and myelodysplastic syndrome with myelofibrosis after chemotherapy against ovarian cancer. Intern Med. 2005;44(5):490-495. Felix CA, Hosler MR, Provisor D, et al. The p53 gene in pediatric therapy-related leukemia and myelodysplasia. Blood. 1996;87(10):4376-4381. Anensen N, Skavland J, Stapnes C, et al. Acute myelogenous leukemia in a patient with Li-Fraumeni syndrome treated with valproic acid, theophyllamine and all-trans retinoic acid: a case report. Leukemia. 2006;20(4):734-736.

21. Alter BP, Giri N, Savage SA, Rosenberg PS. Cancer in dyskeratosis congenita. Blood. 2009;113(26):6549-6557. 22. 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. 23. Song WJ, Sullivan MG, Legare RD, et al. Haploinsufficiency of CBFA2 causes familial thrombocytopenia with propensity to develop acute myelogenous leukaemia. Nat Genet. 1999;23(2):166-175. 24. Field JJ, Mason PJ, An P, et al. Low frequency of telomerase RNA mutations among children with aplastic anemia or myelodysplastic syndrome. J Pediatr Hematol Oncol. 2006;28(7):450-453. 25. Wlodarski MW, Hirabayashi S, Pastor V, et al. Prevalence, clinical characteristics, and prognosis of GATA2-related myelodysplastic syndromes in children and adolescents. Blood. 2016;127(11):1387-1397. 26. Ghemlas I, Li H, Zlateska B, et al. Improving diagnostic precision, care and syndrome definitions using comprehensive next-generation sequencing for the inherited bone marrow failure syndromes. J Med Genet. 2015;52(9):575-584. 27. Zhang J, Walsh MF, Wu G, et al. Germline Mutations in predisposition genes in pediatric cancer. N Engl J Med. 2015;373(24): 2336-2346. 28. Scheinberg P, Cooper JN, Sloand EM, Wu CO, Calado RT, Young NS. Association of telomere length of peripheral blood leukocytes with hematopoietic relapse, malignant transformation, and survival in severe aplastic anemia. JAMA. 2010;304(12):13581364. 29. Du HY, Pumbo E, Manley P, et al. Complex inheritance pattern of dyskeratosis congenita in two families with 2 different mutations in the telomerase reverse transcriptase gene. Blood. 2008;111(3):11281130.

haematologica | 2016; 101(11)

ARTICLE

Acute Myeloid Leukemia

An operational definition of primary refractory acute myeloid leukemia allowing early identification of patients who may benefit from allogeneic stem cell transplantation

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Paul Ferguson,1 Robert K. Hills,2 Angela Grech,2 Sophie Betteridge,2 Lars Kjeldsen,3 Michael Dennis,4 Paresh Vyas,5 Anthony H. Goldstone,6 Donald Milligan,7 Richard E. Clark,8 Nigel H. Russell9 and Charles Craddock10 on behalf of the UK NCRI AML Working Group

Queen Elizabeth Hospital Birmingham NHS Foundation Trust, UK; Cardiff University School of Medicine, UK; 3Copenhagen University Hospital, Denmark; 4 The Christie NHS Foundation Trust, Manchester, UK; 5University of Oxford and Oxford University Hospitals NHS Trust, UK; 6University College Hospital, London, UK; 7 Birmingham Heartlands Hospital, UK; 8Royal Liverpool University Hospital, UK; 9 Nottingham University Hospital, UK; and 10University of Birmingham, UK 1 2

Haematologica 2016 Volume 101(11):1351-1358

ABSTRACT

p to 30% of adults with acute myeloid leukemia fail to achieve a complete remission after induction chemotherapy - termed primary refractory acute myeloid leukemia. There is no universally agreed definition of primary refractory disease, nor have the optimal treatment modalities been defined. We studied 8907 patients with newly diagnosed acute myeloid leukemia, and examined outcomes in patients with refractory disease defined using differing criteria which have previously been proposed. These included failure to achieve complete remission after one cycle of induction chemotherapy (RES), less than a 50% reduction in blast numbers with >15% residual blasts after one cycle of induction chemotherapy (REF1) and failure to achieve complete remission after two courses of induction chemotherapy (REF2). 5year overall survival was decreased in patients fulfilling any criteria for refractory disease, compared with patients achieving a complete remission after one cycle of induction chemotherapy: 9% and 8% in patients with REF1 and REF 2 versus 40% (P<0.0001). Allogeneic stem cell transplantation improved survival in the REF1 (HR 0.58 (0.46-0.74), P=0.00001) and REF2 (HR 0.55 (0.41-0.74), P=0.0001) cohorts. The utilization of REF1 criteria permits the early identification of patients whose outcome after one course of induction chemotherapy is very poor, and informs a novel definition of primary refractory acute myeloid leukemia. Furthermore, these data demonstrate that allogeneic stem cell transplantation represents an effective therapeutic modality in selected patients with primary refractory acute myeloid leukemia.

Correspondence: charles.craddock@uhb.nhs.uk

Received: May 9, 2016. Accepted: August 10, 2016. Pre-published: August 18, 2016. doi:10.3324/haematol.2016.148825

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1351

Introduction Up to 30% of adults with newly diagnosed acute myeloid leukemia (AML) fail to achieve a morphological complete remission (CR) after one or two courses of induction chemotherapy (IC).1 Although the outcome of AML refractory to IC is known to be poor, the optimal management of this important cause of treatment failure remains undetermined. Whilst a number of recent registry studies have demonstrated long-term survival after allogeneic stem cell transplantation (SCT) in patients with primary refractory AML (PREF AML), interpretation of these data are complicated by multiple factors, haematologica | 2016; 101(11)

1351

P. Ferguson et al.

including limited cohort size, selection bias and lead time reporting errors.2-7 Nonetheless, the advent of reduced intensity conditioning (RIC) regimens, coupled with increased numbers of alternative stem cell donors, has resulted in allogeneic SCT becoming an increasingly deliverable treatment option in PREF AML, emphasizing the importance of defining its curative potential in this setting. At the same time there remains considerable debate concerning the definition of primary refractory disease.8 The International Working Group (IWG) and the European LeukemiaNet (ELN) both define resistant disease after induction therapy as persistent leukemic blasts in either the peripheral blood or the bone marrow in a patient alive seven days or more following treatment.9,10 However, most studies investigating the impact of allogeneic SCT in AML refractory to induction therapy have defined refractoriness as a failure to achieve CR following two courses of chemotherapy.2,3,5,7,11,12 A number of reports have demonstrated that failure to achieve CR after one course of IC is an adverse prognostic indicator; however, this has not been universally reported.13 The UK Medical Research Council (MRC) data have previously demonstrated that patients who had between 5-15% residual leukemic blasts following their first cycle of IC had similar relapse rates to those who achieved a CR, although they demonstrated a reduced overall survival (OS).14 Schlenk et al. analyzed 223 patients enrolled on the HD93 trial and defined those with a <50% reduction in bone marrow blasts following one course of IC as having refractory disease. In this relatively small study, patients with refractory disease defined using this criterion demonstrated a lower OS than patients in CR.15 Previous studies which have defined refractory disease as failure to achieve a CR after two courses of IC, have consistently demonstrated an extremely poor survival rate in this sizeable proportion of newly diagnosed patients.5,16,17 Importantly, to our knowledge, there have been no systematic comparisons of outcome according to different definitions of putative refractoriness in a large cohort of patients, nor has the impact of allogeneic SCT been systematically evaluated. We have therefore analyzed the outcome of patients with AML resistant to induction therapy, utilizing different definitions of PREF AML, in order to generate diagnostic criteria and examine whether patients genuinely refractory to IC can be identified earlier in their treatment path-

way. This has allowed us to study the role of allogeneic SCT in the management of PREF AML - an important but largely ignored disease entity.

Methods We performed a retrospective analysis of patient data on 8907 patients with non-promyelocytic AML treated with intensive chemotherapy regimens on the MRC/NCRI AML 10, 11, 12, 14, 15 and 16 trials. The AML 11, 14 and 16 trials were predominantly for older AML patients (>60 years), and their treatment intensity was reduced compared with trials for younger AML patients. The trial chemotherapy regimens used have been previously outlined,1,18-21 and are summarized in the Online Supplementary Figure S1. Trials were conducted in accordance with the declaration of Helsinki, were approved by the Wales multi-center research ethics committee and participating institutions ethical review committees, and patients provided written informed consent for their inclusion in each trial and for the use of their clinical data in the outcome analysis. Karyotype risk stratification was designated according to Grimwade et al.22 Bone marrow blasts were analyzed for FLT3 internal tandem duplications (ITD) and NPM1 mutations as previously described.1 Response to IC was assessed by bone marrow evaluation performed 14-21 days after completion of chemotherapy. Complete response was defined as the presence of less than 5% blasts in the bone marrow. In patients who failed to achieve a CR after their first course (C1) of IC response assessment was repeated after a second course of IC (C2). CR after a second course of IC was defined as CR occurring within 42 days of commencing C2, or 75 days after trial entry, if the date of administration of C2 was not available. Patients failing to achieve a CR after two courses of IC were typically treated off study. Failure to respond to either the first or second course of IC was defined according to four definitions of refractoriness, namely: RES: resistant disease with failure to achieve CR after C1, PR: those deemed to have had a partial response to IC with failure to achieve CR after C1 and fewer than 15% blasts or a greater than 50% proportional reduction in blast percentage, REF1: those deemed to have had a minor or no response to IC with more than 15% blasts and a less than 50% proportional reduction in blast percentage after C1, and REF2: failure to achieve CR after two courses of IC (Figure 1). 371 patients were deemed to be refractory but their blast percentage was not available, and for the purpose of this analysis these patients were

Table 1. Factors predicting the presence of refractory disease in RES, REF1 and REF 2 cohorts.

RES Effect Karyotype Diagnostic WBC Secondary disease Older protocol Year of diagnosis Age Male sex Blast %

OR per unit

95% CI

Effect

3.01 2.68-3.39 <0.0001 Karyotype 1.003 1.001-1.004 <0.0001 Diagnostic WBC 1.8 1.53-2.12 <0.0001 Age 1.22 1.03-1.44 <0.0001 Year of diagnosis 0.98 0.97-0.99 <0.0001 Male sex 1.007 1.001-1.012 0.01 1.14 1.02-1.27 0.02 1.003 1.000-1.005 0.03

REF1 OR per 95% CI unit 4.11 1.004 1.012 0.98 1.21

REF2 OR per 95% CI unit

Effect

3.48-4.85 <0.0001 Karyotype 1.002-1.006 <0.0001 Diagnostic WBC 1.006-1.018 0.0008 Older protocol 0.97-1.00 0.01 Secondary disease 1.02-1.43 0.03 Year of diagnosis Male sex

3.75 1.6 1.58 1.72 0.97 1.35

3.02-4.65 1.34-1.92 1.26-1.96 1.28-2.30 0.95-0.99 1.09-1.67

P <0.0001 <0.0001 <0.0001 0.0003 0.0008 0.007

WBC: white blood cell count; CI: confidence interval; OR: odds ratio; IC: induction chemotherapy; C1: course 1; RES: resistant disease with failure to achieve a complete remission after C1; REF1: those deemed to have had a minor or no response to IC with more than 15% blasts and a less than 50% proportional reduction in blast percentage after C1; REF2: failure to achieve a complete remission after two courses of IC.

1352

haematologica | 2016; 101(11)

Primary refractory AML and transplantation

classified as PR. Separately, patients fulfilling either REF1 or REF2 criteria were combined and analyzed separately in a compendious cohort, REF1/2. Survival in patients fulfilling the different criteria of refractoriness was measured from the time at which refractoriness was ascertained according to the defined criteria. Survival percentages are measured using the method of Kaplan-Meier, or that of Mantel-Byar for the analyses of transplant versus not. In comparing allograft with no transplant, patients receiving other types of transplant were censored on the date of transplant. The outcomes of patients allografted after the year 2000 were analyzed according to age (greater or less than 50 years), to take into account the introduction of reduced intensity conditioning (RIC) regimens in older patients from this date. Models for risk of refractoriness or prognosis after being defined as refractory were built using Cox proportional hazards regression with forward selection; because molecular data were not uniformly available, this was performed in 2 stages â&#x20AC;&#x201C; first using clinical variables, and then adding the presence of mutations in the FLT3 or NPM1 genes to the model.

Results Characterization of induction failure cohorts 8907 patients were treated with intensive chemotherapy and form the subject of this study (Figure 2). 5480

Figure 1. Diagram illustrating the four definitions of refractory AML (acute myeloid leukemia) used in this study. C1: course 1; C2: course 2; CR: complete remission; RES: resistant disease with failure to achieve a complete remission after C1; REF1: those deemed to have had a minor or no response to IC with more than 15% blasts and a less than 50% proportional reduction in blast percentage after C1; REF2: failure to achieve a complete remission after two courses of IC; PR: those deemed to have had a partial response after C1 with fewer than 15% blasts or a greater than 50% proportional reduction in blast percentage.

Table 2. Prognostic factors for survival of the defined cohorts of patients studied.

Effect

RES HR 95% per CI unit

Karyotype

1.96

Age

1.02

WHO PS

1.17

Secondary disease

1.37

REF1

1.68

Diagnostic WBC 1.13 Older protocol 1.28

FLT3-ITD

1.6

NPM1

0.72

1.772.18 1.011.02 1.101.24 1.201.56

PR P

Effect

HR per unit

<0.001 Karyotype

2.01

<0.001

Age

1.02

<0.001

WHO PS

1.11

<0.001 Secondary disease

1.51

1.52- <0.001 Diagnostic 1.87 WBC 1.05- 0.0009 Older 1.23 protocol 1.10- 0.002 Male 1.50 sex Year diagnosed

1.13

1.29- <0.001 1.99 0.56- 0.01 0.93

NPM1

1.24 1.13 0.99

95% CI

Effect

REF1 HR 95% per CI unit

Model without molecular factors 1.78- <0.001 Karyotype 1.97 2.26 1.01- <0.001 Age 1.03 1.03 1.03- <0.001 WHO PS 1.19 1.19 1.31- <0.001 Year 0.99 1.74 diagnosed 1.031.24 1.041.48 1.011.27 0.98-1.00

Effect

1.67- <0.001 Karyotype 2.32 1.02- <0.001 Age 1.03 1.08- <0.001 WHO PS 1.32 0.97- 0.03 Diagnostic 1.00 WBC

REF2 HR 95% per CI unit 1.58 1.02 1.23 1.29

1.26 -1.99 1.011.03 1.081.40 1.071.56

<0.001 <0.001 <0.001 0.007

0.0006 0.01 0.05 0.05

Additional significant molecular factors FLT3-ITD 2.09 1.44- <0.001 FLT3-ITD 3.02 0.55 0.43-0.72 <0.001

1.68

1.082.262

0.02

Effect of year (P) if not on model above 0.06

0.4

WBC: white blood cell count; WHO PS: World Health Organization performance status; HR: hazards ratio; CI: confidence interval; IC: induction chemotherapy; C1: course 1; RES: resistant disease with failure to achieve a complete remission after C1; REF1: those deemed to have had a minor or no response to IC with more than 15% blasts and a less than 50% proportional reduction in blast percentage after C1; REF2: failure to achieve a complete remission after two courses of IC; PR; those deemed to have had a partial response after C1 with fewer than 15% blasts or a greater than 50% proportional reduction in blast percentage.

haematologica | 2016; 101(11)

1353

P. Ferguson et al.

Figure 2. Consort diagram of patient cohorts and treatment outcomes. CR: Complete remission; MAC: myeloablative conditioning; RIC: reduced intensity conditioning; RES: resistant disease with failure to achieve a complete remission after C1; REF1: those deemed to have had a minor or no response to IC with more than 15% blasts and a less than 50% proportional reduction in blast percentage after C1; PR: those deemed to have had a partial response after C1 with fewer than 15% blasts or a greater than 50% proportional reduction in blast percentage; REF2: failure to achieve a complete remission after two courses of IC; REF 1/2; all patients in groups REF1 and REF2.

patients achieved a CR following C1 and there were 879 induction deaths. A total of 2548 patients did not achieve remission with C1 (RES) of whom 802 fulfilled the criteria for refractoriness according to definition REF1. Of those not in CR post C1, 1059 patients achieved a CR after C2, with 100 patients dying during C2. 473 patients fulfilled the criteria for REF2. Of 802 patients fulfilling the criteria for REF1, 204 achieved remission after C2. The total number of patients who received an allogeneic SCT was 498. Of these, 351 underwent a myeloablative conditioning regimen whilst 147 received a RIC regimen. The demographics of the patients with refractory disease as defined by these criteria are outlined in the Online Supplementary Table S1.

Patient outcomes according to category of refractory disease

Factors predicting resistance to induction chemotherapy

Factors predicting long term survival in refractory disease

The factors determining the presence of refractory disease after IC, according to the studied definitions, are summarized in Table 1. Factors common to patients fulfilling REF1, PR and REF2 criteria included the year of diagnosis, presentation of white blood cell count (WBC) and karyotype.

The prognostic factors associated with survival for each of the defined populations with refractory disease are outlined in Table 2, and are broadly similar to those which predicted the presence of refractory disease. Karyotype, age and performance status were predictive of survival across all cohorts. When we included the mutational sta-

1354

The 5-year OS for patients in RES, REF1 , PR , REF2 and REF1/2 cohorts was 17%, 9%, 21%, 8% and 9%, respectively, compared with 40% for patients achieving a CR after one course of IC (P<0.0001). REF1 criteria identify a distinct sub-population of patients who fail to achieve CR after course 1, with significantly worse 5year OS compared with PR patients (P<0.0001) (Figure 3A). The 5-year OS for REF1 patients (9%) was equivalent to REF2 patients (8%). The 5-year OS for the minority (204) of REF1 patients who achieved CR with C2 was markedly reduced compared with patients achieving CR with their first course of IC (HR 1.39 (1.15-1.69) P=0.0008) (Figure 3B).

haematologica | 2016; 101(11)

Primary refractory AML and transplantation

tus for FLT3 ITD and NPM1, we found that NPM1 mutations predicted for survival in the REF1 cohort with FLT3 ITD being predictive for survival in the REF1, REF2 and REF1/2 cohorts.

Identification of treatment factors determining long term survival We next studied the impact of allogeneic transplantation on outcome in the defined groups of primary refractory disease using a Mantel-Byar approach. Analyses are presented as Forest plots stratified by age (Figure 4). Mantel-Byar analysis demonstrated that OS in allografted patients was significantly improved compared with nontransplant patients in REF1, REF2 and REF1/2 cohorts, with roughly equivalent estimates of the hazard ratio for the benefit of transplantation: REF1 (HR 0.58 (0.46-0.74), P=0.00001), REF2 (HR 0.55 (0.41-0.74), P=0.0001), and REF1/2 (HR 0.58 (0.49-0.69), P<0.00001). In the RES cohort patients over 50 years of age (HR 0.75 (0.62-0.91), haematologica | 2016; 101(11)

Figure 3. Survival from first being identified as refractory according to the definitions studied or entering complete remission (CR) after one course (C1) of induction chemotherapy a) CR post C1, RES, REF1, PR , REF 2; b) comparison of REF1 patients who achieve CR after course two of IC with REF 2 patients not identified in the REF1 cohort (REF2 not REF1), REF1 patients included in the REF 2 cohort (REF1 & REF2) and REF1 patients whose REF 2 status is unknown (REF1, REF2 U/K). RES: resistant disease with failure to achieve a complete remission after C1; REF1: those deemed to have had a minor or no response to IC with more than 15% blasts and a less than 50% proportional reduction in blast percentage after C1; PR: those deemed to have had a partial response after C1 with fewer than 15% blasts or a greater than 50% proportional reduction in blast percentage; REF2: failure to achieve a complete remission after two courses of IC; REF 1/2; all patients in groups REF1 and REF2.

P=0.003), allogeneic transplantation improved survival, although there was no difference in RES patients under the age of 50 (HR 1.06 (0.87-1.28), P=0.6; test for interaction P=0.01). When analysis was restricted to PR patients, there was no benefit for transplantation in either age group. In the minority of patients in REF1 who achieved a CR (204/802) with further courses of chemotherapy, there was a trend towards improved OS after allografting, but this did not achieve statistical significance (HR 0.77 (0.57-1.05) P=0.09). In patients with REF2 disease survival after allogeneic transplant was improved in patients who had achieved a CR with subsequent courses of chemotherapy (n=49), compared with those transplanted with active disease (n=37) (38% vs. 17%), although numbers were small. In analyses of the REF1/2 group censored at stem cell transplant, there was no evidence of improvement in survival over time (P=0.3), implying that improved survival is likely to be related to the use of transplantation. 1355

P. Ferguson et al.

Discussion This analysis, performed in a large and coherently treated population of adults, confirms previous reports that failure to achieve CR after one course of IC is associated with decreased survival. Furthermore, the presence of more than 15% blasts and a less than 50% reduction in blast percentage after the first course of IC identifies a population of patients whose survival is significantly worse than those who achieve a CR after course one, and

equivalent to patients who fail to achieve CR after two courses of IC. Reasoning that the definition of refractoriness is failure to achieve long-term survival if treated with chemotherapy alone, our data support a novel operational definition of PREF AML based either on a minimal response to the first course of IC, defined as a less than 50% proportional reduction in blasts and the presence of more than 15% blasts, or a failure to achieve CR after two courses of IC. In other words, the outcomes for patients fulfilling either REF1 or REF2 criteria, if treated with fur-

Figure 4. Mantel-Byar analysis of impact of allogeneic transplant on survival according to different definitions of PREF AML (primary refractory acute myeloid leukemia). SCT: stem cell transplant; O.R.: odds ratio; CI: confidence intervals; Var.: variance; 2P: 2-sided P value; NS: non-significant; RES: resistant disease with failure to achieve a complete remission after C1; PR: those deemed to have had a partial response after C1 with fewer than 15% blasts or a greater than 50% proportional reduction in blast percentage; REF1: those deemed to have had a minor or no response to IC with more than 15% blasts and a less than 50% proportional reduction in blast percentage after C1; REF2: failure to achieve a complete remission after two courses of IC; REF 1/2; all patients in groups REF1 and REF2.

1356

haematologica | 2016; 101(11)

Primary refractory AML and transplantation

ther intensive chemotherapy, is very poor, consistent with chemorefractoriness. Importantly, our data do not support the continued use of the RES or PR criteria to define PREF AML. Our analysis has identified a number of factors including karyotype, age, sex and diagnostic white cell count as predicting refractoriness, consistent with previous studies of high-risk AML. Interestingly, the use of the MRC risk score designed for risk stratification of younger patients with AML in conjunction with REF1 criteria identifies more than 90% of patients within the REF1/2 group.14 Whilst it has been reported that allogeneic SCT may represent an important treatment modality in patients with PREF AML, the absence of a consensus concerning the definition of refractory disease and the selection bias inherent in registry studies has led to skepticism and therapeutic uncertainty. By applying different definitions of primary refractoriness it has been possible, for the first time, to examine the impact of allogeneic transplant in four different clinical settings. These data demonstrate that allografting confers a marked survival advantage in patients fulfilling REF1 and REF2 criteria. There are a number of limitations in the interpretation of our data. Firstly, it is not possible to quantify the degree to which selection bias contributed to the observed improved outcome in the population of patients who proceeded to transplant. Equally, the impact of an allogeneic transplant may have been underestimated because patients often proceeded to transplant after multiple courses of IC, which has previously been shown to compromise the outcome of patients allografted for PREF AML.11,12,23 It is perhaps of no surprise that the outcome of patients fulfilling REF2 criteria who subsequently achieved a CR prior to transplant appeared to be improved compared with those who never achieved CR, but nonetheless our data demonstrate that allografting represents the only curative option for a proportion of REF2 patients; although the degree to which this benefit is restricted to those who achieve a CR with further chemotherapy will require further study. Since time to transplant is an important predictor of outcome in refractory AML, the use of REF1 criteria to identify patients with refractory disease represents an opportunity to improve

References 1. Burnett AK, Russell NH, Hills RK, et al. Addition of gemtuzumab ozogamicin to induction chemotherapy improves survival in older patients with acute myeloid leukemia. J Clin Oncol. 2012;30(32):39243931. 2. Oyekunle AA, Kroger N, Zabelina T, et al. Allogeneic stem-cell transplantation in patients with refractory acute leukemia: a long-term follow-up. Bone Marrow Transplant. 2006;37(1):45-50. 3. Forman SJ, Schmidt GM, Nademanee AP, et al. Allogeneic bone marrow transplantation as therapy for primary induction failure for patients with acute leukemia. J Clin Oncol. 1991;9(9):1570-1574. 4. Mehta J, Powles R, Horton C, et al. Bone marrow transplantation for primary refractory acute leukaemia. Bone Marrow Transplant. 1994;14(3):415-418. 5. Othus M, Appelbaum FR, Petersdorf SH, et al. Fate of patients with newly diagnosed

haematologica | 2016; 101(11)

transplant outcomes by shortening the time from diagnosis to transplant. More prosaically, these data also underline the importance of tissue typing newly diagnosed adult patients and the commencement of an urgent donor search as a cornerstone of the management of adult AML. An important potential determinant of chemorefractoriness in AML is the intensity of induction chemotherapy. In this study, it was observed that older patients, for whom lower intensity therapy was felt more appropriate, were at a higher risk of having refractory disease after two courses of IC. This underlines the importance of the development of either more effective, but well tolerated, novel chemotherapeutic agents, or improved delivery strategies such as the use of liposomal preparations.24 This is particularly pertinent given the higher incidence of PREF AML in older patients.25 A weakness of this study is that we have analyzed the outcome in patients treated with standard doses of induction chemotherapy only, and it will be important to repeat this analysis in patients receiving high dose cytosine arabinoside regimens. Our data does, however, support the further exploration of sequential conditioning regimens which incorporate a cycle of intensive chemotherapy as an integral component of the preparative regimen, such as those developed by Kolb and Schmid.12,26 In this context, it is of interest to note the particularly encouraging results reported by these authors using the sequential FLAMSA regimen in patients with PREF AML. Taken together our data support a clarification of the criteria used to define refractoriness to IC in adult AML. Furthermore, we demonstrate the ability of allogeneic transplantation to improve long-term survival in selected patients with PREF AML. Adoption of the proposed criteria will assist in the early identification of patients with PREF AML who have the potential to benefit from allogeneic transplantation. Such an approach has the potential to reduce transplant toxicity and prevent potential selection of chemotherapy resistant sub-clones.27 Funding This study was supported (in part) by research funding from Bloodwise UK, the Medical Research Council and Cancer Research UK to CC, RH, NR.

acute myeloid leukemia who fail primary induction therapy. Biol Blood Marrow Transplant. 2015;21(3):559-564. Jabbour E, Daver N, Champlin R, et al. Allogeneic stem cell transplantation as initial salvage for patients with acute myeloid leukemia refractory to high-dose cytarabine-based induction chemotherapy. Am J Hematol. 2014;89(4):395-398. Fung HC, Stein A, Slovak M, et al. A longterm follow-up report on allogeneic stem cell transplantation for patients with primary refractory acute myelogenous leukemia: impact of cytogenetic characteristics on transplantation outcome. Biol Blood Marrow Transplant. 2003;9(12):766771. Ravandi F. Primary refractory acute myeloid leukaemia - in search of better definitions and therapies. Br J Haematol. 2011;155(4):413-419. Cheson BD, Bennett JM, Kopecky KJ, et al. Revised recommendations of the International Working Group for

10.

11.

12.

13.

Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia. J Clin Oncol. 2003;21(24):4642-4649. Dohner H, Estey EH, Amadori S, et al. Diagnosis and management of acute myeloid leukemia in adults: recommendations from an international expert panel, on behalf of the European LeukemiaNet. Blood. 2010;115(3):453-474. Craddock C, Labopin M, Pillai S, et al. Factors predicting outcome after unrelated donor stem cell transplantation in primary refractory acute myeloid leukaemia. Leukemia. 2011;25(5):808-813. Schmid C, Schleuning M, Schwerdtfeger R, et al. Long-term survival in refractory acute myeloid leukemia after sequential treatment with chemotherapy and reducedintensity conditioning for allogeneic stem cell transplantation. Blood. 2006;108(3):1092-1099. Rowe JM, Kim HT, Cassileth PA, et al.

1357

P. Ferguson et al.

14.

15.

16.

17.

1358

Adult patients with acute myeloid leukemia who achieve complete remission after 1 or 2 cycles of induction have a similar prognosis: a report on 1980 patients registered to 6 studies conducted by the Eastern Cooperative Oncology Group. Cancer. 2010;116(21):5012-5021. Wheatley K, Burnett AK, Goldstone AH, et al. A simple, robust, validated and highly predictive index for the determination of risk-directed therapy in acute myeloid leukaemia derived from the MRC AML 10 trial. United Kingdom Medical Research Council's Adult and Childhood Leukaemia Working Parties. Br J Haematol. 1999;107(1):69-79. Schlenk RF, Benner A, Hartmann F, et al. Risk-adapted postremission therapy in acute myeloid leukemia: results of the German multicenter AML HD93 treatment trial. Leukemia. 2003;17(8):1521-1528. Revesz D, Chelghoum Y, Le QH, Elhamri M, Michallet M, Thomas X. Salvage by timed sequential chemotherapy in primary resistant acute myeloid leukemia: analysis of prognostic factors. Ann Hematol. 2003;82(11):684-690. Ravandi F, Cortes J, Faderl S, et al. Characteristics and outcome of patients with acute myeloid leukemia refractory to 1 cycle of high-dose cytarabine-based induction chemotherapy. Blood. 2010;116(26):5818-5823.

18. Hann IM, Stevens RF, Goldstone AH, et al. Randomized comparison of DAT versus ADE as induction chemotherapy in children and younger adults with acute myeloid leukemia. Results of the Medical Research Council's 10th AML trial (MRC AML10). Adult and Childhood Leukaemia Working Parties of the Medical Research Council. Blood. 1997;89(7):2311-2318. 19. Goldstone AH, Burnett AK, Wheatley K, Smith AG, Hutchinson RM, Clark RE. Attempts to improve treatment outcomes in acute myeloid leukemia (AML) in older patients: the results of the United Kingdom Medical Research Council AML11 trial. Blood. 2001;98(5):1302-1311. 20. Burnett AK, Hills RK, Milligan DW, et al. Attempts to optimize induction and consolidation treatment in acute myeloid leukemia: results of the MRC AML12 trial. J Clin Oncol. 2010;28(4):586-595. 21. Burnett AK, Milligan D, Goldstone A, et al. The impact of dose escalation and resistance modulation in older patients with acute myeloid leukaemia and high risk myelodysplastic syndrome: the results of the LRF AML14 trial. Br J Haematol. 2009;145(3):318-332. 22. Grimwade D, Hills RK, Moorman AV, et al. Refinement of cytogenetic classification in acute myeloid leukemia: determination of prognostic significance of rare recurring chromosomal abnormalities among 5876

23.

24.

25.

26.

27.

younger adult patients treated in the United Kingdom Medical Research Council trials. Blood. 2010;116(3):354-365. Biggs JC, Horowitz MM, Gale RP, et al. Bone marrow transplants may cure patients with acute leukemia never achieving remission with chemotherapy. Blood. 1992;80(4):1090-1093. Cortes JE, Goldberg SL, Feldman EJ, et al. Phase II, multicenter, randomized trial of CPX-351 (cytarabine:daunorubicin) liposome injection versus intensive salvage therapy in adults with first relapse AML. Cancer. 2015;121(2):234-242. Burnett AK, Hills RK, Milligan D, et al. Identification of patients with acute myeloblastic leukemia who benefit from the addition of gemtuzumab ozogamicin: results of the MRC AML15 trial. J Clin Oncol. 2011;29(4):369-377. Pfeiffer T, Schleuning M, Mayer J, et al. Influence of molecular subgroups on outcome of acute myeloid leukemia with normal karyotype in 141 patients undergoing salvage allogeneic stem cell transplantation in primary induction failure or beyond first relapse. Haematologica. 2013;98(4):518525. Ding L, Ley TJ, Larson DE, et al. Clonal evolution in relapsed acute myeloid leukaemia revealed by whole-genome sequencing. Nature. 2012;481(7382):506510.

haematologica | 2016; 101(11)

ARTICLE

Acute Myeloid Leukemia

Effect of age and body weight on toxicity and survival in pediatric acute myeloid leukemia: results from NOPHO-AML 2004

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Ditte J. A. Løhmann,1 Jonas Abrahamsson,2 Shau-Yin Ha,3 Ólafur G. Jónsson,4 Minna Koskenvuo,5 Birgitte Lausen,6 Josefine Palle,7 Bernward Zeller, 8 and Henrik Hasle 1

Department of Pediatrics, Aarhus University Hospital Skejby, Denmark; 2Institution for Clinical Sciences, Department of Pediatrics, Queen Silvia Children’s Hospital, Gothenburg, Sweden; 3Department of Pediatrics, Queen Mary Hospital and Hong Kong Pediatric Hematology & Oncology Study Group (HKPHOSG), Hong Kong, China; 4 Department of Pediatrics, Landspitalinn, Reykjavik, Iceland; 5Division of HematologyOncology and Stem Cell Transplantation, Children’s Hospital and Helsinki University Central Hospital, Finland; 6Department of Pediatrics and Adolescent Medicine, Rigshospitalet, University of Copenhagen, Denmark; 7Department of Woman’s and Children’s Health, Uppsala University, Sweden; and 8Department of Pediatric Medicine, Oslo University Hospital, Norway

Haematologica 2016 Volume 101(11):1359-1367

ABSTRACT

reatment for pediatric acute myeloid leukemia is very toxic and the association between outcome and age and Body Mass Index is unclear. We investigated effect of age and Body Mass Index on toxicity and survival in pediatric acute myeloid leukemia. We studied all patients who completed first induction course of NOPHO-AML 2004 (n=318). Toxicity following induction and consolidation courses (n=6) was analyzed. The probabilities of toxicity and death were determined using time-to-event analyses with Cox multivariate proportional hazard regression for comparative analyses. Age 10-17 years was associated with sepsis with hypotension [hazard ratio 2.3 (95% confidence interval 1.1-4.6)]. Being overweight (>1 standard deviation) was associated with requiring supplemental oxygen [1.9 (1.0-3.5)]. The 5-year event-free and overall survival were 47% and 71%. Children aged 10-17 years showed a trend for inferior 5-year overall survival compared to children aged 29 (64% vs. 76%; P=0.07). Infants showed a trend for superior 5-year event-free survival (66% vs. 43%; P=0.06). Overweight children aged 10-17 years showed a trend for superior survival [5-year event-free survival 59% vs. 40% (P=0.09) and 5-year overall survival 78% vs. 56% (P=0.06)] compared to healthy weight children aged 10-17 years. In conclusion, children aged 10-17 years and overweight children had a higher risk of grade 3-4 toxicity. Children aged 10-17 years showed inferior survival, but, unexpectedly, in this age group overweight children tended to have increased survival. This suggests different pharmacokinetics of chemotherapeutic drugs in adolescents and warrants further studies.

Correspondence: dadolfsen@gmail.com

Received: March 17, 2016. Accepted: July 25, 2016. Pre-published: July 28, 2016. doi:10.3324/haematol.2016.146175

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1359

Introduction Despite good overall treatment results for childhood acute myeloid leukemia (AML), 30%-35% of patients die from resistant disease, relapse, or treatment-related toxicity.1-3 Due to the high-intensity treatment of pediatric AML, most patients develop severe toxicity, and about 10% die from treatment-related toxicity.2,4-7 The most common toxicity is infection, often leading to life-threatening sepsis. Almost all patients with AML experience an infection during the first induction course.8,9 Older age and being overweight at diagnosis have been associated with inferior survival in children with AML10-15 and older age has been associated with treatmentrelated mortality (TRM).14,16-18 Infants and older children treated for AML have more severe infections18,19 and being overweight has been associated with more severe haematologica | 2016; 101(11)

©2016 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. A permission in writing by the publisher is required for any other use.

1359

D.J.A. LĂ¸hmann et al.

abdominal pain, hypotension, pulmonary toxicity and coagulopathy.13 Despite pediatric AML treatment causing significant morbidity and mortality, no previous studies have thoroughly reviewed the numerous treatment-related grade 3-4 toxicities or investigated if age and body weight at diagnosis is associated with risk of toxicity. We aimed to describe all grade 3-4 toxicities in the Nordic Society of Pediatric Hematology and Oncology (NOPHO) AML 2004 protocol and investigate associations between toxicity, survival, age and body weight at diagnosis.

Methods

The randomization to gemtuzumab ozogamicin (GO) or no further therapy after completion of consolidation has been reported previously.21 Patients received prophylactic sulfamethoxazole/trimethoprim 2-3 days per week. Prophylactic fluconazole was recommended until one month after the last course. No other prophylactic antibacterial or antiviral drugs were recommended. Use of prophylactic granulocyte colony-stimulating factor was not recommended. Patients were discharged to their own homes when clinically stable regardless of neutrophil count.

Toxicity registration Fourteen grade 3-4 toxicities were collected and graded after each treatment block by the treating physician or the local data manager. The toxicities were defined according to the World Health Organization (WHO).22

Patients The NOPHO-AML 2004 protocol (clinicaltrials.gov identifier: 00476541) included all children below 15 years of age (for some centers all children below 18 years) diagnosed with AML in the five Nordic countries (Denmark, Finland, Iceland, Norway, and Sweden) from 2004-2013 and in Hong Kong from 2007-2013. Children with Down syndrome, acute promyelocytic leukemia, isolated granulocytic sarcoma, or secondary AML were excluded. The number of newly diagnosed pediatric patients with AML during the study period determined the sample size of this study. The national ethics committees in the six participating countries approved the protocol. Children who did not complete first induction course were excluded from these analyses.

Treatment plan Figure 1 illustrates the treatment of NOPHO-AML 2004 including number of patients receiving each course. Protocol details including drug doses have been published previously.20 Children below one year of age or with body weight below 10 kg received doses calculated according to body weight instead of body surface with 1 m2 = 30 kg. No dose adjustment was recommended in overweight patients.

Definitions and statistics Patients were divided into age and Body Mass Index (BMI) zscore groups. Children below two years of age were excluded from the weight analyses. Standard deviations (SD) of BMI for age and sex were calculated according to WHO criteria.23 Being underweight was defined as BMI below -2 SD, overweight as BMI above +1 SD [pooling overweight (+1 to +2 SD, n=40) and obesity (>+2 SD, n=16)], and healthy weight as BMI between -2 and +1 SD.23 Cumulative incidence for first episode of a grade 3-4 toxicity or TRM during therapy was calculated; and for three toxicities (general condition, infection and hypoxia) a separate cumulative incidence of grade 4 toxicity was calculated. The seven most common grade 3-4 toxicities were used for the comparative analyses. Missing data (<1%) were coded as no grade 3-4 toxicity. Interaction of age and body weight group as risk factors for toxicity was calculated by relative excess risk caused by interaction (RERI), using the algorithm of Andersson.24 Treatment-related mortality, event-free survival (EFS) and overall survival (OS) are defined in Online Supplementary Table S1. Differences in toxicity and survival in age and body weight

Figure 1. Flow chart of the 318 patients who completed the first course of the NOPHO-AML 2004 protocol. AIET: cytarabine, 6-thioguanine, etoposide, and idarubicin; AM: cytarabine and mitoxantrone; HA1M: high-dose cytarabine and mitoxantrone; HA2E: high-dose cytarabine and etoposide; HA3: high-dose cytarabine; FLA: fludarabine 30 mg/m2 day 1-5 and cytarabine 2000 mg mg/m2 day 1-5; G: G-CSF 200 mg/m2 day 0-5; Dx: liposomal daunorubicin 60 mg/m2 day 2, 4 and 6. *Protocol deviations: one patient received HA2E-HA1M- HA2E after induction, one patient skipped HA1M and had an extra HA3 after the second HA2E, one patient skipped HA1M, and one patient had etoposide, idarubicin and 6-thioguanine as the fourth consolidation.

1360

haematologica | 2016; 101(11)

Effect of age and BMI on toxicity and survival in pediatric AML

Table 1. Baseline characteristics according to age group (n=318) and weight group (n=239).

0 yr. N (%) Patients Sex Age group Weight group Ethnicity WBC (109/L) FAB subtype

FLT3 Cytogenetics

Achieved CR1

Male Female 2-9 years 10-17 years Underweight (<-2 SD) Healthy (-2-+1 SD) Overweight (>+1 SD) White Asian Other* 0-9.9 10-99.9 â&#x2030;Ľ100 M0 M1 M2 M4 M5 M6 M7 Unclassified/unknown ITD Wild-type/other Unknown t(8;21) inv(16) MLL rearrangements None of the above No After one induction After two inductions After three inductions

HSCT in CR1

34 (11) 9 (26) 25 (74) 27 (79) 5 (15) 2 (6) 11 (32) 16 (47) 7 (21) 4 (12) 3 (9) 5 (15) 7 (21) 10 (29) 0 (0) 4 (12) 1 (3) 0 (0) 23 (68) 11 (32) 0 (0) 2 (6) 15 (44) 17 (50) 1 (3) 27 (79) 6 (18) 0 (0) 4 (12)

Age group (n=318) 1 yr. 2-9 yr. N (%) N (%) 44 (14) 23 (52) 21 (48) 35 (80) 8 (18) 1 (2) 11 (25) 28 (64) 5 (11) 3 (7) 1 (2) 2 (5) 8 (18) 13 (30) 3 (7) 11 (25) 3 (7) 0 (0) 28 (64) 16 (36) 0 (0) 2 (5) 16 (36) 26 (59) 3 (7) 35 (80) 6 (14) 0 (0) 6 (14)

126 (40) 68 (54) 58 (46) 4 (3) 103 (82) 19 (15) 87 (69) 22 (17) 17 (13) 48 (38) 58 (46) 20 (16) 9 (7) 15 (12) 38 (30) 20 (16) 21 (17) 2 (2) 11 (9) 10 (8) 9 (7) 87 (69) 30 (24) 27 (21) 12 (10) 27 (21) 60 (48) 5 (4) 89 (71) 29 (23) 3 (2) 21 (17)

10-17 yr. N (%)

Weight group (n=239) <-2 SD -2-+1 SD >+1 SD N (%) N (%) N (%)

114 (36) 79 (69) 35 (31) 8 (7) 68 (60) 37 (33) 75 (66) 29 (25) 10 (9) 40 (35) 57 (50) 17 (15) 5 (4) 22 (19) 33 (29) 19 (17) 25 (22) 0 (0) 0 (0) 10 (9) 14 (12) 59 (52) 41 (36) 20 (18) 12 (11) 16 (14) 66 (58) 5 (4) 70 (61) 32 (28) 7 (6) 24 (21)

12 (5) 10 (83) 2 (17) 4 (33) 8 (67) 5 (42) 6 (50) 1 (8) 4 (33) 6 (50) 2 (17) 0 (0) 1 (8) 3 (25) 1 (8) 3 (25) 1 (8) 0 (0) 3 (25) 1 (8) 6 (50) 5 (42) 0 (0) 0 (0) 4 (33) 8 (67) 2 (17) 4 (33) 5 (42) 1 (8) 2 (17)

171 (72) 92 (54) 79 (46) 103 (60) 68 (40) 118 (69) 33 (19) 20 (12) 61 (36) 82 (48) 28 (16) 11 (6) 27 (16) 49 (29) 29 (17) 31 (18) 1 (1) 9 (5) 14 (8) 19 (11) 102 (60) 50 (29) 32 (19) 15 (9) 27 (16) 97 (57) 7 (4) 115 (67) 42 (25) 7 (4) 33 (19)

56 (23) 44 (79) 12 (21) 19 (34) 37 (66) 38 (68) 12 (21) 6 (11) 23 (41) 26 (46) 7 (13) 3 (5) 9 (16) 19 (34) 8 (14) 12 (21) 0 (0) 2 (4) 3 (5) 3 (5) 37 (66) 16 (29) 15 (27) 9 (16) 12 (21) 20 (36) 1 (2) 40 (71) 13 (23) 2 (4) 10 (18)

*The other category consists of children with African, Arab and mixed ancestry and 2 children where only the ethnicity of one of the parents was known. yr.: year; SD: standard deviation; N: number; WBC: white blood cell count; FAB: French-American-British classification; CR1: first complete remission; HSCT: hematopoietic stem cell transplantation; WBC: white blood count.

groups were analyzed using Cox regression and adjusted for potential confounders. OS and EFS were estimated using the Kaplan-Meier method. Differences in survival were compared using log rank tests. All tests of significance were two-sided. Statistical significance was defined as P<0.05.

Results Patients' characteristics In total, 323 patients from the six participating countries were treated according to the NOPHO-AML 2004 protocol. Five patients who died within the first seven days after diagnosis due to aggressive/progressive AML were excluded, leaving 318 for analysis. The five excluded patients were 0, 2, 2, 2, and 7 years old; 3 were underweight, one of healthy weight and in one BMI was not calculated due to the patient being under 2 years old. The median age at diagnosis was 6.4 years (range 0.117.9) and 37 patients were aged between 15-17 years. There were more females in the infant group and more males in the older and overweight groups. Older children were more overweight, had more FAB type M1 and M2, needed more induction courses to achieve remission, had haematologica | 2016; 101(11)

more FLT3-ITD mutations, inv(16), or t(8;21) and more were of Asian ancestry. Younger children had more FAB type M5 and M7 and MLL arrangements. Overweight patients were older, had less frequently FLT3-ITD mutations, and more often inv(16) and t(8;21) (Table 1). The higher rate of inv(16) and t(8;21) in the overweight group was not due to older age; the frequencies of inv(16) or t(8;21) were higher in overweight children age 2-9 years (10 of 19, 53%) than in overweight children age 10-17 years (14 of 37, 38%). Very few patients were underweight (n=12) and this group was thus excluded from the comparative toxicity and survival analyses. The two induction and four consolidation courses were completed by 237 (75%) of the 318 patients; Figure 1 shows the therapy courses completed. The median duration of therapy (time from start of AIET to last day with ANC below 0.5x109/L after HA2E2) for those who completed six courses was 182 days (range 128-281 days).

Toxicity The 318 patients completed 1670 courses and 14 different toxicities were requested for registration after each course. Toxicities were registered 23,206 times (99.3% complete). 1361

D.J.A. LĂ¸hmann et al. Table 2. The 14 toxicities registered after each treatment course, the definitions of grade 3 and 4 toxicities from the toxicity registration form, number of first grade 3-4 toxicities and the cumulative incidence of grade 3-4 toxicity.

Toxicity Any grade 3 or 4 toxicity General condition Infection Hypoxia Abdominal pain/constipation Abdominal symptom Renal Allergic reaction Hyperglycemia Bilirubin Thrombosis Hemorrhage Cardiac function Central neurotoxicity Peripheral neurotoxicity/ myopathy

Definition of grade 3 and 4

N. of first events

Cumulative incidence % (95%CI)

283 204* 58 246* 41 70* 33 86*

90 (87-94) 65 (59-70)* 19 (14-23) 79 (74-84)* 13 (10-17) 23 (18-27)* 11 (7-14) 28 (23-33)*

8 8*

2.5 (0.8-4.3) 2.6 (0.8-4.4)*

1.6 (0.2-3.0)*

4 13*

1.3 (0.03-2.5) 4.2 (1.9-6.4)*

2.2 (0.6-3.8)*

9 11*

2.9 (1.0-4.8) 3.8 (1.6-6.1)*

15*

4.9 (2.5-7.3)*

2.0 (0.4-3.5)*

3: Bedridden, in need of care 4: Intensive care, very sick 3: Pathogen identified, intravenous antibiotics given 4: Septic shock/hypotension 3: Decreased O2 sat at rest, requiring supplemental oxygen 4: Decreased O2 sat requiring CPAP or assisted ventilation 3: Severe pain or analgesics severely interfering with activities of daily life 4: Paralytic ileus or intestinal obstruction 3: No registration 4: Leading to laparotomy 3: Creatinine > 3.0-6.0 x UNL 4: Creatinine > 6.0 x UNL 3: Bronchospasm, requiring parenteral medication 4: Anaphylaxis 3: No registration 4: Need of insulin 3: 3-10 x UNL 4: >10 x UNL 3: Requiring systemic anticoagulation 4: Severe thrombosis causing organ dysfunction 3: No registration 4: Catastrophic bleeding requiring non-elective intervention 3: Mild congestive heart failure, therapeutically compensated 4: Severe/refractory congestive heart failure 3: Somnolence > 50%/day or severe disorientation or hallucinations 4: Coma or seizures 3: Unbearable paresthesia or pronounced deficit in motor functions 4: Paralysis

*Combination of grade 3 and 4 toxicity; UNL: upper normal limit. CPAP: continuous positive airway pressure.

The first induction (AIET) and the first consolidation (HA1M) were most toxic with longer median time to ANC recovery, more infections requiring intravenous administration of antibiotics for AIET (including fever with both known and unknown pathogen), and more grade 3-4 toxicity (Online Supplementary Table S2). All 318 children received antibiotics due to a febrile episode at least once during treatment. The treatment caused a high degree of toxicity (Table 2). Almost all patients (90%) had at least one grade 3 or 4 toxicity with infection with a verified pathogen (infection grade 3 or 4) being the most common. Moreover, severe abdominal pain, being bedridden/needing care (general condition grade 3), admission to the intensive care unit (general condition grade 4), need of supplemental oxygen, assisted ventilation, and sepsis with hypotension were frequent (cumulative incidence above 10%). Abdominal symptoms leading to laparotomy, creatinine above three times the upper normal limit (UNL), allergic reactions, hyperglycemia, bilirubin above three times the UNL, thrombosis, hemorrhage, cardiac failure, central and peripheral neurotoxicity were rare (cumulative incidence below 10%) and were not analyzed further [except for being included in any grade 3 or 4 toxicity (Table 3)]. The cumulative incidence of TRM was 4.6% (95%CI: 2.2%-7.0%). Nine of the 14 TRM cases occurred during induction (AIET, AM or FLAG). Nine patients died from or with an infection (seven bacterial, one viral and one 1362

fungal). Further details on TRM are provided in Online Supplementary Table S3. The risk of grade 3-4 toxicity correlated with age and body weight groups (Table 3). Confounders were defined as factors, which could influence both independent variables (age and weight at diagnosis) and dependent variables (toxicity and survival) and were selected a priori. Consequently, age analyses were adjusted for ethnicity and sex. The age analyses were not adjusted for body weight group, since this does not influence age at diagnosis. The body weight analyses were adjusted for age group, sex and ethnicity. Children aged 10-17 years had an increased risk of sepsis with hypotension compared to children aged 2-9 years. Children aged 10-17 years also showed a trend for increased risk of a number of toxicities, e.g. admission to the intensive care unit and severe abdominal pain. Infants did not seem to experience increased toxicity during chemotherapy. In the 30 infants and 99 children aged 2-9 years who completed the four high-dose cytarabine consolidation courses, 16 infants compared to 59 children aged 2-9 years (53 vs. 59%) had at least one infection with a verified pathogen; no infant was admitted to the intensive care unit or needed assisted ventilation compared to 9 (9%) and 3 (3%) of the children aged 2-9 years; one infant (3%) had liver toxicity compared to no children aged 2-9 years and no infant had central neurotoxicity compared to 5 children aged 29 years (5%). Overweight children had higher risk of being bedridden haematologica | 2016; 101(11)

Effect of age and BMI on toxicity and survival in pediatric AML

and requiring supplemental oxygen. This group also showed a trend for higher risk of several other toxicities, e.g. sepsis with hypotension and severe abdominal pain. Of the 56 overweight patients (298 courses), only 5 (9%) had dose reductions of one or two courses and none had dose reductions of more than two courses. Because we did not have information on change in body weight during treatment, and therefore do not know if patients who were overweight at diagnosis remained overweight during treatment, we performed the toxicity analyses again including toxicity after the first course (AIET) only (Online Supplementary Table S4). After the first course, overweight children had an increased risk of requiring supplemental oxygen and severe abdominal pain. The subgroup RERI analysis for interaction of body weight and age rendered the groups small and confidence intervals were thus wide (Online Supplementary Table S5). No conclusions can be made on the basis of these calculations, but being both older and overweight seemed to markedly increase the risk of sepsis.

Discussion The aim of this study of 318 children treated according to the NOPHO-AML 2004 protocol during a 10-year period was to describe grade 3-4 toxicity and investigate asso-

Outcome The median follow up time for patients alive at last follow up was 5.0 years (range 0.8-10.7 years). The 5-year EFS and OS for the 318 patients were 47% (95%CI: 42%53%) and 71% (95%CI: 65%-76%), respectively. The main cause of decreased EFS was relapse; 125 (78%) of 161 events were due to relapse and 16 (10%) were due to TRM in first complete remission (2 after HSCT). Outcome was associated with age (Figure 2A and B). Infants (<1 year) showed a trend for superior EFS compared with children aged 2-9 years (5-year EFS 66% vs. 43%, sex and ethnicity adjusted HR 0.5, 95%CI: 0.3-1.0). OS differed between children aged 2-9 and aged 10-17 years with the older children showing a trend for inferior OS (5-year OS 64% vs. 76%, adjusted HR 1.5, 95%CI: 0.9-2.4). Children aged 1 year also had a trend for inferior OS (5-year OS 65% vs. 76%, adjusted HR 1.7, 95%CI: 0.9-3.2). For children aged 2-9 years, being overweight (BMI> +1 SD for age) at diagnosis did not appear to influence prognosis (Figure 3A and C). In children aged 10-17 years, there was a trend for improved outcome in overweight patients, EFS (5-year EFS 59% vs. 40%, adjusted HR 0.6, 95%CI: 0.3-1.1) and OS (5-year OS 78% vs. 56%, adjusted HR 0.5 95%CI: 0.2-1.0) (Figure 3B and D). To test if this difference was influenced by the high frequency of t(8;21) in overweight patients, the weight-group analysis of children aged 10-17 years was stratified by t(8;21). A similar trend of superior survival in overweight patients was found in patients with t(8;21) (n=20, 5-year EFS 56% vs. 45% and 5-year OS 67% vs. 48%) and without t(8;21) (n=94, 5-year EFS 60% vs. 39% and 5-year OS 82% vs. 56%). As a sensitivity analysis, we performed the survival analysis using the 2000 Center for Disease Control (CDC) Growth charts25 to investigate if differences in results between our and previous studies12,13 could be explained by difference in growth standard. Sixteen children were reclassified (11 overweight as healthy weight and 5 healthy weight as underweight) using this approach, but the tendency for overweight children aged 10-17 years to have a better outcome remained (5-year EFS 57% vs. 43%, adjusted HR 0.6 95%CI: 0.3-1.2 and 5-year OS 77% vs. 58%, adjusted HR 0.5 95%CI: 0.2-1.2). haematologica | 2016; 101(11)

Figure 2. Event-free (A) and overall survival (B) according to age groups (n=318).

1363

D.J.A. LĂ¸hmann et al.

ciations between age and body weight at diagnosis and severe toxicity and survival. The cumulative incidence of grade 3-4 toxicity was high (90%). Children aged 10-17 years and overweight children were at higher risk of several grade 3-4 toxicities. To our knowledge, no previous studies have reported such a thorough review of toxicities and associations with age and body weight at diagnosis, though many studies suggest similar associations.12,13,18,19,26,27

We found a trend for age 10-17 years being associated with poorer survival. In contrast to previous studies,12,13 we found a trend for being overweight being associated with improved survival in children aged 10-17 years. Our patient cohort is complete and unselected including all newly diagnosed pediatric AML cases from the Nordic countries between 2004 and 2013, and Hong Kong between 2007 and 2013. All patients were treated according to the same protocol during a relatively short time

Table 3. Analysis of differences in toxicities according to age and weight groups.

Toxicity Any grade 3 or 4 toxicity

Age/body weight group

0 1 2-9 10-17 Healthy weight Overweight Bedridden and in need of care 0 (general condition grade 3 and 4) 1 2-9 10-17 Healthy weight Overweight Intensive care unit (general condition grade 4) 0 1 2-9 10-17 Healthy weight Overweight Requiring supplemental oxygen (hypoxia grade 3 and 4) 0 1 2-9 10-17 Healthy weight Overweight Continuous positive airway pressure 0 or assisted ventilation (hypoxia grade 4) 1 2-9 10-17 Healthy weight Overweight Infection with an identified pathogen 0 (infection grade 3 and 4) 1 2-9 10-17 Healthy weight Overweight Sepsis/hypotension (infection grade 4) 0 1 2-9 10-17 Healthy weight Overweight Abdominal pain severely interfering 0 with activities of daily life 1 (abdominal pain grade 3 and 4) 2-9 10-17 Healthy weight Overweight

N. of first events

Crude hazard ratio (95%CI)

Adjusted* hazard ratio (95%CI)

28 41 114 100 152 51 20 31 75 78 99 45 8 6 18 26 26 15 9 4 26 31 34 19 4 4 9 16 17 6 26 36 98 86 130 45 3 3 12 23 19 14 9 9 30 38 44 23

0.9 (0.6-1.3) 1.1 (0.8-1.6) 1 0.9 (0.7-1.2) 1 1.1 (0.8-1.6) 1.0 (0.6-1.6) 1.3 (0.8-1.9) 1 1.2 (0.9-1.6) 1 1.6 (1.1-2.2) 1.8 (0.8-4.0) 1.0 (0.4-2.5) 1 1.7 (0.9-3.1) 1 1.9 (1.0-3.5) 1.3 (0.6-2.8) 0.4 (0.2-1.2) 1 1.4 (0.8-2.3) 1 1.8 (1.0-3.2) 1.7 (0.5-5.5) 1.3 (0.4-4.3) 1 2.0 (0.9-4.6) 1 1.1 (0.4-2.8) 1.1 (0.7-1.7) 1.1 (0.8-1.7) 1 1.0 (0.7-1.3) 1 1.2 (0.9-1.7) 0.9 (0.3-3.3) 0.7 (0.2-2.6) 1 2.3 (1.1-4.6) 1 2.4 (1.2-4.8) 1.1 (0.5-2.4) 0.9 (0.4-1.9) 1 1.5 (0.9-2.4) 1 1.7 (1.0-2.8)

0.5 0.6 0.7 0.4 0.9 0.3 0.3 0.01 0.2 1.0 0.09 0.06 0.5 0.1 0.2 0.04 0.4 0.6 0.09 0.9 0.7 0.5 0.8 0.3 0.9 0.6 0.02 0.01 0.8 0.8 0.1 0.04

0.9 (0.6-1.3) 1.1 (0.7-1.5) 1 0.9 (0.7-1.2) 1 1.2 (0.9-1.7) 1.0 (0.6-1.6) 1.3 (0.8-1.9) 1 1.2 (0.9-1.7) 1 1.6 (1.1-2.3) 1.6 (0.7-3.8) 1.0 (0.4-2.5) 1 1.8 (1.0-3.3) 1 1.9 (0.9-3.6) 1.2 (0.6-2.6) 0.4 (0.1-1.2) 1 1.4 (0.8-2.4) 1 1.9 (1.0-3.5) 1.4 (0.4-4.7) 1.2 (0.4-4.1) 1 2.2 (1.0-5.0) 1 1.1 (0.4-2.8) 1.1 (0.7-1.7) 1.1 (0.8-1.6) 1 1.0 (0.7-1.3) 1 1.2 (0.9-1.8) 1.0 (0.3-3.5) 0.7 (0.2-2.7) 1 2.3 (1.1-4.6) 1 2.1 (1.0-4.3) 1.0 (0.5-2.2) 0.9 (0.4-1.8) 1 1.6 (1.0-2.6) 1 1.7 (1.0-2.9)

0.5 0.7 0.6 0.3 0.9 0.3 0.2 0.02 0.3 1.0 0.06 0.07 0.6 0.1 0.2 0.04 0.6 0.7 0.07 0.9 0.7 0.6 0.8 0.3 1.0 0.6 0.02 0.052 0.9 0.7 0.07 0.06

*The age analysis was adjusted for sex and ethnicity. The Body Mass Index analysis was adjusted for sex, age group and ethnicity and only included children 2 years or older. The underweight group was very small (n=12) and was therefore not included in these analyses. N: number.

1364

haematologica | 2016; 101(11)

Effect of age and BMI on toxicity and survival in pediatric AML

period and toxicity registration was more than 99% complete. The grade 3 and 4 toxicities were defined by the WHO, but we cannot exclude minor institutional difference in registration practices. Another limitation to our study is an inadequate number of patients to provide confident estimates for the more rare toxicities. Furthermore, weight change during treatment has been shown to have prognostic value in pediatric ALL,28,29 but we did not have information on weight change in our cohort and could thus not examine this further. Under-reporting of toxicities in pediatric AML studies is a problem,30 so the cumulative incidences of toxicities might be under-estimated, but we have no reason to believe that this should differ across age or body weight groups. Not all the toxicity end points were ideal. The toxicities were selected at conception of the protocol, but it became evident that not all were AML-relevant based on the low prevalence. Some of the end points were not specific enough (e.g. abdominal pain which could be caused by a number of underlying conditions). The knowledge gained from this study can guide the planning of future toxicity registrations in pediatric AML protocols. The baseline characteristics of the cohort were as expected. The age distribution of FAB groups, cytogenetics, and FLT3-ITD mutations were comparable to previous rapports.11,26,31 The higher frequency of inv(16) and t(8;21) in overweight patients was not the result of older age, and

Age 2-9

higher frequency of t(8;21), t(9;11) and inv(16) in overweight patients had also been found in a previous study,13 suggesting that being overweight may be associated with certain AML subtypes. The cumulative incidence of grade 3-4 toxicity, especially verified infection, was excessive, similar to the St Jude AML02 trial,3 reflecting the acute toxicity to be expected from modern pediatric AML protocols. The St Jude group have introduced antibiotic prophylaxis with vancomycin and ciprofloxacin for children treated for AML,9 but there is no international consensus on prophylactic antibiotics for children treated for AML.32 Specific microbiological organisms were not required for registration in the NOPHO-AML 2004 study, but have been collected from Danish patients treated on the protocol8 showing that viridians group streptococci was the most common cause of bloodstream infections and fungal infections were rare. Toxicity did not increase during the course of treatment (Online Supplementary Table S2). The second and fourth consolidation courses were identical (HA2E) and no more toxicity followed the second HA2E compared to the first, indicating that the bone marrow does not become exhausted during treatment. The course with the highest dose of cytarabine (HA3) given as monotherapy was the least toxic. Infants have been reported to have lower cytarabine clearance33 and other collaborative groups have considerably reduced cytarabine doses for infants.19,34 Infants did

Age 10-17

Figure 3. Event-free (A and B) and overall survival (C and D) according to body weight group and stratified for age group (n=239). The underweight group was very small (n=12) and was therefore not included in these analyses.

haematologica | 2016; 101(11)

1365

D.J.A. Løhmann et al.

not seem to experience more toxicity compared to 2-9year-olds, but the infant group in our study was small (n=34) so we might have missed small differences. The NOPHO-AML 2004 protocol did not seem to be more toxic for infants compared to the three previous BFM protocols (AML-BFM-98, -98-Interim and -2004),19 despite the higher doses of cytarabine. In particular, the rate of central neurotoxicity (a known cytarabine toxicity35) following the high-dose cytarabine consolidation courses was low, and during consolidation infants did not have more toxicity than children aged 2-9 years. The outcome for infants treated according to NOPHO-AML 2004 was excellent (5-year EFS: 66%, 95%CI: 46%-79%, 5-year OS: 82%, 95%CI: 63%-91%) and confirm previous reports showing favorable outcome for infants.19,34 High-dose cytarabine seems safe and effective in treating infants with AML, and our results do not support further dose reductions in infants as recommended by others.14 Children aged 10-17 years were at higher risk of toxicity. In particular, the risk of sepsis with hypotension was higher in children aged 10-17 years in agreement with previous findings.18 The older children also had higher risk of severe abdominal pain, which we speculate could be due to increased mucosal barrier injury in this group; mucosal barrier injury leads to infection with more virulent pathogens and a stronger host immune response.36 A review of pharmacology in adolescent cancer patients showed slower clearance of etoposide in adolescents compared to younger children.37 If older children have decreased clearance of antineoplastic drugs used in AML, increased exposure to toxic metabolites could lead to increased toxicity. We found a trend for OS was better for children aged 2-9 years. The 14 TRM cases on this protocol meant the study was not sufficiently powered to demonstrate if the increased cumulative incidence of toxicity in children aged 10-17 years translated into increased TRM, as shown by others.14,16-18 Children overweight at diagnosis were also at higher risk of several grade 3-4 toxicities both after the first course and during the entire course of treatment. Chemotherapy dose reduction in overweight children was not recommended and very few received reduced doses. In contrast to this, studies in adults show that, despite dose reductions not being recommended, overweight patients often receive reduced doses.38 In our study, overweight

References 1. Lie SO, Abrahamsson J, Clausen N, et al. Long-term results in children with AML: NOPHO-AML Study Group – report of three consecutive trials. Leukemia. 2005;19(12):2090-2100. 2. Gibson BES, Webb DKH, Howman AJ, et al. Results of a randomized trial in children with Acute Myeloid Leukaemia: medical research council AML12 trial. Br J Haematol. 2011;155(3):366-376. 3. Rubnitz JE, Inaba H, Dahl G, et al. Minimal residual disease-directed therapy for childhood acute myeloid leukaemia: results of the AML02 multicentre trial. Lancet Oncol. 2010;11(6):543-552. 4. Creutzig U, Zimmermann M, Reinhardt D, Dworzak M, et al. Early deaths and treatment-related mortality in children undergo-

1366

children received chemotherapy doses based on actual weight (and not “ideal weight”), which could lead to increased toxicity. This speculation is in part contrasted by Hijiya et al.39 who have shown that in pediatric ALL, there is no statistical difference in pharmacokinetics of cytarabine between normal and overweight patients. Previous studies have found increased TRM in overweight12 and obese13 patients with AML resulting in poorer outcome for this group; in this cohort, however, none of 9 TRM cases in children over one year of age were overweight. In contrast, we surprisingly found a trend for being overweight at diagnosis being associated with superior outcome in children aged 10-17 years [unrelated to t(8;21) status], similar to what has been reported in adults with AML.40-42 The effect of being overweight in the oldest children with AML may be more similar to the effect of being overweight in adults. In addition, the overweight group may have benefitted from the therapy without dose reductions. In our cohort, few were obese (n=16, 29% of the overweight group) compared to all and 57% of the patients in the overweight group in two American studies.12,13 This difference might partly explain the difference in results. In conclusion, we found the toxicity of the NOPHOAML 2004 protocol to be considerable. The high doses of cytarabine given were safe for infants in our setting and resulted in excellent outcome for the youngest patients. Age 10-17 years was associated with increased toxicity and a trend for poorer survival. Further studies on the pharmacology of AML drugs in adolescence are needed. Being overweight at diagnosis was associated with increased toxicity, but also with a trend for improved survival in children aged 10-17 years. Dose reduction in overweight patients does not seem justified in those patients with appropriate supportive care. Acknowledgments The authors would like to thank all the participating patients and their parents, clinicians and research staff at the NOPHO institutions for reporting data, and Peter Haubjerg Asdahl for statistical assistance. Funding This study was supported by grants from the Danish Childhood Cancer Foundation, the Danish Cancer Society and the Novo Nordisk Foundation.

ing therapy for acute myeloid leukemia: analysis of the multicenter clinical trials AML-BFM 93 and AML-BFM 98. J Clin Oncol. 2004;22(21):4384-4393. 5. Riley LC, Hann IM, Wheatley K, Stevens RF. Treatment-related deaths during induction and first remission of acute myeloid leukaemia in children treated on the Tenth Medical Research Council acute myeloid leukaemia trial (MRC AML10). The MCR Childhood Leukaemia Working Party. Br J Haematol. 1999;106(2):436-444. 6. Rubnitz JE, Lensing S, Zhou Y, et al. Death during induction therapy and first remission of acute leukemia in childhood: the St. Jude experience. Cancer. 2004;101(7):1677-1684. 7. Molgaard-Hansen L, Möttönen M, Glosli H, et al. Early and treatment-related deaths in childhood acute myeloid leukaemia in the Nordic countries: 1984-2003. Br J

Haematol. 2010;151(5):447-459. 8. Johannsen KH, Handrup MM, Lausen B, Schrøder H, Hasle H. High frequency of streptococcal bacteraemia during childhood AML therapy irrespective of dose of cytarabine. Pediatr Blood Cancer. 2012; 60(7):1154-1160. 9. Inaba H, Gaur AH, Cao X, et al. Feasibility, efficacy, and adverse effects of outpatient antibacterial prophylaxis in children with acute myeloid leukemia. Cancer. 2014;120(13):1985-1992. 10. Razzouk BI, Estey E, Pounds S, et al. Impact of age on outcome of pediatric acute myeloid leukemia: a report from 2 institutions. Cancer. 2006;106(11):2495-2502. 11. Creutzig U, Büchner T, Sauerland MC, et al. Significance of age in acute myeloid leukemia patients younger than 30 years: a common analysis of the pediatric trials

haematologica | 2016; 101(11)

Effect of age and BMI on toxicity and survival in pediatric AML

12.

13.

14.

15. 16.

17.

18.

19.

20.

21.

22.

AML-BFM 93/98 and the adult trials AMLCG 92/99 and AMLSG HD93/98A. Cancer. 2008;112(3):562-571. Inaba H, Surprise HC, Pounds S, et al. Effect of body mass index on the outcome of children with acute myeloid leukemia. Cancer. 2012;118(23):5989-5996. Lange BJ, Gerbing RB, Feusner J, et al. Mortality in overweight and underweight children with acute myeloid leukemia. JAMA. 2005;293(2):203-211. Tomizawa D, Tawa A, Watanabe T, et al. Appropriate dose reduction in induction therapy is essential for the treatment of infants with acute myeloid leukemia: a report from the Japanese Pediatric Leukemia/Lymphoma Study Group. Int J Hematol. 2013;98(5):578-588. Hossain MJ, Xie L, Caywood EH. Cancer Epidemiology. Cancer Epidemiol. 2015; 39(5):720-726. Rubnitz JE, Pounds S, Cao X, et al. Treatment outcome in older patients with childhood acute myeloid leukemia. Cancer. 2012;118(24):6253-6259. Canner J, Alonzo TA, Franklin J, et al. Differences in outcomes of newly diagnosed acute myeloid leukemia for adolescent/young adult and younger patients. Cancer. 2013;119(23):4162-4169. Sung L, Buxton A, Gamis A, Woods WG, Alonzo TA. Life-threatening and fatal infections in children with acute myeloid leukemia: a report from the Children's Oncology Group. J Pediatr Hematol Oncol. 2012;34(1):e30-35. Creutzig U, Zimmermann M, Bourquin J-P, et al. Favorable outcome in infants with AML after intensive first- and second-line treatment: an AML-BFM study group report. Leukemia. 2012;26(4):654-661. Abrahamsson J, Forestier E, Heldrup J, et al. Response-Guided Induction Therapy in Pediatric Acute Myeloid Leukemia With Excellent Remission Rate. J Clin Oncol. 2011;29(3):310-315. Hasle H, Abrahamsson J, Forestier E, et al. Gemtuzumab ozogamicin as postconsolidation therapy does not prevent relapse in children with AML: results from NOPHOAML 2004. Blood. 2012;120(5):978-984. World Health Organization. WHO handbook for reporting results of cancer treatment. Geneva, Switzerland: World Health

haematologica | 2016; 101(11)

Organization, offset publication. 1979;48. 23. de Onis M, Lobstein T. Defining obesity risk status in the general childhood population: Which cut-offs should we use? Int J Pediatr Obes. 2010;5(6):458-460. 24. Andersson T, Alfredsson L, KĂ¤llberg H, Zdravkovic S, Ahlbom A. Calculating measures of biological interaction. Eur J Epidemiol. 2005;20(7):575-579. 25. Ogden CL, Kuczmarski RJ, Flegal KM, et al. Centers for Disease Control and Prevention 2000 growth charts for the United States: improvements to the 1977 National Center for Health Statistics version. Pediatrics. 2002;109(1):45-60. 26. Webb DK, Harrison G, Stevens RF, et al. Relationships between age at diagnosis, clinical features, and outcome of therapy in children treated in the Medical Research Council AML 10 and 12 trials for acute myeloid leukemia. Blood. 2001;98(6):17141720. 27. Sung L, Lange BJ, Gerbing RB, Alonzo TA, Feusner J. Microbiologically documented infections and infection-related mortality in children with acute myeloid leukemia. Blood. 2007;110(10):3532-3539. 28. Hoed den MAH, Pluijm SMF, de GrootKruseman HA, et al. The negative impact of being underweight and weight loss on survival of children with acute lymphoblastic leukemia. Haematologica. 2014; 100(1):62-69. 29. Orgel E, Sposto R, Malvar J, et al. Impact on Survival and Toxicity by Duration of Weight Extremes During Treatment for Pediatric Acute Lymphoblastic Leukemia: A Report From the Children's Oncology Group. J Clin Oncol. 2014;32(13):13311337. 30. Miller TP, Li Y, Kavcic M, et al. Accuracy of Adverse Event Ascertainment in Clinical Trials for Pediatric Acute Myeloid Leukemia. J Clin Oncol. 2016;34(13):15371543. 31. Zwaan CM, Meshinchi S, Radich JP, et al. FLT3 internal tandem duplication in 234 children with acute myeloid leukemia: prognostic significance and relation to cellular drug resistance. Blood. 2003;102(7): 2387-2394. 32. Lehrnbecher T, Ethier M-C, Zaoutis T, et al. International variations in infection supportive care practices for paediatric patients

33.

34.

35. 36.

37.

38.

39.

40.

41.

42.

with acute myeloid leukaemia. Br J Haematol. 2009;147(1):125-128. Periclou AP, Avramis VI. NONMEM population pharmacokinetic studies of cytosine arabinoside after high-dose and after loading bolus followed by continuous infusion of the drug in pediatric patients with leukemias. Cancer Chemother Pharmacol. 1996;39(1-2):42-50. Masetti R, Rondelli R, Fagioli F, et al. Infants with acute myeloid leukemia treated according to the Associazione Italiana di Ematologia e Oncologia Pediatrica 2002/01 protocol have an outcome comparable to that of older children. Haematologica. 2014;99(8):e127-129. Stentoft J. The toxicity of cytarabine. Drug Saf. 1990;5(1):7-27. van der Velden WJFM, Herbers AHE, Netea MG, Blijlevens NMA. Mucosal barrier injury, fever and infection in neutropenic patients with cancer: introducing the paradigm febrile mucositis. Br J Haematol. 2014;167(4):441-452. Veal GJ, Hartford CM, Stewart CF. Clinical pharmacology in the adolescent oncology patient. J Clin Oncol. 2010;28(32):47904799. Lyman GH, Sparreboom A. Chemotherapy dosing in overweight and obese patients with cancer. Nat Rev Clin Oncol. 2013; 10(8):451-459. Hijiya N, Panetta JC, Zhou Y, et al. Body mass index does not influence pharmacokinetics or outcome of treatment in children with acute lymphoblastic leukemia. Blood. 2006;108(13):399-4002. Wenzell CM, Gallagher EM, Earl M, et al. Outcomes in obese and overweight acute myeloid leukemia patients receiving chemotherapy dosed according to actual body weight. Am J Hematol. 2013; 88(10):906-909. Ando T, Yamazaki E, Teshigawara H, et al. Body Mass Index Is a Prognostic Factor in Adults with Newly Diagnosed Acute Myeloid Leukemia: A Retrospective MultiInstitutional Study in Japan [abstract]. Blood. 2015;126(23):1316-1316. Medeiros BC, Othus M, Estey EH, et al. Impact of body-mass index on the outcome of adult patients with acute myeloid leukaemia. Haematologica. 2015;97(9): 1401-1404.

1367

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Lymphoblastic Leukemia

Ferrata Storti Foundation

Optimal interleukin-7 receptor-mediated signaling, cell cycle progression and viability of T-cell acute lymphoblastic leukemia cells rely on casein kinase 2 activity Alice Melão, Maureen Spit, Bruno A. Cardoso, and João T. Barata

Haematologica 2016 Volume 101(11):1368-1379

Instituto de Medicina Molecular, Faculdade de Medicina, Universidade de Lisboa, Lisbon, Portugal

ABSTRACT

Correspondence: joao_barata@medicina.ulisboa.pt

Received: December 18, 2015. Accepted: July 26, 2016. Pre-published: July 28, 2016. doi:10.3324/haematol.2015.141143

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1368

1368

nterleukin-7 and interleukin-7 receptor are essential for normal T-cell development and homeostasis, whereas excessive interleukin7/interleukin-7 receptor-mediated signaling promotes leukemogenesis. The protein kinase, casein kinase 2, is overexpressed and hyperactivated in cancer, including T-cell acute lymphoblastic leukemia. Herein, we show that while interleukin-7 had a minor but significant positive effect on casein kinase 2 activity in leukemia T-cells, casein kinase 2 activity was mandatory for optimal interleukin-7/ interleukin-7 receptor-mediated signaling. Casein kinase 2 pharmacological inhibition impaired signal transducer and activator of transcription 5 and phosphoinositide 3-kinase/v-Akt murine thymoma viral oncogene homolog 1 pathway activation triggered by interleukin-7 or by mutational activation of interleukin-7 receptor. By contrast, forced expression of casein kinase 2 augmented interleukin-7 signaling in human embryonic kidney 293T cells reconstituted with the interleukin-7 receptor machinery. Casein kinase 2 inactivation prevented interleukin-7-induced B-cell lymphoma 2 upregulation, maintenance of mitochondrial homeostasis and viability of T-cell acute lymphoblastic leukemia cell lines and primary leukemia cells collected from patients at diagnosis. Casein kinase 2 inhibition further abrogated interleukin-7-mediated cell growth and upregulation of the transferrin receptor, and blocked cyclin A and E upregulation and cell cycle progression. Notably, casein kinase 2 was also required for the viability of mutant interleukin-7 receptor expressing leukemia T-cells. Overall, our study identifies casein kinase 2 as a major player in the effects of interleukin-7 and interleukin-7 receptor in T-cell acute lymphoblastic leukemia. This further highlights the potential relevance of targeting casein kinase 2 in this malignancy. Introduction T-cell acute lymphoblastic leukemia (T-ALL) is an aggressive hematological cancer that results from the transformation of thymic T-cell precursors and accounts for 10–15% of pediatric ALL cases. Although the 5-year event-free survival rate has significantly improved for these patients, reaching up to 80%, they still present an increased risk for early relapse with very poor prognosis.1 Moreover, current intensive therapies have considerable long-term side effects. Thus, it is critical to better define the underlying mechanisms involved in leukemogenesis and resistance to treatment, in order to develop improved therapeutic strategies that minimize toxicities and the probability of relapse. Interleukin-7 (IL-7) is a cytokine essential for normal T-cell development and homeostasis in humans and mice.2,3 IL-7 is present in the microenvironments where T-cell precursors reside, and are secreted by a variety of cells, amongst which strohaematologica | 2016; 101(11)

CK2 is essential for IL-7 effects in T-ALL

mal cells are involved, in the thymus and bone marrow. In the last few years several studies have provided new insights into the relevance of this cytokine and its receptor (IL-7R) for the development of autoimmune and chronic inflammatory diseases.4 Moreover, activation of the IL7/IL-7R signaling axis has been shown to contribute to Tcell leukemogenesis,5-10 whereas IL-7 deficiency leads to decreased in vivo expansion of leukemia T-cells and delayed leukemia-associated death in mice transplanted with human T-ALL cells.11 Notably, we and others revealed that IL7R (encoding the IL-7Ra subunit, also known as CD127) is a bona fide oncogene. Around 10% of pediatric T-ALL patients display IL7R gain-of-function mutations, which lead to constitutive activation of downstream signaling and subsequent promotion of cell transformation and tumorigenesis.12-16 Casein kinase 2 (CK2) is a ubiquitously expressed serine/threonine kinase, that is involved in the regulation of numerous cellular processes (e.g. cell cycle, gene expression and proliferation), through the modulation of the crosstalk between multiple signaling pathways.17 Many of the CK2 described substrates are proteins involved in the regulation of cell survival, with compiled evidence that the reduction of CK2 activity or expression leads to cell death, in such a way that CK2 is considered to have mainly a pro-survival and proliferative function. In agreement with these features, CK2 is significantly and consistently overexpressed in solid18 and hematological19-22 tumor cells, including T-ALL.23 Primary T-ALL cells collected from diagnostic patients display basal hyperactivation of the PI3K/Akt signaling pathway.23 Although gene inactivation of PTEN, the main negative regulator of the pathway, can occur in up to 25% of T-ALL cases,23,24 PI3K/Akt signaling pathway activation results most frequently from PTEN post-translational inhibition mediated by oxidation via reactive oxygen species and by phosphorylation due to high CK2 activity in the leukemia cells.23 More recently, it has been shown that CK2 also regulates the JAK/STAT pathway by interacting with JAKs, thereby facilitating the activation of STATs.25 These observations highlight the ability of CK2 to positively modulate JAK/STAT and PI3K/Akt pathways in the context of cancer. Notably, PI3K/Akt/mTOR and JAK/STAT signaling pathways are also activated by IL-7, and have a pivotal role in leukemia development.26

However, whether CK2 is involved in IL-7-mediated signaling, particularly in the context of T-cell leukemia, remains to be elucidated. Although CK2 has constitutive kinase activity and is viewed as largely refractory to ‘vertical’ stimulation by growth factors, playing mostly a ‘horizontal’ role as a modulator of the activity of diverse signaling pathways,27 there is evidence that CK2 can play an important function downstream from external stimuli.28,29 In the study herein, we evaluated the possible involvement of CK2 in IL-7-mediated effects on T-ALL cells. Our results indicate that CK2 activity is essential for optimal IL-7/IL-7R-dependent signaling via PI3K/Akt and JAK/STAT pathways in leukemia T-cells. Moreover, inhibition of CK2 prevents IL7/IL-7R-mediated viability and cell cycle progression of TALL cells. Our results indicate that CK2 partakes in T-cell leukemia development, not only via its basal impact on key oncogenic signaling pathways, but also by being a major regulator of IL-7/IL-7R-mediated signaling in T-ALL.

Methods Cells Primary leukemia cells were obtained from the bone marrow and/or peripheral blood of diagnostic pediatric T-ALL cases, and were classified according to the European Group for the Immunological Classification of Leukemias (EGIL) criteria30 (Table 1). Informed consent was obtained in accordance with the Declaration of Helsinki and under the ethical review board approval of Instituto Português de Oncologia (Lisbon, Portugal). The IL-7–dependent cell line TAIL7, which shares significant similarities with primary leukemia samples,31 DND-41, HPB-ALL and HEK293T cell lines were cultured as described in the Online Supplementary Methods.

In vitro CK2 kinase assay CK2 activity was measured using the Casein Kinase 2 Assay Kit (Millipore) according to the manufacturer's instructions, as previously described.19 Kinase activity was calculated by subtracting the substrate-less background for each sample

Transfection of HEK293T cells

Vectors bearing human JAK3, γC, IL-7Ra, and mouse Stat5a were used to reconstitute the IL-7 signaling machinery in

Table 1. Immunophenotypic characteristics, maturation stage and IL7R mutational status of T-ALL primary cells and cell lines.

Primary Samples T-ALL#1 T-ALL#2 T-ALL#3 T-ALL#4 T-ALL#5 T-ALL#6 T-ALL#7 T-ALL#8 Cell Lines TAIL7 DND-41 HPB-ALL

CD1a

CD2

CD3

cCD3

CD5

CD7

CD4

CD8

Maturation stage

IL7R mutation

+ ND + ND + +

+ + + + + +

ND +

+ + + + + ND + ND

+ + + + + + +

+ ND + + + + + +

ND + +

ND + + ND

III (cortical) II or IV (non-cortical) II (pre-T) III (cortical) II (pre-T) or III (cortical) III (cortical) II (pre-T) III (cortical)

ND +

− + +

+ + +

− + +

+ + +

+ +

II (pre-T) III (cortical) III (cortical)

+ -

ND: not determined; T-ALL: T-cell acute lymphoblastic leukemia; +: positive; -:negative.

haematologica | 2016; 101(11)

1369

A. Melão et al.

HEK293T cells (which express only JAK1 endogenously). Vectors bearing the CK2a and a’ subunit were kindly provided by D.W. Litchfield.32 Cells were transfected using Lipofectamine 2000 (Invitrogen) according to the instructions of the manufacturer. Transfected cells were stimulated with IL-7 (100ng/ml) for 15 minutes and 6 hours at 37°C. Reactions were stopped by placing samples on ice.

Actin (Santa Cruz Biotechnology, Inc.), p-STAT5a/b (Y694/Y699) (Millipore), p-JAK1 (Y1022/1023), p-Akt (S473), p-PTEN (S380), Akt, PTEN (Cell Signaling Technology), p27Kip1 (BD Biosciences), PARP (Novus Biologicals) and p-Akt (S129) (Abgent). Densitometry analysis was performed using Adobe Photoshop CS5 Extended software (Adobe Systems). Results were normalized to the loading control.

Immunoblotting

Analysis of cell growth, activation and viability

Lysates were prepared as described,33 resolved by SDS-PAGE, and immunoblotted with antibodies against p-JAK3 (Y980), JAK3, JAK1, STAT5, Cyclin A, Cyclin E, Cyclin D2, CK2a, CK2a’, CK2β,

Cell growth was determined as described.34 The activation marker CD71 was measured using FITC-conjugated anti-CD71 (eBioscience) antibody. Results were expressed as the percentage

Figure 1. CK2 activity modulates IL-7-mediated signaling in T-ALL cells. (A) TAIL7 cells were pre-incubated with the CK2-specific inhibitor CX-4945 (6μM) for 2 hours and stimulated with IL-7 (50ng/ml) for 15 minutes. Data represent normalized mean±sem from two independent experiments. *P<0.05, ***P<0.001 (One-way ANOVA, with Tukey’s post-test). (B) TAIL7 cells were incubated with IL7 for the indicated time periods. (C) TAIL7 cells were pre-incubated with a pan-Jak inhibitor (10μM) for 2 hours and stimulated with IL-7 for 15 minutes or 48 hours. CK2 activity was determined as described in ‘Methods’. (D) TAIL7 cells were incubated with or without IL-7 for the indicated time points. Immunoprecipitation of IL-7Rα, CK2α and IgG (negative control) was followed by immunoblotting with anti-IL-7Rα and anti-CK2α antibodies. Ponceau staining of IgG heavy chain was used as loading control. (E) TAIL7 and HPB-ALL cells were pre-incubated for 2 hours with CX-4945 and stimulated with IL-7 for 15 minutes. Total protein extracts were resolved by SDS-PAGE, and total and phosphorylated proteins were detected by immunoblot. (F) HEK293T cells were transfected with CK2α and/or CK2α’ subunits or mock control (θ) and stimulated with IL-7 (100ng/ml) for 15 minutes. (E,F) Values correspond to load control-normalized densitometric ratios of phospho-proteins. CK2: casein kinase 2; IL-7: interleukin 7; IgG: immunoglobulin G; p-Jak1: phospho-janus kinase 1; p-STAT5: phospho-signal transducer and activator of transcription 5; p-Akt: phospho-protein kinase B; T-ALL: T-cell acute lymphoblastic leukemia; SDS-PAGE: sodium dodecyl sulfate polyacrylamide gel electrophoresis.

1370

haematologica | 2016; 101(11)

CK2 is essential for IL-7 effects in T-ALL

of positive cells and as the specific mean intensity of fluorescence (MIF).34 The determination of cell viability was performed by flow cytometry analysis of FSCxSSC distribution, and by Annexin V (eBioscience) and 7-AAD (BD Biosciences) staining.

a cell permeabilization kit according to the manufacturer's instructions (ADG Bio Research GmbH), and detected by flow cytometry with FITC-conjugated anti–Bcl-2 antibody (Dako).

Assessment of mitochondrial membrane potential (Dym)

Proliferation assays Proliferation was assessed as described.5

Intracellular staining Bcl-2 expression was determined by intracellular staining using

Cells were harvested, stained in RPMI1640 medium with TMRE (Sigma-Aldrich) to a final concentration of 100nM, incubated for 15 minutes at 37°C with 5% CO2, and analyzed by flow cytometry.

F P<0.001

Figure 2. CK2 activity is required for IL-7-mediated Bcl-2 upregulation, mitochondrial homeostasis and prevention of T-ALL cell apoptosis. TAIL7 cells were cultured with IL-7 (10ng/ml) in the presence of increasing concentrations of the CK2 inhibitor CX-4945, as indicated, and analyzed for (A) apoptosis at 96 hours; (B,C) mitochondrial membrane potential (DΨm) at 72 hours, gated in the whole population (B) as an additional measure of overall apoptosis, or within the live cell population (C) as a measure of a very early event in apoptosis; and (D) PARP cleavage at 24 hours. (E,F) Bcl-2 levels were analyzed by flow cytometry at 72 hours. Mean fluorescence intensity (MFI) are indicated for each condition. Data are representative (E) of the mean±sem (F) of two independent experiments. CK2: casein kinase 2; IL-7: interleukin 7; Bcl-2: B-cell lymphoma 2; T-ALL: T-cell acute lymphoblastic leukemia; PARP: poly ADP (Adenosine, Diphosphate)-ribose polymerase; TMRE: tetramethylrhodamine ethyl ester.

haematologica | 2016; 101(11)

1371

A. Melão et al.

Cell cycle analysis

Results

DNA content after staining with propidium iodide was measured as described previously.5 Cell cycle distribution was determined using ModFit LT software (Verity Software House).

Optimal IL-7-mediated signaling requires CK2 activity in T-ALL cells

Statistical analysis FlowJo Software (Tree Star Inc.) was used to analyze flow cytometry data. GraphPad Prism software was used for statistical analysis. Differences between mean values were calculated using two-tailed Student's t-test and two-way ANOVA, as appropriate. P<0.05 was considered significant. Additional methods are available in the Online Supplementary Methods.

We previously showed that CK2 is overexpressed and hyperactivated in T-ALL23 (Online Supplementary Figure S1A), and that T-ALL cells can benefit from IL-7 in vitro5,34 and in vivo.11 We now questioned whether a link could exist between IL-7-mediated signaling and CK2 activity. We determined CK2 activity in human IL-7-dependent TAIL7 T-ALL cells, which display the biological and signaling properties of primary leukemia cells.31 IL-7 induced rapid (within 15 minutes) yet very mild

H-thymidine incorporation

1372

Figure 3. CK2 activity is essential for IL-7-mediated T-ALL cell growth and cell cycle progression. TAIL7 and HPB-ALL cells were cultured in medium alone or with IL-7 (20ng/ml), in the presence or absence of CX-4945 (6mM), and analyzed at the time points indicated below. Results were similar for both cell lines and representative data are presented. (A) TAIL7 cell size increase (cell growth) was determined by analysis of FSC distribution by flow cytometry after 72 hours. (B) Expression of the ‘activation’ marker CD71 (transferring receptor) at the surface of HPB-ALL cells was measured by flow cytometry at 72 hours. (C) Proliferation of TAIL7 cells was assessed at 72 hours by 3H-thymidine incorporation as described in ‘Methods’. (D) Cell cycle profile of TAIL7 cells was assessed by PI staining of fixed cells and analyzed by flow cytometry at 48 hours. (E) The corresponding expression profile of the indicated cell cycle regulators was measured by immunoblot. Values denote load control-normalized densitometric ratios. Results are representative of 2 to 3 independent experiments. CK2: casein kinase 2; IL-7: interleukin 7; T-ALL: Tcell acute lymphoblastic leukemia; PI: propidium iodide; MFI: mean fluorescence intensity.

haematologica | 2016; 101(11)

CK2 is essential for IL-7 effects in T-ALL

upregulation of CK2 activity, which was completely inhibited by pre-treatment with CX-4945, a clinical-grade, highly selective CK2 inhibitor35 (Figure 1A). This transient early response, which was no longer detectable 30 minutes after IL-7 stimulation, was followed by late, more robust increased CK2 activation at 24-48 hours (Figure 1B). IL-7 did not significantly alter the expression of the transcript (Online Supplementary Figure S1B) or protein (Online

Supplementary Figure S1C) levels of any of the CK2 isoforms (a, a’ or β), suggesting that IL-7-mediated upregulation of CK2 activity was independent of the regulation of CK2 expression levels. JAK1 and JAK3 associate with IL-7Ra and the other IL-7R subunit (γc), respectively.36 The blockade of IL-7 signaling using a pan-JAK inhibitor (Pyridone 6) abrogated CK2 activity upon short- and long-term IL-7 stimulation

H-thymidine incorporation

Figure 4. CK2 inhibition abolishes IL-7-mediated viability, activation and proliferation of primary T-ALL blasts from diagnostic pediatric patients. Primary T-ALL cells were cultured for 24 hours (T-ALL#6) or 48 hours (T-ALL#7) in medium alone or with IL-7 (20ng/ml) in the presence of increasing concentrations of CX-4945, as indicated. Viability and apoptosis were assessed by Annexin V/7-AAD staining (A), TMRE (B) and Bcl-2 expression (C). ‘Activation’ status was assessed by cell size determination (D) and CD71 surface expression (E). Proliferation was determined by 3H-thymidine incorporation at 72 hours (F) as described in ‘Methods’. CK2: casein kinase 2; IL7: interleukin 7; T-ALL: T-cell acute lymphoblastic leukemia; TMRE: tetramethylrhodamine ethyl ester; Bcl-2: B-cell lymphoma 2; MFI: mean fluorescence intensity.

haematologica | 2016; 101(11)

1373

A. Melão et al.

(Figure 1C), indicating that JAKs are required for IL-7Rmediated CK2 activation in T-ALL cells. To further characterize the mechanisms linking IL-7 and CK2, we performed co-immunoprecipitation experiments involving

IL-7Ra and CK2a. We found that IL-7Ra co-immunoprecipitated CK2a and vice-versa in both non-stimulated and IL-7-treated TAIL7 cells (Figure 1D). This indicates that IL-7Ra and CK2a physically interact at steady-state, and

p-Akt p-Akt Akt

E F

H-thymidine incorporation

Figure 5. CK2 inhibition abolishes constitutive IL-7R-mediated signaling and survival of T-ALL cells expressing mutant IL-7Rα. (A) DND-41 cells were cultured for 24 hours in the presence or absence of CX-4945 (10μM). Lysates were resolved by SDS-PAGE and analyzed for the levels of phosphorylation of the indicated proteins. Total Akt was used as loading control. (B-D) DND-41 cells were cultured in medium alone or in the presence of the indicated doses of CX-4945, and their viability was analyzed at 48 hours by FSCxSSC flow cytometry discrimination (B, C) and Annexin V staining (D). Results in (B) are summarized as mean±sem of four independent experiments. Data in (A-D) are representative of 4 independent experiments. (E-H). Primary mutant IL7R (p.Leu242_Leu243insAsnProCys) leukemia cells from patient #8 (see Table 1) were cultured in medium alone or in the presence of the indicated doses of CX-4945. (E) Viability was assessed by Annexin V/ 7-AAD staining at 72 hours. (F) Mitochondrial membrane potential (DΨm) was evaluated at 48 hours after TMRE staining, gated on the live population. (G) Cell growth was determined by analysis of FSC distribution by flow cytometry at 72 hours. (H) Proliferation was determined by 3H-thymidine incorporation. Results in (F) and (H) are summarized as mean±sem of 2 replicate analyses. CK2: casein kinase 2; IL-7: interleukin 7; T-ALL: T-cell acute lymphoblastic leukemia; TMRE: tetramethylrhodamine ethyl ester; p-Jak1: phospho-janus kinase 1; p-STAT5: phospho-signal transducer and activator of transcription 5; p-Akt: phospho-protein kinase B/murine thymoma viral oncogene homolog 1; (p-)PTEN: (phospho)-phosphatase and tensin homolog deleted on chromosome ten; SDS-PAGE: sodium dodecyl sulfate polyacrylamide gel electrophoresis.

1374

haematologica | 2016; 101(11)

CK2 is essential for IL-7 effects in T-ALL

that this interaction is maintained after IL-7 stimulation. Next, we assessed whether CK2 activity could impact on IL-7-mediated signaling in T-ALL. We pre-treated TAIL7 and HPB-ALL (IL-7-responsive) T-ALL cells with CX-4945 for 2 hours, stimulated each cell line with IL-7 for 15 minutes, and determined the phosphorylation status of known IL-7-activated signaling pathways (Figure 1E). Akt is phosphorylated at S129 by CK2.37 In accordance with the CK2 kinase activity assay, IL-7 promoted only a minor increase in Akt phospho-S129 levels, which were strongly downregulated by CX-4945. IL-7-dependent activation of the JAK/STAT pathway, measured by the increase in phospho-STAT5 and phospho-JAK1 levels, was significantly downregulated upon abrogation of CK2 activity (Figure 1E and Online Supplementary Figure S2). Likewise, IL-7-induced PI3K/Akt pathway activation was significantly prevented by CK2 inhibition, as determined by the levels of Akt S473 phosphorylation (Figure 1E and Online Supplementary Figure S2). To further confirm the relevance of CK2 for optimal IL-7-mediated signaling, we reconstituted all the elements of the IL-7 receptor signaling machinery in HEK293T cells,13 in the presence or absence of forced expression of CK2a and/or aâ&#x20AC;&#x2122; subunits. We found that overexpression of CK2 significantly augmented IL-7-mediated signaling as evaluated by STAT5 phospho-

rylation (Figure 1F). Overall, these experiments indicate that CK2 activity is essential for maximal IL-7/IL-7R-mediated signaling, impacting both PI3K/Akt and JAK/STAT pathways.

CK2 activity is essential for IL-7-mediated T-ALL cell viability IL-7 has the ability to promote leukemic T-cell viability and proliferation through PI3K/Akt34 and JAK/STAT5 pathways.38 Therefore, we next sought to evaluate the consequences of CK2 inhibition on the functional outcomes of IL-7 upon T-ALL cells. The culture of TAIL7 or HPB-ALL cells in the presence of IL-7 with or without concomitant treatment with CX-4945, demonstrated that IL7-mediated upregulation of leukemia cell viability was completely prevented by CK2 inhibition (Figure 2A, Online Supplementary Figures S3 and S4A). This effect was dose-dependent (Figure 2), and mainly due to increased apoptosis, as shown by 7-AAD and Annexin V staining (Figure 2A and Online Supplementary Figure S4A), mitochondrial transmembrane potential (Figure 2B,C and Online Supplementary Figures S4B and S4C) and PARP cleavage (Figure 2D). In agreement, IL-7-mediated upregulation of Bcl-2, which is mandatory for the pro-survival effects of IL-7,5,34,39 was significantly abrogated by CX-4945 (Figure

haematologica | 2016; 101(11)

Figure 6. CK2 and JAK inhibitors synergize in inducing cell death of IL-7R mutant-expressing and IL-7-stimulated T-ALL cells. TAIL7 (A,C) and DND41 (B,D) cells were cultured in the presence of increasing amounts of CX4945 alone or combined with pan-JAK inhibitor (JAKi) or JAK1/2 inhibitor Ruxolitinib (Ruxo) for 72 hours. (A,B) Viability was determined by FSCxSSC discrimination. (C,D) Synergistic effect determination, combination index (CI) and median-effect dose (Dm) calculation was performed as described in the Online Supplementary Methods. CK2: casein kinase 2; IL-7: interleukin 7; IL-7R: interleukin 7 receptor; T-ALL: T-cell acute lymphoblastic leukemia.

1375

A. MelĂŁo et al.

2E,F and Online Supplementary Figure S4D). These results, which were corroborated using the unrelated CK2 small molecule inhibitor 4,5,6,7-tetrabromobenzotriazole (TBB)23 (Online Supplementary Figure S5), strongly suggest that CK2 activity is required for IL-7-mediated viability of T-ALL cells.

CK2 activity is essential for IL-7-mediated T-ALL cell growth and cell cycle progression IL-7 promotes hypertrophy of T-ALL cells, which is associated with augmented metabolism as measured by increased glucose uptake.5,34 To understand the effect of CK2 on IL-7-mediated leukemia cell growth, we incubated TAIL7 cells with IL-7 and CX-4945 or TBB, and determined cell size by FSCxSSC flow cytometry discrimination, where bigger cells (higher FSC) tend to be more metabolically active and proliferating. Our analyses demonstrate that the increase in cell size triggered by IL-7 was abolished by co-treatment with CX-4945 (Figure 3A) in a dose-dependent manner (Online Supplementary Figure S6A). Likewise, TBB prevented IL-7-mediated T-ALL cell growth (Online Supplementary Figure S7). Moreover, IL-7dependent surface upregulation of the transferrin receptor (CD71), which associates with T-ALL cell growth,34 was reversed by CK2 inhibition (Figure 3B and Online Supplementary Figure S6B). Because growth of lymphoid cells is often associated with cell division, we next evaluated the impact of CK2 activity on IL-7-dependent T-ALL cell proliferation. In

agreement with our previous reports,11,31,34 IL-7 increased 3 H-thymidine incorporation, indicative of cell cycle progression into S-phase. This effect was reversed in the presence of CX-4945 (Figure 3C and Online Supplementary Figure S6C) and TBB (Online Supplementary Figure S8). In agreement with these data, analysis of the cell cycle profile of TAIL7 cells demonstrated that IL-7 led to an increase in the frequency of cells in S-phase, which was completely abolished by CK2 inhibition (Figure 3D and Online Supplementary Figure S9). At the molecular level, IL7 promoted an increase in the expression of cyclins A and E, which are involved in S-phase entry and progression towards G2/M, and a decrease in the cyclin-dependent kinase p27Kip1, whose expression contributes to prevent cell cycle progression past G1. In agreement with progression towards S-phase, the G1-associated cyclin D2 was mildly downregulated by IL-7. Upon inhibition of CK2 activity with CX-4945, these effects were completely reversed (Figure 3E), in accordance with the accumulation of TAIL7 cells in G1 (Figure 3D). These data demonstrate that CK2 kinase activity is required for IL-7 to promote growth and cell cycle progression of T-ALL cells.

CK2 activity is mandatory for IL-7-mediated viability, growth and proliferation of primary T-ALL cells from diagnostic patients Because T-ALL cell lines may accumulate alterations that do not necessarily reflect primary disease, we next sought to extend the pathophysiological relevance of our

Figure 7. Model for CK2 involvement in IL-7/IL-7R-mediated signaling in T-ALL. In addition to its known role on PTEN posttranslational inactivation in T-ALL, CK2 appears to bind to the IL-7RÎą chain and to be required for optimal IL-7R-mediated signaling, both in response to IL-7 and as a result of mutational activation of the receptor. Although the exact mechanisms by which CK2 may be regulated by and impact on JAK activity require further research, the consequence is that the activation of IL-7R downstream effectors, namely STAT5 and Akt, relies on CK2 activity. Thus, the use of CK2 inhibitors (such as CX-4945 or TBB) may be therapeutically relevant not only because they block CK2-mediated PTEN inactivation but also, as reported herein, because they are able to abrogate IL-7/IL-7R-mediated proleukemia signaling. CK2: casein kinase 2; IL-7: interleukin 7; IL-7R: interleukin 7 receptor; T-ALL: T-cell acute lymphoblastic leukemia; PTEN: phosphatase and tensin homolog deleted on chromosome ten; TBB: 4,5,6,7-tetrabromobenzotriazole; PIP2: phosphatidylinositol 4,5-bisphosphate; PIP3: phosphatidylinositol 3,4,5-trisphosphate; PI3K: phosphoinositide 3-kinase.

1376

haematologica | 2016; 101(11)

CK2 is essential for IL-7 effects in T-ALL

findings by determining whether these were reproduced in T-ALL blasts collected from patients at diagnosis. We confirmed that IL-7 had a major positive impact on the viability of all the primary T-ALL patient samples we analyzed, with significant downregulation of spontaneous apoptosis. This effect was completely reversed by CK2 inhibition in all cases (n=8; Figure 4A, Online Supplementary Figure S10A and data not shown). Analysis of mitochondrial transmembrane potential confirmed that CK2 modulates the ability of IL-7 to prevent this very early sign of apoptosis (Figure 4B and Online Supplementary Figure S10B). This process, similar to that which was observed in TAIL7 and HPB-ALL cells, likely reflects IL-7triggered fluctuations in Bcl-2 expression that are dependent on CK2 activity (Figure 4C and Online Supplementary Figure S10C). The similarities between T-ALL cell lines and primary leukemia cells extended beyond the impact on cell survival. Upon CK2 inhibition T-ALL blasts were no longer able to augment their size (Figure 4D, Online Supplementary Figures S7 and S10D) or CD71 surface expression (Figure 4E and Online Supplementary Figure S10E) in the presence of IL-7. Accordingly, the well known proliferative effect of IL-7 on primary T-ALL cells9 was blocked by CX-4549 (Figure 4F and Online Supplementary Figure S10F) in a dose-dependent fashion (Figure 4F). Likewise, TBB prevented IL-7-mediated proliferation of primary T-ALL cells (Online Supplementary Figure S8). These data taken together indicate that CK2 is a key element in IL-7-mediated promotion of viability, growth and proliferation of T-ALL blasts.

CK2 inhibition abrogates constitutive signaling and viability of mutant IL7R-expressing T-ALL cells We and others have shown that around 10% of pediatric T-ALL cases display gain-of-function mutations in the a chain of the IL-7R, leading to constitutive activation of downstream signaling, namely PI3K/Akt and STAT5, with the consequent promotion of cell proliferation, transformation12,13,40 and tumorigenesis.13,14,16 Therefore, we next sought to determine whether CK2 is also required in the context of the signals elicited by mutant IL7R. We found that CK2 inhibition prevented constitutive signaling downstream from mutated IL7R in DND-41 T-ALL cells,16 as determined by the levels of phosphorylation of Akt and STAT5 (Figure 5A). Accordingly, CX-4945 induced DND41 cell death in a time- and dose-dependent manner (Figure 5B,C), which was associated with high levels of apoptosis (Figure 5D). Next, we extended our analysis to a primary T-ALL diagnostic sample (T-ALL#8, Table 1) displaying IL7R mutational activation, as previously characterized (patient P1 in reference 13). Remarkably, the CK2 inhibitor promoted apoptosis (Figure 5E,F) and atrophy (Figure 5G), and prevented proliferation (Figure 5H) of leukemia blasts at concentrations even lower than those required for DND-41 cells. Overall, these results indicate that, similar to T-ALL cells stimulated with IL-7, cells displaying mutant IL7R remain sensitive to abrogation of CK2 activity.

Combined CK2 and JAK inhibition is synergistic against both IL-7-dependent and mutant IL7R-expressing T-ALL cells To further generate preliminary evidence of the clinical potential of our observations we next investigated whether the combination of CX-4945 with JAK inhibitors haematologica | 2016; 101(11)

would result in more efficient elimination of IL-7/IL-7Rmediated T-ALL cell viability. Treatment of TAIL7 cells with a combination of CX-4945 and either the pan-JAK inhibitor or the JAK1/2 clinical-stage inhibitor ruxolitinib, synergized in preventing IL-7-mediated viability (Figure 6A and 6C). The same combinations also displayed a synergistic effect in inducing cell death of mutant IL-7Rexpressing DND-41 cells (Figure 6B and 6D). These results suggest that inhibiting concomitantly CK2 and JAK may be particularly effective in targeting both IL-7- and mutant IL7R-dependent T-ALL cells.

Discussion Throughout the years, considerable evidence has accumulated pinpointing the importance of the IL-7/IL-7R axis for T-cell leukemogenesis.26 The pro-oncogenic role of IL7 and IL-7R in human T-ALL has been clearly highlighted by recent data, revealing that IL-7 significantly accelerates T-ALL disease progression in vivo11 and that gain-of-function mutations in the IL-7R exist in T-ALL patients, including in poor prognosis cases.12,13,40 On the other hand, CK2 is frequently overexpressed and hyperactivated in T-ALL,23 driving PI3K/Akt pathway activation by posttranslationally inhibiting PTEN.23 In the present studies we sought to determine whether CK2 and IL-7/IL-7R could be functionally linked, by evaluating whether CK2 is involved in the mechanisms underlying IL-7/IL-7R-mediated effects in T-ALL cells. We demonstrated that IL-7 upregulates CK2 activity in a minor but significant manner, with clear biological impacts. The underlying mechanisms remain to be elucidated. CK2 has been shown to bind to JAK kinases,25 which in turn are known to associate with the IL-7R. In agreement, we have demonstrated that CK2 interacts with IL-7Ra. Thus, it is possible that CK2 activity is upregulated in the context of multimeric complexes involving the IL-7 receptor and JAK1. Consistent with this was the demonstration that JAK inhibition impeded IL-7mediated CK2 activation. Given that Akt was recently shown to phosphorylate and thereby regulate CK2,41 it is also possible that CK2 may be involved in complex loops in which it is both regulated by and a regulator of PI3K/Akt signaling downstream from IL-7/IL-7R. This has a precedent in mTOR, which is activated downstream of Akt within the mTORC1 complex and is responsible for Akt activation as part of mTORC2.42 Interestingly, mTOR is also an important component of the IL-7 signaling network in ALL.5,43,44 These considerations apart, we found that CK2 activity is absolutely required for maximal IL-7mediated signaling. This was demonstrated by using two distinct CK2-specific pharmacological inhibitors, TBB and CX-4945 (Silmitasertib), the latter of which has entered phase I clinical trials for refractory solid tumors and multiple myeloma,45 and is a well-characterized, highly specific inhibitor of CK2.35 We also tried silencing the expression of CK2a and Î˛ subunits. Notably, although we were able to efficiently knockdown CK2 in T-ALL cells by up to 80%, we only partially eliminated CK2 activity, which was sufficient to maintain normal levels of Akt S129 phosphorylation and the viability of T-ALL cells (data not shown). Moreover, CRISPR/Cas9-mediated deletion of CK2 failed to produce viable T-ALL cells, suggesting that minimal CK2 expression is sufficient to maintain biologi1377

A. Melão et al.

cally relevant kinase activity that is absolutely required for T-ALL cell viability. Thus, we used the opposite strategy and forced CK2a and/or CK2a’ expression in HEK293T cells ectopically expressing the IL-7 receptor signaling machinery, thereby demonstrating that CK2 overexpression augments IL-7/IL-7R-mediated STAT5 phosphorylation. This indicates that CK2 is effectively involved in IL7-mediated signaling. Moreover, using DND-41 T-ALL cells, which display a cysteine-introducing IL-7Ra mutation, we demonstrated that CK2 is also required for constitutive signaling downstream from mutationally activated IL7R. We further characterized the extent to which CK2 impacts on IL-7/IL-7R-mediated effects on T-ALL cells. We found that CK2 activity is required for IL-7-induced viability, cell size increase and T-ALL cell cycle progression past G0/G1. In accordance, IL-7-mediated leukemia T-cell proliferation also depends on CK2. These results are consistent with the fact that CK2 regulates IL-7-triggered JAK/STAT pathway activation, which is essential for IL-7mediated leukemogenesis in mice,8 as well as PI3K/Akt signaling, which is fundamental for IL-7-mediated effects on human T-ALL cells.26,34,44 Whether CK2 is involved in the regulation of other microenvironmental signals that promote T-ALL expansion remains to be explored. Our preliminary data demonstrate that IL-4-mediated T-ALL cell viability and proliferation46 was prevented by CK2 inhibition (Online Supplementary Figure S11). These observations are consistent with the possibility that CK2 may have a broader role in regulating different extracellular (pro-leukemogenic) stimuli. Evidently, this needs to be seen in light of the fact that T-ALL cells also rely on high constitutive, cell-intrinsic activation of CK2.23 Accordingly, CK2 pharmacological inhibition decreased viability and promoted apoptosis of T-ALL cells cultured in medium alone, as previously reported.23 Importantly, in agreement with the requirement of CK2 for optimal IL-7-mediated signaling, IL-7 was not able to fully reverse this effect (Online Supplementary Figures S3 and S12). CK2 inhibition also leads to clear cell death of the mutant IL7R T-ALL cell line DND-41. This appears to be a natural corollary from both of these pathways being constitutively activated downstream of mutant IL-7R,12,13 and is in agreement with the fact that Ba/F3 cells stably expressing mutant IL-7Ra are sensitive to CK2 inhibition (data not shown). Of note, our observations indicate that clinical grade CK2 inhibitors, such as CX-4945 (Silmitasertib)23,35 or CIGB-300, may constitute valid thera-

References 1. Pui CH, Mullighan CG, Evans WE, et al. Pediatric acute lymphoblastic leukemia: where are we going and how do we get there? Blood. 2012;120(6):1165-1174. 2. Peschon JJ, Morrissey PJ, Grabstein KH, et al. Early lymphocyte expansion is severely impaired in interleukin 7 receptor-deficient mice. J Exp Med. 1994;180(5):1955-1960. 3. Puel A, Ziegler SF, Buckley RH, et al. Defective IL7R expression in T(-)B(+)NK(+) severe combined immunodeficiency. Nat Genet. 1998;20(4):394-397. 4. Mazzucchelli RI, Riva A, Durum SK. The

1378

peutic tools against T-ALL patients displaying IL-7Ra mutations, including a significant fraction of very poor prognosis ETP-ALL cases.40 Interestingly, IL-7 has also been shown to activate both PI3K/Akt and STAT5 in normal T-cell precursors, which are implicated in thymocyte proliferation and differentiation.47,48 Whether CK2 plays a role in normal T-cell development and homeostasis or, alternatively, this is a feature that is restricted to leukemia cells, in association with their high levels of CK2 expression, remains an open question. Curiously, CK2 was shown to be a key element in the process of receptor internalization through the formation of clathrin-coated pits,49 and we previously demonstrated that IL-7 promotes IL-7R internalization in both normal and leukemia T-cells, which is required for optimal IL-7mediated signaling.50 Hence, it is possible that CK2 may act as an IL-7R internalization regulator, and by this upstream effect modulate IL-7-mediated signaling, which could explain the effects observed in JAK1 protein phosphorylation status after CK2 inhibition. This possibility warrants further investigation. Overall, our present study contributes to a better understanding of the regulation of IL-7 and IL-7R-mediated signals, identifying CK2 as a critical modulator of IL-7 functional effects on T-ALL cells. There is an increasing recognition of the relevance of IL-7 and its receptor for T-cell leukemogenesis and leukemia maintenance, especially after the identification of IL7R as a bona fide T-cell oncogene,11-13,40 and clear evidence that CK2 is critical for the viability of T-ALL cells.23,35 Our data add to this knowledge by placing CK2 at the center point of both basal and IL-7Rdependent activation of pro-survival and proliferative pathways (Figure 7), and strongly supporting the rationale for the testing of CK2 inhibitors in the context of T-ALL. Funding This work was supported by the grants PTDC/SAUOBD/104816/2008 and PTDC/SAU-ONC/122428/2010 from Fundação para a Ciência e a Tecnologia and by the consolidator grant ERC CoG-648455 from the European Research Council. JTB is an FCT investigator (consolidator). AM had an FCT-SFRH PhD fellowship. Acknowledgments We thank Dr. Litchfield for generously providing the CK2 plasmids. We especially thank the generosity of patients and their families, and the collaboration of all the team from the Pediatrics Service of Instituto Português de Oncologia de Lisboa.

human IL-7 receptor gene: deletions, polymorphisms and mutations. Semin Immunol. 2012;24(3):225-230. 5. Barata JT, Cardoso AA, Nadler LM, et al. Interleukin-7 promotes survival and cell cycle progression of T-cell acute lymphoblastic leukemia cells by down-regulating the cyclin-dependent kinase inhibitor p27(kip1). Blood. 2001;98(5):1524-1531. 6. Gonzalez-Garcia S, Garcia-Peydro M, Martin-Gayo E, et al. CSL-MAML-dependent Notch1 signaling controls T lineagespecific IL-7R{alpha} gene expression in early human thymopoiesis and leukemia. J Exp Med. 2009;206(4):779-791. 7. Goossens S, Radaelli E, Blanchet O, et al.

ZEB2 drives immature T-cell lymphoblastic leukaemia development via enhanced tumour-initiating potential and IL-7 receptor signalling. Nat Commun. 2015;6:5794. 8. Rich BE, Campos-Torres J, Tepper RI, et al. Cutaneous lymphoproliferation and lymphomas in interleukin 7 transgenic mice. J Exp Med. 1993;177(2):305-316. 9. Touw I, Pouwels K, van Agthoven T, et al. Interleukin-7 is a growth factor of precursor B and T acute lymphoblastic leukemia. Blood. 1990;75(11):2097-2101. 10. Scupoli MT, Perbellini O, Krampera M, et al. Interleukin 7 requirement for survival of Tcell acute lymphoblastic leukemia and human thymocytes on bone marrow stro-

haematologica | 2016; 101(11)

CK2 is essential for IL-7 effects in T-ALL

ma. Haematologica. 2007;92(2):264-266. 11. Silva A, Laranjeira AB, Martins LR, et al. IL7 contributes to the progression of human Tcell acute lymphoblastic leukemias. Cancer Res. 2011;71(14):4780-4789. 12. Shochat C, Tal N, Bandapalli OR, et al. Gainof-function mutations in interleukin-7 receptor-alpha (IL7R) in childhood acute lymphoblastic leukemias. J Exp Med. 2011;208 (5):901-908. 13. Zenatti PP, Ribeiro D, Li W, et al. Oncogenic IL7R gain-of-function mutations in childhood T-cell acute lymphoblastic leukemia. Nat Genet. 2011;43(10):932-939. 14. Treanor LM, Zhou S, Janke L, et al. Interleukin-7 receptor mutants initiate early T cell precursor leukemia in murine thymocyte progenitors with multipotent potential. J Exp Med. 2014;211(4):701-713. 15. Vicente C, Schwab C, Broux M, et al. Targeted sequencing identifies associations between IL7R-JAK mutations and epigenetic modulators in T-cell acute lymphoblastic leukemia. Haematologica. 2015;100(10): 1301-1310. 16. Yokoyama K, Yokoyama N, Izawa K, et al. In vivo leukemogenic potential of an interleukin 7 receptor alpha chain mutant in hematopoietic stem and progenitor cells. Blood. 2013;122(26):4259-4263. 17. Piazza F, Manni S, Ruzzene M, et al. Protein kinase CK2 in hematologic malignancies: reliance on a pivotal cell survival regulator by oncogenic signaling pathways. Leukemia. 2012;26(6):1174-1179. 18. Duncan JS, Litchfield DW. Too much of a good thing: the role of protein kinase CK2 in tumorigenesis and prospects for therapeutic inhibition of CK2. Biochim Biophys Acta. 2008;1784(1):33-47. 19. Martins LR, Lucio P, Silva MC, et al. Targeting CK2 overexpression and hyperactivation as a novel therapeutic tool in chronic lymphocytic leukemia. Blood. 2010; 116(15):2724-2731. 20. Kim JS, Eom JI, Cheong JW, et al. Protein kinase CK2alpha as an unfavorable prognostic marker and novel therapeutic target in acute myeloid leukemia. Clin Cancer Res. 2007;13(3):1019-1028. 21. Piazza FA, Ruzzene M, Gurrieri C, et al. Multiple myeloma cell survival relies on high activity of protein kinase CK2. Blood. 2006;108(5):1698-1707. 22. Song C, Gowda C, Pan X, et al. Targeting casein kinase II restores Ikaros tumor suppressor activity and demonstrates therapeutic efficacy in high-risk leukemia. Blood. 2015;126(15):1813-1822. 23. Silva A, Yunes JA, Cardoso BA, et al. PTEN posttranslational inactivation and hyperactivation of the PI3K/Akt pathway sustain primary T cell leukemia viability. J Clin Invest. 2008;118(11):3762-3774.

haematologica | 2016; 101(11)

24. Gutierrez A, Sanda T, Grebliunaite R, et al. High frequency of PTEN, PI3K, and AKT abnormalities in T-cell acute lymphoblastic leukemia. Blood. 2009;114(3):647-650. 25. Zheng Y, Qin H, Frank SJ, et al. A CK2dependent mechanism for activation of the JAK-STAT signaling pathway. Blood. 2011;118(1):156-166. 26. Ribeiro D, Melao A, Barata JT. IL-7R-mediated signaling in T-cell acute lymphoblastic leukemia. Adv Biol Regul. 2013;53(2):211222. 27. Meggio F, Pinna LA. One-thousand-and-one substrates of protein kinase CK2? FASEB J. 2003;17(3):349-368. 28. Hauck L, Harms C, An J, et al. Protein kinase CK2 links extracellular growth factor signaling with the control of p27(Kip1) stability in the heart. Nat Med. 2008;14(3):315-324. 29. Ji H, Wang J, Nika H, et al. EGF-induced ERK activation promotes CK2-mediated disassociation of alpha-Catenin from beta-Catenin and transactivation of beta-Catenin. Mol Cell. 2009;36(4):547-559. 30. Bene MC, Castoldi G, Knapp W, et al. Proposals for the immunological classification of acute leukemias. European Group for the Immunological Characterization of Leukemias (EGIL). Leukemia. 1995; 9(10):1783-1786. 31. Barata JT, Boussiotis VA, Yunes JA, et al. IL7-dependent human leukemia T-cell line as a valuable tool for drug discovery in T-ALL. Blood. 2004;103(5):1891-1900. 32. Vilk G, Saulnier RB, St Pierre R, et al. Inducible expression of protein kinase CK2 in mammalian cells. Evidence for functional specialization of CK2 isoforms. J Biol Chem. 1999;274(20):14406-14414. 33. Silva A, Jotta PY, Silveira AB, et al. Regulation of PTEN by CK2 and Notch1 in primary T-cell acute lymphoblastic leukemia: rationale for combined use of CK2- and gamma-secretase inhibitors. Haematologica. 2010;95(4):674-678. 34. Barata JT, Silva A, Brandao JG, et al. Activation of PI3K is indispensable for interleukin 7-mediated viability, proliferation, glucose use, and growth of T cell acute lymphoblastic leukemia cells. J Exp Med. 2004;200(5):659-669. 35. Buontempo F, Orsini E, Martins LR, et al. Cytotoxic activity of the casein kinase 2 inhibitor CX-4945 against T-cell acute lymphoblastic leukemia: targeting the unfolded protein response signaling. Leukemia. 2014;28(3):543-553. 36. Mackall CL, Fry TJ, Gress RE. Harnessing the biology of IL-7 for therapeutic application. Nat Rev Immunol. 2011;11(5):330-342. 37. Di Maira G, Salvi M, Arrigoni G, et al. Protein kinase CK2 phosphorylates and upregulates Akt/PKB. Cell Death Differ. 2005;12(6):668-677.

38. Abraham N, Ma MC, Snow JW, et al. Haploinsufficiency identifies STAT5 as a modifier of IL-7-induced lymphomas. Oncogene. 2005;24(33):5252-5257. 39. Karawajew L, Ruppert V, Wuchter C, et al. Inhibition of in vitro spontaneous apoptosis by IL-7 correlates with bcl-2 up-regulation, cortical/mature immunophenotype, and better early cytoreduction of childhood T-cell acute lymphoblastic leukemia. Blood. 2000;96(1):297-306. 40. 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. 41. Nguyen le XT, Mitchell BS. Akt activation enhances ribosomal RNA synthesis through casein kinase II and TIF-IA. Proc Natl Acad Sci USA. 2013;110(51):2068120686. 42. Laplante M, Sabatini DM. mTOR signaling in growth control and disease. Cell. 2012; 149(2):274-293. 43. Brown VI, Fang J, Alcorn K, et al. Rapamycin is active against B-precursor leukemia in vitro and in vivo, an effect that is modulated by IL-7-mediated signaling. Proc Natl Acad Sci USA. 2003; 100(25):15113-15118. 44. Silva A, Girio A, Cebola I, et al. Intracellular reactive oxygen species are essential for PI3K/Akt/mTOR-dependent IL-7-mediated viability of T-cell acute lymphoblastic leukemia cells. Leukemia. 2011;25(6):960967. 45. Marschke RF, Borad MJ, McFarland RW, et al. Findings from the phase I clinical trials of CX-4945, an orally available inhibitor of CK2. 2011 ASCO Annual Meeting. May 20 Supplement ed. Boston. J Clin Oncol. 2011:3087. 46. Cardoso BA, Martins LR, Santos CI, et al. Interleukin-4 stimulates proliferation and growth of T-cell acute lymphoblastic leukemia cells by activating mTOR signaling. Leukemia. 2009;23(1):206-208. 47. Pallard C, Stegmann AP, van Kleffens T, et al. Distinct roles of the phosphatidylinositol 3kinase and STAT5 pathways in IL-7-mediated development of human thymocyte precursors. Immunity. 1999;10(5):525-535. 48. Wofford JA, Wieman HL, Jacobs SR, et al. IL-7 promotes Glut1 trafficking and glucose uptake via STAT5-mediated activation of Akt to support T-cell survival. Blood. 2008; 111(4):2101-2111. 49. Galovic M, Xu D, Areces LB, et al. Interplay between N-WASP and CK2 optimizes clathrin-mediated endocytosis of EGFR. J Cell Sci. 2011;124(Pt 12):2001-2012. 50. Henriques CM, Rino J, Nibbs RJ, et al. IL-7 induces rapid clathrin-mediated internalization and JAK3-dependent degradation of IL-7Ralpha in T cells. Blood. 2010; 115(16): 3269-3277.

1379

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Non-Hodgkin Lymphoma

Ferrata Storti Foundation

Haematologica 2016 Volume 101(11):1380-1389

Alterations of microRNA and microRNA-regulated messenger RNA expression in germinal center B-cell lymphomas determined by integrative sequencing analysis Kebria Hezaveh,1,* Andreas Kloetgen,1,2* Stephan H Bernhart,3,4,5* Kunal Das Mahapatra,1 Dido Lenze,6 Julia Richter,7 Andrea Haake,7 Anke K Bergmann,7 Benedikt Brors,8,9,10 Birgit Burkhardt,11 Alexander Claviez,12 Hans G Drexler,13 Roland Eils,14,15 Siegfried Haas,16 Steve Hoffmann,3,4 Dennis Karsch,17 Wolfram Klapper,18 Kortine Kleinheinz,14 Jan Korbel,19 Helene Kretzmer,3,4 Markus Kreuz,20 Ralf Küppers,21 Chris Lawerenz,14 Ellen Leich,22 Markus Loeffler,20 Luisa Mantovani-Loeffler,23 Cristina López,7 Alice C McHardy,2,24 Peter Möller,25 Marius Rohde,26 Philip Rosenstiel,27 Andreas Rosenwald,22 Markus Schilhabel,27 Matthias Schlesner,14 Ingrid Scholz,14 Peter F Stadler,3,4,5,28,29,30 Stephan Stilgenbauer,31 Stéphanie Sungalee,19 Monika Szczepanowski,18 Lorenz Trümper,32 Marc A Weniger,21 Reiner Siebert,7§ Arndt Borkhardt,1§ Michael Hummel,6,§ and Jessica I. Hoell,§,1 on behalf of the ICGC MMML-Seq Project‡

Department of Pediatric Oncology, Hematology and Clinical Immunology, Heinrich-HeineUniversity, Medical Faculty, Düsseldorf, Germany; 2Department of Algorithmic Bioinformatics, Heinrich-Heine University, Duesseldorf, Germany; 3Transcriptome Bioinformatics Group, LIFE Research Center for Civilization Diseases, University of Leipzig, Germany; 4Bioinformatics Group, Department of Computer Science, University of Leipzig, Germany; 5Interdisciplinary Center for Bioinformatics, University of Leipzig, Germany; 6 Institute of Pathology, Charité – University Medicine Berlin, Germany; 7Institute of Human Genetics, University Hospital Schleswig-Holstein Campus Kiel/ Christian-Albrechts University Kiel, Germany; 8Division Applied Bioinformatics, German Cancer Research Center (DKFZ), Heidelberg, Germany; 9National Center for Tumor Diseases (NCT), Heidelberg, Germany; 10German Cancer Consortium (DKTK), Heidelberg, Germany; 11 Department of Pediatric Hematology and Oncology, University Hospital Münster, Germany; 12 Department of Pediatrics, University Hospital Schleswig-Holstein, Campus Kiel, Germany; 13 Department of Human and Animal Cell Cultures, German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany; 14Division of Theoretical Bioinformatics (B080), German Cancer Research Center (DKFZ), Heidelberg, Germany; 15Department of Bioinformatics and Functional Genomics, Institute for Pharmacy and Molecular Biotechnology and Bioquant, Heidelberg University, Germany; 16Friedrich-Ebert Hospital Neumünster, Clinics for Hematology, Oncology and Nephrology, Neumünster, Germany; 17 Department of Internal Medicine II: Hematology and Oncology, University Medical Centre, Campus Kiel, Germany; 18Hematopathology Section, University Hospital Schleswig-Holstein Campus Kiel/ Christian-Albrechts University Kiel, Germany; 19EMBL Heidelberg, Genome Biology, Heidelberg, Germany; 20Institute for Medical Informatics Statistics and Epidemiology, Leipzig, Germany; 21Institute of Cell Biology (Cancer Research), University of Duisburg-Essen, Essen, Germany; 22Institute of Pathology, University of Würzburg, and Comprehensive Cancer Center Mainfranken, Würzburg, Germany; 23Hospital of Internal Medicine II, Hematology and Oncology, St-Georg Hospital Leipzig, Germany; 24Computational Biology of Infection Research, Helmholtz Center for Infection Research, Braunschweig, Germany 25Institute of Pathology, Medical Faculty of the Ulm University, Germany; 26 Department of Pediatric Hematology and Oncology University Hospital Giessen, Germany; 27 Institute of Clinical Molecular Biology, University Hospital Schleswig-Holstein Campus Kiel/ Christian-Albrechts University Kiel, Germany; 28RNomics Group, Fraunhofer Institute for Cell Therapy and Immunology IZI, Leipzig, Germany; 29Max-Planck-Institute for Mathematics in Sciences, Leipzig, Germany; 30Santa Fe Institute, NM, USA; 31Department of Internal Medicine III, University of Ulm, Germany; and 32Department of Hematology and Oncology, Georg-August-University of Göttingen, Germany; 1

Correspondence: jessica.hoell@med.uni-duesseldorf.de

Received: February 4, 2016. Accepted: July 1, 2016. Pre-published: July 6, 2016. doi:10.3324/haematol.2016.143891

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1380

1380

*These authors contributed equally to this work. ‡A full list of members of the ICGC MMML-Seq Project and their affiliations appears in the Online Supplementary Information.

ABSTRACT

icroRNA are well-established players in post-transcriptional gene regulation. However, information on the effects of microRNA deregulation mainly relies on bioinformatic prediction of potential targets, whereas proof of the direct physical microRNA/target messenger RNA interaction is mostly lacking. Within the International Cancer Genome Consortium Project “Determining haematologica | 2016; 101(11)

miRNA-mRNA target regulation in B-cell lymphomas

Molecular Mechanisms in Malignant Lymphoma by Sequencing”, we performed miRnome sequencing from 16 Burkitt lymphomas, 19 diffuse large B-cell lymphomas, and 21 follicular lymphomas. Twentytwo miRNA separated Burkitt lymphomas from diffuse large B-cell lymphomas/follicular lymphomas, of which 13 have shown regulation by MYC. Moreover, we found expression of three hitherto unreported microRNA. Additionally, we detected recurrent mutations of hsa-miR-142 in diffuse large B-cell lymphomas and follicular lymphomas, and editing of the hsa-miR-376 cluster, providing evidence for microRNA editing in lymphomagenesis. To interrogate the direct physical interactions of microRNA with messenger RNA, we performed Argonaute-2 photoactivatable ribonucleoside-enhanced cross-linking and immunoprecipitation experiments. MicroRNA directly targeted 208 messsenger RNA in the Burkitt lymphomas and 328 messenger RNA in the non-Burkitt lymphoma models. This integrative analysis discovered several regulatory pathways of relevance in lymphomagenesis including Ras, PI3KAkt and MAPK signaling pathways, also recurrently deregulated in lymphomas by mutations. Our dataset reveals that messenger RNA deregulation through microRNA is a highly relevant mechanism in lymphomagenesis.

Introduction B-cell lymphomas account for approximately 85% of all lymphomas and form a heterogeneous group of lymphoid neoplasms arising at different stages of B-cell development.1 They are classified according to morphological and immunophenotypic features, supplemented by characteristic genomic translocations (WHO 2008). Although these features allow the diagnosis of different histological subtypes of B-cell lymphomas, molecular subtypes remain largely indistinguishable.2 Presumably due to this molecular heterogeneity, many patients do not respond well to common therapy regimens.3 New biomarkers and therapeutic targets need, therefore, to be identified in order to improve the accuracy of lymphoma diagnosis and subsequent therapy selection. One potential class of biomarkers and/or therapeutic targets is a subset of RNA molecules named microRNA (miRNA). These are small non-coding RNA (17–25 nucleotides in length) that bind mostly to target sequences within the 3’ untranslated region of messenger RNA (mRNA). MiRNA regulate the expression of thousands of mRNA including those with key roles in cell differentiation and cancer pathogenesis.4 MiRNA influence immune cell differentiation and play crucial roles in both early and late B-cell differentiation5 and lymphomagenesis.6 Mechanisms of miRNA dysregulation in lymphomas include copy number alterations (e.g. the miR-17~92 polycistron7), chromosomal translocation (e.g. hsa-miRNA-1258) and mutations (e.g. hsa-miR-1429). Several molecular profiling studies have tried to assess differential miRNA expression in Bcell lymphomas, as recently described.5,6,10 It has been reported that a signature of 38 miRNA containing MYCregulated and nuclear factor-κB pathway-associated miRNA differentiates Burkitt lymphoma (BL) from diffuse large B-cell lymphoma (DLBCL).11 Available data on miRNA expression profiling in B-cell lymphomas is, however, still preliminary, as published profiles are either mostly not derived from large collections of samples, do not compare subtypes or originate from either quantitative real time polymerase chain reaction (qRT-PCR)-based approaches or microarrays. Nextgeneration sequencing is able to overcome the disadvantages of previous methods such as probe cross-hybridization12 and the limitations of qRT-PCR, such as restricting haematologica | 2016; 101(11)

the analysis to previously known miRNA. Furthermore, sequencing-based approaches allow for the discovery of novel miRNA and large-scale identification of mutated miRNA. The present study was performed within the framework of the International Cancer Genome Consortium Project “Determining Molecular Mechanisms in Malignant Lymphoma by Sequencing” (ICGC MMMLSeq). Our aim was to identify, using next-generation sequencing, miRNA signatures in three common subtypes of B-cell lymphomas, BL, DLBCL and follicular lymphoma (FL), and to correlate these to mRNA expression and genomic mutations. Moreover, by performing photoactivatable ribonucleoside-enhanced cross-linking and immunoprecipitation (PAR-CLIP) experiments13 and intersecting the results with the patient-derived mRNA/miRNA expression profiles, we aimed at identifying specific miRNA-mRNA target pairs in BL and DLBCL.

Methods Patients’ samples The ICGC MMML-Seq project was approved by the Institutional Review Board of the Medical Faculty of Kiel University (A150/10) and by the recruiting centers. Informed consent was obtained from all patients (or, in the case of children, from their legal guardians). Histopathological, immunophenotypic and genetic characterization of the tumor samples and initial diagnosis (tumor cell content ≥60%) was performed as described recently.14

Next-generation sequencing Nucleic acids were extracted as previously detailed.14 Libraries for miRNA sequencing were prepared using TruSeq Small RNA sample prep kits (Illumina, San Diego, CA, USA) according to the manufacturer’s instructions. with 100 ng - 1 µg total RNA as input. Libraries were size-fractionated on 6 x TBE gels (Life Technologies, Carlsbad, CA, USA). DNA concentration and sizes were analyzed on a 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA). Seven pmol of DNA of each library were loaded onto a flow cell (multiplexing up to four libraries per lane), 50-cycle sequencing was performed using the TruSeq SBS Kit v3 on the HiSeq 2500 (Illumina). Whole genome sequencing data of tumors and matched con1381

K. Hezaveh et al.

trols and transcriptome sequencing data of tumors were generated by the ICGC MMML-Seq project as previously described.14 All sequencing data have been deposited at the European Genomephenome Archive (EGA, http://www.ebi.ac.uk/ega/, accession number EGAS00001001394).

Quantitative real-time polymerase chain reactions Undiluted reverse transcriptase reactions (20 ng of RNA per sample) were combined with TaqMan Universal Master Mix II (no UNG) (Life Technologies) and amplified (7500HT Real-Time PCR System, Life Technologies) with RNU24 and RNU48 as housekeeping genes. Experiments were performed in triplicate and analyzed using the 2-DDCT method.

AGO2 immunoprecipitation AGO2-PAR-CLIP was carried out as previously described13 with modifications (mostly relating to the washing steps during immunoprecipitation) because of the use of monoclonal antiAGO2 antibody (#4-642, EMD Millipore, Billerica, MA, USA).15 In brief, following the addition of 4-thiouridine, an immunoprecipitation using a monoclonal anti-AGO2 antibody isolated the RNAprotein complexes. After protein digestion, sequencing adapters were ligated to the purified RNA fragments. Following reverse transcription, PAR-CLIP libraries were sequenced on a HiSeq2500 (Illumina).16

Bioinformatic methods Genome and transcriptome analysis Bioinformatic analyses of the genome and transcriptome data were performed as described recently, employing the various pipelines established in the ICGC MMML-Seq14 (additional information is provided in the Online Supplementary Methods).

MicroRNA and immunoprecipitation analysis Following adapter removal, reads were mapped onto the human genome (1000 genomes project, hs37d5100) using segemehl.17 Novel miRNA prediction was performed using miRanalyzer 0.318 (default parameters), target prediction using miRanda19 (miRsvr-score < -1.2). After filtering and trimming the PAR-CLIP reads, we obtained a total of 62,281,382 single-end reads, which were aligned with BWA20 with up to two mismatches between a read sequence and the reference sequence (hg19). All reads failing this mapping were aligned against the transcriptome database (Ensembl Genes 75). Aligned reads were piled into clusters by PARA-suite (Kloetgen et al., submitted). As PAR-CLIP reads contain thymidine to cytidine (T-C) conversions at the sites of crosslinks, all identified clusters were filtered to receive the most confident target regions. Excluding clusters containing <5 reads and <25% T-C conversions (excluding 100% T-C conversion sites as these might result from single nucleotide variants) resulted in (prior to pooling) 1,329 clusters for SU-DHL-4, 1,517 clusters for SU-DHL-6, 1,209 clusters for Namalwa and 425 clusters for Raji. Further details (including miRNA-mRNA correlation analyses) are given in the Online Supplementary Methods.

Results Molecular classification of Burkitt lymphomas versus diffuse large B-cell and follicular lymphomas using a 25 miRNA classifier We profiled tumor samples from 56 patients including 16 with BL (based on a molecular classifier; all patients ≤18 1382

years), 19 with DLBCL (including 7 with germinal center DLBCL, 10 activated B-cell DLBCL and 2 with type III DLBCL) and 21 with FL (mainly grade 1/2) (Online Supplementary Table S1). We obtained 1,169,752,727 sequencing reads in total (average of 20,888,442 reads per sample, Online Supplementary Table S2). Following normalization of miRNA reads, we performed an unsupervised hierarchical clustering. Unexpectedly (and differently to what we observed at the transcriptomic level, data not shown), no clear distinction between BL, DLBCL and FL was achieved based on miRNA expression profiles (Online Supplementary Figure S1A). We then ranked the miRNA by mean expression and, discarding those that showed little expression variability, chose the top ten miRNA for validating our next-generation sequencing data by qRT-PCR. A correlation analysis showed the consistency of miRNA expression levels regardless of the method of quantification employed [Spearman’s rank correlation test, 10/10, high correlation (R>0.7), P-values for the correlation between qRT-PCR expression and next-generation sequencing expression ≤0.05 in 7/10 cases; details on all Pvalue calculations are given in the Online Supplementary Bioinformatic Methods)] (Online Supplementary Figure S1B, Online Supplementary Table S3). To recognize subtler molecular differences that escape unsupervised clustering approaches, we performed a differential gene expression analysis between BL versus DLBCL, BL versus FL and DLBCL versus FL using edgeR (Online Supplementary Table S4 and Online Supplementary Bioinformatic Methods). Clustering of the top 25 differentially expressed miRNA between each two lymphoma subtypes (BL/DLBCL, BL/FL, and FL/DLBCL) revealed separation according to the subtypes (Figure 1A). Employing this approach, BL and FL separated clearly, whereas the discrimination between BL/DLBCL and FL/DLBCL was less pronounced, most likely due to the molecular heterogeneity of DLBCL.21,22 As there were no patients with dual-hit lymphomas and no DLBCL cases with MYC breaks as single events in our cohort, we were not able to test, whether our classifier was able to single out those cases. Interestingly, 7/25 miRNA differentially expressed between BL/DLBCL (hsa-miRs23a/29b/130b/146a/155/196b/222) were also part of a recently published, 27-miRNA qRT-PCR-derived classifier for the differentiation of those two subtypes22 (6.9-fold enrichment, one-sided Fisher exact test, P-value for the overlap of the two classifiers 2.322x10-5). In a previous array-based study, we established a classifier consisting of 38 miRNA, which differentiated BL from DLBCL.11 From the top 25 miRNA differentially expressed herein, eight overlapped with those 38 miRNA (hsa-miRs23a/29b/146a/155/193a/221/222/339) (5.1-fold, P-value for this overlap 5.900x10-5). In summary, five miRNA (hsamiR-23a/29b/146a/155/222) seem to be robustly able to differentiate BL from DLBCL irrespective of the collection of cases and the method used for analysis. We additionally analyzed previously published microarray data11 for 64 BL cases and 86 DLBCL cases to validate the predictive power of our classifiers on an independent dataset (see the Online Supplementary Methods for further details). We predicted correct class labels for 126/128 cases with a majority vote of at least 80% (recall = 98.44%; 57/58 BL cases and 69/70 DLBCL cases; overall accuracy = 84.0%) on our 25miRNA-classifier for BL versus DLBCL. To address the question of how to distinguish BL from haematologica | 2016; 101(11)

miRNA-mRNA target regulation in B-cell lymphomas

the other investigated histological subtypes, we merged DLBCL and FL and clustered the top 25 differentially expressed miRNA between BL and DLBCL/FL inferred with edgeR. This resulted - with the exception of two BL cases - in a clear separation between BL and DLBCL/FL (Figure 1B). Of those top 25 differentially expressed miRNA, 14 were upregulated and 11 were downregulated in BL compared to DLBCL/FL (Table 1). As our analysis takes both “5p” and “3p” versions (previously referred to as mature miRNA and star strand) of each miRNA into

account, our classifier consists of 22 unique miRNA. Interestingly, for a total of 13 of these miRNA, regulation by MYC was reported in the literature.23-28

Hsa-miRNA-143 is highly abundant in germinal center B-cell lymphomas Contrary to earlier reports,29 hsa-miR-143 showed a very high expression across most lymphoma samples (Figure 1C). Expression ranged from 0.8% to 68.2% (mean 8.9%) of all reads mapping to miRNA for this miRNA alone with

Figure 1. The miRnome of B-cell lymphomas. (A) Clustering according to the top 25 differentially expressed miRNA inferred with edgeR between FL (light blue), DLBCL (blue), and BL (gray), in pairwise comparisons. (B) Clustering according to the top 25 differentially expressed miRNA inferred with edgeR between BL and DLBCL/FL. (C) Hsa-miR-143 expression across all patients’ samples. (D) Visualization of the genomic mutations of those miRNA, which show alterations in their mature sequences. The mature sequences of the respective miRNA are shown; the seed sequences are highlighted by black boxes. The positions of the mutations are also indicated. (E) Predicted folding of the three biochemically validated novel miRNA.

haematologica | 2016; 101(11)

1383

K. Hezaveh et al.

no significant differences between subtypes (means 10.8%, 7.4% and 8.9% for BL, FL and DLBCL, respectively). The extremely high expression of this miRNA (68.2%) in patient 4146289 (BL) was confirmed by qRT-PCR, as was the lower expression (0.8%) in patient 4142267 (BL) (Online Supplementary Figure S1B). As hsa-miR-143 forms a bicistronic cluster on chromosomal region 5q33.1 with hsa-miR-145, we also investigated the latter’s expression. MiRNA in bicistronic clusters are transcribed simultaneously and thus show similar expression patterns. The correlation analysis (P-value 0.0034 for the correlation between hsa-miR-143 and hsa-miR-145 expression, R=0.39) confirmed the validity of the hsa-miR-143 expression with similar expression patterns (Online Supplementary Figure S1C). Whole-genome-derived copy number analysis of all patients’ samples did not reveal any relevant alterations in either the promoter or the genomic region of hsamiR-143/145. The reason for the observed high expression of the hsa-miR-143/145 cluster thus remained unclear. To identify molecular pathways associated with the high expression of hsa-miR-143, we performed a target prediction and investigated which of the predicted targets were downregulated in the respective RNASeq data. This resulted in 186 predicted hsa-miR-143/mRNA interaction pairs (Online Supplementary Table S5). Gene Ontology (GO) analysis employing Gorilla30 revealed that the GO term “ubiquitin-protein transferase activity” (GO:0004842) showed the highest enrichment (5.33-fold, P-value <0.001). The associated target genes are listed in Online Supplementary Table S6, the entire GO output in Online Supplementary Table S7.

Hsa-miR-142 is recurrently mutated in its mature sequence in diffuse large B-cell lymphomas and follicular lymphomas Next, we searched for mutated miRNA, which were detectable at both DNA and RNA levels. Mutations in miRNA located in the IGH gene locus were excluded. We identified ten mutations (Table 2) in eight patients (6 mutations in 5 DLBCL patients, 4 mutations in 3 FL patients) with a total of four miRNA affected (hsa-miR142/-612/-3655/-4322). In two miRNA (hsa-miR-142/612), the mutations were within the mature sequence (Figure 1D). Hsa-miR-142 was the most frequently mutated miRNA with six different mutations in 5/40 DLBCL/FL patients. Two of those were located within the seed sequence. Looking at the subgroups, this broke up into 3/19 in DLBCL and 2/21 in FL. A recent publication9 reported mutation of hsa-miR-142 in 11/56 DLBCL cases. Our data therefore confirm the mutation frequency in DLBCL and extend this finding to FL.

The hsa-miR-376 cluster is recurrently edited in germinal center B-cell lymphoma subtypes RNA editing is a process in which (most commonly) adenosine deaminases perform the site-specific hydrolytic deamination of adenosine to inosine.31 When an RNA molecule contains an inosine, the sequence change is usually A-to-G. We searched for mutations exclusive to the miRNA data (not seen at the genomic level), which thus represented bona fide miRNA editing events. Starting with all single nucleotide variants, we restricted our search to those in the seed regions and discarded known single nucleotide variants as reported in dbSNP_135 including 1384

Table 1. The 25-miRNA classifier separating BL from DLBCL/FL

miRNA

P value

FDR

cpm BL

cpm DLBCL/FL

hsa-miR-17-3p hsa-miR-18a-3p hsa-miR-19a-3p hsa-miR-20a-3p hsa-miR-25-5p hsa-miR-29c-5p hsa-miR-93-3p hsa-miR-106b-3p hsa-miR-106b-5p hsa-miR-130b-3p hsa-miR-150-3p hsa-miR-150-5p hsa-miR-155-5p hsa-miR-184 hsa-miR-196b-5p hsa-miR-151b hsa-miR-211-5p hsa-miR-221-3p hsa-miR-296-3p hsa-miR-335-3p hsa-miR-339-5p hsa-miR-664-3p hsa-miR-664-5p hsa-miR-573 hsa-miR-4521

1.4 E-14 4.8 E-12 5.7 E-12 7.2 E-28 1.7 E-21 6.6 E-12 4.7 E-10 7.6 E-11 3.9 E-11 1.1 E-18 9.3 E-12 8.4 E-13 3.2 E-10 1.6 E-10 2.9 E-10 6.1 E-11 1.4 E-11 9.6 E-15 2.1 E-12 1.9 E-11 1.9 E-13 1.1 E-10 4.1 E-10 5.3 E-11 1.9 E-13

1.6 E-12 2.7 E-10 3.0 E-10 4.1 E-25 5.0 E-19 3.1 E-10 1.1 E-08 2.3 E-09 1.4 E-09 2.1 E-16 4.1 E-10 6.0 E-11 7.9 E-09 4.4 E-09 7.6 E-09 1.9 E-09 5.9 E-10 1.4 E-12 1.3 E-10 7.2 E-10 1.6 E-11 3.1 E-09 9.8 E-09 1.8 E-09 1.6 E-11

2177.0 88.9 1852.6 28.9 79.2 12.6 153.4 1010.8 617.7 701.2 3.5 649.6 1152.3 0.9 4.5 15.5 0.1 461.7 6.6 566.5 98.8 14.7 4.2 3.3 35.6

279.2 16.1 349.2 3.9 11.2 51.0 26.0 381.5 211.0 148.6 36.5 7980.4 10989.1 123.6 53.7 146.0 1.7 3018.5 1.7 100.8 20.3 97.9 29.1 0.4 4.6

MiRNA for which regulation by MYC has been shown are in bold. cpm indicates counts per million; FDR, false discovery rate.

rare variants. The remaining 40 candidates were manually evaluated (correct position of single nucleotide variants in sequence reads, A-to-G change, sequencing quality of errors), narrowing the list to four single nucleotide variants (Table 3). These mapped to hsa-miR-1260b, hsa-miR376a1/2, and hsa-miR-376c, with the hsa-miR-376 family belonging to the same genomic cluster on 14q32. Editing frequencies (edited reads versus all reads) ranged from 3586% across miRNA in the lymphoma samples showing this phenomenon. The editing “efficiency” (percent alternative base) and the expression of ADAR, one of the main enzymes responsible for RNA editing,31 per case (with observed editing) showed a weak correlation (P-value 0.044; R=0.30), possibly pointing to the mechanism behind the observed miRNA editing.

Discovery of three hitherto unreported microRNA expressed in germinal center B-cell lymphomas We employed miRanalyzer to predict hitherto unreported miRNA,18 then choosing a subset of 20 (Online Supplementary Table S8), and observed the correct processing of three candidates (Table 4) by northern blotting (Online Supplementary Figure S2A). Secondary structures of these three hitherto unreported miRNA as predicted by RNAfold32 are depicted in Figure 1E. Novel-miR-1 was moderately expressed in SU-DHL-4 and weakly expressed in Namalwa and Raji. Novel-miR-2 haematologica | 2016; 101(11)

miRNA-mRNA target regulation in B-cell lymphomas

was expressed in Raji and SU-DHL-4, and novel-miR-3 was expressed in Raji and Namalwa (Online Supplementary Figure S2A). We next assessed publicly available RNASeq data of 16 cell lines (details in the Online Supplementary Material) across a variety of tissues/diseases for expression of our three novel miRNA. Novel-miR-2 and novel-miR-3 were broadly expressed (in 16/16 cell lines and 12/16 cell lines, respectively), whereas novel-miR-1 showed restricted expression and was only detected in the B-lymphoblastoid cell line GM12878 (data not shown). We then focused on novel-miR-1 (restricted expression) and novel-miR-2 (broad expression) for further experiments. We performed overexpression/knockdown studies in SU-DHL-4 (novel-miR-1) and Raji (novel-miR-2) followed by RNASeq (Online Supplementary Figure S2B). In order to identify only those mRNA whose differential expression was due to direct targeting effects, we searched for mRNA that carried the respective seed sequence, had a significant miRanda score and were inversely regulated (false discovery rate for all further calculations <0.05). Downregulation of novel-miR-1 and novel-miR-2 resulted in two (EIF3C, MPEG1) and three (HLA-DRB5, PFKFB4, PPP1R35) upregulated mRNA, respectively. Upregulation of novel-miR-1 led to the downregulation of 55 coding mRNA (Online Supplementary Table S9), whereas

overexpression of novel-miR-2 only resulted in two downregulated mRNA (SLCO2B1, UPP1). Interestingly, there were many genes previously reported in the context of lymphomagenesis among those mRNA, which carried novel-miR-1 seed sequences. These genes represent its bona fide direct targets and included CARD11, E2F1, MCM2 and MCM7. Novel-miR-1 thus potentially represents a new player in lymphomagenesis. Sequences of novel-miR-1/-2/-3 have been submitted to miRBase.

AGO2 immunoprecipitation identifies direct mRNA-miRNA interactions in lymphoma subtypes To identify those mRNA that were physically targeted by miRNA in Argonaute-miRNA-mRNA complexes (rather than using bioinformatic predictions to identify putative interactions only), we performed PAR-CLIP experiments of endogenous AGO213 (Figure 2A) in two BL cell lines (Namalwa, Raji) and two non-BL cell lines (SUDHL-4, SU-DHL-6; both t(14;18)-positive). Merging the BL and the non-BL sequencing reads resulted in 1,587 and 2,532 clusters, respectively (individual read numbers are shown in Figure 2B). Combining these miRNA-target sites with the transcriptome data also available for each patient (Figure 2C) led to 302 (BL)/540 (non-BL) miRNA-mRNA interactions with negative correlations, with several genes

Table 2. Genomically mutated miRNA.

miRNA hsa-mir-142 hsa-mir-142 hsa-mir-142 hsa-mir-142 hsa-mir-142 hsa-mir-142 hsa-mir-612 hsa-mir-3655 hsa-mir-4322 hsa-mir-4322

Chromosome Genomic position (hg 19) chr17 chr17 chr17 chr17 chr17 chr17 chr11 chr5 chr19 chr19

56408624 56408616 56408630 56408620 56408621 56408612 65211962 140027478 10341090 10341109

Mature

Reference/alternative

PID

Subtype

y y n y y y y n n n

C>T A>C C>T A>T A>G A>T G>A A>G C>A C>T

4102009 4112447 4120193 4160468 4160468 4176133 4135099 4177376 4134434 4135099

DLBCL FL DLBCL FL FL DLBCL DLBCL FL DLBCL DLBCL

chr: chromosome; gen. pos.: genomic position (hg19); mature, whether or not the sequenced alteration is located within the mature miRNA sequence; PID: personal identifier.

Table 3. RNA editing events across lymphoma subtypes.

miRNA hsa-miR-376a1/2 hsa-miR-376c hsa-miR-1260b

Chromosome

Genomic position (hg 19)

Reference/alternative

Mean % alternative

N. of samples with editing

chr14 chr14 chr11

101506460 101506074 96074619

A>G A>G A>G

86.2% 45.2% 35.3%

39 19 11

The table lists the numbers of samples showing the editing events at the indicated genomic positions with at least ten sequenced reads at this position.mean % alternative: mean % of reads differing from the reference sequence.

Table 4. Novel miRNA in B-cell lymphomas.

Northern blot

Probe

Chromosome

Genomic position (hg 19)

Mature miRNA sequence

Positive (novel-miR-1) Positive (novel-miR-2) Positive (novel-miR-3)

NB-5 NB-19 NB-20

10 M 12

50035510-50035603 3363-3463 52453530-52453613

GCACACTGACACAGAGAGAGAGA CCAACGTTGTAGGCCCCTACGGGCTACT TCACTGCAGGGCCCTAGCAATA

haematologica | 2016; 101(11)

1385

K. Hezaveh et al.

being targeted by more than one miRNA (Online Supplementary Table S10). At the individual gene level the numbers were 208 (BL) and 328 (non-BL). Many of the genes showing direct regulation by miRNA have well-known roles in lymphomagenesis (Figure 2D). These genes fell into different functional categories, some for which expression was correlated to prognosis33 [B2M33 (targeted by hsa-miR-106b), MDM234 (hsa-miR-361)], for which differential expression was shown [CCR635 (hsamiR-296)] or were correlated to treatment resistance [e.g. THY136 (hsa-miR-149)]. For other targeted genes, mutations [ID314 (hsa-miR-4424), NPAT37 (hsa-miR-4518), SMARCA438 (hsa-miR-2467), TCF339 (hsa-miR-184)] or translocations [e.g. CDK640 (hsa-miR-148b)] have been

described in several types of lymphomas. Significantly enriched and lymphoma-relevant targeted KEGG pathways (Table 5) showing a differential expression between BL and non-BL included “miRNA in cancer” (hsa05206, 10 genes, P-value 7.56x10-7, enrichment 7.7), “MAPK signaling” (hsa04010, 11 genes, P-value 1.79x10-8, enrichment 9.7), “Ras signaling” (hsa04014, 8 genes, P-value 7.56x10-6, enrichment 8.0), and “PI3K-Akt signaling” (hsa04151, 8 genes, P-value 1.48x10-4, enrichment 5.2). Total numbers of genomically detected mutations in the four mentioned pathways were in that order 122 (297 genes in the pathway), 111 (257 genes), 83 (228 genes) and 124 (347 genes). As these overlaps were not statistically significant (P-values 0.221 to 0.409), this sug-

Figure 2. Direct miRNA-mRNA regulation in B-cell lymphomas. (A) PAR-CLIP principle. Following the addition of 4-thiouridine, an immunoprecipitation with subsequent protein digestion is performed. Purified RNA fragments are reverse transcribed and cDNA libraries are sequenced on a HiSeq2500 followed by bioinformatic analysis (adapted from Hafner et al.13). (B) PAR-CLIP library statistics. The left y-axis shows the number of aligned reads, the right y-axis the number of high quality PAR-CLIP clusters. The cell lines employed are indicated. (C) Flow chart of the integrative miRNA-mRNA regulation analysis (adapted from Farazi et al.15). (D) List of lymphoma relevant genes for which regulation by distinct miRNA could be elucidated.

1386

haematologica | 2016; 101(11)

miRNA-mRNA target regulation in B-cell lymphomas

gests that the respective pathways are targeted and deregulated either by virtue of miRNA interference or by mutations.

Discussion We here report a deep sequencing analysis to identify differences in miRNA expression in samples from patients with BL, FL and DLBCL collected within the ICGC MMML-Seq Consortium. Comparing our miRNA classifiers separating the three entities to previous array- and qRT-PCR based studies, five miRNA (hsa-miRs23a/29b/146a/155/222) were recurrently identified to dif-

ferentiate BL from DLBCL11,22 and two miRNA (hsa-miR92/150) to robustly separate FL from DLBCL.41,42 Of note, 13 of those miRNA differentiating BL/DLBCL were previously reported to be regulated by MYC,23-28 emphasizing the role of MYC in the pathogenesis of BL. The greater discriminative power between BL, DLBCL and FL based on unsupervised analysis of the RNA-Seq data likely comes from less variation among the patients, which might be partly due to the higher number of analyzed genes when compared to miRNA-Seq as well as overlapping effects of some miRNA. Supervised analysis based on differentially expressed miRNA did, however, have a similar discriminative power as the supervised analysis of differentially expressed mRNA.

Table 5. Targeted KEGG pathways and associated miRNA-mRNA regulation pairs.

KEGG pathway hsa05206: microRNA in cancer hsa05206: microRNA in cancer hsa05206: microRNA in cancer hsa05206: microRNA in cancer hsa05206: microRNA in cancer hsa05206: microRNA in cancer hsa05206: microRNA in cancer hsa05206: microRNA in cancer hsa05206: microRNA in cancer hsa05206: microRNA in cancer hsa04014: Ras signaling pathway hsa04014: Ras signaling pathway hsa04014: Ras signaling pathway hsa04014: Ras signaling pathway hsa04014: Ras signaling pathway hsa04014: Ras signaling pathway hsa04014: Ras signaling pathway hsa04014: Ras signaling pathway hsa04151: PI3K-Akt signaling pathway hsa04151: PI3K-Akt signaling pathway hsa04151: PI3K-Akt signaling pathway hsa04151: PI3K-Akt signaling pathway hsa04151: PI3K-Akt signaling pathway hsa04151: PI3K-Akt signaling pathway hsa04151: PI3K-Akt signaling pathway hsa04151: PI3K-Akt signaling pathway hsa04010: MAPK signaling pathway hsa04010: MAPK signaling pathway hsa04010: MAPK signaling pathway hsa04010: MAPK signaling pathway hsa04010: MAPK signaling pathway hsa04010: MAPK signaling pathway hsa04010: MAPK signaling pathway hsa04010: MAPK signaling pathway hsa04010: MAPK signaling pathway hsa04010: MAPK signaling pathway hsa04010: MAPK signaling pathway haematologica | 2016; 101(11)

Gene APC2 CCND1 E2F3 MDM2 MMP16 NOTCH4 PAK4 PDGFA PRKCB ZFPM2 FLT4 MRAS PAK1 PAK4 PAK6 PDGFA PLA2G4A PRKCB CCND1 COL6A6 FLT4 LPAR1 MDM2 PDGFA PPP2R1B PPP2R3A CACNB1 ECSIT MRAS PAK1 PDGFA PLA2G4A PPM1A PRKCB RAPGEF2 TAB1 TGFBR2

Targeting miRNA

Mutations detected

hsa-miR-663b, hsa-miR-3648 hsa-miR-27b-5p, hsa-miR-590-5p hsa-miR-141-5p hsa-miR-361-3p BL4112512 hsa-miR-151a-3p BL4190495 hsa-miR-573 FL4178655 hsa-miR-2355-5p hsa-miR-181b-3p, hsa-miR-4420 hsa-miR-577 DLBCL4131257 hsa-miR-127-5p, hsa-miR-181b-3p, hsa-miR-4420 DLBCL4134434,FL4112447 hsa-miR-17-3p hsa-miR-181b-3p, hsa-miR-1304-3p hsa-miR-424-5p hsa-miR-2355-5p hsa-miR-125a-3p DLBCL4135099 hsa-miR-181b-3p, hsa-miR-4420 hsa-miR-3940-3p hsa-miR-577 DLBCL4131257 hsa-miR-27b-5p, hsa-miR-590-5p hsa-miR-135b-5p, hsa-miR-140-3p, hsa-miR-4424, hsa-miR-4999-5p hsa-miR-17-3p hsa-miR-3194-5p, hsa-miR-3940-3p hsa-miR-361-3p BL4112512 hsa-miR-181b-3p, hsa-miR-4420 hsa-miR-140-3p BL4127766 hsa-miR-708-5p hsa-miR-3622a-5p hsa-miR-34a-5p, hsa-miR-3605-3p hsa-miR-181b-3p, hsa-miR-1304-3p hsa-miR-424-5p hsa-miR-181b-3p, hsa-miR-4420 hsa-miR-3940-3p hsa-miR-199a-3p, hsa-miR-199b-3p hsa-miR-577 DLBCL4131257 hsa-miR-641, hsa-miR-3613-3p, hsa-miR-4517 DLBCL4177376 hsa-miR-361-3p hsa-miR-4487 DLBCL4108101 1387

K. Hezaveh et al.

We identified hsa-miR-143 as highly expressed (compared to all other miRNA) across all three subtypes. This miRNA has hitherto mostly been discussed as a tumor suppressor in (mainly) epithelial malignancies.43 However, a recent study in colorectal cancer found hsa-miR-143 overexpression correlated to short overall survival.44 Earlier publications also reported a downregulation (mostly associated with its deletion) of the hsa-miR143/145 cluster in some leukemias and lymphomas.45,46 Examples of other miRNA, which have - depending on the tumor type - been described as both tumor suppressors and oncogenes include hsa-miR26a, and the hsa-miR-141/200a-cluster.4 The high expression of hsa-miR-143 raises the possibility of a new and more general role for this miRNA in lymphomagenesis. We describe recurrent mutations in hsa-miR-142 in FL at a frequency of 9.5%. Additionally, we confirm recurrent mutations of hsa-miR-142, at a frequency of 12.5% in DLBCL compared to 19.6% as previously published.9 HsamiR-142 mutations lead to the generation of new target sites as well as abolishing originally canonical ones in lymphoma-relevant genes, suggesting that hsa-miR-142 mutations act as a pathogenic mechanism across lymphoma subtypes. Other - albeit non-recurrent - seed sequence mutations affected hsa-miR-612, which was previously shown to suppress local invasion and distant colonization of hepatocellular carcinoma47 but has not been linked to lymphoid malignancies yet. RNA editing as a post-transcriptional modification is the site-specific alteration of an RNA transcript. The most frequently observed form is adenosine to inosine (A-to-I) editing, catalyzed by ADAR enzymes. Both the splicing and the translation machinery recognize inosines as guanosines. RNA editing occurs in a tissue-specific manner and increases the diversity of protein products in the case of mRNA editing. The specific deamination of miRNA affects the stability of their precursors and thus the processing efficacy as well as results in the generation of novel mRNA targets sites in addition to altering existing ones.48 The hsa-mir-376 family was previously shown to be subject to miRNA editing in different cancer types, although not yet in lymphoma. This resulted in an altered mRNA target profile with both the loss of regulation of previous targets as well as the gain of new targets.49,50 Both aspects promoted the respective cancers. We provide here evidence of miRNA editing (hsa-miR-1260b, hsa-miR376a1/2, and hsa-miR-376c) in lymphomas. Only sequencing data allow larger scale identification of novel miRNA. The current release 21 of miRBase lists 1,881 human miRNA. Similar to previous studies,10 we identified hundreds of putative novel miRNA candidates. Through northern blot experiments, we provide experimental evidence of the correct processing of three novel

References 1. Lenz G, Staudt LM. Aggressive lymphomas. N Engl J Med. 2010;362(15):1417-1429. 2. Campo E, Swerdlow SH, Harris NL, et al. The 2008 WHO classification of lymphoid neoplasms and beyond: evolving concepts and practical applications. Blood. 2011;117(19):5019-5032. 3. Sinha R, Nastoupil L, Flowers CR.

1388

miRNA. Novel-miR-1 emerged as the most interesting candidate, only being detectable in SU-DHL-4, Namalwa and a B-lymphoblastoid cell lines. Our analysis showed that it regulates many well-known lymphoma genes including CARD11, E2F1, MCM2 and MCM7, thus presenting itself as a potential novel player in lymphomagenesis. Through our integrative analysis of miRNA and mRNA profiles in patients’ samples in combination with AGO2 PAR-CLIP data, for the first time it is possible to pinpoint individual, biochemically defined miRNA/mRNA target interactions in lymphomas as well as functional consequences of miRNA dysregulation. We focused our analysis on those target pairs (208 in BL, 328 in DLBCL/FL) with consistent expression changes (presumably due to aberrant miRNA expression) in the respective patients’ RNASeq data. Just performing a correlation analysis between differentially expressed miRNA and mRNA in patients’ samples coupled with a miRanda target prediction would have resulted in a much greater number of predicted interaction pairs (2,151 predicted pairs, data not shown). We described associated regulatory pathways including “Ras signaling”, “PI3K-Akt signaling”, and “MAPK signaling”. As there was very little overlap between those mRNA that are targeted by miRNA and those genes for which genomic mutations were detected (in those pathways), we suggest miRNAmRNA targeting with subsequent deregulation as an additional oncogenic mechanism. We also provide evidence of miRNA regulation of many genes with already established roles in lymphomagenesis, including ID3, CDK6, MDM2, SMARCA4, and TCF3. Our miRNA expression profiles uncovered subtype-specific differences in miRNA expression, evidence of recurrent hsa-miR-142 mutations in FL and DLBCL as well as miRNA editing and revealed distinct miRNA/mRNA target interaction pairs with roles in lymphomagenesis. Thus, we confirm and extend the important role that miRNA play in lymphomagenesis. Acknowledgments The project was funded by the Federal Ministry of Education and Research in Germany (BMBF) within the Program for Medical Genome Research (01KU1002A through 01KU1002J). The authors also acknowledge funding from Heinrich-Heine University Duesseldorf (Forschungskommission 17/2013), the Deutsche Forschungsgemeinschaft (DFG, HO 5456/3-1) and the Duesseldorf School of Oncology (funded by the Comprehensive Cancer Center Düsseldorf/Deutsche Krebshilfe and the Medical Faculty HHU Duesseldorf). We would like to thank H. Lammert for excellent technical assistance and M. Gombert for assistance with sequencing and data handling.

Treatment strategies for patients with diffuse large B-cell lymphoma: past, present, and future. Blood Lymphat Cancer. 2012;2012(2):87-98. 4. Farazi TA, Spitzer JI, Morozov P, Tuschl T. miRNAs in human cancer. J Pathol. 2011; 223(2):102-115. 5. Di Lisio L, Martinez N, Montes-Moreno S, et al. The role of miRNAs in the pathogenesis and diagnosis of B-cell lymphomas. Blood. 2012;120(9):1782-1790.

6. Musilova K, Mraz M. MicroRNAs in B-cell lymphomas: how a complex biology gets more complex. Leukemia. 2015;29(5):10041017. 7. He L, Thomson JM, Hemann MT, et al. A microRNA polycistron as a potential human oncogene. Nature. 2005;435(7043):828-833. 8. Enomoto Y, Kitaura J, Hatakeyama K, et al. Emu/miR-125b transgenic mice develop lethal B-cell malignancies. Leukemia. 2011;25(12):1849-1856.

haematologica | 2016; 101(11)

miRNA-mRNA target regulation in B-cell lymphomas

9. Kwanhian W, Lenze D, Alles J, et al. MicroRNA-142 is mutated in about 20% of diffuse large B-cell lymphoma. Cancer Med. 2012;1(2):141-155. 10. Lim EL, Trinh DL, Scott DW, et al. Comprehensive miRNA sequence analysis reveals survival differences in diffuse large Bcell lymphoma patients. Genome Biol. 2015;16:18. 11. 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. 12. Creighton CJ, Reid JG, Gunaratne PH. Expression profiling of microRNAs by deep sequencing. Brief Bioinform. 2009;10(5):490497. 13. Hafner M, Landthaler M, Burger L, et al. Transcriptome-wide identification of RNAbinding protein and microRNA target sites by PAR-CLIP. Cell. 2010;141(1):129-141. 14. 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. 15. Farazi TA, Ten Hoeve JJ, Brown M, et al. Identification of distinct miRNA target regulation between breast cancer molecular subtypes using AGO2-PAR-CLIP and patient datasets. Genome Biol. 2014;15(1):R9. 16. Spitzer J, Hafner M, Landthaler M, et al. PAR-CLIP (Photoactivatable ribonucleosideenhanced crosslinking and immunoprecipitation): a step-by-step protocol to the transcriptome-wide identification of binding sites of RNA-binding proteins. Methods Enzymol. 2014;539:113-161. 17. Hoffmann S, Otto C, Kurtz S, et al. Fast mapping of short sequences with mismatches, insertions and deletions using index structures. PLoS Comput Biol. 2009;5(9): e1000502. 18. Hackenberg M, Sturm M, Langenberger D, Falcon-Perez JM, Aransay AM. miRanalyzer: a microRNA detection and analysis tool for next-generation sequencing experiments. Nucleic Acids Res. 2009;37(Web Server issue):W68-76. 19. Enright AJ, John B, Gaul U, et al. MicroRNA targets in Drosophila. Genome Biol. 2004; 5(1):R1. 20. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25(14):17541760. 21. Alizadeh AA, Eisen MB, Davis RE, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature. 2000;403(6769):503-511. 22. Iqbal J, Shen Y, Huang X, et al. Global

haematologica | 2016; 101(11)

23. 24. 25.

26. 27.

28.

29.

30.

31.

32. 33. 34.

35.

36.

microRNA expression profiling uncovers molecular markers for classification and prognosis in aggressive B-cell lymphoma. Blood. 2015;125(7):1137-1145. Chang TC, Yu D, Lee YS, et al. Widespread microRNA repression by Myc contributes to tumorigenesis. Nat Genet. 2008;40(1):43-50. Di Lisio L, Sanchez-Beato M, Gomez-Lopez G, et al. MicroRNA signatures in B-cell lymphomas. Blood Cancer J. 2012;2(2):e57. Liu Z, Mai C, Yang H, et al. Candidate tumour suppressor CCDC19 regulates miR184 direct targeting of C-Myc thereby suppressing cell growth in non-small cell lung cancers. J Cell Mol Med. 2014;18(8):16671679. Tao J, Zhao X. c-MYC-miRNA circuitry: a central regulator of aggressive B-cell malignancies. Cell Cycle. 2014;13(2):191-198. Xiong L, Jiang W, Zhou R, Mao C, Guo Z. Identification and analysis of the regulatory network of Myc and microRNAs from highthroughput experimental data. Comput Biol Med. 2013;43(9):1252-1260. Zhao ZN, Bai JX, Zhou Q, et al. TSA suppresses miR-106b-93-25 cluster expression through downregulation of MYC and inhibits proliferation and induces apoptosis in human EMC. PLoS One. 2012;7(9): e45133. Akao Y, Nakagawa Y, Kitade Y, Kinoshita T, Naoe T. Downregulation of microRNAs-143 and -145 in B-cell malignancies. Cancer Sci. 2007;98(12):1914-1920. Eden E, Navon R, Steinfeld I, Lipson D, Yakhini Z. GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists. BMC Bioinformatics. 2009;10:48. Tomaselli S, Locatelli F, Gallo A. The RNA editing enzymes ADARs: mechanism of action and human disease. Cell Tissue Res. 2014;356(3):527-532. Denman RB. Using RNAFOLD to predict the activity of small catalytic RNAs. Biotechniques. 1993;15(6):1090-1095. Hagberg H, Killander A, Simonsson B. Serum beta 2-microglobulin in malignant lymphoma. Cancer. 1983;51(12):2220-2225. Solenthaler M, Matutes E, Brito-Babapulle V, Morilla R, Catovsky D. p53 and mdm2 in mantle cell lymphoma in leukemic phase. Haematologica. 2002;87(11):1141-1150. Durig J, Schmucker U, Duhrsen U. Differential expression of chemokine receptors in B cell malignancies. Leukemia. 2001;15(5):752-756. Ishiura Y, Kotani N, Yamashita R, et al. Anomalous expression of Thy1 (CD90) in Bcell lymphoma cells and proliferation inhibition by anti-Thy1 antibody treatment. Biochem Biophys Res Commun. 2010;396

(2):329-334. 37. Kuppers R. NPAT mutations in Hodgkin lymphoma. Blood. 2011;118(3):484-485. 38. Love C, Sun Z, Jima D, et al. The genetic landscape of mutations in Burkitt lymphoma. Nat Genet. 2012;44(12):1321-1325. 39. Schmitz R, Young RM, Ceribelli M, et al. Burkitt lymphoma pathogenesis and therapeutic targets from structural and functional genomics. Nature. 2012;490(7418):116-120. 40. Chen D, Law ME, Theis JD, et al. Clinicopathologic features of CDK6 translocation-associated B-cell lymphoproliferative disorders. Am J Surg Pathol. 2009;33(5):720729. 41. Lawrie CH, Chi J, Taylor S, et al. Expression of microRNAs in diffuse large B cell lymphoma is associated with immunophenotype, survival and transformation from follicular lymphoma. J Cell Mol Med. 2009;13(7):1248-1260. 42. Roehle A, Hoefig KP, Repsilber D, et al. MicroRNA signatures characterize diffuse large B-cell lymphomas and follicular lymphomas. Br J Haematol. 2008;142(5):732-744. 43. Kent OA, McCall MN, Cornish TC, Halushka MK. Lessons from miR-143/145: the importance of cell-type localization of miRNAs. Nucleic Acids Res. 2014;42(12): 7528-7538. 44. Schou JV, Rossi S, Jensen BV, et al. miR-345 in Metastatic colorectal cancer: a non-invasive biomarker for clinical outcome in nonKRAS mutant patients treated with 3rd line cetuximab and irinotecan. PLoS One. 2014; 9(6):e99886. 45. Dou L, Zheng D, Li J, et al. Methylationmediated repression of microRNA-143 enhances MLL-AF4 oncogene expression. Oncogene. 2012;31(4):507-517. 46. Liu C, Iqbal J, Teruya-Feldstein J, et al. MicroRNA expression profiling identifies molecular signatures associated with anaplastic large cell lymphoma. Blood. 2013;122(12):2083-2092. 47. Tao ZH, Wan JL, Zeng LY, et al. miR-612 suppresses the invasive-metastatic cascade in hepatocellular carcinoma. J Exp Med. 2013;210(4):789-803. 48. Blow MJ, Grocock RJ, van Dongen S, et al. RNA editing of human microRNAs. Genome Biol. 2006;7(4):R27. 49. Choudhury Y, Tay FC, Lam DH, et al. Attenuated adenosine-to-inosine editing of microRNA-376a* promotes invasiveness of glioblastoma cells. J Clin Invest. 2012;122 (11):4059-4076. 50. Mizuguchi Y, Mishima T, Yokomuro S, et al. Sequencing and bioinformatics-based analyses of the microRNA transcriptome in hepatitis B-related hepatocellular carcinoma. PLoS One. 2011;6(1):e15304.

1389

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Plasma Cell Disorders

Ferrata Storti Foundation

Upfront autologous stem cell transplantation for newly diagnosed elderly multiple myeloma patients: a prospective multicenter study Laurent Garderet,1 Eric Beohou,2 Denis Caillot,3 Anne Marie Stoppa,4 Cyrille Touzeau,5 Marie Lorraine Chretien,3 Lionel Karlin,6 Philippe Moreau,5 Jean Fontan,7 Didier Blaise,4 Emmanuelle Polge,2 Mor Seny Gueye,8 Souhila Ikhlef,8 Zora Marjanovic,8 Myriam Labopin,2 and Mohamad Mohty1

INSERM, UMR_S 938, Proliferation and Differentiation of Stem Cells, AP-HP, Hôpital Saint Antoine, Département d’Hématologie et de Thérapie Cellulaire, F-75012, Paris, Université Pierre et Marie Curie-Paris 6; 2EBMT Registry Office, Paris; 3 CHRU Dijon; 4Department of Hematology, Institut Paoli Calmettes, Marseille; 5 Department of hematology, University Hospital Hotel Dieu, Nantes; 6Department of Hematology, CHU Lyon Sud, Pierre Bénite; 7Department of Hematology, CHU Besançon; and 8Hôpital Saint Antoine, Département d’Hématologie et de Thérapie Cellulaire, F-75012, Paris, France

Haematologica 2016 Volume 101(11):1390-1397

ABSTRACT

Correspondence: laurent.garderet@aphp.fr

Received: May 31, 2016. Accepted: September 6, 2016. Pre-published: September 9, 2016. doi:10.3324/haematol.2016.150334

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1390

1390

he feasibility and efficacy of high-dose melphalan followed by autologous hematopoietic stem cell transplantation in newly diagnosed elderly patients with multiple myeloma was analyzed prospectively. Fifty-six multiple myeloma patients, aged 65 years or over, from 6 French centers were studied. The induction therapy was bortezomib-based in combination with dexamethasone and either thalidomide, cyclophosphamide or lenalidomide, for 4-6 cycles. Peripheral blood stem cells were collected after high-dose cyclophosphamide plus G-CSF or G-CSF alone, with plerixafor if needed. The conditioning regimen consisted of melphalan at 140 mg/m² in 18 patients (36%) and 200 mg/m2 in 32 (64%). Three months post autologous hematopoietic stem cell transplantation, a 2-month consolidation phase with either lenalidomide plus dexamethasone or bortezomibbased combination therapy was allowed, but maintenance treatment was not given. All but 6 patients underwent autologous hematopoietic stem cell transplantation and 3 had tandem transplantations. The treatment-related mortality was 0% at 100 days post transplantation. Sixtyeight percent received consolidation therapy following transplantation. The best response achieved was 40% complete response, 36% very good partial response, and 18% partial response. After a median follow up of 21 months (range 6-31), the estimated progression-free and overall survival rates at two years were 76% [95%CI: (61.6-94.1)] and 88% [95%CI: (76.7-100)], respectively. The higher dose of melphalan (200 mg/m2) afforded superior progression-free and overall survival rates. This prospective study provides evidence for the safety and efficacy of autologous hematopoietic stem cell transplantation as a first-line treatment approach in elderly multiple myeloma patients. (clinicaltrials.gov identifier: 01671826)

Introduction Two-thirds of multiple myeloma (MM) patients are over 65 years of age at the time of diagnosis. As the general population becomes older, this proportion is destined to increase. Autologous stem cell transplantation (ASCT) is a standard form of treatment for myeloma patients under the age of 65 years1 but is a controversial procedure for patients over this age, mostly because of a suspected increase in toxicity.2-5 haematologica | 2016; 101(11)

Autologous stem cell transplantation in elderly myeloma patients In elderly patients, only two randomized studies have compared a transplant versus a no transplant approach. Palumbo et al. first reported a benefit of intermediate-dose (100 mg/m2) melphalan (HDT) plus ASCT for patients aged between 65 and 70 years.6 In contrast, the IFM 99-06 study did not show any benefit of transplantation after melphalan (100 mg/m2) as compared to a combination of melphalan, prednisone and thalidomide.7 However, many subsequent studies, mostly retrospective or registry-based and performed before the latest drugs became available, have shown encouraging results with ASCT in patients over 65 years of age.8-13 Some investigators even reported a successful outcome for patients over the age of 70.14-17 A recent European Society for Blood and Marrow Transplantation (EBMT) study showed that, over the past few years, ASCT was performed more often, especially in the elderly population, and with better outcomes.18 There has been a considerable decrease in toxicity due to better patient selection and improved supportive care. Nowadays, geriatric assessment is routinely performed in the clinic, which helps the treatment decision-making process.19,20 Furthermore, new drugs have emerged, such as the immunomodulatory drugs lenalidomide and pomalidomide and proteasome inhibitors like carfilzomib and ixazomib.21 These, used as single agents or more often in combination, together with the previous standard treatment, are stimulating a new interest in ASCT for elderly patients.22 Therefore, we initiated a multicenter prospective observational study, from 2013 to 2015 in 6 French centers, which included 56 myeloma patients aged 65 years or over, 50 of whom underwent ASCT after bortezomibbased induction.

Methods

Induction regimen The induction regimen was bortezomib-based, either bortezomib plus dexamethasone (VD), bortezomib plus thalidomide plus dexamethasone (VTD), bortezomib plus cyclophosphamide plus dexamethasone (VCD), bortezomib plus lenalidomide plus dexamethasone (VRD), or melphalan plus prednisone plus bortezomib (MPV). Patients received 4-6 21-day cycles according to the local guidelines of each center.

Stem cell mobilization and collection Peripheral blood hematopoietic stem cells were mobilized using the procedure in routine practice at each center. The cells were collected either after administration of high-dose cyclophosphamide plus G-CSF or in the steady state after administration of G-CSF alone, plus plerixafor if needed.

Conditioning regimen and supportive care To be eligible for transplantation, the patient had to have adequate organ function and no uncontrolled infection. The conditioning regimen consisted of melphalan (140 or 200 mg/m2), given over one or two days, according to the physician’s choice. Tandem ASCT was allowed and supportive care was given according to the current protocol in each institution.

Consolidation/maintenance A short 2-month consolidation phase three months post ASCT was allowed (lenalidomide-dexamethasone, VD, VTD, VCD or VRD). No maintenance treatment was given.

Engraftment and disease response The date of neutrophil engraftment was defined as the first of three consecutive days when the absolute neutrophil count was over 0.5x109/L. The date of platelet engraftment was defined as the first of seven consecutive days when the platelet count was over 20x109/L, independent of any platelet transfusions. Response, disease progression and relapse were defined according to the International Myeloma Working Group uniform response criteria.26

Patients Patients were eligible if they were over 65 years of age and presented with symptomatic, measurable, newly diagnosed multiple myeloma. Between September 2012 and September 2014, a total of 56 newly diagnosed elderly MM patients were treated in 6 institutions in France. The diagnosis, clinical staging and prognostic score of MM were based on the Durie and Salmon staging system and the International Staging System (ISS).23,24 The Seattle group‘s hematopoietic cell transplantation specific comorbidity index (HCT-CI) was used to score the comorbidities.25 Baseline demographics, clinical and laboratory data at diagnosis, and information on treatment and response were collected prospectively and recorded in the EBMT Promise (Med B) database. Patients gave their informed consent to the study. This prospective observational study was approved by the Ethics Committee/Institutional Review Board of Paris Île de France V and registered as clinicaltrials.gov identifier: 01671826.

Treatment The primary objective was to assess patient outcome and especially any treatment toxicity. In each case, therapy was decided by the physician responsible for the patient. The shortterm use of dexamethasone for emergent disease control was not considered as conventional chemotherapy. ASCT was performed as upfront therapy after induction, provided the disease was not progressive.

haematologica | 2016; 101(11)

Assessment of transplant-related toxicity Transplant-related mortality (TRM) was defined as the percentage of patients dying without relapse or progression within a given time interval following transplantation. Non-hematologic toxicity was assessed by the local physician. Variables analyzed included bacterial and viral infections, gastro-enteric, renal (serum creatinine) and hepatic (bilirubin, alanine transaminase and aspartate transaminase) function, and cardiotoxicity.

Statistical analysis Patients’ demographic and clinical characteristics were summarized using the median and range for continuous variables, and counts and percentages for categorical variables. Progression-free survival (PFS) was defined as the time from the date of starting treatment to the date of disease progression or death from any cause. Overall survival (OS) was defined as the time from the date of starting treatment to death from any cause. PFS and OS curves were calculated using the Kaplan-Meier method. We examined the relationship between outcomes and potential prognostic factors. The differences between the curves were evaluated using the log-rank test. Variables included baseline patient factors, and prognostic and treatment-related factors. The selection rule for multivariate analysis was a threshold of 20%. A multivariate Cox proportional hazards model was used to determine the independent predictors associated with extended OS.

1391

L. Garderet et al. Table 1. Patients' characteristics at diagnosis and transplantation.

Variable At diagnosis Median age, years Median (range) Sex, n. (%) Male Female SD staging, n. (%) I II III ISS staging, n. (%) I II III Type of M protein, n. (%) Ig G Ig A Light chain Other Bone marrow aspirate: % plasmacytosis Median (range) Creatinine, mmol/L Median (range) At transplantation Median age, years Median (range) Diagnosis to transplantation, months Median (range) Sorror score, n. (%) 0 1 2 3 6 Sorror score II, n. (%) 0 â&#x2030;Ľ1

Melphalan dose 140 mg/m2 200 mg/m2 (n=16) (n=34)

All patients (n=56)

68.7(64.3-73.4)

66.5(64.5-74)

67.4(64.3-74)

0.29

7(43.8%) 9(56.2%)

21(61.8%) 13(38.2%)

30(53.6%) 26(46.4%)

0.36

2(12.5%) 1(6.2%) 13(81.2%)

0(0%) 3(9.4%) 29(90.6%)

2(3.8%) 4(7.5%) 47(88.7%)

0.12

5(38.5%) 3(23.1%) 5(38.5%)

12(37.5%) 14(43.8%) 6(18.8%)

18(35.3%) 19(37.3%) 14(27.5%)

0.28

9(56.2%) 6(37.5%) 1(6.2%) 0(0%)

14(42.4%) 9(27.3%) 9(27.3%) 1(3%)

29(52.7%) 15(27.3%) 10(18.2%) 1(1.8%)

0.3

40(7-90)

36(4-72)

37(4-90)

0.65

105(54-371)

79.5(42-442)

84(42-442)

0.093

69.2(66.5-74.3)

67(65-74.5)

68.1(65-74.5)

0.07

5.1(4.2-14.9)

5.4(3.4-16)

0.97

8(50%) 4(25%) 1(6.2%) 3(18.8%) 0(0%)

27(79.4%) 2(5.9%) 1(2.9%) 3(8.8%) 1(2.9%)

35(70%) 6(12%) 2(4%) 6(12%) 1(2%)

0.18

8(50%) 8(50%)

27(79.4%) 7(20.6%)

35(70%) 15(30%)

0.074

ISS: International Staging System; n: number.

Statistical analysis was performed with a 2-sided a = 0.05 and a 95% confidence interval. Data were analyzed using R software, v.2.15.1, and IBM SPSS statistics v.22.

Results Patients' characteristics Patients' demographics and disease characteristics are summarized in Table 1. At the time of diagnosis, median age was 67 years (range 64-74) with 23% of patients over 70 years of age; 30 males and 26 females. The myeloma immunoglobulin subtypes were: IgG (n=29), IgA (n=15), light chain (n=10), other (n=2). The Salmon and Durie stage was III in 89% of cases (n=47) and ISS scores were I (n=18, 35%), II (n=19, 37%), and III (n=14, 27%). Highrisk cytogenetic features [t (4;14) and/or del17p] were found in 9 cases (16%). Although 10% of patients had a serum creatinine level of more than 176 micromol/L, none 1392

underwent hemodialysis. Twenty-eight patients (5%) received VTD, 9 (17%) VCD, 9 (17%) VD, 4 (7%) MPV and 3 (6%) VRD, with 11 patients (21%) requiring two lines of induction and one three lines. At transplantation, the HCT-CI comorbidity scores were 0 (n=34), 1 (n=6), 2 (n=2), 3 (n=6), 6 (n=1), and unknown (n=1). Median age at the time of ASCT was 68 years and the median time from diagnosis to ASCT was five months. Median follow up was 21 months (range 631).

Mobilization A median of 5.31x106/kg CD34+ cells were collected. Thirty-two patients (57%) were mobilized with cyclophosphamide + G-CSF, 13 (23%) with G-CSF alone, 6 (10%) with G-CSF + plerixafor, and one with cyclophosphamide + G-CSF + plerixafor. The number of mobilization courses was 1 (n=39), 2 (n=10), and haematologica | 2016; 101(11)

Autologous stem cell transplantation in elderly myeloma patients unknown (n=2); there were 2 failed mobilizations. There was no ex vivo manipulation of the autologous graft. Median number of CD34+ cells infused was 4.1x106/kg (range 1.7-7.6x106/kg).

Patients unable to proceed to ASCT In an intention to treat analysis, 6 of the 56 patients could not proceed to ASCT due to an early infectious death (n=1), serious comorbidity (n=2), disease refractoriness to the induction regimen (n=1), or failure to collect an adequate peripheral blood stem cell (PBSC) graft (n=2).

Univariate analysis We performed a univariate analysis to identify the predictors independently associated with PFS and OS using the Cox proportional hazards model. Variables included in the analysis were: baseline patients' characteristics (age, sex, type of myeloma protein), prognostic factors (albumin, β2 microglobulin, ISS stage), disease status, and melphalan dose at transplantation. We found the dose of the conditioning regimen to be the only significant prognostic factor for both PFS and OS. ISS stage was only prognostic for OS (Table 3).

Engraftment The conditioning regimen consisted of 140 mg/m² melphalan in 18 patients (36%) and 200 mg/m² melphalan in 32 (64%). Five patients received bortezomib in combination with melphalan (melphalan 200 mg/m²), while 3 patients (6%) underwent tandem ASCT. Median time to neutrophil and platelet engraftment was 12 days (range 9-56). There was no significant difference in the time to neutrophil or platelet engraftment between the two doses of melphalan.

Consolidation Consolidation treatment (three months post ASCT) was given in 38 patients (68%). Thirteen (34%) received VTD, 6 (16%) VRD, 6 (16%) VCD, 5 (13%) VD, 4 (10%) RD, 2 (5%) lenalidomide, 1 (3%) MPV, and 1 pomalidomide (3%). In 12 cases, the physician decided to administer no consolidation therapy.

Treatment-related toxicity The day-100 post ASCT treatment-related mortality (TRM) was 0%. There was no significant difference in TRM between the two doses of melphalan. Table 2 summarizes the non-hematologic toxicities appearing after ASCT. Infection within the first 100 days post ASCT occurred in 18 patients (36%) and non-infectious complications in 24 (48%). Gastrointestinal toxicities were frequent, the most common being oral mucositis (n=18, 36%) and diarrhea (n=3, 6%). Pulmonary infection occurred in 7 patients (14%). Malnutrition was noted in 5 patients and thrombosis in 2, while one had a hemorrhage and another a cardiac complication. The incidence of infectious complications post ASCT and the response rate were comparable between the two doses of melphalan (P=0.28).

Discussion Over the past decade, the use of HDT followed by ASCT in combination with new drugs has substantially improved the outcome of younger patients with MM. However, the safety and efficacy of HDT in patients over 65 years of age remain uncertain. In this prospective study,

Table 2. Non-hematopoietic toxicities.

Toxicity Bacteremia Pneumonia Gastrointestinal Malnutrition Cystitis Septicemia Thrombosis Skin rash Peripheral neuropathy Other

Patients (n=56) 10 8 11 5 2 2 2 2 2 12

Response and survival Disease status at the time of ASCT was defined as: complete response (CR) (n=12, 24%), very good partial response (VGPR) (n=19, 38%), partial response (PR) (n=17, 34%), or stable disease (SD)/non-responsive (n=2, 4%). The overall response rate on day 100 was 96% (CR: 34%, VGPR: 47%, PR: 15%, and SD/non-responsive: 4%). At three months post ASCT, 68% of the patients were able to receive the planned consolidation treatment. The best responses were: CR (n=20, 40%), VGPR (n=18, 36%), PR (n=9, 18%), progression (n=1, 2%), and unknown (n=2, 4%) (Figure 1). After a median follow up of 21 months (range 6-31), the PFS and OS rates at two years were 76% [95%CI: (61.694.1)] and 88% [95%CI: (76.7-100)], respectively (Figure 2). There was a trend to a better rate of PFS in the 200 mg/m2 melphalan group (Figure 3). haematologica | 2016; 101(11)

Figure 1. Response rates before and after autologous stem cell transplantation (ASCT). CR: complete response; VGPR: very good partial response; PR: partial response; SD: stable disease; PD: progressive disease.

1393

L. Garderet et al. Table 3. Univariate analysis of prognostic factors.

PFS Variables Regimen 140 mg/m2 200 mg/m2 Age <70 yrs ≥70 yrs Diagnosis to transplantation ≤6 mo >6 mo Ig Ig G Ig A ISS stage at diagnosis 1 2 3 Creatinine at diagnosis ≤140 mmol/L >140 mmol/L Hemoglobin at diagnosis ≤10 g/dL >10 g/dL Albumin at diagnosis <35 g/L ≥35 g/L Bone marrow plasma cells at diagnosis ≤36% >36% Status at transplantation CR VGPR PR Sorror score at transplantation 0 ≥1

1 yr

2 yrs

1 yr

2 yrs

67.5 [41.7 - 100] 100 [100 - 100]

NA 82.4 [66.1 - 100]

0.022

67 [41.2 - 100] 100 [100 - 100]

NA 94.1 [83.6 - 100]

0.012

93.4 [84.9 - 100] 90.9 [75.4 - 100]

73.8 [55.5 - 98.1] 79.5 [57.7 - 100]

0.842

93.2 [84.6 - 100] 90.9 [75.4 - 100]

93.2 [84.6 - 100] 79.5 [57.7 - 100]

0.285

96.2 [89 - 100] 85.9 [ 69 - 100]

71.1 [52.3 - 96.8] 85.9 [69 - 100]

0.53

96.2 [89 - 100] 83.7 [64.5 - 100]

89.3 [75.8 - 100] 83.7 [64.5 - 100]

0.251

87.5 [72.7 - 100] 100 [100 - 100]

70 [48.8 - 100] 85.7 [63.3 - 100]

0.646

86.5 [70.7 - 100] 100 [100 - 100]

77.9 [58.4 - 100] 100 [100 - 100]

0.175

92.3 [78.9 - 100] 100 [100 - 100] 77.9 [54.6 - 100]

92.3 [78.9 - 100] 78.8 [56.4 - 100] 51.9 [26.6 - 100]

0.158

90.9 [75.4 - 100] 100 [100 - 100] 77.9 [54.6 - 100]

90.9 [75.4 - 100] 100 [100 - 100] 64.9 [39.2 - 100]

0.0308

91.1 [82 - 100] 100 [100 - 100]

81 [66.6 - 98.5] NA

0.611

90.5 [80.6 - 100] 100 [100 - 100]

90.5 [80.6 - 100] 80 [51.6 - 100]

0.776

92.9 [80.3 - 100] 92.6 [83.1 - 100]

61.9 [38.1 - 100] 86 [71.7 - 100]

0.419

92.9 [80.3 - 100] 91.8 [81.4 - 100]

82.5 [62.8 - 100] 91.8 [81.4 - 100]

0.606

85.6 [68.8 - 100] 95.7 [87.7 - 100]

68.4 [42 - 100] 78.3 [61.3 - 99.9]

0.538

85.1 [68 - 100] 95.2 [86.6 - 100]

68.1 [41.6 - 100] 95.2 [86.6 - 100]

0.0456

100 [100 - 100] 85 [70.7 - 100]

80 [62.1 - 100] 72.9 [51.1 - 100]

0.333

100 [100 - 100] 82.5 [65.8 - 100]

92.9 [80.3 - 100] 82.5 [65.8 - 100]

0.0974

100 [100 - 100] 100 [100 - 100] 90.9 [75.4 - 100]

83.3 [58.3 - 100] 87.5 [67.3 - 100] 68.2 [43.8 - 100]

0.186

100 [100 - 100] 100 [100 - 100] 90 [73.2 - 100]

100 [100 - 100] 87.5 [67.3 - 100] 90 [73.2 - 100]

0.677

97.1 [91.8 - 100] 86.7 [71.1 - 100]

73.9 [58.6 - 93.3] 86.7 [71.1 - 100]

0.606

97.1 [91.5 - 100] 86.7 [71.1 - 100]

93.3 [84.7 - 100] 86.7 [71.1 - 100]

0.305

PFS: progression-free survival; OS: overall survival; Ig: immunoglobulin; ISS: International Staging System; VGPR: very good partial remission; PR: partial remission; CR: complete remission; NA: not available ; yrs: years; mo: months.

the relatively low toxicity of the ASCT procedure for this patient population is very encouraging, with 0% TRM at 100 days post transplantation. In comparison, patients under 65 years of age have a 100-day TRM of approximately 1%. This is particularly striking considering that 10% of the patients had renal impairment at diagnosis with serum creatinine levels of more than 176 micromol/L, while 16% had high-risk cytogenetic features. It is also important to note that two-thirds of the patients received melphalan at a dose of 200 mg/m2. In this setting, patient selection is important.27,28 Six of the 56 patients (10%) could not proceed to ASCT; these frail individuals were nevertheless not excluded from the post-transplant analysis, which was performed on the basis of the intention-to-treat. Moreover, the comorbidity as measured by the Sorror score was low: 40 in 50 patients (80%) had no or only one comorbidity factor at transplantation. This patient selection could partly explain the low TRM. An improvement in post-transplant care may also have contributed to the lack of early toxicity following transplantation. In this study, an adequate number of stem cells to support ASCT was obtained in all but 2 patients (3.5%).There was no difference in the numbers of stem 1394

cells mobilized compared to those collected in younger patients, in accordance with previously published results.14 We confirmed that the inclusion of novel drugs, namely a bortezomib-based induction regimen, improved both response and outcome, and should be incorporated into the HDT approach for elderly patients. In terms of response, 34% CR on day 100 post ASCT is similar to results published in the literature. Palumbo et al. reported that bortezomib-based induction plus ASCT led to 38% CR in patients aged 65-75 years.29 Similarly, Mertz et al. obtained 43% CR + near CR after ASCT.22 The results are even better following post ASCT consolidation, reaching 40% CR in our study. Palumbo et al. even reported a CR rate of up to 66% with lenalidomide plus dexamethasone consolidation and post ASCT maintenance.29 Post ASCT maintenance is, however, still a controversial issue in young patients, and more data will be needed before it can be implemented in an older patient population. The relationship between melphalan dose and outcome has been demonstrated previously. In a report from the Mayo Clinic, in which 33 patients aged 70 years or older undergoing high-dose therapy were compared with a cohort of matched patients aged 65 years old or under, haematologica | 2016; 101(11)

Autologous stem cell transplantation in elderly myeloma patients

OS in the entire cohort

A 1.0 0.8 0.6 0.4 0.2 0.0

Time from transplant (years)

Number of patients at risk

PFS in the entire cohort B 1.0 0.8 0.6 0.4 0.2 0.0

Time from transplant (years)

Number of patients at risk

toxicity and survival were comparable.15 Although a dose reduction to 140 mg/m2 was required for 10 patients in the elderly group, the majority received conditioning with 200 mg/m2 melphalan, and the response rate was similar in the two groups. On the other hand, in a report from the University of Arkansas, 200 mg/m2 melphalan was associated with excessive early mortality (16%) in patients aged aged 70 years or older.14 In the latter study, all subsequent patients were treated with 140 mg/m2 melphalan, which resulted in a TRM of 2%. In our work, there was a better outcome using 200 mg/m2 melphalan compared to 140 mg/m2. Considering that there was no increased toxicity, 200 mg/m2 melphalan could be an appropriate regimen for patients aged 65-70 years. Our results should also be compared to those of nontransplant approaches, and in particular to the data obtained using new drugs. In the past, for patients aged 65-75 years, a combination of melphalan plus prednisone and thalidomide yielded a median PFS of 27.5 months and a median OS of 51.6 months, which was superior to the PFS of 19.4 months achieved using VAD plus double ASCT (IFM 99-06).7 A combination of thalidomide plus doxorubicine and dexamethasone (Thal DD) plus thalidomide maintenance was not inferior to Thal DD plus highdose therapy and ASCT in elderly patients with de novo MM.30 After a median follow up of 36 months, there was haematologica | 2016; 101(11)

Figure 2. (A) Overall survival and (B) progression-free survival in the entire cohort after autologous stem cell transplantation (ASCT) (n=56). The dotted lines represent confidence intervals.

no significant difference in the median time to progression (TTP) between the group of patients who underwent ASCT and those patients receiving Thal DD plus maintenance (32 vs. 31 months, P=0.962; 32 vs. 29 months, P=0.726, respectively). The 5-year OS was 49% in the first group and 46% in the second (P=0.404). In the Velcade as Initial Standard Therapy in Multiple Myeloma (Vista) study, the TTP among patients receiving bortezomib plus melphalanâ&#x20AC;&#x201C;prednisone was 24.0 months.31 In the Frontline Investigation of Revlimid and Dexamethasone versus Standard Thalidomide (First) trial, the median PFS was 25.5 months under continuous lenalidomide plus dexamethasone and the OS at 4 years was 59%.19 In our study, the estimated PFS and OS rates at two years were 76% and 88%, respectively, which is encouraging. Moreover, these data are almost superimposable on those of Palumbo et al. using PAD induction followed by ASCT with lenalidomide consolidation and maintenance: after a median follow up of 21 months, their 2-year PFS and OS rates were 69% and 86%, respectively.29 In the younger myeloma patients (aged <65 years), the combination of bortezomib and lenalidomide and dexamethasone as induction and consolidation post ASCT along with a 1-year lenalidomide maintenance gave even better results; with a median follow up of 39 months, estimated 3-year PFS and OS were 77% and 100%, respectively.32 1395

L. Garderet et al. A

OS 1.0 0.8 0.6 0.4 0.2 0.0

Time from transplant (years) 140 mg/m2 200 mg/m2 Number of patients at risk

PFS 1.0 0.8 0.6 0.4 0.2 0.0 Figure 3. Outcome, overall survival (A) and progression-free survival (B) according to the conditioning regimen: 140 mg/m2 versus 200 mg/m2melphalan.

Time from transplant (years) 140 mg/m22 200 mg/m Number of patients at risk

Nevertheless, cross trial comparisons should be viewed with caution on account of the patient selection bias. This implies that selected elderly patients could benefit from auto-SCT, which might be superior to chemotherapy or new drug combinations. Although we looked for prognostic factors, the only two significant factors detected in our univariate analysis of OS were the dose of the conditioning regimen and the ISS stage. This could be related to the small number of patients and the relatively short follow up. However, the β2 microglobulin level before transplantation, which is a confirmed prognostic variable in many studies, may lack significance in this elderly population.14 β2 microglobulin levels are higher in the elderly. This probably reflects an age-related decrease in creatinine clearance, rather than a high tumor burden. The weaknesses of our study lie in the non-randomized nature of the trial and the highly selected patient population included, as reflected by the low Sorror score in most of our patients. Therefore, the data concerning ASCT may not be relevant to all newly diagnosed elderly myeloma patients. We also acknowledge that the induction, conditioning and consolidation regimens were very heteroge-

References 1. Blade J, Rosinol L, Cibeira MT, et al. Hematopoietic stem cell transplantation for multiple myeloma beyond 2010. Blood. 2010;115(18):3655-3663.

1396

neous, which makes it more difficult to draw conclusions. Follow up was also relatively short. Other groups are currently studying the feasibility and efficacy of high-dose melphalan in elderly patients, such as the DSMM group in Germany and the Freiburg team.13 Specifically, the Freiburg team has proposed a revised Myeloma Comorbidity Index for future frailty measurements which could help to identify those patients fit enough to undergo stem cell transplantation.33 In conclusion, these prospective multicenter results indicate that ASCT is a safe and effective mode of treatment for elderly and fit MM patients in the present era of novel induction agents. One may note that patients over 70 years of age did not have a worse prognosis. Thus, age per se should not be used as an exclusion criterion for ASCT. These results provide a framework for a randomized comparison with non-transplant approaches in this patient subgroup. Funding The study was supported by a grant from the “Association for Training, Education and Research in Hematology, Immunology and Transplantation” (ATERHIT, Nantes, France).

2. Jantunen E. Autologous stem cell transplantation beyond 60 years of age. Bone Marrow Transplant. 2006;38(11):715-720. 3. Gertz MA. Too old for transplantation: think again. Blood. 2004;104(10):30003001.

4. Wildes TM, Rosko A, Tuchman SA. Multiple Myeloma in the Older Adult: Better Prospects, More Challenges. J Clin Oncol. 2014;32(24):2531-2540. 5. Ozaki S and Shimizu K. Autologous stem cell transplantation in elderly patients with

haematologica | 2016; 101(11)

Autologous stem cell transplantation in elderly myeloma patients

10.

11.

12.

13.

14.

15.

multiple myeloma: past, present, and future. Biomed Res Int. 2014;394792. Palumbo A, Bringhen S, Petrucci MT, et al. Intermediate-dose melphalan improves survival of myeloma patients aged 50 to 70: results of a randomized controlled trial. Blood. 2004;104(10):3052-3057. Facon T, Mary JY, Hulin C, et al. Melphalan and prednisone plus thalidomide versus melphalan and prednisone alone or reduced-intensity autologous stem cell transplantation in elderly patients with multiple myeloma (IFM 99â&#x20AC;&#x201C;06): a randomised trial. Lancet. 2007; 370(9594):1209-1218. Sirohi B, Powles R, Treleaven J, et al. The role of autologous transplantation in patients with multiple myeloma aged 65 years and over. Bone Marrow Transplant. 2000;25(5):533-539. Reece D, Bredeson C, Perez WS, et al. Autologous stem cell transplantation in multiple myeloma patients <60 vsâ&#x2030;Ľ 60 years of age. Bone Marrow Transplant. 2003;32(12):1135-1143. Jantunen E, Kuittinen T, Penttila K, et al. High-dose melphalan (200 mg/m2) supported by autologous stem cell transplantation is safe and effective in elderly (>65 years) myeloma patients: comparison with younger patients treated on the same protocol. Bone Marrow Transplant. 2006; 37(10):917-922. Lenhoff S, Hjorth M, Westin J, et al. Impact of age on survival after intensive therapy for multiple myeloma: a population-based study by the Nordic Myeloma Study Group. Br J Haematol. 2006;133(4):389-396. Muta T, Miyamoto T, Fujisaki T, et al. Evaluation of the Feasibility and Efficacy of Autologous Stem Cell Transplantation in Elderly Patients with Multiple Myeloma. Intern Med. 2013;52(1):63-70. Mertz M, Jansen L, Castro FA, et al. Survival of elderly patients with multiple myeloma-effect of upfront autologous stem cell transplantation. Eur J Cancer. 2016;62;1-8. Badros A, Barlogie B, Siegel E, et al. Autologous stem cell transplantation in elderly multiple myeloma patients over the age of 70 years. Br J Haematol. 2001; 114(3):600-607. Kumar SK, Dingli D, Lacy MQ, et al.

haematologica | 2016; 101(11)

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

Autologous stem cell transplantation in patients of 70 years and older with multiple myeloma: Results from a matched pair analysis. Am J Hematol. 2008;83(8):614617. Bashir Q, Shah N, Parmar S, et al. Feasibility of autologous hematopoietic stem cell transplant in patients aged â&#x2030;Ľ 70 years with multiple myeloma. Leuk Lymphoma. 2012;53(1):118-122. Wildes TM, Finney JD, Fiala M, et al. Highdose therapy and autologous stem cell transplant in older adults with multiple myeloma. Bone Marrow Transplant. 2015;50(8):1075-1082. Auner HW, Szydlo R, Hoek J, et al. Trends in autologous hematopoietic cell transplantation for multiple myeloma in Europe: increased use and improved outcomes in elderly patients in recent years. Bone Marrow Transplant. 2015;50(2):209-215. Palumbo A, Bringhen S, Mateos MV, et al. Geriatric assessment predicts survival and toxicities in elderly myeloma patients: an International Myeloma Working Group report. Blood. 2015;125(13):2068-2074. Palumbo A, Sezer O, Kyle R, et al. International Myeloma Working Group guidelines for the management of multiple myeloma patients ineligible for standard high-dose chemotherapy with autologous stem cell transplantation. Leukemia. 2009;23(10):1716-1730. Benboubker L, Dimopoulos MA, Dispenzieri A, et al. Lenalidomide and dexamethasone in transplant-ineligible patients with myeloma. N Engl J Med. 2014;371(10):906-917. Merz M, Neben K, Raab MS, et al. Autologous stem cell transplantation for elderly patients with newly diagnosed multiple myeloma in the era of novel agents. Ann Oncol. 2014;25(1):189-195. Durie BG, Salmon SE. A clinical staging system for multiple myeloma. Correlation of measured myeloma cell mass with presenting clinical features, response to treatment, and survival. Cancer. 1975;36(3):842-854. Greipp PR, San Miguel J, Durie BG, et al. International staging system for multiple myeloma. J Clin Oncol. 2005;23(15):34123420. Sorror ML, Maris MB, Storb R, et al. Hematopoietic cell transplantation (HCT)-

26.

27.

28.

29.

30.

31.

32.

33.

specific comorbidity index: a new tool for risk assessment before allogeneic HCT. Blood. 2005;106(8):2912-2919. Durie BG, Harousseau JL, Miguel JS, et al. International uniform response criteria for multiple myeloma. Leukemia. 2006; 20(9):1467-1473. Palumbo A, Bringhen S, Ludwig H, et al. Personalized therapy in multiple myeloma according to patient age and vulnerability: a report of the European Myeloma Network (EMN). Blood. 2011;118(17):4519-4529. Bringhen S, Mateos MV, Zweegman S, et al. Age and organ damage correlate with poor survival in myeloma patients: metaanalysis of 1435 individual patient data from 4 randomized trials. Haematologica. 2013;98(6):980-987. Palumbo A, Gay F, Falco P, et al. Bortezomib As Induction Before Autologous Transplantation, Followed by Lenalidomide As ConsolidationMaintenance in Untreated Multiple Myeloma Patients. J Clin Oncol. 2010; 28(5):800-807. Offidani M, Leoni P, Corvatta L, et al. ThaDD plus high dose therapy and autologous stem cell transplantation does not appear superior to ThaDD plus maintenance in elderly patients with de novo multiple myeloma. Eur J Haematol. 2010; 84(6):474-483. San Miguel JF, Schlag R, Nuriet K, et al. Bortezomib plus Melphalan and Prednisone for Initial Treatment of Multiple Myeloma. N Engl J Med. 2008;359(9):906917. Roussel M, Lauwers-Cances V, Robillard N, et al. Front-Line Transplantation Program With Lenalidomide, Bortezomib, and Dexamethasone Combination As Induction and Consolidation Followed by Lenalidomide Maintenance in Patients With Multiple Myeloma: A Phase II Study by the Intergroupe Francophone du Myelome. J Clin Oncol. 2014;32(25):27122717. Engelhardt M, Dold SM, Ihorst G, Zober A, et al. Geriatric assessment in multiple myeloma patients: validation of the International Myeloma Working Group (IMWG) score and comparison with other common comorbidity scores. Haematologica. 2016;101(9):1110-1119.

1397

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Cell Therapy and Immunotherapy

Ferrata Storti Foundation

Haematologica 2016 Volume 101(11):1398-1406

Autotransplant with and without induction chemotherapy in older multiple myeloma patients: long-term outcome of a randomized trial

Christian Straka,1,2 Peter Liebisch,3 Hans Salwender,4 Burkhard Hennemann,5 Bernd Metzner,6 Stefan Knop,7,8 Sigrid Adler-Reichel,2 Christian Gerecke,9 Hannes Wandt,10 Martin Bentz,11 Tim Hendrik Bruemmendorf,12 Marcus Hentrich,13 Michael Pfreundschuh,14 Hans-Heinrich Wolf,15 Orhan Sezer,16 Ralf Bargou,8,16 Wolfram Jung,17 Lorenz Trümper,17 Bernd Hertenstein,18 Else Heidemann,19 Helga Bernhard,20 Nicola Lang,21 Norbert Frickhofen,22 Holger Hebart,23 Ralf Schmidmaier,2 Andreas Sandermann,24 Tobias Dechow,20 Albrecht Reichle,5 Brigitte Schnabel,1,2 Kerstin Schäfer-Eckart,10 Christian Langer,3 Martin Gramatzki,25 Axel Hinke,24 Bertold Emmerich,2 and Hermann Einsele7,8

Schön Klinik Starnberger See, Berg; 2Medizinische Klinik und Poliklinik IV, Klinikum der Universität München (LMU); 3Universitätsklinikum Ulm; 4Asklepios Klinik Altona, Hamburg; 5Universitätsklinikum Regensburg; 6Klinikum Oldenburg, Oldenburg; 7 Universitätsklinikum Tübingen; 8Universitätsklinikum Würzburg; 9HELIOS Klinikum Berlin-Buch, Berlin; 10Klinikum Nürnberg Nord, Nürnberg; 11Städtisches Klinikum Karlsruhe; 12Universitätsklinikum Hamburg-Eppendorf, Hamburg; 13Städtisches Klinikum München-Harlaching, München; 14Universitätsklinikum des Saarlandes, Homburg; 15 Universitätsklinikum Halle; 16Universitätsklinikum Charité, Berlin; 17Universitätsklinikum Göttingen; 18Klinikum Bremen-Mitte, Bremen; 19Diakonie Klinikum Stuttgart; 20Klinikum rechts der Isar, Technische Universität München; 21Medizinische Klinik und Poliklinik III, Klinikum der Universität München (LMU); 22HSK Dr. Horst Schmidt Klinik, Wiesbaden; 23 Stauferklinikum Schwäbisch Gmünd, Mutlangen; 24WISP Research Institute, Langenfeld; and 25Universitätsklinikum Kiel, Germany

ABSTRACT

Correspondence: cstraka@schoen-kliniken.de

Received: July 1, 2016. Accepted: August 3, 2016. Pre-published: August 4, 2016. doi:10.3324/haematol.2016.151860

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1398

1398

utologous transplantation is controversial for older patients with multiple myeloma. The role of age-adjusted high-dose melphalan and the impact of induction chemotherapy cycles is still unclear. A total of 434 patients aged 60-70 years were randomly assigned to 4 cycles of standard anthracycline-based induction chemotherapy or no induction. For all patients, double autologous transplantation after melphalan 140 mg/m2 (MEL140) was planned. The primary end point was progression-free survival. Of 420 eligible patients, 85% received a first transplant and 69% completed double transplantation. Treatment duration was short with a median of 7.7 months with induction chemotherapy cycles and 4.6 months without induction. On an intention-to-treat basis, median progression-free survival with induction chemotherapy cycles (207 patients) was 21.4 months versus 20.0 months with no induction cycles (213 patients) (hazard ratio 1.04, 95% confidence interval 0.84-1.28; P=0.36). Per protocol, progression-free survival was 23.7 months versus 23.0 months (P=0.28). Patients aged 65 years or over (55%) did not have an inferior outcome. Patients with low-risk cytogenetics [absence of del17p13, t(4;14) and 1q21 gains] showed a favorable overall survival and included the patients with sustained first remission. MEL140 was associated with a low rate of severe mucositis (10%) and treatment-related deaths (1%). Based on hazard ratio, the short treatment arm consisting of mobilization chemotherapy and tandem MEL140 achieved 96% of the progression-free survival, demonstrating its value as an independent component of therapy in older patients with multiple myeloma who are considered fit for autologous transplantation. (clinicaltrials.gov identifier: 02288741) haematologica | 2016; 101(11)

Autotransplant in older multiple myeloma patients

Introduction High-dose chemotherapy with autologous stem cell transplantation (ASCT) is considered the standard treatment for younger patients with multiple myeloma (MM).13 In older patients, however, its role is less clear and remains controversial.4-6 As a consequence, many older but otherwise fit patients are excluded from the procedure. This may have contributed to a survival disadvantage.7 Concerns about toxicity or an inferior outcome in comparison to younger patients continue to inhibit the application of high-dose therapy with ASCT in older patients, although its use has increased considerably in recent years.8,9 Only limited data are available as these patients are under-represented or missing entirely from the large recent prospective multicenter transplantation trials which mostly apply an age cut off of 65 years.10-17 The little information there is concerning higher age groups is based on retrospective single center experience. These mostly report limited numbers of patients.18-23 Registry data are lacking details of toxicity and outcome but demonstrate improvements in overall survival (OS) over the years.8,9 Melphalan 200 mg/m2 (MEL200) represents the standard high-dose regimen for the younger patient population.24 The frequent reluctance to apply MEL200 in patients over the age of 65 years is related to concerns about potential higher toxicity. As an alternative to MEL200, an ageadjusted melphalan dose of 140 mg/m2 (MEL140)25 can be given in patients over 60 years of age; the intention is to decrease severe mucositis and other toxicities and thereby to enable older patients who are not considered eligible for MEL200 to proceed to ASCT. Reports show that this strategy has now become part of current clinical routine.9,18,20,21,23 The proportion of patients receiving an ageadjustment of the melphalan dose is steadily increasing within the higher age groups when considering patients over the age of 60, over 65 and over 70 years.9,20,21,23 If ageadjustment were to be applied consistently, many more older patients could be considered candidates for ASCT. Two randomized clinical trials investigated intermediate-dose melphalan (MEL100) with ASCT in older patients.5,26 Data from a prospective randomized trial specifically reporting the efficacy and toxicity of MEL140 in older patients has been lacking; this study aims to provide this missing information. Historically, stem cell therapy is preceded by 3-6 cycles of induction chemotherapy. This strategy is considered to be important but since the progression-free survival (PFS) achieved after ASCT is achieved from the complete treatment, the contribution of induction chemotherapy alone still has to be defined. In the prospective phase III trial presented here, we addressed: 1) the role of conventional induction chemotherapy cycles prior to high-dose chemotherapy by randomization between anthracycline-based induction (the standard therapeutic approach when the trial started) and no induction cycles; and 2) the real toxicity and efficacy of tandem MEL140 with and without induction chemotherapy in a large older patient population.

Methods Patients and study design This randomized multicenter trial was planned by the German Multiple Myeloma Study Group (DSMM) and was conducted at haematologica | 2016; 101(11)

40 sites. Eligible patients had newly diagnosed stage II or III MM according to Durie and Salmon and were 60-70 years of age. Additional criteria for inclusion were an Eastern Cooperative Oncology Group performance status of 0-2, adequate organ function, and absence of uncontrolled infection. Enrollment began in August 2001 and ended in August 2006. Patients with no previous chemotherapy or a maximum of one cycle were randomly assigned between conventional induction chemotherapy cycles and a short course of dexamethasone only (Figure 1A). The study was performed in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines, and was approved by the local ethics committees at each participating center. Patients were required to provide written informed consent before enrollment. The study was registered at clinicaltrials.gov identifier: 02288741.

Treatment plan In the induction arm, patients received 4 cycles of conventional anthracycline-dexamethasone-based regimens: vincristine-doxorubicin-dexamethasone (VAD),27 idarubicin-dexamethasone (ID),28 cyclophosphamide-doxorubicin-dexamethasone (CAD).29 In the no induction arm, patients received only 40 mg oral dexamethasone on days 1-4 and 8-11 for symptom control. For the subsequent stem-cell mobilization, age-adjusted (75% dose) ifosfamide-epirubicin-etoposide (IEV) with granulocyte-colony stimulating factor (G-CSF) was recommended.30 The target dose for stem cell collection was 6x106 CD34-positive cells/kg (two transplants and one back-up). The standard dose for each transplantation was 2x106 CD34-positive cells/kg. High-dose melphalan at a total dose of 140 mg/m2 (MEL140) was given in two doses of 70 mg/m2 on days -3 and -2. ASCT was performed on day 0. A second MEL140 course was planned two months after the first. No maintenance treatment was given but regular bisphosphonate administration was recommended.

Sample size and statistical aspects The primary study end point was PFS calculated from the time point of randomization. To detect a 10-month advantage in PFS for the induction arm [24-34 months, corresponding to a hazard ratio (HR) of 0.71] with a power of 80% and based on a one-sided type I error rate of 0.05, at least 132 patients were required per randomization arm. Toxicity of treatment was evaluated using Common Toxicity Criteria (v.2.0, 1999), the definition of remission followed European Group for Blood and Marrow Transplantation (EBMT) criteria.31 Except for the primary end point, all analyses were descriptive or explorative in nature, providing two-sided P-values without referring to a specified error level. No adjustments were made for multiple testing. Proportions were eventually compared using Fisher's exact or Ď&#x2021;Â˛ test. All time-to-event end points were calculated by the Kaplan-Meier method. Survival curves were compared using the log-rank test. HR with confidence intervals were derived from Cox models.

Results Patients Figure 1 shows the study design and the consort diagram. A total of 434 patients were enrolled into the study protocol and were randomized. Fourteen patients (3%) were excluded from analysis. Accordingly, 420 patients could be analyzed with respect to the primary end point: PFS. The median follow-up period was 5.2 years (range 010.1 years). Details of consecutive treatment steps were 1399

C. Straka et al.

documented in 416 patients. Baseline characteristics of patients are shown in Table 1.

Induction chemotherapy In the induction arm (n=207 patients), 98% received one of the recommended induction regimens (idarubicin-dexamethasone 67%, vincristine-doxorubicin-dexamethasone 25%, cyclophosphamide-doxorubicin-dexamethasone 6%), 2% received dexamethasone alone. A median of 4 cycles (range 1-7 cycles) were given for a median of

3.9 months (range 0.3-12.3 months). In the no induction arm, the 2 cycles of 4x40 mg dexamethasone were given for a median of 0.7 months (range 0-5.7 months) before stem cell mobilization was initiated.

Stem-cell mobilization and ASCT A total of 385 patients (92%) were treated with stem cell mobilization chemotherapy: ifosfamide-epirubicinetoposide in 89%, cyclophosphamide-doxorubicin-dexamethasone in 5%, cyclophosphamide in 4%, cyclophos-

Randomized (n=434)

Analyzed for primary objective (n=207)

Analyzed for primary objective (n=213)

Figure 1. Study design and consort diagram. (A) Study design. Patients were randomized (R) to the two study arms: 1) induction chemotherapy cycles; and 2) no induction cycles. (B) Consort diagram. Inclusion, randomization, treatment and follow up of enrolled patients. PBSCs: peripheral blood stem cells; ICC: induction chemotherapy cycles; Dex: dexamethasone; Mob: mobilization chemotherapy; MEL140: high-dose melphalan.

1400

haematologica | 2016; 101(11)

Autotransplant in older multiple myeloma patients Table 1. Patients’ characteristics.

Randomization Characteristic

Age, years Median Range Sex, n. (%) Male Female Durie and Salmon stage, n. (%) I II III A B Undefined M-component, n. (%) IgG IgA IgM IgD Kappa light chain only Lambda light chain only Non-secretory myeloma Hemoglobin, g/dL Mean ± SD Median Range Platelets, x109/L Mean ± SD Median Range Serum creatinine, mg/dL Mean ± SD Median Range β2-microglobulin, mg/L Mean ± SD Median Range < 3.0 ≥ 3.0 Unknown ≥ 5.5 (*ISS stage III) Lytic bone lesions, n. (%) Present Absent Unknown Cytogenetics,° n./total (%) Adverse Deletion 17p13 Translocation (4;14) Translocation (14;16) Amplification +1q21.2 Other Translocation (11;14) † Hyperdiploid myeloma ǂ Deletion 13q14 Risk profile ll High $ Intermediate ¶ Low

Induction chemotherapy (n=207)

No induction (n=213)

65 (60-72)

119 (57) 88 (43)

117 (55) 96 (45)

0 (0) 38 (18) 169 (82) 185 (89) 22 (11) -

3 (1) 45 (21) 164 (77) 173 (81) 39 (18) 1 (1)

135 (65) 40 (19) 2 (1) 2 (1) 21 (10) 5 (3) 2 (1)

132 (62) 49 (23) 0 (0) 1 (1) 18 (8) 10 (5) 3 (1)

11.3±1.8 11.3 6.3 - 16.1

10.9±1.8 11.0 5.7-15.8

245±93 240 69-757

237±94 227 54-540

1.1±0.4 1.0 0.6-3.3

1.4±1.1 1.0 0.4-9.6

4.6±5.0 3.3 0.2-46.7 85 (41) 110 (53) 12 (6) 45 (22)

5.1±5.4 3.8 0.9-54.6 71 (33) 124 (58) 18 (9) 52 (24)

168 (81) 37 (18) 2 (1)

165 (77) 46 (22) 2 (1)

9/116 (8) 17/112 (15) 2/ 98 (2) 35/116 (30)

10/104 (10) 2/100 (12) 5/ 93 (5) 43/102 (42)

18/100 (18) 66/116 (57) 56/161 (35)

14/ 93 (15) 64/102 (63) 53/148 (36)

25/111 (23) 26/111 (23) 60/111 (54)

19/99 (19) 32/99 (32) 48/ 99 (48)

*ISS: International Staging System. †Hyperdiploid myeloma was assessed by the presence of gains of 9q34.2 Data from either central or local cytogenetic laboratory. llHigh: presence of t (4;14) and/or del 17p13. §Intermediate: presence of +1q21.2 and absence of t(4;14) and del 17p13. ¶Low: absence of t (4;14), del 17p13 and +1q21.2 in patients with analysis of all these three characteristics. °Cut-off of aberrant cells: 20%; SD: standard deviation.

haematologica | 2016; 101(11)

1401

C. Straka et al.

A Progression-free survival rate

phamide-etoposide in 1%, ifosfamide in 1%. Stem cells were collected in 376 patients. The recommended target value (â&#x2030;Ľ6x106 CD34+ cells/kg) and the required dose for a double transplantation (â&#x2030;Ľ4x106 CD34+ cells/kg) were achieved in 80% and 90% of patients, respectively. A total of 357 patients (85%) then went on to receive at least one transplant and 289 patients (69%) completed a double transplant. The main reasons for drop-out in the 68 patients (16%) who did not receive the second transplant were: progression of disease (21%), toxicity (22%), death (7%), patient refusal (18%), insufficient stem cell collection (7%).

Toxicity, deaths and second primary malignancies

Response rates Response rates reflected differences in treatment and treatment progress between patients in the two arms (Table 3). There was an initial lag in response rate in the no induction arm. However, this recovered in the following treatment steps, and after the second MEL140 response rates were similar in the two arms.

Long-term outcomes after randomization Median PFS for patients in the induction arm was 21.4 months compared to 20.0 months for patients in the no induction arm [hazard ratio (HR) for progression or death 1.04, 95% confidence interval (CI) 0.84-1.28; P=0.36] (Figure 2A). Therefore, based on this HR, 96% of the duration of PFS was already achieved by ASCT alone; induction chemotherapy cycles contributed only 4%. Treatment duration was short with a median of 7.7 months in patients in the induction arm and a median of 4.6 months in patients in the no induction arm. For double transplant recipients (per protocol), median PFS was 23.7 months for patients in the induction arm compared to 23.0 months for patients in the no induction arm (P=0.28) (Figure 2B). In the intention-to-treat analysis, median OS for patients in the induction arm was 53.4 months compared to 55.9 months for patients in the no induction arm (HR for death 1.01, 95%CI: 0.77-1.32; P=0.95). Per protocol (double transplants), the median OS 1402

B Progression-free survival rate

The major grade III/IV non-hematologic toxicities were infection and mucositis (Table 2). Deaths up to 100 days from the last treatment occurred in 25 patients (6.0%) and death was due to: disease progression in 5 patients (1.2%), toxicity (infection, sepsis, renal failure, cardiac) in 20 patients (4.8%). In 20 of 25 cases, death occurred before the first MEL140. Transplant-related mortality (TRM) was very low: 1.4% after the first MEL140 and, notably, 0% following the second MEL140. Some cases of grade III/IV mucositis were seen during induction chemotherapy cycles (4%), but none were observed during the short dexamethasone pre-phase. Comparison of the induction and the no induction arms showed: 2% versus 6% grade III/IV mucositis after mobilization, 11% versus 16% after the first MEL140 course, 5% versus 7% for the second MEL140, respectively. Moderate rates of grade III/IV infections occurred during induction chemotherapy cycles or the dexamethasone pre-phase (17% vs. 4%), after mobilization (17% vs. 30%), and at a higher frequency after the first MEL140 (35% vs. 44%) and second MEL140 (38% vs. 34%). A second primary malignancy was reported (2 solid tumors, 2 acute myeloid leukemias) in 4 of 420 patients (1%).

Figure 2. Long-term outcomes after randomization between induction chemotherapy cycles and no induction cycles. The curves for progression-free survival with induction chemotherapy cycles versus no induction cycles are shown based on intention-to-treat (A) and per protocol (tandem transplants) (B).

for patients in the induction arm was 68.5 months compared to 64.4 for patients in the no induction arm (P=0.98). A subgroup of 27 patients (6.4%) from both arms were survivors in first remission at five years. Among these 27 patients, characterized by the presence of low-risk cytogenetics (100%), 96% received MEL140, 85% received a double transplant, and 70% were not in ISS stage III. In 16 of these patients, no relapses were seen beyond five years.

Treatment effects according to age subgroups The discontinuation rate before the first high-dose melphalan course was higher for patients aged 65-70 years (18%) than for those aged 60-64 years (9%). The drop-out rate after the first MEL140 was similar for the younger (18%) and the older (17%) patients. More patients aged 60-64 years than aged 65-70 years completed tandem MEL140 (73% vs. 65%). Deaths within 100 days from the last treatment were more frequent during the induction phase in patients aged 65-70 years (6.1% vs. 3.1%), but were similar in both age groups following the first transplant (1.6% vs. 1.1%) and did not occur in either age haematologica | 2016; 101(11)

Autotransplant in older multiple myeloma patients

Table 2. Hematologic and non-hematologic toxicity during treatment steps.

Induction phase Induction No induction Hematologic (%) Anemia Leukopenia Neutropenia Grade III/IV Thrombocytopenia Grade III/IV Non-hematologic (%) Nausea Grade III/IV Vomiting Diarrhea Constipation Mucositis Grade III/IV Infection Grade III/IV Dyspnea Creatinine

Mobilization Induction No induction

1. MEL 140 Induction No induction

2. MEL 140 Induction No induction

84 59 47 22 27 5

64 26 15 5 16 2

89 86 77 63 77 40

95 88 82 75 84 44

100 100 100 100 100 100

32 2 15 11 15 17 4 33 17 11 18

10 0 1 2 5 3 0 14 4 6 22

45 1 18 13 7 23 2 31 17 5 17

45 4 23 18 9 28 6 40 30 11 23

70 7 36 41 10 51 11 49 35 13 25

72 14 46 42 9 51 16 58 44 10 19

71 5 39 31 12 46 5 56 38 10 16

72 9 43 34 8 49 7 49 34 6 16

MEL140: melphalan 140 mg/m2.

Table 3. Responses with consecutive treatment steps according to EBMT criteria.

Induction phase Induction No induction *

ORR (%) CR (%) ‡ PR (%) § MR (%) NC (%) ¶ PD (%) †

48 0 48 19 25 7

30 1 29 14 52 4

Mobilization Induction No induction 58 2 56 20 20 2

1. MEl140 Induction No induction

46 2 44 19 32 3

78 6 72 13 6 3

79 9 70 12 8 2

2. MEL140 Induction No induction 87 17 70 7 5 2

87 12 75 8 5 0

EBMT: European Society for Blood and Marrow Transplantation; MEL140: melphalan 140 mg/m2.*ORR: overall response rate (CR+PR); †CR: complete response; ‡PR: partial response; §MR: minimal response; ǁNC: no change; ¶PD: progressive disease.

group after the second transplant. On the intention-totreat basis, median PFS in younger versus older patients (19.5 and 22.1 months; P=0.23) and median OS (56.3 and 53.1 months; P=0.58) were similar (Figure 3A).

Cytogenetic risk and outcome Of the 210 evaluated patients, 108 patients (51%) fulfilled the IMWG32,33 low-risk cytogenetics criteria: absence of translocation (4;14), deletion 17p13, amplification of 1q21.2) (Table 1). High-risk cytogenetics [presence of translocation (4;14) and/or deletion 17p13] were found in 21%, intermediate-risk cytogenetics [presence of amplification of 1q21.2, absence of translocation (4;14) and deletion (17p13)] were present in 28% of cases. A highly significant difference between low-risk and high-risk cytogenetics was seen for PFS (median 23.5 vs. 14.9 months; P<0.0001) and OS (median 74.7 vs. 32.9 months; P<0.0001) (Figure 3B). The patients with low-risk cytogenetics who completed double transplantation (78%) had an excellent outcome (median PFS 26.7 months and median OS 87.4 months). Patients still alive in sustained first remission beyond five years were considered to be in the low-risk cytogenetics group. haematologica | 2016; 101(11)

Discussion Our trial abandoned induction chemotherapy cycles before ASCT in one arm of the study in order to test the relationship between induction chemotherapy and ASCT, and did not use consolidation or maintenance therapy. This must be considered when comparing the outcome of our trial to others which applied the common strategy of incorporating all available drugs into treatment phases before and after ASCT with the aim of achieving a maximum duration of PFS after first-line treatment in MM patients. Unexpectedly, we found that a large part of the anti-tumor effect (96%) was achieved with high-dose chemotherapy and ASCT alone, and that the 4 cycles of anthracycline-based induction chemotherapy only made a small contribution. By 3-4 months of anthracycline-based induction therapy, the PFS was improved by only two months. Based on the HR of the comparison between the two arms, such induction chemotherapy cycles achieved only 4% of the PFS. Anthracycline-based induction chemotherapy was the standard approach when our trial started but has since been replaced by induction regimens incorporating novel agents. In this respect, both arms of 1403

C. Straka et al.

our trial may be regarded as a 'baseline' from which the achievements of novel agent-based induction therapy can be evaluated. The current use of bortezomib-based induction regimens,13,11,15,34-36 were found to increase the PFS over the typical anthracycline-based regimen VAD13,15 by approximately six months. Despite this improvement, the estimated impact of such induction therapy on the overall PFS appears to be limited when we consider our 'baseline' results and those from other trials demonstrating a high efficacy of post-transplant treatments with novel agents.10,12,14 Our study is the first prospective randomized trial to characterize the 'real' toxicity and efficacy of MEL140 with ASCT in older MM patients. The results highlight an independent role for ASCT in older patients. MEL140 was well-tolerated. The rate of severe mucositis was approximately 10%, TRM approximately 1%. It should be emphasized that following the treatment pause after the first MEL140, the non-hematologic toxicity of the second MEL140 appeared to be a little lower compared to the first MEL140 and the TRM was 0%. Importantly,

the long-term outcome identified a subgroup of patients who did not relapse after MEL140 even over a number of years. This may indicate some curative potential with ageadjusted high-dose melphalan. For more than 200 patients aged 65-70 years, the outcome was at least as good compared to the younger patients in this study, demonstrating that an age cut off at 65 years for MEL140 was not relevant. The long-term outcome for the tandem MEL140 component in conjunction with the large number of patients treated may serve as a point of reference for future trials. Geriatric assessment and comorbidity scores will be helpful to encourage autologous transplantation in many MM older patients.37,38 The low rate of second primary malignancies following tandem MEL140 in comparison to that in the published literature39 is noteworthy and could be related to the limited first-line treatment in the study population. Due to the significant efforts of a central laboratory, we can present up-to-date cytogenetic data for around 200 patients. Many publications define high-risk versus stan-

Figure 3. Outcome for age and cytogenetic risk groups. The curves for progression-free survival and overall survival are shown for (A) the age groups 60-64 years and 65-70 years, and (B) for high-risk, intermediate-risk and low-risk cytogenetics.

1404

haematologica | 2016; 101(11)

Autotransplant in older multiple myeloma patients

dard-risk groups but very few reports indicate a low-risk group as defined by the IMWG.33,34 This requires investigation of 1q21 gains together with analysis of translocation (4;14) and deletion 17p. We found that 50% of the MM patients of this age group who had a low-risk cytogenetic profile showed excellent survival rates with front-line ageadjusted high-dose melphalan. Importantly, the patients with low-risk cytogenetics included those with sustained unmaintained first remission, and some of these could be considered 'cured'. Recently, treatment with novel agents plus ASCT has been found to be more effective than novel agents plus conventional chemotherapy in patients up to 65 years of age.14,16,17 Similar randomized clinical trials are also needed in the older patient population and treatment should be based on tandem MEL140. Another important aspect to consider is that the concept of age-adjusted high-dose melphalan includes double transplantation, which compensates the dose reduction of single melphalan. The tandem MEL100 regimen divided the standard MEL200 dose in two parts and enabled older patients to proceed to ASCT with reduced toxicity.5,26,40 A tandem application of MEL200 in older patients, however, was found to decrease steeply with higher age due to the well-known risks associated with these patients.22 In contrast, in our trial, a rapid improvement in performance status post transplant allowed the second MEL140 course to be given after two months in approximately 80% of patients. Therefore, as far as feasibility is concerned, tandem MEL140 (cumulative melphalan dose 280 mg/m2) represents an alternative to single MEL200 (200 mg/m2) in the older patient population. Such a comparison should be investigated in a prospective randomized trial for patients aged 60-70 years. In fact, our trial shows that the treatment arm consisting of stem-cell mobilization chemotherapy followed by tandem MEL140 with ASCT, despite its short treatment time of 4-5 months, is extremely effective, with a PFS of 20-23 months. Current non-transplant regimens used in this age group, such as VMP41 or MPT42 or MPR43 or Rd,44 provide similar PFS rates (median 22, 20, 14 and 21 months,

References 1. Attal M, Harousseau JL, Stoppa AM, et al. A prospective, randomized trial of autologous bone marrow transplantation and chemotherapy in multiple myeloma. Intergroupe Français du Myélome. N Engl J Med. 1996;335(2):91-97. 2. Barlogie B, Jagannath S, Desikan KR, et al. Total therapy with tandem transplants for newly diagnosed multiple myeloma. Blood. 1999;93(1):55-65. 3. Child JA, Morgan GJ, Davies FE, et al. High-dose chemotherapy with hematopoietic stem-cell rescue for multiple myeloma. N Engl J Med. 2003;348(19):1875-1883. 4. Dumontet C, Ketterer N, Espinouse D, et al. Reduced progression-free survival in elderly patients receiving intensification with autologous peripheral blood stem cell reinfusion for multiple myeloma. Bone Marrow Transplant. 1998;21(10):10371041. 5. Facon T, Mary JY, Hulin C, et al. Melphalan

haematologica | 2016; 101(11)

respectively) but are associated with the development of neuropathy or thrombosis and thromboembolism and require a prolonged treatment time that often does exceed one year. On the other hand, our trial demonstrates that age-adjusted high-dose therapy is not necessarily 'aggressive', nor is this 'aggressiveness' observed in all patients. Older fit patients may, therefore, benefit from an ageadjusted transplant program that would allow longer unmaintained remissions following transplantation during which patients can enjoy freedom-from-therapy. When lenalidomide is continued as maintenance therapy in the MPRR43 or continues Rd44 regimens, the PFS can be extended. Similarly, lenalidomide maintenance can also be used after MEL140.10,12,14,40 The preference for MEL140 or MEL200 in subgroups of older patients is beyond the scope of the present paper as this has been discussed extensively in previous publications. The known arguments center around the specific features of MEL200 (higher toxicity, potentially higher efficacy, single transplantation, upper age limit around 70 years) versus MEL140 (lower toxicity, potentially lower efficacy, tandem transplantation possible, upper age limit around 75 years). Obviously, any valid recommendation can only be made on the basis of the availability of results from prospective randomized trials in specific age groups which, however, are completely lacking. Therefore, for the moment, MEL140 and MEL200 represent complementary rather than competing options for older patients. In conclusion, short-term tandem MEL140 with ASCT can lead to long-lasting unmaintained remission, and should be considered an independent component of myeloma therapy. A sustained first remission is associated with low-risk cytogenetics present in approximately 50% of patients. Funding Novartis provided financial support for conducting the study, but was not involved in data collection, data analysis or writing the manuscript. Molecular cytogenetic analysis was supported by the Deutsche José Carreras Leukämie-Stiftung.

and prednisone plus thalidomide versus melphalan and prednisone alone or reduced-intensity autologous stem cell transplantation in elderly patients with multiple myeloma (IFM 99-06): a randomised trial. Lancet. 2007;370(9594): 1209-1218. Siegel DS, Desikan KR, Mehta J, et al. Age is not a prognostic variable with autotransplants for multiple myeloma. Blood. 1999;93(1):51-54. 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. Auner HW, Szydlo R, Hoek J, et al. Trends in autologous hematopoietic cell transplantation for multiple myeloma in Europe: increased use and improved outcomes in elderly patients in recent years. Bone Marrow Transplant. 2015;50(2):209-215. Sharma M, Zhang MJ, Zhong X. Older patients with myeloma derive similar benefit from autologous transplantation. Biol

10.

11.

12.

13.

Blood Marrow Transplant. 2014;20(11): 1796-1803. Attal M, Lauwers-Cances V, Marit G, et al. Lenalidomide maintenance after stem-cell transplantation for multiple myeloma. N Engl J Med. 2012;366(19):1782-1791. Cavo M, Tacchetti P, Patriarca F, et al. Bortezomib with thalidomide plus dexamethasone compared with thalidomide plus dexamethasone as induction therapy before, and consolidation therapy after, double autologous stem-cell transplantation in newly diagnosed multiple myeloma: a randomised phase 3 study. Lancet. 2010;376(9758):2075-2085. McCarthy PL, Owzar K, Hofmeister CC, et al. Lenalidomide after stem-cell transplantation for multiple myeloma. N Engl J Med. 2012;366(19):1770-1781. Harousseau JL, Attal M, Avet-Loiseau H, et al. Bortezomib plus dexamethasone is superior to vincristine plus doxorubicin plus dexamethasone as induction treatment prior to autologous stem-cell transplantation in newly diagnosed multiple myeloma:

1405

C. Straka et al.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

1406

results of the IFM 2005-01 phase III trial. J Clin Oncol. 2010;28(30):4621-4629. Palumbo A, Cavallo F, Gay F, et al. Autologous transplantation and maintenance therapy in multiple myeloma. N Engl J Med. 2014;371(10):895-905. Sonneveld P, Schmidt-Wolf IG, van der Holt B, et al. Bortezomib induction and maintenance treatment in patients with newly diagnosed multiple myeloma: results of the randomized phase III HOVON-65/ GMMG-HD4 trial. J Clin Oncol. 2012;30(24):2946-2955. Attal M, Lauwers-Cances V, Hulin C, et al. Autologous Transplantation for Multiple Myeloma in the Era of New Drugs: A Phase III Study of the Intergroupe Francophone Du Myelome (IFM/DFCI 2009 Trial). Blood. 2015;126(23):391. Gay F, Magarotto V, Petrucci MT, et al. Autologous Transplantation Versus Cyclophosphamide-LenalidomidePrednisone Followed By LenalidomidePrednisone Versus Lenalidomide Maintenance in Multiple Myeloma: LongTerm Results of a Phase III Trial. Blood. 2015;126(23):392. Badros A, Barlogie B, Siegel E, et al. Autologous stem cell transplantation in elderly multiple myeloma patients over the age of 70 years. Br J Haematol. 2001;114(3):600-607. Bashir Q, Shah N, Parmar S, et al. Feasibility of autologous hematopoietic stem cell transplant in patients aged ≥70 years with multiple myeloma. Leuk Lymphoma. 2012;53(1):118-122. El Cheikh J, Kfoury E, Calmels B, et al. Age at transplantation and outcome after autologous stem cell transplantation in elderly patients with multiple myeloma. Hematol Oncol Stem Cell Ther. 2011;4(1):30-36. Gertz MA, Lacy MQ, Dispenzieri A, et al. Impact of age and serum creatinine value on outcome after autologous blood stem cell transplantation for patients with multiple myeloma. Bone Marrow Transplant. 2007;39(10):605-611. Merz M, Neben K, Raab MS, et al. Autologous stem cell transplantation for elderly patients with newly diagnosed multiple myeloma in the era of novel agents. Ann Oncol. 2014;25(1):189-195. Kumar SK, Dingli D, Lacy MQ, et al. Autologous stem cell transplantation in patients of 70 years and older with multiple myeloma: Results from a matched pair analysis. Am J Hematol. 2008;83(8):614617. Moreau P, Facon T, Attal M, et al. Comparison of 200 mg/m(2) melphalan

25.

26.

27.

28.

29.

30.

31.

32.

34.

and 8 Gy total body irradiation plus 140 mg/m(2) melphalan as conditioning regimens for peripheral blood stem cell transplantation in patients with newly diagnosed multiple myeloma: final analysis of the Intergroupe Francophone du Myélome 9502 randomized trial. Blood. 2002; 99(3):731-735. Selby PJ, McElwain TJ, Nandi AC, et al. Multiple myeloma treated with high dose intravenous melphalan. Br J Haematol. 1987;66(1):55-62. Palumbo A, Bringhen S, Petrucci MT, et al. Intermediate-dose melphalan improves survival of myeloma patients aged 50 to 70: results of a randomized controlled trial. Blood. 2004;104(10):3052-3057. Barlogie B, Smith L, Alexanian R. Effective treatment of advanced multiple myeloma refractory to alkylating agents. N Engl J Med. 1984;310(21):1353-1356. Clark AD, Douglas KW, Mitchell LD, et al. Dose escalation therapy in previously untreated patients with multiple myeloma following Z-Dex induction treatment. Br J Haematol. 2002;117(3):605-612. Szelényi H, Kreuser ED, Keilholz U, et al. Cyclophosphamide, adriamycin and dexamethasone (CAD) is a highly effective therapy for patients with advanced multiple myeloma. Ann Oncol. 2001;12(1):105-108. Straka C, Hebart H, Adler-Reichel S, Werding N, Emmerich B, Einsele H. Blood stem cell collections after mobilization with combination chemotherapy containing ifosfamide followed by G-CSF in multiple myeloma. Oncology. 2003;65(Suppl 2):94-98. Bladé J, Samson D, Reece D, et al. Criteria for evaluating disease response and progression in patients with multiple myeloma treated by high-dose therapy and haemopoietic stem cell transplantation. Myeloma Subcommittee of the EBMT. European Group for Blood and Marrow Transplant. Br J Haematol. 1998; 102(5):1115-1123. Chng WJ, Dispenzieri A, Chim CS, et al. IMWG consensus on risk stratification in multiple myeloma. Leukemia. 2014; 28(2):269-277. Avet-Loiseau H, Attal M, Campion L, et al. Long-term 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. Moreau P, Avet-Loiseau H, Facon T, et al. Bortezomib plus dexamethasone versus reduced-dose bortezomib, thalidomide plus dexamethasone as induction treatment

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

before autologous stem cell transplantation in newly diagnosed multiple myeloma. Blood. 2011;118(22):5752-5758. Rosiñol L, Oriol A, Teruel AI, et al. Superiority of bortezomib, thalidomide, and dexamethasone (VTD) as induction pretransplantation therapy in multiple myeloma: a randomized phase 3 PETHEMA/GEM study. Blood. 2012;120(8):15891596. Richardson PG, Weller E, Lonial S, et al. Lenalidomide, bortezomib, and dexamethasone combination therapy in patients with newly diagnosed multiple myeloma. Blood. 2010;116(5):679-686. Engelhardt M, Dold SM, Ihorst G, et al. Geriatric assessment in multiple myeloma patients: validation of the International Myeloma Working Group (IMWG) score and comparison with other common comorbidity scores. Haematologica. 2016 [In press]. Kleber M, Ihorst G, Gross B, et al. Validation of the Freiburg Comorbidity Index in 466 multiple myeloma patients and combination with the International Staging System are highly predictive for outcome. Clin Lymphoma Myeloma Leuk. 2013;13(5):541-551. Engelhardt M, Ihorst G, Landgren O, et al. Large registry analysis to accurately define second malignancy rates and risks in a wellcharacterized cohort of 744 consecutive multiple myeloma patients followed-up for 25 years. Haematologica. 2015; 100(10):1340-1349. Gay F, Magarotto V, Crippa C, et al. Bortezomib induction, reduced-intensity transplantation, and lenalidomide consolidation-maintenance for myeloma: updated results. Blood. 2013;122(8):1376-1383. San Miguel JF, Schlag R, Khuageva NK, et al. Bortezomib plus melphalan and prednisone for initial treatment of multiple myeloma. N Engl J Med. 2008;359(9):906917. Fayers PM, Palumbo A, Hulin C, et al. Thalidomide for previously untreated elderly patients with multiple myeloma: metaanalysis of 1685 individual patient data from 6 randomized clinical trials. Blood. 2011;118(5):1239-1247. Palumbo A, Hajek R, Delforge M, et al. Continuous lenalidomide treatment for newly diagnosed multiple myeloma. N Engl J Med. 2012;366(19):1759-1769. Benboubker L, Dimopoulos MA, Dispenzieri A, et al. Lenalidomide and dexamethasone in transplant-ineligible patients with myeloma. N Engl J Med. 2014;371(10):906-917.

haematologica | 2016; 101(11)

ARTICLE

Cell Therapy and Immunotherapy

Splenic pooling and loss of VCAM-1 causes an engraftment defect in patients with myelofibrosis after allogeneic hematopoietic stem cell transplantation

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Christina Hart,1* Sabine Klatt,1* Johann Barop,1 Gunnar Müller,1 Roland Schelker,1 Ernst Holler,1 Elisabeth Huber,2 Wolfgang Herr,1 and Jochen Grassinger1

Department of Hematology and Oncology, Internal Medicine III, University Hospital Regensburg; and 2Institute of Pathology, University Hospital of Regensburg, Germany

*CH and SK contributed equally to this work.

Haematologica 2016 Volume 101(11):1407-1416

ABSTRACT

yelofibrosis is a myeloproliferative neoplasm that results in cytopenia, bone marrow fibrosis and extramedullary hematopoiesis. Allogeneic hematopoietic stem cell transplantation is the only curative treatment but is associated with a risk of delayed engraftment and graft failure. In this study, patients with myelofibrosis (n=31) and acute myeloid leukemia (n=31) were analyzed for time to engraftment, graft failure and engraftment-related factors. Early and late neutrophil engraftment and late thrombocyte engraftment were significantly delayed in patients with myelofibrosis as compared to acute myeloid leukemia, and graft failure only occurred in myelofibrosis (6%). Only spleen size had a significant influence on engraftment efficiency in myelofibrosis patients. To analyze the cause for the engraftment defect, clearance of hematopoietic stem cells from peripheral blood was measured and immunohistological staining of bone marrow sections was performed. Numbers of circulating CD34+ were significantly reduced at early time points in myelofibrosis patients, whereas CD34+CD38– and colony-forming cells showed no significant difference in clearance. Staining of bone marrow sections for homing proteins revealed a loss of VCAM-1 in myelofibrosis with a corresponding significant increase in the level of soluble VCAM-1 within the peripheral blood. In conclusion, our data suggest that reduced engraftment and graft failure in myelofibrosis patients is caused by an early pooling of CD34+ hematopoietic stem cells in the spleen and a bone marrow homing defect caused by the loss of VCAM-1. Improved engraftment in myelofibrosis might be achieved by approaches that reduce spleen size and cleavage of VCAM-1 in these patients prior to hematopoietic stem cell transplantation.

Correspondence: christina.hart@ukr.de

Received: March 29, 2016. Accepted: August 3, 2016. Pre-published: August 4, 2016. doi:10.3324/haematol.2016.146811

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1407

Introduction Myelofibrosis (MF) is a rare chronic myeloproliferative neoplasm with an incidence of 0.22 to 0.99 per 100,000.1 MF appears de novo or as a progression of polycythemia vera (PV) or essential thrombocythemia (ET). MF is characterized by bone marrow (BM) fibrosis, extramedullary hematopoiesis with splenomegaly, and severe constitutional symptoms.2 As the prognosis of MF is heterogeneous, the Dynamic International Prognostic Scoring System (DIPSS) is widely used to stratify newly diagnosed MF patients prior to evaluation of therapy.3 Treatment options are conventional drugs including anti-proliferative medication, immunomodulatory drugs (iMiDs), Janus kinase (JAK) inhibitors and hematopoietic haematologica | 2016; 101(11)

1407

C. Hart et al.

growth factors. However, allogeneic hematopoietic stem cell transplantation (HSCT) is the only curative treatment that is recommended to transplant-eligible intermediate 2 and high-risk patients.4 A number of studies have reported on the successful use of HSCT following reduced-intensity conditioning (RIC) that induces lower treatment-related mortality than myeloablative conditioning (MAC).5,6 However, graft failure

(GF) of up to 10% in MF patients after RIC HSCT is a critical contributor to morbidity and mortality.7 Factors favorably affecting engraftment were shown to be splenectomy before transplantation, human leukocyte antigen (HLA) matched sibling donor, peripheral stem cell use and absence of pre-transplant thrombocytopenia.8,9 The reconstitution of the BM after transplantation depends on the successful homing of transplanted hematopoietic stem

Table 1. Patients' characteristics at the time of transplant.

Myelofibrosis (MF) patients (n=31) Median age (range), years Female, n (%) Disease, n (%) Primary MF Secondary MF JAK2V617F mutation status* Positive Negative Pre-treatment** Hydroxyurea only Hydroxyurea + additional therapy Other pre-treatment No cytotoxic treatment Response at time of transplant Progressive disease Stable disease Partial remission DIPPS, n (%) Low Intermediate-1 Intermediate-2 High Unknown Spleen size, n (%), median 21 cm (measured by ultrasonography) Greater than median Smaller than median Splenectomy, n (%) Bone marrow fibrosis according to Bauermeister scoring, n (%) Grade 1 Grade 2 Grade 3 Grade 4 Median number of transplanted CD34+cells/kg/BWx106, n (range) Conditioning, n (%) Myeloablative Reduced intensity Antithymoglobulin Donor, n (%) (on the basis of high HLA resolution testing: HLA-A, -B, -C, -DRB1 and-DQB1) Related Unrelated matched Unrelated mismatched Major ABO mismatch, n (%) GvHD prophylaxis, n (%) CSA + MTX CSA + MMF CSA alone

Acute myeloid leukemia patients (n=31) 54 (36-67) 13 (42) 17 (55) 14 (45)

Median age (range), years Female, n (%) Time of allo HSCT, n (%) Upfront after induction/consolidation therapy Relapse

56 (37-68) 16 (52)

Median number of transplanted CD34+cells/kg/BWx106, n (range) Conditioning, n (%) Myeloablative Reduced intensity Antithymoglobulin Donor, n (%)

5.98 (2.71-8.57)

15 (48) 16 (52)

5 (16) 5 (16) 13 (42) 7 (23) 3 (10) 8 (25) 16 (52) 6 (19) 1 (3) 3 (10) 3 (10) 22 (71) 2 (6) 1 (3) 13 (42) 13 (42) 5 (16) 3 (10) 4 (13) 6 (16) 18 (58) 5.81 (3.53-8.99) 4 (13) 27 (87) 29 (94)

8 (26) 16 (52) 7 (22) 11 (35) 18 (58) 10 (32) 3 (10)

Related Unrelated matched Unrelated mismatched Major ABO mismatch, n (%) GvHD prophylaxis, n (%) CSA + MTX CSA + MMF Everolimus

1 (3) 30 (97) 27 (87)

4 (13) 18 (58) 9 (29) 8 (26) 20 (65) 10 (32) 1 (3)

HSCT: hematopoietic stem cell transplantation; DIPPS: Dynamic International Prognostic Scoring System; BW: body weight; GvHD: graft-versus-host disease; CSA: cyclosporin; MTX: methotrexate.*JAK2V617F mutation was routinely tested after 2006. ** No pre-treatment with JAK2 inhibitor (JAK2 inhibitor was not available at this time).

1408

haematologica | 2016; 101(11)

Engraftment defect in patients with myelofibrosis

cells (HSC).10 After leaving the peripheral blood (PB), the HSC lodge into the stem cell niche that was shown to regulate the HSC pool.11 Several cellular components were identified to regulate the hematopoietic homeostasis, among them endothelial cells, mesenchymal stroma cells and osteoblasts.12,13 The latter were shown to express stromal cell-derived factor (SDF)-1 (CXCL12) and osteopontin (OPN) within the BM that control homing, quiescence and proliferation of HSC after transplantation.12,14 In MF, the BM microenvironment is modified by fibrosis, osteosclerosis and neo-angiogenesis. BM fibrosis is the result of the abnormal deposition of collagen produced by fibroblasts that are stimulated by pro-inflammatory cytokines and growth factors.15,16 Therefore, it can be speculated that the disarrangement of the BM niche in MF is one aspect of GF.17 In this study, we investigated neutrophil and platelet engraftment in patients with MF and acute myeloid leukemia (AML) following RIC-HSCT. In addition, factors that affect engraftment were evaluated. By measuring the number of circulating HSC at defined time points after transplantation we assessed the homing efficiency in patients with MF and AML. Finally, we analyzed BM extracellular components including chemokines and their

receptors expressed on HSC within fibrotic and nonfibrotic BM.

Methods Patients’, disease and transplantation characteristics A total of 31 patients diagnosed with MF and 31 age- and gender-matched patients with AML who underwent allogeneic HSCT in our department between 2000 and 2011 were retrospectively analyzed. Patients’ and disease characteristics, as well as donor and transplant procedures are summarized in Table 1. This study was approved by the local ethics committee (n. 02/220). For further information see the Online Supplementary Appendix.

Clearance of HSC and colony-forming cells from the PB after HSCT Clearance of CD34+, CD34+CD38– cells and colony-forming cells (CFC) from PB was measured at defined time points after infusion by flow cytometry and CFC assay from 5 MF and 5 AML patients. PB samples (2.5 mL) were taken prior to and 10, 20, 40, 80, 160 minutes (min), 6 and 22 hours (h), 3 days (d), and in some cases 5 d after transplantation.

P=0.001 P<0.001

P=0.09

P=0.01

Figure 1. Myelofibrosis (MF) patients show significant delayed early and late engraftment as compared to acute myeloid leukemia (AML) patients. Cumulative incidence of early (A) and late (B) neutrophil and early (C) and late (D) platelet engraftment in MF and AML patients.

haematologica | 2016; 101(11)

1409

C. Hart et al. Table 2. Characteristics of the myelofibrosis patients with primary or secondary graft failure.

Primary graft failure Patient # 1 Patient # 2

Conditioning

Type of donor

ABO blood group

Spleen size (cm)

RIC RIC

MMUR MUR

Minor mismatch Ident

11 > 20

MAC RIC RIC RIC RIC

MUR MUR MRD MUR MRD

Major mismatch Major mismatch Ident Major mismatch Major mismatch

Secondary graft failure Patient # 1 Patient # 2 Patient # 3 Patient # 4 Patient # 5

26 40 28 20 - (splenectomy)

MAC: myeloablative conditioning; RIC: reduced-intensity conditioning; MRD: matched related donor; MUR: matched unrelated donor; MMUR: mismatched unrelated donor. ; Ident: identical.

Flow cytometric analysis of circulating HSC To analyze the CD34+ and CD34+CD38– cell count, PB was lysed in NH4CL lysis buffer and cells were then stained for 30 min at 4°C with combinations of anti-CD45-FITC, anti-CD34-APC and anti-CD38-PE monoclonal antibodies. Analysis was made using a Becton Dickinson CALIBUR flow cytometer (BD, East Rutherford, NJ, USA). Detailed information is provided in the Online Supplementary Appendix.

ed. Flow cytometric data are presented as average + standard error of the mean (SEM). Student’s t-test, Mann-Whitney U-test or ANOVA were used when appropriate for determining statistical significance. Analyses were performed using SPSS (IBM, Ehningen, Germany) or PRISM (Graphpad, San Diego, CA, USA) statistical software.

Results CFC assay To measure the colony forming ability of transplanted cells, 1 mL of PB was processed as described above. Burst-forming and colony-forming units erythrocyte (BFU-E, CFU-E), CFU granulocyte-macrophage (CFU-GM) and CFU granulocyte-erythrocytemonocyte-macrophage (CFU-GEMM) were assayed as described before.18

Analysis of homing receptors on allogeneic HSC Mononuclear cells (MNC) from granulocyte-colony stimulating factor (G-CSF) mobilized allogeneic donors were isolated as described before.18 Expression of homing receptors (CD44, CD184, CD49d, CD49e and a9β1 integrin) was measured by flow cytometry. For further information see the Online Supplementary Appendix.

Immunohistochemistry Immunohistochemistry (IHC) on BM sections was performed with the Histofine Simple Stain MAX PO (Nichirei Biosciences INC, Tokyo, Japan) and DAB chromogen (ImmunoLogic, Duiven, The Netherlands) according to the manufacturer`s instructions. Expression of OPN, anti-intercellular adhesion molecule (ICAM)1, vascular cell adhesion molecule (VCAM)-1, CD34 and SDF-1 was analyzed using an immunohistological score. More information is available in the Online Supplementary Appendix.

ELISA for soluble VCAM-1 Soluble (s)VCAM-1 was analyzed in the serum of MF, AML and healthy controls as described in the Online Supplementary Appendix.

Statistical analysis Data are presented as median (range) and count (percentage). The probabilities of neutrophil and platelet engraftment, overall survival (OS) and non-relapse mortality (NRM) were calculated from date of transplant, according to the Kaplan-Meier productlimit method. NRM was defined as death without relapse. To determine factors affecting these end points, a log-rank test was performed and variables were significant at P≤0.05. Because of the small number of patients only univariable analyses were conduct1410

Engraftment The cumulative incidence of early and late neutrophil and platelet engraftment is shown in Figure 1. The mean time to early neutrophil engraftment in MF patients was 22 days (range 11-48) and in AML patients 17 days (range 10-23) (P=0.001) (Figure 1A). MF patients showed late neutrophil engraftment after a mean time of 35 days (range 13-65) and AML patients needed 25 days (range 12-55) (Figure 1B). Thirty percent of MF versus 7% of AML patients did not reach neutrophil engraftment by day 100 (P<0.001). The mean time to early platelet engraftment in MF patients was 31 days (range 9-92) and in AML patients 23 days (range 11-55) (P=0.094) (Figure 1C). MF patients showed late platelet engraftment after 33 days (range 11-96), whereas AML patients needed 27 days (range 12-73). Fifty-six percent of MF versus 23% of AML patients did not reach platelet engraftment by day 100 (P=0.01) (Figure 1D). GF only occurred in patients with MF. Primary GF was seen in 2 MF patients (6%) and secondary GF in 5 MF patients (16%) (Table 2). There was no significant difference in early and late neutrophil engraftment in MF patients with regard to donor type (Figure 2E and F). However, AML patients showed a significantly faster early (P<0.001) and late (P<0.01) neutrophil engraftment using MRD (Figure 2G and H). MF patients who had splenectomy before HSCT and MF patients with a spleen size smaller than the median of 21 cm (range 11-40 cm) showed a significantly faster early neutrophil engraftment (P=0.01 and P=0.03, respectively) (Figure 2A); however, spleen size or splenectomy had no impact on late neutrophil engraftment (Figure 2B). Regarding early platelet engraftment, there was no significant difference between splenectomy and spleen size smaller or larger than the median (Figure 2C), whereas for late platelet engraftment it was shown that patients after splenectomy had a significantly faster engraftment compared to patients with spleen size smaller (P=0.02) or larger than the median (P=0.03) (Figure 2D). For 19 MF haematologica | 2016; 101(11)

Engraftment defect in patients with myelofibrosis

patients, data on spleen size could be collected 2-4 months after transplantation and were reduced on average by 20.5% (5.4 cm, range 0.3-13 cm). Higher numbers of transplanted cells [> 6x106 cells/kg/body weight (BW)], blood group (ABO) match status, age, GvHD, BM fibrosis grade, and number of blasts were not associated with neutrophil engraftment.

Survival, relapse and non-relapse mortality With a follow up of 24 months after HSCT, 2-year OS

haematologica | 2016; 101(11)

was 54% (95%CI: 36-72) in MF patients and 58% (95%CI: 41-76) in AML patients (P>0.05) (Figure 3A). A total of 14 MF patients died. Causes of death were: graft failure (n=1), infection (n=6; including 2 patients with secondary GF), hemorrhage (n=1), GvHD (n=2), multi-organ failure (n=2) and relapse (n=1). In the AML group, 13 patients died due to relapse (n=6), infection (n=1), GvHD (n=3) and multi-organ failure (n=3). Two-year NRM is 42% (95%CI: 18-60) in MF and 23% (95%CI: 7-40) in AML patients, respectively.

Figure 2. Spleen size and splenectomy but not donor type are associated with improved engraftment in myelofibrosis (MF) patients. Cumulative incidence of engraftment according to spleen and donor characteristics. Early (A) and late (B) neutrophil and early (C) and late (D) platelet engraftment in MF patients with regard to spleen size. Early and late neutrophil engraftment in MF (E and F) and acute myeloid leukemia (G and H) patients with regard to donor type. MRD: matched related donor; MUD: matched unrelated donor; MMUD: mismatched unrelated donor.

1411

C. Hart et al.

Transfusions The average number of erythrocyte transfusions after 28 and 100 days was significantly higher in MF patients than in AML patients (12.2±1.1 and 20.7±2.7 vs. 6.6±0.6 and 9.1±1.0, respectively; P<0.05) (Figure 3B). Accordingly, average numbers of platelet transfusion (Figure 3C) was significantly higher in patients with MF as compared to AML (11.5±1.3 and 18.8±3.2 vs. 7.7±0.8 and 10.3±1.4, respectively; P<0.05).

Clearance of HSC after HSCT Our clinical data indicate that early and late neutrophil engraftment, as well as late but not early platelet engraftment, is significantly delayed in patients with MF as compared to AML. Therefore, we analyzed the number of circulating HSC within the PB by flow cytometry in 10 patients shortly before and after transplantation as a surrogate marker for the homing efficiency. Average age of these patients was 64.6±1.6 years (MF) and 50.0±1.8 years (AML). Spleen size was 18.7±3.1 cm (MF) and 12.2±2.0 cm (AML) (P<0.05). Figure 4A shows representative dotplots of CD34+ cells within the PB of one recipient. After 40 min, more than 75% of the transplanted CD34+ cells were cleared from the PB (Figure 4B), indicating a high homing efficiency. However, significant differences in the number of circulating cells were seen between MF and AML. After 10, 20 and 40 min 1475±244, 1066±127 and 456±61 CD34+ cells/mL blood circulate in AML as compared to 682±186, 489±162 and 306±81 cells in MF patients (P<0.05). Interestingly, the numbers equalize after 6 h and thereafter. At 22 h, 27±12 and 11±2 CD34+ cells/mL and after 3-5 days 24±10 and 10±3 CD34+ cells/mL were detectable for MF and AML patients, respectively. In addition, the number of CD34+CD38– cells in the PB was analyzed. There was no significant difference in circulating CD34+CD38– in MF compared to AML patients (Online Supplementary Figure S1). The proportion of CD38- cells within the CD34+ cell population was significantly (approx. 3.5 times) higher up to 160 min in MF patients as compared to AML patients (Figure 4C).

compared to AML patients cannot be sufficiently explained solely by the spleen size, we analyzed the expression of homing-related proteins within the BM prior to conditioning chemotherapy. As expected, increased BM cellularity and increased numbers of vessels visualized by CD34 staining was detected in MF patients. Further immunohistochemical analyses revealed no significant difference with regard to SDF-1, OPN or ICAM-1 expression between MF and AML BM (Online Supplementary Figure S2). However, a significant loss of VCAM-1 expression in MF patients was detected (Figure 5B). To determine whether this VCAM-1 loss persists after transplantation, 11 MF patient samples were scored before and 5 samples after HSCT. There was a significantly lower VCAM-1 expression in MF patients before HSCT (median IRS=1) as compared to corresponding samples of MF patients after transplantation (median IRS=6; P<0.05) (Figure 5C). AML patients (n=4, median IRS=6; P<0.05) and healthy controls (n=6) showed VCAM-1 expression levels in BM similar to post-transplant MF patients. We also studied VCAM-1 expression in the bone marrow of PV (n=5) and ET (n=4) patients. IRS values ranged from 1 to 9 (PV) and 1 to 4 (ET), respectively, and thus were similar to levels observed in AML patients and healthy controls (data not shown).

Soluble VCAM-1 in serum of MF and AML patients One possible explanation for the loss of VCAM-1 expression in MF patients is the cleavage by proteases

Clearance of CFC As the expression of CD34 and CD38 on HSC is of limited use with regard to the biological property of the cells, we compared the number of CFC between MF and AML patients at the same time points using a methylcellulose assay. (Figure 4D shows the representative example of white and red colonies; there is no significant difference in CFC numbers as shown in Figure 4E).

Expression of homing molecules on CD34+ cells To determine the expression of homing molecules on transplanted allogeneic PB-HSC, we analyzed the expression of crucial homing receptors, namely CD44, CD184, CD49d, CD49e and a9β1 integrin before infusion. Figure 5A shows one representative flow analysis. Taken together, more than 99% of CD34+ and CD34+CD38– cells express CD44, CD49d and CD49e; 93.6±3% and 87.5±6% of the cells express a9β1 and 40.4±6% and 40.2±10% express CXCR4 (CD184), respectively.

Immunohistochemistry of homing-related niche proteins Since our clinical and experimental data suggest that the decreased long-term engraftment characteristics in MF 1412

Figure 3. Overall survival is equal in myelofibrosis (MF) and acute myeloid leukemia (AML) patients but transfusion needs significantly differ. (A) KaplanMeier estimate of survival in MF and AML patients with a follow up of 24 months after allogeneic stem cell transplantation. Average number of erythrocyte (B) and platelet (C) transfusions in MF and AML patients on day 28 and 100 after allogeneic stem cell transplantation.

haematologica | 2016; 101(11)

Engraftment defect in patients with myelofibrosis

within the modified MF BM. To test this hypothesis, the level of sVCAM-1 within the MF patient serum was analyzed in comparison to AML patients and healthy controls. sVCAM-1 level was significantly higher in the serum of MF patients (n=8, 1672±288 ng/mL) as compared to AML patients (n=3, 747±106 ng/mL) (P<0.05) and healthy controls (n=8, 595±56 ng/mL) (P<0.05), respectively (Figure 5D).

Discussion This study presents retrospective data on neutrophil and platelet engraftment after allogeneic HSCT from 62 matched MF and AML patients. Our data indicate that MF patients show significantly delayed early and late neutrophil, as well as late platelet engraftment, compared to the AML cohort. Analysis of engraftment-related factors

E Figure 4. Clearance of CD34+ cells in myelofibrosis (MF) patients is significantly different to acute myeloid leukemia (AML) patients. Clearance of CD34+ cells after allogeneic stem cell transplantation in MF and AML patients at defined time points. (A) Representative dotplot of CD34 and CD38 stained HSC within the peripheral blood of one recipient. Clearance of CD34+ cells (B) and proportion of CD38– cells within the CD34+ cell fraction (C) in MF (n=5) and AML (n=5) patients. (D) Representative white (CFU-M) and red (BFU-E) colonies after hematopoietic stem cell transplantation. (E) Clearance of colony-forming cells (CFC) in MF and AML patients at defined time points after transplantation.

haematologica | 2016; 101(11)

1413

C. Hart et al.

revealed no correlation to the blood group (ABO), age, GvHD, BM fibrosis grade, number of blasts prior to transplantation, or donor source. Interestingly, Robin et al.8 and Rondelli et al.19 reported a significant difference in engraftment between MRD and MUD. This divergence might be

due to the somewhat smaller number of patients within our study. Interestingly, primary GF was seen in 2 MUD transplanted MF patients, whereas no difference with regard to donor source was detected in the rate of secondary GF.

1414

Figure 5. Peripheral blood (PB) hematopoietic stem cells (HSC) express common homing receptors but loss of VCAM-1 is detected in the bone marrow of myelofibrosis patients prior to hematopoietic stem cell transplantation (HSCT). (A) Flow cytometric analysis of the homing receptors CD44, CD184, CD49d, CD49e and a9Î˛1 integrin on CD34+CD38+ and CD34+CD38â&#x20AC;&#x201C; cells. (B) Representative immunohistology staining of VCAM-1 on acute myeloid leukemia (AML) and MF patients derived BM sections shortly before conditioning chemotherapy as compared to isotype control and one representative MF BM section eight months after HSCT. (C) Immunohistochemical rating score (IRS) for VCAM-1 expression in BM of MF patients (before HSCT n=11, after HSCT n=5), AML patients (n=4) and healthy controls (n=6). Each data point represents the individual VCAM-1 IRS and bar represents the median of sample groups. (D) Expression of soluble VCAM-1 in the serum of MF patients (n=8) as compared to AML patients (n=3) and healthy controls (n=8). Each data point represents the mean individual sVCAM-1 concentration assayed in duplicate and bar represents the mean of sample groups. *P<0.05

haematologica | 2016; 101(11)

Engraftment defect in patients with myelofibrosis

Previous transplantation data suggest that the time to engraftment is dependent on the number of transplanted CD34+ cells. In the allogeneic setting 2.5-11.0x106 CD34+ cells/kg/BW are considered safe.20 The median number of transplanted cells in our study was 5.81 (MF) and 5.98 (AML)x106 CD34+ cells per kg/BW, respectively, and did not correlate with the neutrophil engraftment. The cumulative incidence of NRM at two years is significantly higher in MF than in AML patients and is slightly higher as previously reported by Claudiani et al.21 after RIC. Fittingly, not only NRM, but also number of blood and platelet transfusions was significantly higher in MF patients implying that HSCT in these patients results in increased morbidity and costs. However, there was no significant difference in 2-year OS between MF and AML patients (54% and 58%, respectively). As allogeneic HSCT provides the only curative treatment option for MF so far, we analyzed factors affecting the transplantation outcome. Effective homing of human HSC into the BM is a prerequisite for successful engraftment after transplantation. After attaching to adhesion proteins on BM vessels, the transplanted HSC trans-migrate through the endothelium and marrow and finally lodge into the stem cell niche. This process is highly regulated by a dynamic interaction of chemokines and adhesion molecules to ensure a purposive homing and engraftment.22 We recently demonstrated that more than 80% of transplanted murine HSC home to the BM within 5 h.23 As our clinical data indicate that early engraftment of neutrophils is significantly delayed in MF patients, we used the clearance of the HSC from the PB as a surrogate marker for the homing of HSC. We show that more than 75% of the transplanted CD34+ cells exited the PB within 40 min and the number of circulating CD34+ cells after 22 h was the same as that prior to transplantation. A similar approach was used by Donmez et al.,24 showing that nearly all autologous transplanted CD34+ cells exit the PB within 24 h. Interestingly, the number of circulating CD34+ cells in MF patients was significantly lower at early time points (up to 80 min) compared to AML patients. This suggests that lineage specific committed CD34+ cells are initially pooled within the spleen of MF patients. This is supported by the fact that splenectomy before HSCT significantly accelerates early neutrophil engraftment. In this regard, the only factors favorably affecting engraftment in MF patients in our study were spleen size smaller than the median of 21 cm and splenectomy, as described before.6,8,9,25 On the other hand, pooling of CD34+ cells within the spleen is a key feature of myelofibrosis.26 Early animal studies proposed that neutrophil pooling strongly depends on the spleen size,27 therefore, low neutrophil count in MF patients is possibly a combined phenomena including reduced engraftment of precursors and pooling of mature neutrophils. In contrast, the CD34+CD38- HSC fraction that is enriched for primitive HSC28 showed no significant clearance difference between AML and MF patients at any time point, but there was a trend towards a higher number of circulating cells in MF. This finding is congruent with early animal data demonstrating that primitive murine HSC display a preferential homing to the BM rather than to the spleen.29 At later time points, the number of circulating HSC was higher in MF than in AML patients, yet this difference was not significant due to the very low numbers of cells within the PB measurable. It can be speculated that, firstly, shortterm engrafting HSC (ST-HSC), that are responsible for haematologica | 2016; 101(11)

early engraftment up to 8-12 weeks, preferentially home to the spleen in MF patients, being less supportive for early reconstitution than the BM. Secondly, long-term engrafting HSC (LT-HSC) show a prolonged circulation due to a homing defect eventually causing a reduced late engraftment. This is supported by the fact that spleen size or splenectomy had no impact on late neutrophil engraftment; however, we cannot prove this assumption as the ST- or LT-HSC phenotype30 of the circulating cells was not analyzed. To determine the cause for the engraftment defect we studied key homing molecules. One essential protein for HSC homing is SDF-1.12,31 Disruption of SDF-1 binding to its receptor CXCR4 expressed on HSC and suppression of SDF-1 in osteoblasts after G-CSF administration leads to a sustained mobilization of HSC into the PB.32,33 SDF-1 is upregulated in MF patient spleens, possibly explaining the preferential homing of ST-HSC to this extramedullary site.34,35 On the other hand, Migliaccio et al. demonstrated in gata-1 deficient MF mice and also MF patients a higher SDF-1 expression within the BM.36 Moreover, we recently demonstrated that OPN, expressed by osteoblasts within the endosteal niche, also has chemotactic activity.37 However, immunohistology of BM sections obtained before transplantation did not show any significant difference in SDF-1 and OPN expression in AML and MF patients. As adherence of circulating HSC to endothelial cells via VCAM-1 and ICAM-138 induces the homing process, we further analyzed the expression of these proteins. Whereas no difference in ICAM-1 expression was seen, there was a significant loss of VCAM-1 in BM samples from MF patients as compared to AML patients. After transplantation and reconstitution of BM, the VCAM-1 expression increased to normal levels. VCAM-1 is commonly expressed by BM stromal cells and endothelial cells,39 and is a key protein for the adhesion of HSC to the endothelium before migration into the BM.38 Cleavage of VCAM-1 by metalloproteases after application of G-CSF results in HSC mobilization.40 In this context, data of Xu et al.41 indicate that a proteolytic environment within the BM of MF patients results in the cleavage of VCAM-1 and increased plasma level of cleaved VCAM-1, which is in accordance with our findings. This leads to the constitutive mobilization of CD34+ cells into the PB. Therefore, one can suggest that the cleavage of VCAM-1 not only results in an increased mobilization of steady state recipient HSC in MF patients, but also causes a homing defect responsible for the reduced engraftment of LT-HSC. As an engraftment defect not only results in increased risk of therapy-related mortality but also higher costs for blood products and antibiotics, approaches to accelerate reconstitution should be discussed. One possible method would be to improve HSC homing by reduction of spleen size and of the cleavage of VCAM-1 by reducing the proteolytic activity within the BM. Ruxolitinib, the only commercially available JAK-2 inhibitor approved as therapy for MF, leads to a modulation of BM microenvironment, reduces fibrosis, and has an anti-inflammatory action.42-44 First results for patients pre-treated with the JAK-2 inhibitor before HSCT were presented by Jaekel et al.45 These showed a significant reduction of spleen size and cytokine-induced clinical side effects. However, primary engraftment failure was seen in 7% of the patients, as compared to 6% in our study, and no data on time to engraftment was given. Thus, further studies evaluating other anti-inflammatory drugs are needed. 1415

C. Hart et al.

References 1. Titmarsh GJ, Duncombe AS, McMullin MF, et al. How common are myeloproliferative neoplasms? A systematic review and metaanalysis. Am J Hematol. 2014;89(6):581587. 2. Mesa RA, Verstovsek S, Cervantes F, et al. Primary myelofibrosis (PMF), post polycythemia vera myelofibrosis (post-PV MF), post essential thrombocythemia myelofibrosis (post-ET MF), blast phase PMF (PMF-BP): Consensus on terminology by the international working group for myelofibrosis research and treatment (IWG-MRT). Leuk Res. 2007;31(6):737-740. 3. Passamonti F, Cervantes F, Vannucchi AM, et al. A dynamic prognostic model to predict survival in primary myelofibrosis: a study by the IWG-MRT (International Working Group for Myeloproliferative Neoplasms Research and Treatment). Blood. 2010;115(9):1703-1708. 4. Tefferi A. How I treat myelofibrosis. Blood. 2011;117(13):3494-3504. 5. Alchalby H, Kroger N. Reduced-intensity conditioning followed by allogeneic hematopoietic stem cell transplantation in myelofibrosis. Curr Hematol Malig Rep. 2010;5(2):53-61. 6. Kroger N, Holler E, Kobbe G, et al. Allogeneic stem cell transplantation after reduced-intensity conditioning in patients with myelofibrosis: a prospective, multicenter study of the Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation. Blood. 2009;114(26):5264-5270. 7. Olsson R, Remberger M, Schaffer M, et al. Graft failure in the modern era of allogeneic hematopoietic SCT. Bone Marrow Transplant. 2013;48(4):537-543. 8. Robin M, Tabrizi R, Mohty M, et al. Allogeneic haematopoietic stem cell transplantation for myelofibrosis: a report of the Societe Francaise de Greffe de Moelle et de Therapie Cellulaire (SFGM-TC). Br J Haematol. 2011;152(3):331-339. 9. Lissandre S, Bay JO, Cahn JY, et al. Retrospective study of allogeneic haematopoietic stem-cell transplantation for myelofibrosis. Bone Marrow Transplant. 2011;46(4):557-561. 10. Papayannopoulou T, Craddock C. Homing and trafficking of hemopoietic progenitor cells. Acta Haematol. 1997;97(1-2):97-104. 11. Haylock DN, Nilsson SK. Stem cell regulation by the hematopoietic stem cell niche. Cell Cycle. 2005;4(10):1353-1355. 12. Sugiyama T, Kohara H, Noda M, Nagasawa T. Maintenance of the hematopoietic stem cell pool by CXCL12CXCR4 chemokine signaling in bone marrow stromal cell niches. Immunity. 2006;25(6):977-988. 13. 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. 14. Nilsson SK, Johnston HM, Whitty GA, et al. Osteopontin, a key component of the hematopoietic stem cell niche and regulator of primitive hematopoietic progenitor cells. Blood. 2005;106(4):1232-1239. 15. Tefferi A. Pathogenesis of myelofibrosis with myeloid metaplasia. J Clin Oncol.

1416

2005;23(33):8520-8530. 16. Tefferi A. Myelofibrosis with myeloid metaplasia. N Engl J Med. 2000;342(17):1255-1265. 17. Lataillade J, Pierre-Louis O, Hasselbalch H, et al. Does primary myelofibrosis involve a defective stem cell niche? From concept to evidence. Blood. 2008;112(8):3026-3035. 18. Grassinger J, Mueller G, Zaiss M, et al. Differentiation of hematopoietic progenitor cells towards the myeloid and B-lymphoid lineage by hepatocyte growth factor (HGF) and thrombopoietin (TPO) together with early acting cytokines. Eur J Haematol. 2006;77(2):134-144. 19. Rondelli D, Goldberg JD, Isola L, et al. MPDRC 101 prospective study of reduced-intensity allogeneic hematopoietic stem cell transplantation in patients with myelofibrosis. Blood. 2014;124(7):1183-1191. 20. Remberger M, Torlen J, Ringden O, et al. Effect of Total Nucleated and CD34(+) Cell Dose on Outcome after Allogeneic Hematopoietic Stem Cell Transplantation. Biol Blood Marrow Transplant. 2015;21(5): 889-893. 21. Claudiani S, Marktel S, Piemontese S, et al. Treosulfan based reduced toxicity conditioning followed by allogeneic stem cell transplantation in patients with myelofibrosis. Hematol Oncol. 2014 Dec 3. [Epub ahead of print]. 22. Heazlewood SY, Oteiza A, Cao H, Nilsson SK. Analyzing hematopoietic stem cell homing, lodgment, and engraftment to better understand the bone marrow niche. Ann N Y Acad Sci. 2014;1310:119-128. 23. Ellis SL, Grassinger J, Jones A, et al. The relationship between bone, hemopoietic stem cells, and vasculature. Blood. 2011;118(6):1516-1524. 24. Donmez A, Ozsan F, Arik B, et al. The clearance time of infused hematopoietic stem cell from the blood circulation. Transfus Apher Sci. 2013;48(2):235-239. 25. Akpek G, Pasquini MC, Logan B, et al. Effects of spleen status on early outcomes after hematopoietic cell transplantation. Bone Marrow Transplant. 2013;48(6):825831. 26. Thiele J, Kvasnicka HM, Czieslick C. CD34+ progenitor cells in idiopathic (primary) myelofibrosis: a comparative quantification between spleen and bone marrow tissue. Ann Hematol. 2002;81(2):86-89. 27. Rohrer C, Arni U, Deubelbeiss KA. Influence of the spleen on the distribution of blood neutrophils. Quantitative studies in the rat. Scand J Haematol. 1983;30(2):103-109. 28. Terstappen LW, Huang S, Safford M, et al. Sequential generations of hematopoietic colonies derived from single nonlineagecommitted CD34+CD38- progenitor cells. Blood. 1991;77(6):1218-1227. 29. Plett PA, Frankovitz SM, Orschell CM. Distribution of marrow repopulating cells between bone marrow and spleen early after transplantation. Blood. 2003; 102(6):2285-2291. 30. Majeti R, Park CY, Weissman IL. Identification of a hierarchy of multipotent hematopoietic progenitors in human cord blood. Cell Stem Cell. 2007;1(6):635-645. 31. Jung Y, Wang J, Schneider A, et al. Regulation of SDF-1 (CXCL12) production by osteoblasts; a possible mechanism for

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

stem cell homing. Bone. 2006;38(4):497508. Uy GL, Rettig MP, Cashen AF. Plerixafor, a CXCR4 antagonist for the mobilization of hematopoietic stem cells. Expert Opin Biol Ther. 2008;8(11):1797-1804. Christopher MJ, Liu F, Hilton MJ, Long F, Link DC. Suppression of CXCL12 production by bone marrow osteoblasts is a common and critical pathway for cytokineinduced mobilization. Blood. 2009;114(7):1331-1339. Miwa Y, Hayashi T, Suzuki S, et al. Up-regulated expression of CXCL12 in human spleens with extramedullary haematopoiesis. Pathology. 2013;45(4):408416. Wang X, Cho SY, Hu CS, et al. C-X-C motif chemokine 12 influences the development of extramedullary hematopoiesis in the spleens of myelofibrosis patients. Exp Hematol. 2015;43(2):100-109 e101. Migliaccio AR, Martelli F, Verrucci M, et al. Altered SDF-1/CXCR4 axis in patients with primary myelofibrosis and in the Gata1 low mouse model of the disease. Exp Hematol. 2008;36(2):158-171. Grassinger J, Haylock DN, Storan MJ, et al. Thrombin-cleaved osteopontin regulates hemopoietic stem and progenitor cell functions through interactions with alpha9beta1 and alpha4beta1 integrins. Blood. 2009;114(1):49-59. Peled A, Kollet O, Ponomaryov T, et al. The chemokine SDF-1 activates the integrins LFA-1, VLA-4, and VLA-5 on immature human CD34(+) cells: role in transendothelial/stromal migration and engraftment of NOD/SCID mice. Blood. 2000;95(11):32893296. Simmons PJ, Masinovsky B, Longenecker BM, et al. Vascular cell adhesion molecule1 expressed by bone marrow stromal cells mediates the binding of hematopoietic progenitor cells. Blood. 1992;80(2):388-395. Levesque JP, Takamatsu Y, Nilsson SK, et al. Vascular cell adhesion molecule-1 (CD106) is cleaved by neutrophil proteases in the bone marrow following hematopoietic progenitor cell mobilization by granulocyte colony-stimulating factor. Blood. 2001; 98(5):1289-1297. Xu M, Bruno E, Chao J, et al. Constitutive mobilization of CD34+ cells into the peripheral blood in idiopathic myelofibrosis may be due to the action of a number of proteases. Blood. 2005;105(11):4508-4515. Verstovsek S, Mesa RA, Gotlib J, et al. A double-blind, placebo-controlled trial of ruxolitinib for myelofibrosis. N Engl J Med. 2012;366(9):799-807. Caocci G, Maccioni A, Murgia F, et al. Modulation of bone marrow microenvironment following ruxolitinib therapy in myelofibrosis. Leuk Lymphoma. 2016; 57(5):1215-1218. Mesa RA, Scherber RM, Geyer HL. Reducing symptom burden in patients with myeloproliferative neoplasms in the era of Janus kinase inhibitors. Leuk Lymphoma. 2015;56(7):1989-1999. Jaekel N, Behre G, Behning A, et al. Allogeneic hematopoietic cell transplantation for myelofibrosis in patients pretreated with the JAK1 and JAK2 inhibitor ruxolitinib. Bone Marrow Transplant. 2014; 49(2):179-184.

haematologica | 2016; 101(11)

ARTICLE

Cell Therapy and Immunotherapy

A prospective randomized trial comparing cyclosporine/methotrexate and tacrolimus/sirolimus as graft-versus-host disease prophylaxis after allogeneic hematopoietic stem cell transplantation

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Johan Törlén,1,2 Olle Ringdén,1 Karin Garming-Legert,3 Per Ljungman,1,4 Jacek Winiarski,5 Kari Remes,6,7 Maija Itälä-Remes,6 Mats Remberger,1,2 and Jonas Mattsson1,2

Center for Allogeneic Stem Cell Transplantation, Karolinska University Hospital, Stockholm, Sweden; 2Department of Oncology-Pathology, Karolinska Institutet, Stockholm, Sweden; 3Division of Oral and Maxillofacial Surgery, Department of Dental Medicine, Karolinska Institutet, Huddinge, Sweden; 4Department of Hematology, Karolinska University Hospital and Division of Hematology, Department of Medicine, Karolinska Institutet, Karolinska University Hospital, Stockholm, Sweden; 5Astrid Lindgren Children's Hospital, Karolinska University Hospital, Stockholm, Sweden; 6 Department of Internal Medicine, Turku University Hospital, Finland; and 7Turku University, Finland

Haematologica 2016 Volume 101(11):1417-1425

ABSTRACT

mprovement of graft-versus-host disease prophylaxis remains an important goal in allogeneic hematopoietic stem cell transplantation. Based on reports of possibly preferential properties of sirolimus, we compared the standard regimen of cyclosporine and methotrexate (n=106) with a combination of tacrolimus and sirolimus (n=103) as graft-versus-host disease prophylaxis after allogeneic hematopoietic stem cell transplantation in a prospective, open, randomized trial. The hypothesis was that the tacrolimus/sirolimus regimen would lead to less acute graft-versus-host disease and reduced transplant-related mortality. There was no significant difference in the cumulative incidence of acute graft-versus-host disease of grades II-IV (41% vs. 51%; P=0.19) or grades III-IV (13% vs. 7%; P=0.09) between the groups. Time to neutrophil engraftment (18 days vs. 17 days; P=0.24) was similar, but time to platelet engraftment was longer in cyclosporine/methotrexate patients (14 vs. 12 days; P<0.01). No significant differences in incidence of oropharyngeal mucositis, time to full donor chimerism, or number of cytomegalovirus infections were seen between the two treatment arms, and transplant-related toxicities were equally distributed. Triglyceride (P=0.005) and cholesterol (P=0.009) levels were higher in tacrolimus/sirolimus patients. Transplant-related mortality (18% vs. 12%; P=0.40) and 5-year overall survival (72% vs. 71%; P=0.71) were similar. Five-year relapse-free survival in patients with malignant diagnoses was 65% in the cyclosporine/methotrexate group and 63% in the tacrolimus/sirolimus group (P=0.73). We conclude that tacrolimus/sirolimus remains a valid and safe alternative to cyclosporine/methotrexate as graft-versus-host disease prophylaxis after allogeneic hematopoietic stem cell transplantation, with comparable transplant-related outcomes. The trial was registered at clinicaltrials.gov identifier: 00993343. haematologica | 2016; 101(11)

Correspondence: johan.karlsson-torlen@karolinska.se

Received: May 13, 2016. Accepted: August 1, 2016. Pre-published: August 4, 2016. doi:10.3324/haematol.2016.149294

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1417

1417

J. TĂśrlĂŠn et al.

Introduction Allogeneic hematopoietic stem cell transplantation (HSCT) is an established treatment for a series of otherwise lethal hematopoietic disorders.1-3 Continuous refinements of transplant procedures, e.g. in the areas of human leukocyte antigen (HLA)-typing, expanded donor registries, conditioning regimens and supportive care, have steadily improved patient outcome after HSCT.4,5 Despite these advances, graft-versus-host disease (GvHD) remains a frequent and serious complication, affecting 30%-50% of matched sibling transplantations and 40%-70% of matched unrelated HSCT recipients. Both acute and chronic GvHD contribute significantly to morbidity, and are associated with high mortality.6,7 Ambitious efforts to address this issue have included substitution of immunosuppressive pharmaceuticals, graft engineering, and cellular therapies.8-10 In parallel, new immunosuppressive strategies have been evaluated in solid organ transplantation, making way for their introduction in HSCT.11,12 A regimen that has shown promising results in prevention of GvHD is the combination of sirolimus and tacrolimus.13,14 This regimen differs in action from the most commonly used GvHD prophylaxis in HSCT today, cyclosporine in combination with methotrexate.15,16 Sirolimus has been of interest to the HSCT field due to its promising mechanisms, which in theory offer potential advantages over the immunosuppressive agents currently in use. Its actions include immunosuppression through inhibition of T-cell and dendritic cell activity, while promoting regulatory

T cells.17,18 Furthermore, sirolimus has antifibrotic, antineoplastic, antiviral, and antifungal activities, as well as synergistic action when combined with tacrolimus, but has limited toxicity in relation to calcineurin inhibitors.19-22 With this knowledge, we performed a prospective randomized trial to determine whether an immunosuppressive regimen of tacrolimus/sirolimus (Tac/Sir) is better than the established prophylaxis with cyclosporine/methotrexate (CsA/Mtx) in HSCT, hypothesizing that the combination of Tac/Sir would lead to less acute GvHD and reduced transplant-related mortality (TRM). To the best of our knowledge, the regimens of CsA/Mtx and Tac/Sir have not been compared head-to-head in any previous prospective randomized trial of GvHD prophylaxis.6,23

Methods Study design The study was designed to compare two GvHD prophylaxis regimens: CsA/Mtx versus Tac/Sir. It was preceded by a safety pilot study using Tac/Sir in 24 HSCT patients,24 and was performed as a prospective, randomized, open-label, phase III, multicenter trial. Study end points are presented in Figure 1. The aim was to include 200 patients, 100 in each arm; this would have sufficient power to detect a difference in incidence of acute GvHD of 24 percentage points between the treatment groups with a probability of P<0.05. The level of effect was set since the incidence of grade II-IV GvHD in our patient material at the time was 34% using CsA/Mtx prophylaxis, and the corresponding incidence in

Figure 1. CONSORT diagram and end points. Flow of patients enrolled in the trial. CR: complete remission; CP: chronic phase; RIC: reduced intensity conditioning; MAC: myeloablative conditioning; MUD: matched unrelated donor; CsA: cyclosporine; Mtx: methotrexate; Tac: tacrolimus; Sir: sirolimus; GvHD: graft-versus-host disease; HSCT: allogeneic hematopoietic stem cell transplantation; TRM: transplant-related mortality; OS: overall survival.

1418

haematologica | 2016; 101(11)

CsA/Mtx vs. Tac/Sir as GvHD prophylaxis in HSCT

the published pilot study of Tac/Sir by Cutler et al. was 10%.13 Patients were enrolled at two participating centers (Stockholm and Turku) between September 2007 and January 2014. Randomization occurred 4-7 days before HSCT graft infusion. It was performed at a ratio of 1:1 with the use of random block sizes, stratified by age (children or adult), hematologic risk group (CR 1, CP or >CR 1, advanced disease; see below for explanation of abbreviations), conditioning regimen [reduced-intensity conditioning (RIC) or myeloablative conditioning (MAC)], and donor type [sibling or matched unrelated donor (MUD)]. Patients with non-malignant disease were included in the low hematologic risk group. No blinding was attempted after randomization. The study protocol was approved by the Ethical Review Boards in Stockholm (DNR 2006/1430-31/3) and Helsinki (#541/2007, DNR 360/E5/07), and the Swedish and Finnish Medical Products Agencies (DNR 151:2007/38987 and KLNR 57/2008, respectively). The study was registered at clinicaltrials.gov identifer: 00993343 and the European Clinical Trials Database n. 2006-006577-25, and performed in accordance with the Declaration of Helsinki. Written informed consent was obtained from each patient, or from parents/guardians of patients who were under 18 years of age, before the start of HSCT conditioning treatment. All the authors vouch for the accuracy of the data reported, and for adherence to the study protocol.

Patient eligibility and random assignment Patients eligible for inclusion were 0.5-75 years of age and subject to treatment with HSCT. Inclusion and exclusion criteria are described in Table 1. Due to the risk of toxicity, the intention to use a MAC regimen of busulphan and cyclophosphamide (BuCy) was added as an exclusion criterion during the trial (see below). After assessment for eligibility, 215 patients were randomized

in the trial. The flow of enrolled patients is shown in Figure 1. Two patients with symptoms of uncontrolled infection before HSCT, 2 patients with delayed notification of previous inclusion in a different trial with another investigational drug, and 2 patients scheduled for BuCy conditioning (at the time point of introduction of the additional exclusion criterion; see below) were excluded from the trial after randomization but before administration of their assigned GvHD prophylaxis. Inclusion of these 6 patients was considered as protocol violations, and they were excluded from further analysis.

Transplant procedures All patients and donors were typed using molecular high-resolution typing (PCR-SSP) for both HLA class I and II antigens.25 Pretransplantation conditioning regimens depended on disease, age, and standard criteria at the participating centers. Myeloablative conditioning consisted of cyclophosphamide (Cy) 50 mg/kg/d for 4 days, or Cy 60 mg/kg/d for 2 days in combination with fractionated total body irradiation (TBI) with 12 Gy given in 3 fractions over 4 days26,27 (Online Supplementary Table S1). Initially, a MAC regimen of Cy, 60 mg/kg/d for 2 days, in combination with busulphan (Bu), 4 mg/kg/d for 4 days (BuCy) was permitted in the trial. Of the first 6 patients who received this regimen in combination with Tac/Sir prophylaxis, 2 developed early signs of excess toxicity and veno-occlusive disease (VOD)28 of the liver, and 2 developed signs of thrombotic microangiopathy (TMA),29 reported as severe adverse events. These findings coincided with the publication of similar discoveries in a comparable trial conducted by Cutler et al.30,31 Following a review of our data at the time, it was decided by the sponsor and the principal investigator to stop further recruitment of patients receiving BuCy conditioning to the trial.

Table 1. Study criteria, patient eligibility. Inclusion criteria Diagnoses with indication for HSCT treatment, including (but not limited to): AML/ALL in complete remission (CR) CML in 1st or 2nd chronic phase (CP) CLL Lymphoma MDS Severe aplastic anemia Hemoglobinopathies Immunodeficiencies Metabolic disorders Exclusion criteria Previous HSCT treatment Relapse or blast crisis of malignant disease Recipient of HLA-A, -B, or -DR antigen mismatched grafts or a cord-blood graft Addiction to drugs or alcohol Uncontrolled infection Pregnancy or breastfeeding within 4 weeks of study entry Impaired kidney function (creatinine clearance <40 mL/min/1.72 mÂ˛ or proteinuria >0.3 g/day) Impaired liver function (most recent bilirubin, ALT, or AST more than twice the upper limit of normal) Lung disease (in adults, FVC or FEV1 <60% of predicted value, and in children, <92% unsupported oxygen saturation within 4 weeks of study entry) Cardiac ejection fraction <45% in adults or <26% shortening fraction in children Cholesterol level >300 mg/dL or triglyceride level >300 mg/dL Karnofsky performance status <70% (Karnofsky/Lansky for children <16 years of age) Requiring voriconazole at time of study entry Prior history of allergy to sirolimus Receiving another investigational drug unless cleared by the principal investigator and sponsor HSCT: allogeneic hematopoietic stem cell transplantation; AML: acute myeloid leukemia; ALL: acute lymphoblastic leukemia; CML: chronic myeloid leukemia; CLL: chronic lymphocytic leukemia; MDS: myelodysplastic syndrome; FVC: forced vital capacity; FEV1: forced expiratory volume in 1 second.

haematologica | 2016; 101(11)

1419

J. Törlén et al.

Reduced-intensity conditioning consisted of fludarabine, 30 mg/m2/d for 3-6 days, in combination with either Cy 60 mg/kg/d for 2 days, 2x3 Gy TBI and Cy 60 mg/kg/d for 2 days, 2 Gy TBI, treosulphan 14 g/m2 for 3 days, or Bu 4 mg/kg/d for 2 days32-34 (Online Supplementary Table S1). Antithymocyte globulin (ATG) was given as part of the conditioning regimen to patients receiving grafts from an unrelated donor, and to patients with non-malignant disorders. A total dose of 4-8 mg/kg was given; 4 mg/kg to malignant diseases receiving RIC, 6 mg/kg to malignant diseases receiving MAC, and 8 mg/kg to non-malignant diseases or HLA-mismatched grafts.33,35 The dose was administered in consecutive doses of 2 mg/kg/day, with the last dose given the day prior to graft infusion. Supportive care followed previously described institutional standards.36 The source of stem cells was peripheral blood progenitor cells (PBSCs) or bone marrow. Graft-versus-host disease prophylaxis consisted of CsA/Mtx or Tac/Sir. Patients in the standard arm started CsA on day −1 (the day before graft infusion) and Mtx 15 mg/m2 was given on day +1, with consecutive doses of 10 mg/m2 given on days +3, +6, and +11 for all diagnoses. CsA was given twice a day (mainly orally). During the first two months, monitored plasma concentration levels were kept between 80-100 ng/mL in patients who received grafts from HLA-identical siblings, and between 150-250 ng/mL in MUD transplants. CsA was discontinued after tapering 3-4 months after HSCT in recipients of HLA-identical sibling grafts, and after six months in recipients of MUD transplants, in the absence of GvHD. Patients in the experimental arm started Tac/Sir in combination on day −3 before graft infusion. Sirolimus was given orally once daily, starting with a bolus dose of 6 mg in adults and 0.1 mg/kg in children, followed by continuous individual adjustment with monitored plasma target levels of 3-12 ng/mL. Tacrolimus was given orally twice a day, starting at 0.15 mg/kg/day, with a target plasma concentration of 5-15 ng/mL. Sirolimus was discontinued after tapering during the third month after HSCT in all patients in the absence of GvHD. In recipients of HLA-identical sibling grafts, tacrolimus was tapered in month 3 and discontinued in month 4 in the absence of GvHD. For MUD transplants, tacrolimus was discontinued after tapering six months after HSCT in the absence of GvHD. Dose modification guidelines for both immunosuppressive regimens included adjustments for toxicity, GvHD, and relapse.

according to standard operating protocols. Chimerism analyses of peripheral blood and/or bone marrow were performed on all patients at regular intervals after HSCT, according to standard operating protocols, or at the discretion of the treating physician. The cell lineages analyzed were T lymphocytes (CD3), B lymphocytes (CD19), myeloid cells (CD33), and hematopoietic precursors (CD34; applicable to bone marrow samples). Analyses were performed with real-time PCR based on single nucleotide polymorphisms.40 Full donor chimerism was defined as more than 95% donor-derived cells in all lineages, and mixed chimerism was assumed when more than 5% but less than 95% donor-derived cells were present in the cell lineages analyzed. Transplant-related mortality was defined as death from any

Definitions and outcome assessment Acute GvHD was diagnosed clinically and graded by the attending physicians from 0 to IV according to previously published criteria.37 Chronic GvHD was graded using the National Institutes of Health consensus criteria for clinical trials.38 Neutrophil engraftment was defined as the first of three consecutive daily measurements with a neutrophil count of more than 0.5x109/L, and platelet engraftment as the first of three consecutive daily measurements with a platelet count of more than 20×109/L without transfusions. Oropharyngeal mucositis was assessed three times a week, and graded according to the International Classification of Diseases from the WHO39 until day +24 or until hospital discharge. Cytomegalovirus (CMV) infection was defined as detection of CMV DNA in whole blood by real-time PCR. A PCR result of more than 1000 CMV DNA copies/mL (changed to more than 2000 CMV DNA copies/mL in 2009) was considered to be a clinically relevant CMV-viremia, and initiation of pre-emptive CMV therapy was used as a basis for analysis between the groups. Patients were monitored with CMV-PCR once a week for three months after HSCT. Subsequent monitoring was individually prescribed for each patient at the discretion of the treating physician, 1420

Figure 2. Graft-versus-host disease (GvHD) outcomes. (A) Cumulative incidence of acute GvHD of grades II-IV. (B) Cumulative incidence of acute GvHD of grades III-IV. (C) Cumulative incidence of chronic GvHD. CsA: cyclosporine; Mtx: methotrexate; Tac: tacrolimus; Sir: sirolimus; GvHD: graft-versus-host disease; HSCT: allogeneic hematopoietic stem cell transplantation.

haematologica | 2016; 101(11)

CsA/Mtx vs. Tac/Sir as GvHD prophylaxis in HSCT

cause without relapse. Relapse was defined as recurrent disease after complete remission, or disease progression after partial remission or stable disease. Relapse-free survival (RFS) was defined as survival without any sign of hematologic relapse.

2001), Splus 6.2 software (Insightful, Seattle, WA, USA) and Statistica software (Statsoft, Tulsa, OK, USA).

Results Statistical analysis Data that were current in February 2015 (7.5 and 1.1 years after the first and last patient enrollment, respectively) were used in the intention-to-treat analysis. Categorical variables were compared using the χ2 method and continuous variables were compared using the Mann-Whitney U test. Overall survival and RFS were calculated using the Kaplan-Meier method and compared using the log rank test. Survival time was calculated from the day of transplantation until death or last follow up. The incidences of GvHD, TRM, and relapse were obtained using an estimator of cumulative incidence curves. Patients were censored at time of death or last follow up. Multivariate analyses for OS and RFS were performed with the Cox proportional hazards model. Subgroup analysis was performed for patients with malignant diagnoses. End point analyses were statistically corrected for patient group characteristics, presented as additional hazard ratios (HR). All analyses were performed using the cmprsk package (Gray,

Patients' characteristics Patients' and transplant characteristics are given in Table 2; 106 patients were analyzed in the standard arm (CsA/Mtx) and 103 in the experimental arm (Tac/Sir). Median follow up of the cohorts was 4.0 and 4.5 years, respectively, with a minimum follow up of 1.0 year in both groups at the time of data assessment. No statistically significant differences were seen between the groups for patients' characteristics regarding age, diagnosis, or conditioning regimen, with the exception of an excess of HLA-C mismatched grafts (P=0.02) and CMV-seropositive donors in the CsA/Mtx arm (P=0.04).

Graft-versus-host disease The cumulative incidence of acute GvHD of grades II-IV in the CsA/Mtx arm was 41% [(95% confidence interval

Table 2. Patients', donor, and transplant characteristics, according to treatment arm.

Variable Number of patients Median age, years (range) Number of children <18 years (%) Sex, number of males/females (%) Median follow up, years (range) Diagnosis, number of patients (%): AML/ALL CLL Lymphoma MDS Other malignancies Non-malignant Disease stage, early/late Conditioning regimen, number of patients (%): MAC/RIC TBI-based ATG Donor, number of patients (%): Sibling MUD (8/8) URD (7/8, HLA-C mismatch) URD (7/8, HLA-DR allele mismatch) Median donor age, years (range) Female to male, number of patients (%) HSCT graft, number (%) BM/PBSCs TNC dose, ×108/kg (range) CD34+ cell dose, ×106/kg (range) CMV status, number of patients (%) Recipient CMV positive Donor CMV positive CMV, negative to negative

CsA/Mtx

Tac/Sir

106 52 (0.6-71) 12 (11) 60/46 (57/43) 4.5 (1.0-7.4)

103 50 (2.8-68) 12 (12) 67/36 (65/35) 4.0 (1.0-7.4)

0.91 0.89 0.27 0.39

30/19 (28/18) 7 (7) 14 (13) 20 (19) 6 (6) 10 (9) 56/50 (53/47)

27/23 (26/22) 15 (15) 13 (13) 14 (14) 8 (8) 3 (3) 40/63 (39/61)

35/71 (33/67) 36 (34) 81 (76)

37/66 (36/64) 47 (46) 73 (71)

29 (27) 58 (55) 15 (14) 4 (4) 29 (4-66) 15 (14)

33 (32) 39 (38) 29 (28) 2 (2) 32 (7-70) 14 (14)

0.02 0.71 0.17 0.91

21/85 (20/80) 9.2 (1.8-34.0) 6.6 (1.2-22.8)

18/85 (18/83) 10.8 (1.8-42.8) 6.3 (1.3-19.7)

0.80 0.11 0.98

82 (77) 50 (47) 16 (15)

78 (76) 65 (63) 14 (14)

0.91 0.04 0.91

0.10

0.06 0.77 0.45 0.56

CsA: cyclosporine; Mtx: methotrexate; Tac: tacrolimus; Sir: sirolimus; AML: acute myeloid leukemia; ALL: acute lymphoblastic leukemia; CLL: chronic lymphocytic leukemia; MDS: myelodysplastic syndrome; MAC: myeloablative conditioning; RIC: reduced-intensity conditioning; TBI: total body irradiation; ATG: anti-thymocyte globulin; MUD: matched unrelated donor; URD: unrelated donor; HLA: human leukocyte antigen; HSCT: allogeneic hematopoietic stem cell transplantation; BM: bone marrow; PBSCs: peripheral blood stem cells; TNC: total nucleated cells; CMV: cytomegalovirus.

haematologica | 2016; 101(11)

1421

J. Törlén et al.

(CI): 32%-50%)] as compared to 51% (95%CI: 41%-61%) in the Tac/Sir arm (P=0.19) (Figure 2A). Corrected (for patients' characteristics) effect of CsA/Mtx versus Tac/Sir showed HR 0.85 (95%CI: 0.55-1.31; P=0.46). There was no difference in the cumulative incidence of acute GvHD of grades III‒IV in the CsA/Mtx group and the Tac/Sir group [13% (7%-19%), and 7% (2%-12%), respectively; P=0.09] (Figure 2B). Corrected effect of CsA/Mtx showed HR 2.31 (95%CI: 0.88-6.10; P=0.09). Median time to development of acute GvHD was 26 days in CsA/Mtx patients and 31 days in Tac/Sir patients (P=0.22). In patients with malignant diagnoses (n=96 in the CsA/Mtx arm and n=100 in the Tac/Sir arm), 14% (7%-21%) in the CsA/Mtx arm and 7% (2%-12%) in the Tac/Sir arm developed acute GvHD of grades III‒IV (P=0.06). Rates of acute GvHD did not differ significantly by donor type (siblings 48%, MUDs 46%), stem cell source (PBSC 47%, bone marrow 44%), HLA-match (7/8 37%, 8/8 50%) or conditioning intensity (RIC 47%, MAC 45%). Multivariate analysis for acute GvHD grades II-IV showed higher risk for patients with malignant diagnoses [(risk ratio (RR) 9.39, 95%CI: 1.30-67.9; P=0.03)] and female donor to male recipient transplants [(RR 2.44), 95%CI: 1.17-5.05; P=0.02)]. In the patients who developed acute GvHD, there were no significant differences in the incidence of skin, gut, and liver manifestations between the two groups (Online Supplementary Figure S1). The 5-year cumulative incidence of chronic GvHD in the CsA/Mtx arm was 41% (95%CI: 31%-51%), compared to 37% (95%CI: 26%-48%) in the Tac/Sir arm (P=0.51) (Figure 2C). Corrected effect of CsA/Mtx showed HR 1.52 (95%CI: 0.94-2.58; P=0.10). Median time to development of chronic GvHD was 214 days in CsA/Mtx patients and 204 days in Tac/Sir patients (P=0.72). Subgroup analysis of patients with malignant diagnoses showed no significant difference in moderate-to-severe chronic GvHD [18% (95%CI: 10-26%) in CsA/Mtx patients and 9% (95%CI: 3-15%) in Tac/Sir patients; P=0.09]. Tapering of GvHD prophylaxis occurred according to protocol, with a median treatment period of 191 days for cyclosporine, 150 days for tacrolimus and 68 days for sirolimus. At the time of last follow up, 61 CsA/Mtx patients and 63 Tac/Sir patients were alive and off their assigned immunosuppression.

Engraftment There was no difference in the time to neutrophil engraftment [18 (10-305) days with CsA/Mtx as opposed to 17 (11‒32) days with Tac/Sir; P=0.24] (Figure 3A). The median time to platelet engraftment was longer in CsA/Mtx patients [14 (0-190) days compared to 12 (0-68) days in CsA/Mtx patients; P=0.008] (Figure 3B).

Transplant-related outcomes After exclusion of the BuCy regimen from the study, no additional patients fulfilled strict VOD criteria. In the CsA/Mtx group, no patient developed TMA during the first three months after HSCT, as compared to 2 in the Tac/Sir group. Thrombotic thrombocytopenic purpura (TTP, defined as TMA with severe acquired ADAMTS13 deficiency) was diagnosed in one CsA/Mtx patient and 2 Tac/Sir patients. Minor liver and renal toxicities were equally distributed. 1422

Figure 3. Engraftment outcomes. (A) Cumulative incidence of neutrophil engraftment. (B) Cumulative incidence of platelet engraftment. CsA: cyclosporine; Mtx: methotrexate; Tac: tacrolimus; Sir: sirolimus; HSCT: allogeneic hematopoietic stem cell transplantation.

In the Tac/Sir group, 7% of the patients developed suspected sirolimus-induced lesions in mouth, which healed after discontinuation of sirolimus administration. There was no significant difference in incidence or severity of oropharyngeal mucositis between the two groups. The percentage of patients with elevated serum lipids was higher in the Tac/Sir arm during the first three months after HSCT. For triglycerides, the percentage of patients with any blood sample reaching a level of twice the upper limit of normal (ULN) was 17% in the CsA/Mtx group and 34% in the Tac/Sir group (P=0.005). For cholesterol, 9% and 22% of patients, respectively, had levels above the ULN (P=0.009). There was no significant difference in the number of days on total parenteral nutrition: median for CsA/Mtx patients was 3 (0-66) days as compared to 1.5 (0194) for Tac/Sir patients (P=0.51). There was no significant difference in mean time to discharge between the two groups: 21 (1-45) days and 20 (3-59) days, respectively (P=0.15). There were no significant differences in the number of patients who received pre-emptive therapy for CMV infection after HSCT in the two treatment arms. The incidence of post-transplant lymphoproliferative disorder,41,42 invasive fungal infections, and bloodstream infections were equal (Online Supplementary Table S2). Chimerism analysis showed no significant differences haematologica | 2016; 101(11)

CsA/Mtx vs. Tac/Sir as GvHD prophylaxis in HSCT

between the groups regarding the number of patients classified as full donor or mixed chimerism at standard time points after HSCT (Online Supplementary Figure S2). Subgroup analysis comparing MAC and RIC patients in the study groups did not show any significant differences in any of the analyzed end points.

Transplant-related mortality, survival, and relapse The transplant-related mortality three years after HSCT was similar: 18% (95%CI: 11%-25%) in CsA/Mtx patients and 12% (95%CI: 6%-18%) in Tac/Sir patients (P=0.40), and corrected effect of CsA/Mtx showed HR 0.63 (95%CI: 0.30-1.31; P=0.70). The overall survival at five years after transplantation was 72% (95%CI: 63%81%) in the CsA/Mtx group and 71% (95%CI: 62%-80%) in the Tac/Sir group (P=0.71) (Figure 4A). Corrected effect of CsA/Mtx showed HR 0.89 (95%CI: 0.49-1.59; P=0.67). Multivariate analysis identified allele-mismatched grafts (RR 8.40, 95%CI: 2.95-23.91; P<0.01) and patient age at the time of HSCT (RR 1.49, 95%CI: 1.22-1.91; P<0.01) as significant risk factors for mortality. Reduced-intensity conditioning was identified as a significant protective factor for mortality (RR 0.41, 95%CI: 0.20-0.81; P=0.01). Relapse of malignant disease was the most frequent cause of death in both groups (39% of deaths in CsA/Mtx patients and 50% of deaths in Tac/Sir patients). Subgroup analysis of patients with malignant diagnoses showed no significant differences in relapse or RFS after HSCT (Figure 4B and Online Supplementary Figure S3). Five-year RFS in this cohort was 65% (95%CI: 55%-75%) in CsA/Mtx patients and 63% (95%CI: 53%-73%) in Tac/Sir patients (P=0.73). Corrected effect of CsA/Mtx showed HR 0.91 (95%CI: 0.54-1.54; P=0.72). Multivariate analysis showed that allele-mismatched grafts (RR 4.16, 95%CI: 1.48-11.69; P<0.01) and patient age at the time of HSCT (RR 1.16, 95%CI: 0.99-1.35; P=0.05) were the strongest predictors of lower RFS. Subgroup analysis of patients receiving ATG as part of their conditioning regimen showed no significant differences in HSCT outcomes between the treatment groups, with the exception of incidence in acute GvHD grades III-IV (13% of ATG treated CsA/Mtx patients, and 4% of ATG treated Tac/Sir patients; P=0.04). No significant differences in outcome were seen in subgroup analysis of graft source.

Discussion In this prospective, randomized, clinical trial we compared CsA/Mtx with Tac/Sir as GvHD prophylaxis after HSCT. We did not find any statistical difference in the cumulative incidence of acute GvHD of grades II-IV in the two groups. Similar results were seen in an equivalent randomized study published in 2014 by Cutler et al., which compared Tac/Sir with Tac/Mtx after matched, related HSCT.31 They found no significant increase in grades II-IV acute GvHD-free survival after HSCT between the two groups, but less oropharyngeal mucositis in the Tac/Sir group. Thus, despite our inclusion of MUD recipients and patients receiving allele-mismatched grafts, we did not see any benefit of Tac/Sir regarding the incidence of acute GvHD. This finding is somewhat contrary to another comparable study in children who received unrelated donor grafts, in which addition of sirolimus was associathaematologica | 2016; 101(11)

Figure 4. Survival outcomes. (A) Overall survival (all patients). (B) Relapse-free survival (malignant diagnoses). CsA: cyclosporine; Mtx: methotrexate; Tac: tacrolimus; Sir: sirolimus; HSCT: allogeneic hematopoietic stem cell transplantation.

ed with a lower incidence of GvHD of grades II-IV.43 One explanation of the discrepancy may have been our relatively heterogeneous patient population, possibly limiting the detection of a beneficial effect for particular diagnoses. Most of our patients also received ATG as part of their conditioning regimen, which reduces acute GvHD and TRM, and hence may have evened out differences in GvHD incidence.44,45 Despite the use of ATG, we observed a higher rate of acute GvHD in this trial when compared to the trial by Cutler et al. One reason might be that in the clinical setting for our patients, the objective was to achieve a grade I (-II) acute GvHD in patients with malignant diagnoses to reduce the risk of relapse. This was primarily achieved by a lower concentration of immunosuppression and ATG in both groups.5 For example, in the study plan, there was a wide concentration range for tacrolimus (5-15 ng/mL), but in the clinical situation, the aim was rather a concentration less than 10 ng/mL (no patient exceeded this level). Hence, the focus was rather to reduce the incidence of grades III-IV acute GvHD, which is reflected by a low incidence of this complication in our trial. In a recent publication by Pidala, prolonged sirolimus administration (â&#x2030;Ľ 1 year) after HSCT was associated with a reduced risk of moderate-to-severe chronic GvHD in Tac/Sir patients compared to Tac/Mtx patients (34% vs. 65%; P<0.01).46 Prolonged treatment with cyclosporine 1423

J. Törlén et al. also reduces chronic GvHD.47 In our study, none of the participants remained on sirolimus for more than six months (median 68 days). A suspected increase in VOD and TMA in patients in the Tac/Sir arm was noted prior to exclusion of patients receiving the MAC regimen of BuCy. Similar findings were published by Cutler et al. in 2008.30 When used with Bu-based conditioning, sirolimus was associated with a higher rate of VOD (OR 8.8; P<0.01). In a prospective randomized study, it was found that patients treated with Bu had a cumulative incidence of VOD of 12%, as compared to 1% in patients who were treated with total body irradiation (P=0.009).26 The use of calcineurin inhibitors and/or sirolimus has also been suggested to be a potential risk factor for TMA, but the effect of the Tac/Sir combination is not well defined. Labrador et al. retrospectively analyzed the incidence of TMA in 102 HSCT recipients receiving Tac/Sir (n=68) or Tac/Mtx ±ATG (n=34) as GvHD prophylaxis, and no significant difference in the incidence of TMA was seen between the groups (7.4% vs. 8.8%; P=0.8).48 Accordingly, TMA in this setting might primarily be an effect of high-dose Bu-treatment, especially in combination with tacrolimus-based GvHD prophylaxis. Monitoring and adjustment of Bu concentrations in vivo, and modulation of the glutathione cellular content in the liver during conditioning, might reduce the risk of these side-effects.49 The number of post-transplant infections was similar in both groups. This is of interest, as development of certain infections after HSCT may be considered to be a surrogate indicator of immune competence. We did not detect any discrepancy in the numbers of clinically relevant CMV infections between the Tac/Sir and CsA/Mtx groups. A previous publication showed that sirolimus administration had little effect on induction of immediate-early gene expression in experimentally latent dendritic cells or cells from naturally latent individuals.50 This suggests that favorable CMV outcomes associated with sirolimus may be attributable to more indirect effects that influence CMV reactivation rather than a direct mechanistic action against CMV itself. We acknowledge that our study has limitations, since it was performed in a relatively small patient cohort with notable patient- and transplant-related variations. Furthermore, no stratification of the patient material was

References 1. Negrin RS. Introduction to the review series on "Advances in hematopoietic cell transplantation". Blood. 2014;124(3):307. 2. Goldstone AH, Rowe JM. Transplantation in adult ALL. Hematology Am Soc Hematol Educ Program. 2009:593-601. 3. Sureda A, Bader P, Cesaro S, et al. Indications for allo- and auto-SCT for haematological diseases, solid tumours and immune disorders: current practice in Europe, 2015. Bone Marrow Transplant. 2015;50(8):1037-1056. 4. Gooley TA, Chien JW, Pergam SA, et al. Reduced mortality after allogeneic hematopoietic-cell transplantation. N Engl J Med. 2010;363(22):2091-2101. 5. Remberger M, Ackefors M, Berglund S, et al.

1424

7. 8.

carried out during randomization for graft source or level of HLA-match, factors recognized as risk factors for GvHD. This could have impaired observations of differences in incidence between grades III-IV GvHD between the study groups. At the same time, GvHD remains a major risk for HSCT patients regardless of diagnosis and/or transplant setting, and the inclusion of MUD recipients in the study make the results more valid, not least since there have been relatively few prospective randomized trials in this research area. In summary, this study did not show any significant advantages of Tac/Sir over CsA/Mtx as GvHD prophylaxis after HSCT. However, our results confirm that Tac/Sir is a valid and safe GvHD prophylaxis option, with transplant-related outcomes comparable to those with CsA/Mtx. Hence, this study expands the randomized trial data in the realm of unrelated donor HSCT, but further studies in more homogenous patient settings, including similar or novel GvHD prophylaxis regimens, are required to investigate if specific patient groups benefit from a certain type of GvHD-prophylaxis, as the further refinement of HSCT procedures to improve outcome continues. Acknowledgments The authors would like to thank Björn Skölving, formerly working at Wyeth AB, for excellent collaboration when initiating the study, and we thank the medical, nursing, and laboratory staff of the Center for Allogeneic Stem Cell Transplantation (CAST), the Division of Hematology, and the Division of Pediatrics, at the Karolinska University Hospital, and the staff of the Department of Hematology at Turku University Hospital for their invaluable contributions to the study through compassionate care of the patients. We also thank the research nurses at CAST for excellent assistance with study records, and especially Eva Martell for first-rate help in retrieval of patient data. Funding JM was supported by grants from the Swedish Cancer Society (CF 2014-2016), the Swedish Children's Cancer Foundation (PR2013-0022 and KF2013-0011), the Marianne and Marcus Wallenberg Foundation (2013.0117), and by grants provided by the Stockholm County Council (ALF project 20140451). This study was supported in part by research funding from Astellas Pharma A/S (SE-09-RG-50) and Wyeth AB/Pfizer AB (#0468x1-3329) to OR and JM.

Improved survival after allogeneic hematopoietic stem cell transplantation in recent years. A single-center study. Biol Blood Marrow Transplant. 2011; 17(11):1688-1697. Ziakas PD, Zervou FN, Zacharioudakis IM, Mylonakis E. Graft-versus-host disease prophylaxis after transplantation: a network meta-analysis. PLoS ONE. 2014; 9(12):e114735. Deeg HJ. How I treat refractory acute GVHD. Blood. 2007;109(10):4119-4126. Perkins J, Field T, Kim J, et al. A randomized phase II trial comparing tacrolimus and mycophenolate mofetil to tacrolimus and methotrexate for acute graft-versus-host disease prophylaxis. Biol Blood Marrow Transplant. 2010;16(7):937-947. Reshef R. Prevention of graft-versus-host disease. Clin Adv Hematol Oncol. 2012;10

(10):663-665. 10. Ringden O, Erkers T, Nava S, et al. Fetal membrane cells for treatment of steroidrefractory acute graft-versus-host disease. Stem Cells. 2013;31(3):592-601. 11. Macdonald AS. Use of mTOR inhibitors in human organ transplantation. Expert Rev Clin Immunol. 2007;3(3):423-436. 12. Knoll GA, Kokolo MB, Mallick R, et al. Effect of sirolimus on malignancy and survival after kidney transplantation: systematic review and meta-analysis of individual patient data. BMJ. 2014;349:g6679. 13. Cutler C, Kim HT, Hochberg E, et al. Sirolimus and tacrolimus without methotrexate as graft-versus-host disease prophylaxis after matched related donor peripheral blood stem cell transplantation. Biol Blood Marrow Transplant. 2004;10(5):328-336.

haematologica | 2016; 101(11)

CsA/Mtx vs. Tac/Sir as GvHD prophylaxis in HSCT

14. Cutler C, Li S, Ho VT, et al. Extended follow-up of methotrexate-free immunosuppression using sirolimus and tacrolimus in related and unrelated donor peripheral blood stem cell transplantation. Blood. 2007;109 (7):3108-3114. 15. Storb R, Deeg HJ, Fisher L, et al. Cyclosporine v methotrexate for graft-vhost disease prevention in patients given marrow grafts for leukemia: long-term follow-up of three controlled trials. Blood. 1988;71(2):293-298. 16. Ringden O, Horowitz MM, Sondel P, et al. Methotrexate, cyclosporine, or both to prevent graft-versus-host disease after HLAidentical sibling bone marrow transplants for early leukemia? Blood. 1993;81(4):10941101. 17. Koenen HJ, Michielsen EC, Verstappen J, Fasse E, Joosten I. Superior T-cell suppression by rapamycin and FK506 over rapamycin and cyclosporine A because of abrogated cytotoxic T-lymphocyte induction, impaired memory responses, and persistent apoptosis. Transplantation. 2003;75(9):1581-1590. 18. Hackstein H, Taner T, Zahorchak AF, et al. Rapamycin inhibits IL-4--induced dendritic cell maturation in vitro and dendritic cell mobilization and function in vivo. Blood. 2003;101(11):4457-4463. 19. Sehgal SN. Sirolimus: its discovery, biological properties, and mechanism of action. Transplant Proc. 2003;35(3 Suppl):7S-14S. 20. Zaytseva YY, Valentino JD, Gulhati P, Evers BM. mTOR inhibitors in cancer therapy. Cancer Lett. 2012;319(1):1-7. 21. Li J, Kim SG, Blenis J. Rapamycin: one drug, many effects. Cell Metab. 2014;19(3):373379. 22. Sheng L, Jun S, Jianfeng L, Lianghui G. The effect of sirolimus-based immunosuppression vs. conventional prophylaxis therapy on cytomegalovirus infection after liver transplantation. Clin Transplant. 2015;29 (6):555-559. 23. Wang L, Gu Z, Zhai R, et al. The efficacy and safety of sirolimus-based graft-versushost disease prophylaxis in patients undergoing allogeneic hematopoietic stem cell transplantation: a meta-analysis of randomized controlled trials. Transfusion. 2015;55(9):2134-2141. 24. Ringden O, Remberger M, Dahllof G, et al. Sirolimus and tacrolimus as immune prophylaxis compared to cyclosporine with or without methotrexate in patients undergoing allogeneic haematopoietic stem cell transplantation for non-malignant disorders. Eur J Haematol. 2011;87(6):503-509. 25. Olerup O, Zetterquist H. HLA-DR typing by PCR amplification with sequence-specific primers (PCR-SSP) in 2 hours: an alternative to serological DR typing in clinical practice including donor-recipient matching in cadaveric transplantation. Tissue Antigens. 1992;39(5):225-235. 26. Ringden O, Ruutu T, Remberger M, et al. A randomized trial comparing busulfan with total body irradiation as conditioning in allogeneic marrow transplant recipients with leukemia: a report from the Nordic Bone

haematologica | 2016; 101(11)

27.

28. 29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

Marrow Transplantation Group. Blood. 1994;83(9):2723-2730. Champlin RE, Perez WS, Passweg JR, et al. Bone marrow transplantation for severe aplastic anemia: a randomized controlled study of conditioning regimens. Blood. 2007;109(10):4582-4585. Carreras E. Veno-occlusive disease of the liver after hemopoietic cell transplantation. Eur J Haematol. 2000;64(5):281-291. Ho VT, Cutler C, Carter S, et al. Blood and marrow transplant clinical trials network toxicity committee consensus summary: thrombotic microangiopathy after hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2005;11(8):571575. Cutler C, Stevenson K, Kim HT, et al. Sirolimus is associated with veno-occlusive disease of the liver after myeloablative allogeneic stem cell transplantation. Blood. 2008;112(12):4425-4431. Cutler C, Logan B, Nakamura R, et al. Tacrolimus/sirolimus vs. tacrolimus/ methotrexate as GVHD prophylaxis after matched, related donor allogeneic HCT. Blood. 2014;124(8):1372-1377. Niederwieser D, Maris M, Shizuru JA, et al. Low-dose total body irradiation (TBI) and fludarabine followed by hematopoietic cell transplantation (HCT) from HLA-matched or mismatched unrelated donors and postgrafting immunosuppression with cyclosporine and mycophenolate mofetil (MMF) can induce durable complete chimerism and sustained remissions in patients with hematological diseases. Blood. 2003;101(4):1620-1629. Slavin S, Nagler A, Naparstek E, et al. Nonmyeloablative stem cell transplantation and cell therapy as an alternative to conventional bone marrow transplantation with lethal cytoreduction for the treatment of malignant and nonmalignant hematologic diseases. Blood. 1998;91(3):756-763. Uzunel M, Remberger M, Sairafi D, et al. Unrelated versus related allogeneic stem cell transplantation after reduced intensity conditioning. Transplantation. 2006; 82(7):913919. Remberger M, Svahn BM, Hentschke P, Lofgren C, Ringden O. Effect on cytokine release and graft-versus-host disease of different anti-T cell antibodies during conditioning for unrelated haematopoietic stem cell transplantation. Bone Marrow Transplant. 1999;24(8):823-830. Forslow U, Mattsson J, Ringden O, Klominek J, Remberger M. Decreasing mortality rate in early pneumonia following hematopoietic stem cell transplantation. Scand J Infect Dis. 2006;38(11-12):970-976. 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. 39. World Health Organization. WHO handbook for reporting results of cancer treatment. Geneva Albany, N.Y.: World Health Organization; sold by WHO Publications Centre USA, 1979. 40. Alizadeh M, Bernard M, Danic B, et al. Quantitative assessment of hematopoietic chimerism after bone marrow transplantation by real-time quantitative polymerase chain reaction. Blood. 2002;99(12):46184625. 41. Loren AW, Porter DL, Stadtmauer EA, Tsai DE. Post-transplant lymphoproliferative disorder: a review. Bone Marrow Transplant. 2003;31(3):145-155. 42. Sundin M, Le Blanc K, Ringden O, et al. The role of HLA mismatch, splenectomy and recipient Epstein-Barr virus seronegativity as risk factors in post-transplant lymphoproliferative disorder following allogeneic hematopoietic stem cell transplantation. Haematologica. 2006;91(8):1059-1067. 43. Pulsipher MA, Langholz B, Wall DA, et al. The addition of sirolimus to tacrolimus/methotrexate GVHD prophylaxis in children with ALL: a phase 3 Children's Oncology Group/Pediatric Blood and Marrow Transplant Consortium trial. Blood. 2014;123(13):2017-2025. 44. Remberger M, Svahn BM, Mattsson J, Ringden O. Dose study of thymoglobulin during conditioning for unrelated donor allogeneic stem-cell transplantation. Transplantation. 2004;78(1):122-127. 45. Finke J, Bethge WA, Schmoor C, et al. Standard graft-versus-host disease prophylaxis with or without anti-T-cell globulin in haematopoietic cell transplantation from matched unrelated donors: a randomised, open-label, multicentre phase 3 trial. Lancet Oncol. 2009;10(9):855-864. 46. Pidala J, Kim J, Alsina M, et al. Prolonged sirolimus administration after allogeneic hematopoietic cell transplantation is associated with decreased risk for moderatesevere chronic graft-versus-host disease. Haematologica. 2015;100(7):970-977. 47. Lonnqvist B, Aschan J, Ljungman P, Ringden O. Long-term cyclosporin therapy may decrease the risk of chronic graft-versus-host disease. Br J Haematol. 1990;74(4):547-548. 48. Labrador J, Lopez-Corral L, Lopez-Godino O, et al. Risk factors for thrombotic microangiopathy in allogeneic hematopoietic stem cell recipients receiving GVHD prophylaxis with tacrolimus plus MTX or sirolimus. Bone Marrow Transplant. 2014;49(5):684-690. 49. Hassan Z, Hellstrom-Lindberg E, Alsadi S, Edgren M, Hagglund H, Hassan M. The effect of modulation of glutathione cellular content on busulphan-induced cytotoxicity on hematopoietic cells in vitro and in vivo. Bone Marrow Transplant. 2002;30(3):141147. 50. Glover TE, Kew VG, Reeves MB. Rapamycin does not inhibit human cytomegalovirus reactivation from dendritic cells in vitro. J Gen Virol. 2014;95(Pt 10):2260-2266.

1425

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Cell Therapy and Immunotherapy

Ferrata Storti Foundation

Haematologica 2016 Volume 101(11):1426-1433

The prognostic value of serum C-reactive protein, ferritin, and albumin prior to allogeneic transplantation for acute myeloid leukemia and myelodysplastic syndromes Andrew S. Artz,1 Brent Logan,2,3 Xiaochun Zhu,2 Gorgun Akpek, 4 Rodrigo Martino Bufarull, 5 Vikas Gupta,6 Hillard M. Lazarus,7 Mark Litzow,8 Alison Loren,9 Navneet S. Majhail,10 Richard T. Maziarz,11 Philip McCarthy,12 Uday Popat,13 Wael Saber,2 Stephen Spellman,14 Olle Ringden,15,16 Amittha Wickrema,1 Marcelo C. Pasquini,2 and Kenneth R. Cooke17 from the Center for International Blood and Marrow Transplantation Research

Section of Hematology/Oncology, University of Chicago School of Medicine, IL, USA; CIBMTR, (Center for International Blood and Marrow Transplant Research), Department of Medicine, Milwaukee, WI, USA; 3Division of Biostatistics, Institute for Health and Society, Medical College of Wisconsin, Milwaukee, WI, USA; 4Stem Cell Transplantation and Cellular Therapy Program, Banner MD Anderson Cancer Center, Gilbert, AZ, USA; 5 Division of Clinical Hematology, Hospital de la Santa Creu I Sant Pau, Barcelona, Spain; 6 Blood and Marrow Transplant Program, Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada; 7Seidman Cancer Center, University Hospitals Case Medical Center, Cleveland, OH, USA; 8Division of Hematology and Transplant Center, Mayo Clinic Rochester, Minneapolis, MN, USA; 9Division of Hematology/Oncology, Department of Medicine, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA; 10Blood & Marrow Transplant Program, Cleveland Clinic Taussig Cancer Institute, OH, USA; 11Adult Blood and Marrow Stem Cell Transplant Program, Knight Cancer Institute, Oregon Health and Science University, Portland, OR, USA; 12Blood & Marrow Transplant Program, Department of Medicine, Roswell Park Cancer Institute, Buffalo, NY, USA; 13MD Anderson Cancer Center, Houston, TX, USA; 14CIBMTR (Center for International Blood and Marrow Transplant Research), National Marrow Donor Program/Be The Match, Minneapolis, MN, USA; 15Division of Therapeutic Immunology, Department of Laboratory Medicine, Karolinska Institute, Stockholm, Sweden; 16Centre for Allogeneic Stem Cell Transplantation, Stockholm, Sweden; and 17Pediatric Blood and Marrow Transplant Program, The Sidney Kimmel Cancer Comprehensive Cancer Center at Johns Hopkins, Baltimore, MD, USA

Correspondence: aartz@medicine.bsd.uchicago.edu

ABSTRACT Received: March 11, 2016. Accepted: July 26, 2016. Pre-published: August 4, 2016. doi:10.3324/haematol.2016.145847

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1426

1426

e sought to confirm the prognostic importance of simple clinically available biomarkers of C-reactive protein, serum albumin, and ferritin prior to allogeneic hematopoietic cell transplantation. The study population consisted of 784 adults with acute myeloid leukemia in remission or myelodysplastic syndromes undergoing unrelated donor transplant reported to the Center for International Blood and Marrow Transplant Research. C-reactive protein and ferritin were centrally quantified by ELISA from cryopreserved plasma whereas each center provided pre-transplant albumin. In multivariate analysis, transplant-related mortality was associated with the pre-specified thresholds of C-reactive protein more than 10 mg/L (P=0.008) and albumin less than 3.5 g/dL (P=0.01) but not ferritin more than 2500 ng/mL. Only low albumin independently influenced overall mortality. Optimal thresholds affecting transplant-related mortality were defined as: Creactive protein more than 3.67 mg/L, log(ferritin), and albumin less than 3.4 g/dL. A 3-level biomarker risk group based on these values separated risks of transplant-related mortality: low risk (reference), intermediate (HR=1.66, P=0.015), and high risk (HR=2.7, P<0.001). One-year survival was 74%, 67% and 56% for low-, intermediate- and high-risk groups. Routinely available pre-transplant biomarkers independently risk-stratify for transplant-related mortality and survival.

haematologica | 2016; 101(11)

Biomarkers and transplant

Introduction Successful outcomes following allogeneic hematopoietic cell transplant (HCT) may be offset by risks of transplant-related morbidity and mortality.1 Traditional factors used to gauge transplant risks include conditioning regimen intensity,2 immunosuppression and HLA matching,3 along with hematopoietic cell source, and graft cell yield.4,5 The importance of patient factors on morbidity and transplant-related mortality (TRM) has gained greater appreciation, particularly in the era of HCT for older and/or more fragile patients.6 Comorbidity determined by the hematopoietic cell transplantation-comorbidity index (HCT-CI),7 predicts for worse survival, generally through higher TRM. Impaired performance status interferes with transplant success.7-9 Detailed inventories of patient health through Geriatric Assessment (GA) enhance risk-stratification10 but are time consuming and mostly apply to older adults. Simple, objective, available, and validated prognostic markers are required to more precisely estimate TRM and ideally inform targeted interventions to mitigate adverse risks. Serum biomarkers hold promise in this regard. Prior single institutional studies have suggested elevated C-reactive protein (CRP),11,12 ferritin13,14 and lower albumin15 prior to allogeneic HCT are associated with greater TRM and worse overall survival (OS). Although inflammation affects all three measures, each biomarker has separate and distinct underlying pathways and clinical uses. CRP is an acute phase reactant clinically applied to quantify inflammation, but is also prognostic for long-term function and mortality outside the HCT setting.16 Serum ferritin estimates iron stores.13 Albumin reflects protein synthesis and levels decline in relation to protein restriction and inflammation.17 Hypoalbuminemia represents a risk factor for mortality in older patients in general.18 Prior studies of these biomarkers in HCT have shown inconsistent results on transplant outcomes. We sought to validate the independent prognostic value of pre-conditioning serum CRP, ferritin, and albumin in a large, homogenous, sample derived from transplant registry data.

lated donor allograft for acute myeloid leukemia (AML) in remission or myelodysplastic syndromes (MDS) with less than 5% blasts at time of transplant and available, cryopreserved, plasma samples collected before HCT conditioning were considered. The study period covered transplant from 2008 to 2010 to allow for comorbidity adjustment as routine capture of the HCT-CI began in 2008. Recipients of T-cell depleted allografts and prior autologous or allogeneic transplant were excluded.

Study end points and definitions For validation of TRM, pre-specified, literature-derived, biomarker thresholds were defined as follows: CRP more than 10 mg/L,11 ferritin more than 2500 ng/mL,14 and albumin less than 3.5 g/dL.15 Time to death without evidence of disease relapse defined TRM. Relapse is the competing risk, and patients surviving in continuous complete remission were censored at last follow up. Secondary end points included cumulative incidence of relapse, progression-free survival (PFS), overall survival (OS), acute graft-versus-host disease (GvHD) grade II-IV and grade III-IV. A secondary analysis was planned to explore the optimal biomarker cut-off points and construct an overall biomarker model for future testing.

Sample collection and processing Peripheral blood plasma samples were obtained prior to conditioning and cryopreserved at -80˚C in the CIBMTR biorepository. The University of Chicago later analyzed the samples by enzymelinked immunosorbent assay (ELISA) for ferritin and high-sensitivity CRP (kits from MP Biomedicals). Each center provided pretransplant patient albumin values as available from the medical record before conditioning started. The limited sample volume precluded automated chemistry machine analysis for biomarkers.

Correlation of biomarkers by automated methods and ELISA

Methods

To correlate automated chemistry derived CRP and ferritin (as routinely performed in clinical laboratories) to ELISA, we obtained de-identified serum samples analyzed on the Roche Cobas© 8000 (Roche Diagnostics, USA) covering the entire biomarker range after institutional review board approval. CRP encompassed four ranges in mg/L: < 3, 3-5, 6- 10, and >10. Likewise, ferritin spanned four groups as follows: < 500, 500-1000, >1000-1500, > 1500 (ng/mL). These samples were cryopreserved, thawed and then simultaneously analyzed on the Roche Cobas© 8000 and by ELISA as above for correlation analysis by each technique.

Data source

Statistical analysis

The Center for International Blood and Transplant Research (CIBMTR) is a voluntary working group of more than 450 transplant centers worldwide who contribute data on consecutive allogeneic HCT to a statistical center housed both at the Medical College of Wisconsin in Milwaukee and the National Marrow Donor Program (NMDP) in Minneapolis. Patients are followed longitudinally with yearly follow up. Computerized checks for errors and on-site audits of participating centers ensure data quality. Physician review of data and additional requested information from reporting centers are included. Observational studies conducted by the CIBMTR are performed with a waiver of informed consent and in compliance with HIPAA regulations, as determined by the Institutional Review Board and the Privacy Officer of the Medical College of Wisconsin.

Univariate probabilities of PFS and OS were calculated using the Kaplan-Meier estimator with variance estimated by Greenwood’s formula. Probabilities of acute GvHD, TRM and relapse were calculated using cumulative incidence curves to accommodate competing risks. Ninety-five percent confidence intervals (95%CI) for all probabilities and P-values of pairwise comparisons were derived from pointwise estimates and calculated using standard techniques. Pearson correlation coefficients were generated and compared for each biomarker and standard clinical characteristics. Multivariate Cox models were fit for each outcome of interest after adjusting for clinical characteristics by stepwise regression, omitting the biomarkers. We modeled biomarkers in several ways. First, the analysis used protocol specified biomarker cut-off points, entering each biomarker individually and generating hazard ratios, 95% confidence intervals (CI) and probability values for each. Second, the functional form of the relationship between each biomarker and TRM was determined using splines by examining Martingale residual plots, and considering potential interac-

Patient selection Patients aged 18 years or over, receiving either an 8/8 or 7/8 human leukocyte antigen matched (i.e. HLA-A, B, C, DRB1) unrehaematologica | 2016; 101(11)

1427

A.S. Artz et al. Table 1. Baseline characteristics for AML and MDS patients undergoing unrelated donor hematopoietic cell transplantation.

Patients' characteristics Number of patients Number of centers Age, median (range), years 18-50 >50 Male Race Caucasian Non-Caucasian or unknown HCT-CI 0 1-2 ≥3 Karnofsky score ≥ 90% <90% Missing Time from diagnosis to treatment, median (range), months <12 months ≥12 months Disease AML MDS AML disease status at transplant AML early AML intermediate MDS transplant indication MDS early MDS advanced HLA match (HLA-A, B, C, DRB1) HLA 7/8 HLA 8/8 Donor/recipient sex match M-M F-M M-F F-F Missing Donor/recipient CMV serological status -/+/+ +/-/+ Unknown Graft type Bone marrow Peripheral blood Conditioning regimen intensity Myeloablative Reduced intensity Non-myeloablative Conditioning regimen MA Flu+Bu+/-other MA Bu+cy+/-other MA Cy+TBI+/-other MA other RIC Flu+Bu+/-other RIC Flu+Mel+/- other RIC other NMA Flu+TBI+/-other 1428

N (%) 784 98 50 (18-78) 400 (51) 384 (49) 402 (51) 743 (95) 41 ( 5) 266 (34) 244 (31) 274 (35) 503 (64) 247 (32) 34 (4) 6 (1-291) 579 (74) 205 (26)

NMA TLI+/-other NMA other GvHD prophylaxis Calcineurin inhibitor +MTX+/-other Calcineurin inhibitor +/-other (no MTX) Other a Biomarkers C-reactive protein, median (range), mg/L (n=783) Ferritin, median (range), ng/mL (n=781) Albumin reported by centers, median (range), g/dL (n=695) Year of transplant 2008 2009 2010 Median follow up of survivors, range, months

11 (1) 17 (2) 505 (64) 265 (34) 14 (2) 5.0 (0.3-316.3) 1148 (51-14298) 3.6 (0.6-5.3) 160 (20) 362 (46) 262 (33) 38 (5-61)

HCT-CI: hematopoietic cell transplant-comorbidity index; AML: acute myeloid leukemia; MDS: myelodysplastic syndromes; HLA: human leukocyte antigen; M: male; F: female; CMV: cytomegalovirus; Flu: fludarabine; Bu: busulfan, Cy: cyclophosphamide; Mel: melphalan; TBI: total body irradiation; TLI: total lymphoid irradiation; MTX: methotrexate; AML early disease: first complete remission; AML intermediate disease: second complete remission or greater; MDS early disease: refractory anemia or refractory anemia with ring sideroblasts; MDS advanced disease: refractory anemia with excess blasts or chronic myelomonocytic leukemia; GvHD: graft-versus-host disease; MA: myeloablative; RIC: reduced intensity conditioning; NMA: non-myeloablative. Blasts < 5% at time of transplant for all MDS patients. aIncluding: corticosteroids, monoclonal antibody, mycophenolate mofetil, sirolimus, ursodiol.

626 (80) 158 (20) 441 (70) 185 (30) 80 (51) 78 (49) 177 (23) 607 (77) 290 (37) 110 (14) 246 (31) 136 (17) 2 (<1) 236 (30) 179 (23) 78 (10) 247 (35) 17 ( 2) 136 (17) 648 (83) 566 (72) 168 (21) 50 (6) 207 (26) 183 (23) 155 (20) 21 (3) 106 (14) 38 (5) 24 (3) 22 (3)

tions, leading to models with an optimal threshold. Finally, we constructed a combined model using optimal biomarker levels in their final functional form entered in the multivariate model simultaneously. Risk factors with P<0.05 were included in the model. All computations were performed using the statistical package SAS v.9.1. Linear regression was used to correlate CRP and ferritin values by ELISA and chemistry procedures. Due to variability at very high values, prior to regression, biomarkers underwent log transformation.

Results Patients’ characteristics Table 1 summarizes baseline characteristics of the study cohort. In brief, the median age was 50 years (range 18-78 years) among the 784 patients across 98 centers. Most subjects were Caucasian (95%), and received myeloablative regimens (72%). The primary disease indication was AML (80%), and the major graft source was peripheral blood (83%). Median follow up was 38 months.

Biomarker levels and correlations Almost all patients were evaluable for CRP (99.9%) and ferritin (99.6%) by ELISA. The transplant centers supplied albumin results in 88.6% of patients. The median values were 5.0 mg/L for CRP, 1148 ng/dL for ferritin and 3.6 g/dL for albumin (Table 1). Table 2 shows the correlation of biomarkers and clinical factors. A modest correlation was found between CRP and ferritin (r=0.35, P<0.001) and inverse association with albumin (r=-0.12, P=0.002). In addition, CRP and albumin were not associated with patient age, HCT-CI, disease or conditioning intensity. Higher log ferritin was significantly associated with greater comorbidity by HCT-CI (P=0.014) and AML relative to MDS (P<0.001). haematologica | 2016; 101(11)

Biomarkers and transplant

Table 2. Association of biomarkers with clinical characteristics.

Characteristic Biomarker Log (ferritin) Albumin Clinical characteristics Older age Higher HCT-CI Disease, AML vs. MDS Conditioning

Log (CRP) P

Log (Ferritin) P

Albumin P

<0.001 0.002*

0.086 -

0.49 0.065 0.399 0.95

0.12 0.014 <0.001 0.98

0.42 0.12 0.89 0.96

ROC for Day 180 TRM, using Grouped Biomarker Panel

HCT-CI: hematopoietic cell transplant-comorbidity index; AML: acute myeloid leukemia; MDS myelodysplastic syndromes.*Inverse correlation.

Overall outcomes The cumulative incidence of acute GvHD grades II-IV was 43% (95%CI: 39%-46%) and grade III-IV was 15% (95%CI: 13%-18%) at 100 days. Transplant-related mortality was 18% (95%CI: 16%-21%) at one year and 25% (95%CI: 22%-28%) at three years. Progression-free survival ranged from 57% (95%CI: 54%-61%) at one year to 45% (95%CI: 43%-49%) at three years, whereas 1-year survival was 65% (95%CI: 62%-68%) and 3-year survival was 49% (95%CI: 45%-52%).

Multivariate analysis of outcomes Table 3 summarizes the multivariate analysis adjusted for clinical factors for TRM and OS for the biomarkers. We validated the association of greater TRM for CRP above 10 mg/L (P=0.008) and albumin less than 3.5 g/L (P=0.01) but not for ferritin above 2500 ng/mL (P=0.27), seen in only a small fraction of patients (12%), ferritin levels exceeded 2500 ng/mL. Only hypoalbuminemia was associated with worse OS as CRP was of borderline significance (P=0.07). The optimally defined biomarker thresholds to predict TRM in unadjusted models specified CRP more than 3.67 mg/dL (P=0.01), linear ferritin (P=0.07) after log transformation and albumin less than 3.4 g/dL (P=0.004). A combined model incorporated all three biomarkers at these optimal cut-off points and adjusted for significant clinical factors. Higher comorbidity (HCT-CI of 3 or more vs. HCT-CI of 0) was of borderline significance for OS (HR=1.27, 95%CI: 0.99-1.6, P=0.06) without a significant effect on TRM (HR=1.14, 95%CI: 0.79-1.6, P=0.48). Inclusion of HCT-CI did not materially influence the biomarkers effects in this model. The optimal biomarker thresholds for TRM also adversely influenced OS (Table 3). None of the biomarkers, either at pre-specified or optimal thresholds, were associated with relapse or incidence of acute GvHD, either grade II-IV or grade III-IV (data not shown).

Figure 1. Receiver operating characteristic (ROC) curves for day 180 transplantrelated mortality for three groups: biomarker panel group alone (panel gp only), clinical characteristics alone, or both [panel group (gp) and clinical characteristics].

(panel>2.0, n=159) biomarker risk panel group were similar to those employing a continuous biomarker panel (data not shown). The area under the ROC curve was 0.61, 0.65, and 0.68 for models using only the biomarker panel, only clinical characteristics, and both, respectively (Figure 1). The high-risk biomarker panel group exacted a similar or greater toll on TRM (HR=2.72, P<0.001) among known factors (Online Supplementary Table S1). Clinical characteristics and biomarkers thus may provide added utility for predicting TRM. The multivariate analyses of this biomarker panel are provided in Table 3 for TRM and OS. Figure 2 depicts TRM (Figure 2A) and overall survival (Figure 2B) by each group showing significantly different survival when grouped according to biomarker alone. Clinically, 1-year survival was 74%, 67% and 56% by low, intermediate and high biomarker scores, respectively. Figure 3 shows the significant association of this biomarker panel on OS (Figure 3A and B) and TRM (Figure 3C and D) for both myeloablative and reduced intensity/nonmyeloablative regimens. Finally, we summarized causes of death for each biomarker category (Online Supplementary Table S2). Most causes of non-relapse death appeared relatively similar including GvHD, except perhaps a greater fraction of â&#x20AC;&#x153;organ failureâ&#x20AC;? for high biomarker panel risk patients.

Biomarker Risk Score

Correlation of biomarkers by automated methods and ELISA

To improve clinical use for prognostication and create a simpler tool for future studies, we next generated a 3-level risk group for TRM based on a biomarker panel alone applying the following formula: Panel =0.44633*I(CRP>3.67) + 0.18330*ln(ferritin) + 0.44730*I(albumin<3.4), where I() is an indicator function which has the value 1 if the expression in parentheses is true, and 0 otherwise. The ROC characteristics of this 3level panel on day 180 TRM of low (panel<=1.5, n=200), intermediate (1.5<panel<=2.0, n=331), and high

We performed a pairwise analysis of 80 samples for CRP and 73 samples for ferritin unrelated to these study samples by ELISA and automated chemistry. The R squared for CRP and ferritin were both 95% and Online Supplementary Figures S1 and S2 depict scatterplots for each. CRP by ELISA of 10 mg/L (protocol-defined) and 3.67 mg/L (optimal threshold) corresponding to a calculated chemistry value of 8.45 mg/L and 3.54 mg/L, respectively. Applying the corresponding chemistry values for ELISA-derived CRP and ferritin would give a slightly mod-

haematologica | 2016; 101(11)

1429

A.S. Artz et al.

ified biomarker risk panel as follows: Panel =0.44633*I(CRP_Chemistry>3.54) + 0.07184+0.19646 *ln(ferritin_Chemistry) + 0.44730*I(albumin<3.4), where I() is an indicator function which has the value 1 if the expression in parentheses is true, and 0 otherwise. CRP converts to more than 3.0 g/dL assuming the values are reported as an integer in mg/L. Thus, for future studies or clinical use for biomarkers analyzed by routine automated chemistry methods, this formula may be preferred.

Discussion We report the first validation study regarding the impact of simple, clinically available, serum biomarkers of CRP, ferritin and albumin obtained prior to allogeneic transplant conditioning transplant on subsequent outcomes. We confirmed that the protocol-specified thresholds of CRP more than 10 mg/L and albumin less than 3.5 g/dL each predicted for approximately 50% greater risks of TRM independent of clinical factors. Hypoalbuminemia showed a stronger association with inferior survival (P=0.002) whereas the significance for CRP more than 10 mg/L was borderline (P=0.072). Ferritin in excess of 2500 ng/mL did not heighten risks of TRM or impair survival. None of the biomarkers significantly affected relapse or acute GvHD suggesting abnormal biomarkers represent a non-specific vulnerability to transplant toxicity. Variability of biomarker thresholds and lack of validation have hampered adoption of biomarkers as prognostic tools in practice. For validation, we initially specified thresholds based on the published literature. The lack of an association of ferritin and outcome contradicts prior reports and underscores the critical importance of confirmatory findings. Ferritin thresholds in prior reports covered a range of 500 ng/mL to above 2600 ng/mL;14,19-21 we elected to approximate the inflection point originally reported by Armand et al. in similar patients. In the study by Vaughn et al. and a meta-analysis, ferritin of 1000 ng/mL best stratified for outcome.22,23 Still, even at a threshold of 1000 ng/mL, we found no association with OS (HR=1.02, P=0.92), arguing against the prognostic value of a discrete ferritin cut-off point in myeloid malignancy patients undergoing allogeneic HCT. Prior studies likely have been biased as testing may have been performed as clinically indicated, enriching for patients at risk of iron overload and/or complications. For example, in the original Armand report, ferritin was available in 64% and a recent study by Vaughn et al. revealed serum ferritin only evaluable in 50%.23 Furthermore, most studies select a cutoff point best fitting the available data which also tends to overestimate effects. We next explored optimal biomarker thresholds and established CRP more than 3.67 mg/L, linear ferritin after log transformation and albumin less than 3.4 g/dL best predicted for TRM. These optimal biomarker thresholds for TRM generally held for an adverse effect on overall survival. Patients with high-risk biomarker scores fared worse after myeloablative or reduced intensity/non-myeloablative regimens, although the effects may have been more pronounced after myeloablative rgimens. The HCT-CI is perhaps the best established factor7 predicting transplant-related mortality, and the report by Vaughn et al. showing ferritin, platelet count and albumin augment HCT-CI reinforces our findings.23 Although the 1430

HCT-CI was only of borderline significance for OS, this was attributable to the less robust sample size as the hazard ratio of 1.27 for higher comorbidity mirrored recent registry studies with larger samples.24 Additional tools to estimate TRM, such as Geriatric Assessment (GA), likely further refine prognostication in older adults but CRP and albumin retain prognostic significance when considering GA.10,25 Biomarkers will likely supplement rather than replace the HCT-CI. Therefore, a wealth of date now support biomarkers as independently prognostic for TRM and survival in the allogeneic HCT setting independent of all clinical factors studied.

A P<0.001

B P<0.001

Figure 2. Outcomes after unrelated donor allogeneic transplant in each biomarker risk score group. (A) Transplant related mortality for low, intermediate and high biomarker risk panel group. (B) Overall survival for low, intermediate and high biomarker risk panel group.

haematologica | 2016; 101(11)

Biomarkers and transplant

A biomarker-based panel for risk stratification would be a clinically valuable tool and thus confirmatory studies are critically important. Objective and accurate estimates of low rates of early TRM would reassure both patients and providers in considering allogeneic HCT especially in older patients where transplant rates are very low.26 Conversely, patients at high risk of TRM should proceed

with caution, contemplate avoiding myeloablative regimens, pursue more thorough evaluation by Geriatric Assessment, or escalate post-transplant monitoring. The ready clinical availability of these laboratory tests position biomarkers well to facilitate comparison of patient populations in prospective studies or even retrospectively when data and/or samples are available.

Table 3. Multivariate analysis of transplant-related mortality and overall survival for biomarkers and Biomarker Panel Risk Score.*

Abnormal, N (%)

Overall survival*

95% CI

184 (24) 93 (12) 290 (42)

1.52 1.26 1.48

1.11-2.07 0.84-1.88 1.10-2.0

0.008 0.27 0.010

1.22 1.15 1.39

0.98-1.53 0.86-1.54 1.13-1.72

0.072 0.35 0.002

497 (64) n/a 203 (29.5) N 200 331 159

1.56 1.20 1.56

1.11-2.19 0.99-1.46 1.15-2.13

0.010 0.07 0.004

1.25 1.20 1.38

1.00-1.56 1.05-1.38 1.10-1.72

0.047 0.009 0.005

1.0 1.66 2.72

1.10-2.49 1.77-4.19

0.015 <0.001

1.0 1.38 2.01

1.06-1.80 1.51-2.70

0.017 <0.001

Biomarker Validation thresholds CRP >10 mg/L Ferritin >2500 ng/mL Albumin < 3.5 g/dL Optimal threshold, combined model** CRP >3.67 mg/L Log (ferritin), linear <3.4 g/dL Biomarker Score** Low (<1.5) Intermediate (1.5-2.0) High (>2.0)

Transplant-related mortality*

CRP: C-reactive protein; n/a: not applicable; HR: Hazards Ratio; CI: confidence interval. *Transplant-related mortality adjusted for Body Mass Index (BMI), disease, HLA mismatch, and D/R CMV; stratified on conditioning regimen. Overall survival adjusted for BMI, age, disease/status, HLA mismatch, and sex match; stratified on conditioning regimen. **Combined models and Biomarker Score includes all biomarkers.

Overall Survival, MA

B P<0.001

Transplant-related Mortality, MA P<0.001

Overall Survival, NMA P=0.03

Transplant-related Mortality, NMA P=0.01

Figure 3. Overall survival for low, intermediate and high biomarker risk panel groups in myeloablative (MA) conditioning cohort (A) and reduced intensity/nonmyeloablative regimen cohorts (NMA) (B). Transplant-related mortality for low, intermediate and high biomarker risk panel groups in MA conditioning cohort (C) and reduced intensity/NMA regimen cohorts (D).

haematologica | 2016; 101(11)

1431

A.S. Artz et al.

This study has several limitations. Although the biomarker panel demonstrated a profound effect on TRM beyond all clinical factors (HR=2.7), this must be validated, especially with different diseases and conditioning regimens. We opted to include ferritin in our final biomarker panel even though there was only a borderline association with TRM (P=0.07) as the association with survival was strong (P=0.009). Future studies should examine the utility of measuring biomarkers earlier, such as at the time of transplant referral. The analysis and cut-off points utilized ELISA-generated values from cryopreserved samples. As data comparing these methods were limited, we initiated a sub-study using entirely different samples and found a very strong correlation of ELISA and automated chemistry, as suggested in prior studies.27,28 As centers begin to perform and use these biomarkers routinely in the clinical laboratory, these data provide reassurance that CRP and ferritin likely approximate the values in this report. The biological drivers for abnormal biomarkers including inflammation, iron burden, nutritional status, infection and genetic polymorphisms require further exploration and are critically important in moving beyond patient selection and toward mitigating adverse outcomes in those at heightened risk. Inflammation may be a central theme, yet the modest correlations of biomarkers to each other and prognostically independent effects of each biomarker support the role of non-inflammatory pathways as well. Even then, inflammation can be influenced by divergent pathways of infection, hepatic sources, the primary disease, or genetic origins. The notion of high ferritin as merely a measure of high iron stores appears too simplistic, as high levels of iron stores evaluated according to liver iron content do not solely contribute to adverse outcomes.22,29 Albumin crudely estimates nutrition and more broadly reflects the equilibrium between protein synthesis and catabolism. Poor nutrition itself can result from a multitude of factors and organ dysfunctions. Finally, low albumin increases bioavailability of drugs and could increase toxicity. Future studies elucidating the pathophysiology of biomarkers before transplant may guide biomarker-driven interventional studies to ameliorate the increased risks of TRM. For example, statins have been tested to dampen inflammation in those with elevated CRP.30 In summary, pre-transplant CRP and albumin independently predict for greater TRM but not high ferritin. The association of a new biomarker risk score based on optimal thresholds for ferritin, CRP and albumin on higher

References 1. Horan JT, Logan BR, Agovi-Johnson MA, et al. Reducing the risk for transplantationrelated mortality after allogeneic hematopoietic cell transplantation: how much progress has been made? J Clin Oncol. 2011;29(7):805-813. 2. Wong R, Giralt SA, Martin T, et al. Reduced-intensity conditioning for unrelated donor hematopoietic stem cell transplantation as treatment for myeloid malignancies in patients older than 55 years. Blood. 2003;102(8):3052-3059. 3. Lee SJ, Klein J, Haagenson M, et al. High-

1432

risks of TRM and inferior survival should be validated in independent populations. Future research should investigate the biology underlying biomarker abnormalities to design studies to mitigate the high risks of TRM. Funding The work was partially supported from CTSA Grant UL1 RR024999 and the National Center for Advancing Translational Sciences of the National Institutes of Health UL1 TR000430 (AA) CIBMTR Support List The CIBMTR is supported by Public Health Service Grant/Cooperative Agreement 5U24-CA076518 from the National Cancer Institute (NCI), the National Heart, Lung and Blood Institute (NHLBI) and the National Institute of Allergy and Infectious Diseases (NIAID); a Grant/Cooperative Agreement 5U10HL069294 from NHLBI and NCI; a contract HHSH250201200016C with Health Resources and Services Administration (HRSA/DHHS); two Grants N00014-13-10039 and N00014-14-1-0028 from the Office of Naval Research; and grants from Alexion; *Amgen, Inc.; Anonymous donation to the Medical College of Wisconsin; Be the Match Foundation; *Bristol Myers Squibb Oncology; *Celgene Corporation; *Chimerix, Inc.; Fred Hutchinson Cancer Research Center; Gamida Cell Ltd.; Genentech, Inc.; Genzyme Corporation; *Gilead Sciences, Inc.; Health Research, Inc. Roswell Park Cancer Institute; HistoGenetics, Inc.; Incyte Corporation; *Jazz Pharmaceuticals, Inc.; Jeff Gordon Children’s Foundation; The Leukemia & Lymphoma Society; The Medical College of Wisconsin; Merck & Co, Inc.; Mesoblast; *Millennium: The Takeda Oncology Co.; *Miltenyi Biotec, Inc.; National Marrow Donor Program; Neovii Biotech NA, Inc.; Novartis Pharmaceuticals Corporation; Onyx Pharmaceuticals; Optum Healthcare Solutions, Inc.; Otsuka America Pharmaceutical, Inc.; Otsuka Pharmaceutical Co, Ltd. – Japan; Oxford Immunotec; Perkin Elmer, Inc.; Pharmacyclics; *Sanofi US; Seattle Genetics; Sigma-Tau Pharmaceuticals; *Spectrum Pharmaceuticals, Inc.; St. Baldrick’s Foundation; *Sunesis Pharmaceuticals, Inc.; Swedish Orphan Biovitrum, Inc.; Telomere Diagnostics, Inc.; TerumoBCT; Therakos, Inc.; University of Minnesota; and *Wellpoint, Inc. The views expressed in this article do not reflect the official policy or position of the National Institute of Health, the Department of the Navy, the Department of Defense, Health Resources and Services Administration (HRSA) or any other agency of the U.S. Government. *Corporate Members

resolution donor-recipient HLA matching contributes to the success of unrelated donor marrow transplantation. Blood. 2007;110(13):4576-4583. 4. Pulsipher MA, Boucher KM, Wall D, et al. Reduced-intensity allogeneic transplantation in pediatric patients ineligible for myeloablative therapy: results of the Pediatric Blood and Marrow Transplant Consortium Study ONC0313. Blood. 2009;114(7):14291436. 5. Gratwohl A, Hermans J, Goldman JM, et al. Risk assessment for patients with chronic myeloid leukaemia before allogeneic blood or marrow transplantation. Chronic Leukemia Working Party of the European

Group for Blood and Marrow Transplantation. Lancet. 1998; 352(9134):1087-1092. 6. Sorror ML, Sandmaier BM, Storer BE, et al. Long-term outcomes among older patients following nonmyeloablative conditioning and allogeneic hematopoietic cell transplantation for advanced hematologic malignancies. JAMA. 2011;306(17):1874-1883. 7. Sorror ML, Maris MB, Storb R, et al. Hematopoietic cell transplantation (HCT)specific comorbidity index: a new tool for risk assessment before allogeneic HCT. Blood. 2005;106(8):2912-2919. 8. Artz AS, Pollyea DA, Kocherginsky M, et al. Performance status and comorbidity pre-

haematologica | 2016; 101(11)

Biomarkers and transplant

10.

11.

12.

13.

14.

15.

dict transplant-related mortality after allogeneic hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2006;12(9):954-964. Sorror ML, Storb RF, Sandmaier BM, et al. Comorbidity-age index: a clinical measure of biologic age before allogeneic hematopoietic cell transplantation. J Clin Oncol. 2014;32(29):3249-3256. Muffly LS, Kocherginsky M, Stock W, et al. Geriatric assessment to predict survival in older allogeneic hematopoietic cell transplantation recipients. Haematologica. 2014;99(8):1373-1379. Pavlu J, Kew AK, Taylor-Roberts B, et al. Optimizing patient selection for myeloablative allogeneic hematopoietic cell transplantation in chronic myeloid leukemia in chronic phase. Blood. 2010;115(20):40184020. Artz AS, Wickrema A, Dinner S, et al. Pretreatment C-reactive protein is a predictor for outcomes after reduced-intensity allogeneic hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2008;14(11):1209-1216. Konijn AM, Hershko C. Ferritin synthesis in inflammation. I. Pathogenesis of impaired iron release. Br J Haematol. 1977;37(1):7-16. Armand P, Kim HT, Cutler CS, et al. Prognostic impact of elevated pretransplantation serum ferritin in patients undergoing myeloablative stem cell transplantation. Blood. 2007;109(10):4586-4588. Wong R, Shahjahan M, Wang X, et al. Prognostic factors for outcomes of patients with refractory or relapsed acute myelogenous leukemia or myelodysplastic syndromes undergoing allogeneic progenitor cell transplantation. Biol Blood Marrow Transplant. 2005;11(2):108-114.

haematologica | 2016; 101(11)

16. Reuben DB, Cheh AI, Harris TB, et al. Peripheral blood markers of inflammation predict mortality and functional decline in high-functioning community-dwelling older persons. J Am Geriatr Soc. 2002;50(4):638-644. 17. Moshage HJ, Janssen JA, Franssen JH, Hafkenscheid JC, Yap SH. Study of the molecular mechanism of decreased liver synthesis of albumin in inflammation. J Clin Invest. 1987;79(6):1635-1641. 18. Corti MC, Guralnik JM, Salive ME, Sorkin JD. Serum albumin level and physical disability as predictors of mortality in older persons. JAMA. 1994;272(13):1036-1042. 19. Sucak GT, Yegin ZA, Ozkurt ZN, Aki SZ, Yagci M. Iron overload: predictor of adverse outcome in hematopoietic stem cell transplantation. Transplant Proc. 42(5):1841-1848. 20. Maradei SC, Maiolino A, de Azevedo AM, Colares M, Bouzas LF, Nucci M. Serum ferritin as risk factor for sinusoidal obstruction syndrome of the liver in patients undergoing hematopoietic stem cell transplantation. Blood. 2009;114(6):1270-1275. 21. Lim ZY, Fiaccadori V, Gandhi S, et al. Impact of pre-transplant serum ferritin on outcomes of patients with myelodysplastic syndromes or secondary acute myeloid leukaemia receiving reduced intensity conditioning allogeneic haematopoietic stem cell transplantation. Leuk Res. 2010; 34(6):723-727. 22. Armand P, Kim HT, Virtanen JM, et al. Iron overload in allogeneic hematopoietic cell transplantation outcome: a meta-analysis. Biol Blood Marrow Transplant. 2014;20(8):1248-1251. 23. Vaughn JE, Storer B, Armand P, et al. Design and Validation of an Augmented Hematopoietic Cell Transplantation-

24.

25.

26.

27.

28.

29.

30.

Comorbidity Index Comprising Pretransplant Ferritin, Albumin, and Platelet Count for Prediction of Outcomes after Allogeneic Transplantation. Biol Blood Marrow Transplant. 2015;21(8):1418-24. Armand P, Kim HT, Logan BR, et al. Validation and refinement of the Disease Risk Index for allogeneic stem cell transplantation. Blood. 2014;123(23):3664-3671. Muffly LS, Boulukos M, Swanson K, et al. Pilot study of comprehensive geriatric assessment (CGA) in allogeneic transplant: CGA captures a high prevalence of vulnerabilities in older transplant recipients. Biol Blood Marrow Transplant. 2013;19(3):429434. Yao S, Hahn T, Zhang Y, et al. Unrelated donor allogeneic hematopoietic cell transplantation is underused as a curative therapy in eligible patients from the United States. Biol Blood Marrow Transplant. 2013;19(10):1459-1464. Simo JM, Joven J, Cliville X, Sans T. Automated latex agglutination immunoassay of serum ferritin with a centrifugal analyzer. Clin Chem. 1994;40(4):625-629. Rifai N, Tracy RP, Ridker PM. Clinical efficacy of an automated high-sensitivity Creactive protein assay. Clin Chem. 1999;45(12):2136-2141. Trottier BJ, Burns LJ, DeFor TE, Cooley S, Majhail NS. Association of iron overload with allogeneic hematopoietic cell transplantation outcomes: a prospective cohort study using R2-MRI-measured liver iron content. Blood. 2013;122(9):1678-1684. Albert MA, Danielson E, Rifai N, Ridker PM. Effect of statin therapy on C-reactive protein levels: the pravastatin inflammation/CRP evaluation (PRINCE): a randomized trial and cohort study. JAMA. 2001;286(1):64-70.

1433

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Immunology & Inflammation

Ferrata Storti Foundation

A score of low-grade inflammation and risk of mortality: prospective findings from the Moli-sani study

Marialaura Bonaccio, Augusto Di Castelnuovo, George Pounis, Amalia De Curtis, Simona Costanzo, Mariarosaria Persichillo, Chiara Cerletti, Maria Benedetta Donati, Giovanni de Gaetano, and Licia Iacoviello on behalf of the Moli-sani Study Investigators*

Haematologica 2016 Volume 101(11):1434-1441

Department of Epidemiology and Prevention, IRCCS Istituto Neurologico Mediterraneo NEUROMED, Pozzilli (IS), Italy *Moli-sani study Investigators are listed in Appendix 1

ABSTRACT

Correspondence: marialaura.bonaccio@neuromed.it

Received: February 5, 2016. Accepted: July 26, 2016. Pre-published: October 14, 2016. doi:10.3324/haematol.2016.144055

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/11/1434

ow-grade inflammation is associated with an increased risk of chronic degenerative disease, but its relationship with mortality is less well explored. We aimed at evaluating, at a large epidemiological level, the possible association of low-grade inflammation, as measured by a composite score, with overall mortality risk. We conducted a population-based prospective investigation on 20,337 adult subjects free from major hematological disease and acute inflammatory status, randomly recruited from the general population of the Moli-sani study. A low-grade inflammation score was obtained from the sum of 10-tiles of plasmatic (C-reactive protein) and cellular (leukocyte and platelet counts, granulocyte/lymphocyte ratio) biomarkers of low-grade inflammation; higher levels indicated increased low-grade inflammation. Hazard ratios were calculated using multivariable Cox proportional hazard models with 95% confidence intervals. At the end of follow-up (median 7.6 years), 837 all-cause deaths were recorded. As compared to subjects in the lowest quartile of the low-grade inflammation score, those in the highest category had a significantly increased risk in overall mortality (HR=1.44; 1.17-1.77), independently of possible confounders, including the presence of chronic diseases and a number of health-related behaviors. The magnitude of the association of low-grade inflammation with mortality was relatively higher in type 2 diabetic patients (HR=2.90; 1.74-4.84) and in individuals with a history of cardiovascular disease (HR=2.48; 1.50-4.11) as compared to their counterparts who were free from the disease. In conclusion, an elevated degree of lowgrade inflammation, as measured by a composite score of inflammatory biomarkers, is an independent risk factor for total mortality in an apparently healthy adult general population. Introduction

1434

Low-grade inflammation is a condition not yet consistently defined or measured. A number of plasmatic (e.g. C-reactive protein) or cellular biomarkers (e.g. white blood cell and platelet counts) have been proposed as reliable indicators of such a condition.1,2 This subclinical disorder has been recognized as a risk factor for a number of chronic diseases including cancer, cardiovascular (CVD) and neurodegenerative disease.3-6 In contrast, its relationship with mortality has been poorly investigated, at least in the general population,7,8 while evidence within high-risk groups is more robust.9-11 Low-grade inflammation has also been proposed as an underlying pathophysiological mechanism linking risk factors or metabolic disorders (e.g. oxidative stress, haematologica | 2016; 101(11)

Low-grade inflammation and mortality

obesity, diabetes, dyslipidemia), to an increased risk of chronic degenerative disease1 as well as a common pathogenic denominator in age-related diseases.12 Pioneering large-scale studies focused on circulating fibrinogen, C-reactive protein (CRP), and white blood cell (WBC) counts as reliable inflammatory biomarkers, mostly in relation to cardiovascular events.6,13-16 More recently, a pro-inflammatory action of blood platelets has been proposed,17 whereas the neutrophil-to-lymphocyte ratio better expresses an early inflammatory cellular response.18,19 Evidence of the individual contribution of each of the above mentioned inflammatory biomarkers to different health outcomes is scarce.7,14,17 Previous data suggest that some inflammation biomarkers are associated with lifestyle modifications (e.g. dietary habits17) or electrocardiographic parameters,19 emphasizing the need for further study on their association with clinical outcomes, such as the incidence of chronic degenerative disease and mortality rates. In this context, a comprehensive approach to measure a low-grade inflammation condition has been proposed in high cardiovascular risk subjects.20,21 Lately, a significant inverse association of a composite low-grade inflammation score with dietary polyphenol intake has been observed by our group in a population-based cohort.22 The purpose of the present study was to evaluate whether this composite low-grade inflammation score would be associated with overall mortality in an adult population cohort with no overt acute inflammation or major hematological diseases. In addition, we also dissected the specific contribution of each component of the score and performed sensitivity analysis, to test whether the association with mortality varied across subgroups at different health risks.

to -1. Being in the deciles 5 or 6 got zero points. In such a way, the INFLA-score ranged from between -16 and 16 and came up as the sum of the four biomarkers. An increase in the score represented an increase in low-grade inflammation intensity. For analysis purposes, quartiles of the INFLA-score were also generated.

Statistical analysis Age and sex adjusted multivariable analysis of variance for continuous or categorical variables was applied to test the associations in Table 1. Potential covariates were included in the multivariable model when they resulted as associated with P<0.20 with both the INFLA-score and mortality. Age and sex adjusted and multivariable hazard ratios with corresponding 95% confidence intervals (95%CI) were calculated using the Cox proportional hazards model, considering subjects in the lowest quartile of the INFLA-score as the reference group. To highlight the specific role of each component within the INFLAscore in relation to mortality, hazard ratios were calculated by removing one component at a time from the original score. The models with and without individual components of the score were compared using Akaike's information criterion (AIC).24 Harrell’s C-index was used to quantify the prediction capacity of the model. Sensitivity analyses were undertaken to estimate possible differences in the magnitude of the association between low-grade inflammation and mortality within subgroups at different health risks. Appropriate interaction terms were added to the Cox regression models to test for a difference of the effect of lowgrade inflammation across subgroups. For subgroup analyses, all variables were dichotomised (no/yes), and missing categories excluded. Tests for a violation of the proportional hazards assumption were conducted through the introduction of linear interaction between categories of the INFLA-score and the time variable (P=0.66). A two-sided P-value of <0.05 was considered as statistically significant.

Methods Study population This study analyzed data from the population-based cohort of 24,325 men and women aged ≥35 years enrolled in the Moli-sani study from March 2005 to April 2010.23 We excluded from the analysis individuals lost to follow-up (1.3%), those reporting unreliable anamnestic questionnaires at baseline (1%), or having hepatitis B or C (2.9%) or any hematological disease (2.2%), or having missing data on platelet count (2.7%), WBC (2.7%), high-sensitivity (hs) CRP (0.12%), granulocytes or lymphocytes (3.2%). Subjects with CRP≥ 10 mg/l (4%) were also excluded to avoid confounding due to an acute inflammatory condition; we also eliminated those included in the percentiles of either highest (1%) or lowest (99%) values for platelet (1.9%) or WBC counts (1.9%). The final sample was of 20,337 (48.2% men) participants. The Moli-sani study was approved by the Ethics Committee of the Catholic University of Rome, Italy. See the Online Supplementary Methods for further information

INFLA-score The low-grade inflammation (INFLA) score had been used previously within the Moli-sani cohort22 and allows one to evaluate the possible synergistic effects of inflammation biomarkers. 10tiles of each biomarker levels (CRP, WBC, platelets, G/L ratio) were generated. For all four components, being in the highest deciles (7 to 10) gave a score which increased from 1 to 4, while being in the lowest deciles (1 to 4) was negatively scored from -4 haematologica | 2016; 101(11)

Results Baseline characteristics of the Moli-sani cohort according to quartiles of the INFLA-score are reported in Table 1. Compared to subjects in the lowest category, those with an increasing INFLA-score were mainly women, had a lower level of education, a higher prevalence of unhealthy lifestyles (smoking, poor leisure-time PA, higher WHratio, lower intake of fruit), and an increased prevalence of CVD, diabetes, hypercholesterolemia and hypertension (P<0.05). During a median follow-up of 7.6 years (interquartile range: 6.7 to 8.6 years; 153,897 person/years), 837 deaths occurred overall. Subjects in the higher quartiles of lowgrade inflammation score had a 44% (23% to 83%) increased risk of all-cause mortality in the multivariable analyses, in comparison with subjects in the lowest quartile (Table 2, Figure 1). Harrell’s C-index for the model was equal to 0.86 (95%CI 0.84-0.86). The INFLA-score was alternately deprived of its components in order to assess the relative contribution of each biomarker. Models deprived of one component at a time have an AIC which is lower than models which have a score that includes all of the components (Table 2). These differences are strong enough to support the hypothesis that the model including the full score is the best, and that each component adds some value to the full score. In par1435

M. Bonaccio et al.

ticular, CRP and G/L ratio contributed most, since their exclusion from the score resulted in a more pronounced AIC decrease (Table 2). Table 3 shows potential predictors of all-cause death in the study population. Major risk factors positively associated with death were age, sex (men), a low level of education, waist to hip ratio, smoking habit, poor leisure-time physical activity, and major chronic diseases at baseline, with the exception of hypercholesterolemia. It is noteworthy that the relative risk associated with the highest category of the INFLA-score is comparable to that of canceror diabetes- related risk. Sensitivity analysis is reported in Table 4. A higher risk linked to an increased INFLA-score was found for subjects with diabetes or CVD as compared to their counterparts (P for interaction = 0.0091 and 0.025, respectively, Figure 1), suggesting a significant interaction between a subclinical chronic inflammation and the presence of major diseases at baseline in relation to the risk of death. As far as all other subgroups are concerned, the strength of the

association of the INFLA-score with mortality did not significantly differ (P>0.05) (Table 4).

Discussion Non-communicable diseases (such as cardiovascular, cancer and neurodegenerative disease) account for 80% of deaths in Europe,25 and each of these conditions is associated with a pro-inflammatory state.26 The concomitant presence of inflammatory biomarkers in diseases with apparently different pathogenesis supports the hypothesis of a common inflammatory soil underlying the pathogenetic mechanisms involved.27,28 Yet, the low-grade inflammation status has not been clearly and uniformly defined, and its association with mortality has principally been explored using a single biomarker approach8,29 rather than considering a panel of combined selected biomarkers.2,7 In the present study, increased low-grade inflammation

Table 1. Baseline characteristics of the study population according to low-grade inflammation.

1st -16 to -5 N of subjects, % Age (years) Sex Women Men High school or higher Waist to hip ratio Current smokers Leisure-time PA (met-h/day) Cardiovascular disease Heart failure Cancer Hypertension Systolic BP Diastolic BP Hypercholesterolemia Total cholesterol (mg/dL) HDL-cholesterol (mg/dL) Triglycerides (mg/dL) Diabetes Fruit intake (gr/day) Vegetables intake(gr/day) Energy intake (kcal/day) C-reactive protein (mg/L)* Leukocyte count (x109/L) Platelet count (x109/L) Granulocyte/Lymphocyte Creatinine (mg/dL)

Quartiles of low-grade inflammation score 2nd 3rd 4th -4 to -1 0 to 3 4 to 16

5054 (24.9) 55.0 (11.4)

5079 (25.0) 55.2 (11.7)

5001 (24.6) 55.6 (11.9)

5203 (25.6) 55.4 (11.9)

2591 (51.3) 2463 (48.7) 2565 (50.8) 0.91 (0.08) 890 (17.6) 3.8 (4.1) 232 (4.6) 18 (0.36) 160 (3.2) 2511 (49.7) 138 (20) 81 (9) 1499 (29.7) 211 (41) 60 (15) 118 (77) 369 (7.3) 362 (210) 160 (73) 1836 (566) 0.72 (0.70-0.73) 4.97 (0.83) 211 (40) 1.51 (0.42) 0.81 (0.17)

2537 (50.0) 2542 (50.0) 2428 (47.8) 0.92 (0.08) 1000 (19.7) 3.6 (4.1) 239 (4.7) 24 (0.47) 142 (2.8) 2714 (53.4) 140 (21) 82 (10) 1614 (31.8) 215 (41) 58 (15) 129 (89) 433 (8.5) 352 (206) 158 (72) 1852 (598) 1.16 (1.13-1.18) 5.71 (0.99) 238 (49) 1.79 (0.55) 0.81 (0.18)

2580 (52.6) 2421 (48.4) 2280 (45.6) 0.92 (0.08) 1181 (23.6) 3.4 (4.0) 276 (5.5) 30 (0.60) 138 (2.8) 2956 (59.1) 142 (20) 83 (9) 1616 (32.3) 215 (42) 57 (14) 134 (87) 500 (10.0) 346 (199) 158 (71) 1858 (598) 1.62 (1.58-1.65) 6.39 (1.11) 257 (51) 2.06 (1.09) 0.81 (0.19)

2830 (54.4) 2373 (45.6) 2257 (43.4) 0.93 (0.08) 1558 (29.9) 3.2 (3.9) 289 (5.6) 35 (0.67) 140 (2.7) 3128 (60.1) 142 (21) 83 (10) 1761 (33.9) 217 (42) 55 (14) 143 (89) 595 (11.4) 339 (203) 157 (73) 1854 (599) 2.75 (2.69-2.80) 7.50 (1.32) 285 (53) 2.52 (0.90) 0.82 (0.22)

P 0.084 0.0003

<0.0001 <0.0001 <0.0001 <0.0001 0.012 0.067 0.09 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <.0001 0.09 0.19 <0.0001 <0.0001 <0.0001 <0.0001 0.31

Numbers are presented as meansÂąSD (continuous variables: age, waist to hip ratio, leisure-time PA, systolic and diastolic blood pressure, total cholesterol, HDL-cholesterol, triglycerides, fruit and vegetable intake, energy intake, inflammatory biomarkers) or numbers and percentage (n, % for categorical variables). Means and P values are adjusted for age and sex. *Values for CRP are reported as geometric means with corresponding 95% confidence intervals. **Numbers do not add up to 100% due to missing values. PA: physical activity; BP: blood pressure; HDL: high-density lipoproteins.

1436

haematologica | 2016; 101(11)

Low-grade inflammation and mortality

as measured by a score including plasmatic and cellular biomarkers, was found to be associated with a 44% increased risk of all-cause mortality in a large population cohort, in the absence of an acute inflammation status or major hematological diseases. The INFLA-score tested in our population included four markers of inflammation that have been previously reported in association with an increased risk of adverse health outcomes. To the best of our knowledge, this study is the first to have explored the risk of total mortality in a general population in association with a clinically silent low-grade inflammation condition measured by a composite score. In the conceptual framework of our analyses, low-grade inflammation was conceived as a subclinical condition (systemic or local, often chronic), characterized by increased levels of plasmatic and/or cellular biomarkers of inflammation without any apparent clinical sign. Of note, the population under study was selected to exclude from the analyses all subjects with acute inflammation or any kind of hematological diseases that may lead to an overestimation of the risk; therefore our findings have to be interpreted as related to variations within the ranges of normality for each biomarker of inflammation.

This approach has been previously proposed and tested within the Moli-sani cohort in relation to dietary polyphenol intake.22 As remarked previously, the advantages of using a composite assessment of low-grade inflammation is to summarize the variability of inflammation as a plasmatic and cellular phenomenon at an epidemiological scale. In addition, the use of a score allowed for the evaluation of a possible synergistic effect of inflammation biomarkers that are usually autocorrelated, and may thus produce multi-collinearity when simultaneously studied in a regression model. We also addressed the question of the specific contribution of each biomarker, in order to exclude a training role of one biomarker over others: both CRP and G/L ratio contributed most to the association of the INLFLA-score with mortality. Nonetheless, the contribution of both leukocyte and platelet counts were not negligible, thus justifying their inclusion in the score. The excess of risks associated with the top quartile of the INFLA-score was comparable to that of common risk factors, thus reinforcing the need to take into account a low-grade inflammation condition when dealing with health outcomes. The nature of the association between inflammatory

Table 2. Risk of all-cause mortality associated with quartiles of low-grade inflammation and after alternate subtraction of each component.

N of deaths/ N of subjects Hazard ratio (95%CI) Age/sex adjusted Multivariable model* INFLA-score minus CRP* INFLA-score minus WBC* INFLA-score minus Platelet* INFLA-score minus G/L ratio*

Quartiles of low-grade inflammation 3rd 4th

1st

2nd

148/5054

180/5079

247/5001

262/5203

1.00 (ref) 1.00 (ref) 1.00 (ref) 1.00 (ref) 1.00 (ref) 1.00 (ref)

1.16 (0.93-1.44) 1.11 (0.90-1.39) 1.10 (0.90-1.36) 1.09 (0.87-1.35) 1.16 (0.92-1.47) 1.18 (0.96-1.45)

1.55 (1.26-1.89) 1.44 (1.17-1.77) 1.30 (1.07-1.58) 1.30 (1.06-1.60) 1.28 (1.04-1.59) 1.30 (1.07-1.57)

1.65 (1.35-2.02) 1.44 (1.17-1.77) 1.39 (1.14-1.69) 1.43 (1.17-1.74) 1.52 (1.23-1.87) 1.41 (1.15-1.73)

AIC

AIC variation

14457.092 14462.682 14460.187 14459.344 14463.778

5.6 3.1 2.3 6.7

*Hazard ratios from the multivariable model adjusted for age, sex, cardiovascular disease, heart failure, cancer, diabetes, hypercholesterolemia, systolic and diastolic BP, leisure-time PA, waist to hip ratio, fruit and vegetables intake, energy intake, smoking, education. AIC: Akaikeâ&#x20AC;&#x2122;s information criterion. AIC variation (multivariable model including INFLA score) â&#x20AC;&#x201C; AIC (multivariable model including INFLA score minus each component at a time). BP: blood pressure; PA: physical activity; CI: confidence intervals; CRP: C-reactive protein; WBC: white blood cell count; G/L ratio: granulocyte to lymphocyte ratio.

Figure 1. Relative risk of all-cause death for each quartile of low-grade inflammation score according to the presence of diabetes at baseline. The reference group is the lowest quartile for each subgroup. Hazard ratios are adjusted for age, sex, cardiovascular disease, heart failure, cancer, hypercholesterolemia, systolic and diastolic BP, leisure-time PA, waist to hip ratio, fruit and vegetables intake, energy intake, smoking, education. Vertical bars indicate 95% confidence intervals. PA: physical activity; BP: blood pressure.

haematologica | 2016; 101(11)

1437

M. Bonaccio et al. markers and health outcomes is still a matter of debate.29 Regarding CVD onset, some studies support the hypothesis that the inflammatory markers, or the underlying inflammatory processes they represent, are directly involved in the development of atherosclerosis and plaque rupture,30,31 and more recently, two large-scale trials have been testing whether targeted inhibition of inflammation may reduce cardiovascular event rates by the use of targeted anti-inflammatory agents for the secondary prevention of myocardial infarction.32 On the other hand, others claim that rather than being independent risk factors, inflammation biomarkers confound the ultimate association with adverse CVD outcomes, due to the strict association of inflammatory markers with many cardiovascular risk factors, such as smoking

or hypertension.33 However, our findings were controlled for a wide panel of possible confounders, including dietary factors and health-related behaviors, that should support the independence of low-grade inflammation in predicting overall mortality. Sensitivity analyses revealed that the magnitude of the association of low-grade inflammation with mortality was higher in high-risk subjects, such as those with type 2 diabetes or a history of cardiovascular disease. To the best of our knowledge, few studies have investigated the likely heterogeneous effect of low-grade inflammation across groups at different health risks. Sanchez et al.34 found that inflammation appears more evident in diabetic than in non-diabetic patients with acute coronary syndrome, and that inflammatory markers constitute independent predic-

Table 3. Risk factors associated with all-cause mortality in the study population.

Low-grade inflammation (quartiles) 1st 2nd 3rd 4th Age (for 1 year increment) Sex Women Men Education Up to middle school High school or higher Waist to hip ratio Smoking No Yes Former Cardiovascular disease No Yes Not ascertained Heart failure No Yes Not ascertained Cancer No Yes Not ascertained Leisure-time PA Systolic BP Diastolic BP Hypercholesterolemia No Pre- hypercholesterolemia Yes Diabetes No Prediabetes Yes Fruit intake Vegetables intake Energy intake

N of deaths/ N of subjects

Incidence Rate (%)

Risk of all-cause death HR (95% CI)

148/5054 180/5079 247/5001 262/5203 -

2.9 3.5 4.9 5.0 -

-11.11 (0.90-1.39) 1.44 (1.17-1.77) 1.44 (1.17-1.77) 1.110 (1.101-1.119)

276/10538 561/9799

2.6 5.7

-11.82 (1.51-2.18)

609/10779 225/9530 -

5.7 2.4 -

-10.79 (0.67-0.93) 1.999 (0.725-5.511)

331/10099 174/4629 331/5586

3.3 3.8 5.9

-11.93 (1.57-2.36) 1.19 (1.00-1.43)

639/18988 171/1036 8/77

3.4 16.5 10.4

-11.78 (1.49-2.13) 1.01 (0.50-2.04)

798/20152 29/107 10/78

4.0 27.1 12.8

-12.00 (1.37-2.94) 0.97 (0.51-1.84)

762/19659 61/580 2/4 -

3.9 10.5 50.0 -

-11.56 (1.20-2.03) 6.89 (1.70-27.90) 0.974 (0.957-0.992) 1.003 (0.998-1.007) 1.006 (0.998-1.015)

313/6648 229/6931 275/6490

4.7 3.3 4.2

-10.79 (0.67-0.94) 0.78 (0.66-0.92)

504/15887 120/2366 200/1897 -

3.2 5.1 10.5 -

-11.08 (0.88-1.32) 1.53 (1.28-1.82) 1.000 (0.999-1.000) 0.999 (0.998-1.000) 1.000 (1.000-1.000)

Hazard ratios from the model including all the listed variables. PA: physical activity; BP: blood pressure; HR: hazard ratio; CI: confidence intervals.

1438

haematologica | 2016; 101(11)

Low-grade inflammation and mortality

tors of cardiovascular death in diabetics with unstable coronary disease. Results from the Hoorn study suggest that endothelial dysfunction and low-grade inflammation were associated with greater risks of cardiovascular mortality, especially in those with diabetes.35 Increased levels of inflammatory markers have been reported in patients with CVD, and it has been shown that the risk of death associated with high levels of IL-6 is dependent on the patients history of CVD, whereas for those without CVD the risk completely disappeared after adjustment for potential confounders.36 The strong relationship of low-grade inflammation with mortality in high-risk individuals is attributable to a procoagulant effect associated with inflammation.36 In addition, pro-inflammatory cytokines may be involved in promoting atherosclerotic plaque development and its disruption, an occurrence that is particularly plausible in CVD or diabetic patients.36-38

Strengths and limitations of this study The major strengths of this study include a large community-based cohort, a prospective design and a quite long follow-up period. In addition, further controls of all analyses by a wide panel of possible confounding factors, including health-related behaviors, should assure consistency with the observed associations. A major limitation of the present study is the lack of information on specific causes of death that prevents us from exploring the association of low-grade inflammation with different health outcomes. Since information about the participants were collected at baseline only, life course changes which possibly occurred during the follow-up may influence the strength of the findings. We acknowledge the lack of other biomarkers of inflammation, for example fibrinogen or interleukin-6,14,16 that were not included in the INFLA-score because of the unavailability of the data.

Table 4. Sensitivity analyses for the association of low-grade inflammation and all-cause mortality.

Quartiles of low-grade inflammation

Whole sample Sex Women Men Age â&#x2030;¤65 >65 Education Low High Smoking habit No Yes Waist to hip ratio Normal At risk Leisure-time PA Below median Above median Diabetes No Yes Hypercholesterolemia No Yes Cardiovascular disease No Yes Cancer No Yes Fruit intake Below median Above median Vegetables intake Below median Above median

N of deaths/ N of subjects

1st

2nd

3rd

4th

P value for interaction

837/20337

1.00 (ref)

1.11 (0.90-1.39)

1.44 (1.17-1.77)

276/10538 561/9799

1.00 (ref) 1.00 (ref)

0.83 (0.56-1.23) 1.25 (0.96-1.64)

1.28 (0.91-1.82) 1.49 (1.15-1.92)

1.19 (0.84-1.68) 1.54 (1.19-2.00)

0.27

239/15757 598/4580

1.00 (ref) 1.00 (ref)

1.19 (0.81-1.76) 1.05 (0.80-1.37)

1.23 (0.84-1.81) 1.49 (1.16 1.90)

1.18 (0.80- 1.74) 1.50 (1.17-1.92)

0.44

609/10779 225/9530

1.00 (ref) 1.00 (ref)

0.98 (0.75-1.26) 1.66 (1.09-2.52)

1.28 (1.01 -1.63) 1.90 (1.27 -2.87)

1.31 (1.03 -1.66) 1.87 (1.24 -2.81)

0.14

662/15685 174/4629

1.00 (ref) 1.00 (ref)

1.05 (0.82-1.34) 1.38 (0.82-2.32)

1.46 (1.17-1.84) 1.28 (0.77-2.11)

1.54 (1.23-1.94) 1.08 (0.66-1.78)

0.11

107/5440 724/14897

1.00 (ref) 1.00 (ref)

0.91 (0.50-1.65) 1.13 (0.89-1.43)

1.70 (0.99 -2.92) 1.40 (1.12-1.75)

1.47 (0.82 -2.63) 1.41 (1.13 -1.76)

0.77

462/10071 3753/10266

1.00 (ref) 1.00 (ref)

1.14 (0.83-1.56) 1.18 (0.82-1.52)

1.69 (1.27-2.26) 1.20 (0.89 -1.63)

1.58 (1.18-2.10) 1.35 (1.00 -1.82)

0.31

624/18253 200/1897

1.00 (ref) 1.00 (ref)

0.94 (0.74-1.21) 2.32 (1.35-3.97)

1.27 (1.01-1.59) 2.50 (1.47-4.23)

1.24 (0.99-1.57) 2.90 (1.74-4.84)

0.0091

542/13579 275/6490

1.00 (ref) 1.00 (ref)

1.15 (0.88-1.51) 1.11 (0.75-1.65)

1.46 (1.13-1.88) 1.37 (0.94-2.00)

1.35 (1.05-1.75) 1.58 (1.09-2.27)

0.56

639/18988 171/1036

1.00 (ref) 1.00 (ref)

1.05 (0.82- 1.34) 1.57 (0.91- 2.71)

1.44 (1.14-1.81) 1.80 (1.07-3.06)

1.24 (0.98-1.57) 2.48 (1.50-4.11)

0.025

762/19659 61/580

1.00 (ref) 1.00 (ref)

1.11 (0.88-1.39) 0.85 (0.35-2.10)

1.41 (1.14-1.75) 1.76 (0.82-3.77)

1.37 (1.11-1.71) 2.13 (0.99-4.59)

0.35

459/10135 378/10134

1.00 (ref) 1.00 (ref)

1.06 (0.79-1.42) 1.16 (0.83-1.62)

1.20 (0.90-1.59) 1.75 (1.29-2.38)

1.32 (1.00-1.74) 1.62 (1.18-2.21)

0.24

521/10129 316/10208

1.00 (ref) 1.00 (ref)

1.09 (0.82-1.45) 1.15 (0.81-1.64)

1.38 (1.06-1.81) 1.52 (1.09-2.11)

1.51 (1.16-1.96) 1.32 (0.94-1.85)

0.49

Hazard ratios from the multivariable model adjusted for age, sex, cardiovascular disease, heart failure, cancer, diabetes, hypercholesterolemia, systolic and diastolic BP, leisure-time PA, waist to hip ratio, fruit and vegetables intake, energy intake, smoking, education. PA: physical activity; BP: blood pressure.

haematologica | 2016; 101(11)

1439

M. Bonaccio et al.

Conclusions A high degree of low-grade inflammation, as measured by a composite score accounting for possible synergistic effects of selected inflammation biomarkers, is an independent risk factor for total mortality in a general population with no acute inflammation at baseline. Despite being associated with all major health risk conditions, a low-grade inflammation retains a significant independent health predictive value per se. The strength of the observed association was higher in diabetic patients and in individuals with a history of CVD. Acknowledgments The Moli-sani research group thanks the Associazione Cuore Sano Onlus (Campobasso, Italy) for its financial support and the Azienda Sanitaria Regionale del Molise (ASReM, Campobasso, Italy), the Offices of vital statistics of the Molise region and the Molise Dati Spa (Campobasso, Italy) for their collaboration and support provided during the follow-up activities. Funding The enrolment phase of the Moli-sani study was supported by research grants from Pfizer Foundation (Rome, Italy), the Italian Ministry of University and Research (MIUR, Rome, Italy)– Programma Triennale di Ricerca, Decreto no.1588 and Instrumentation Laboratory, Milan, Italy. Marialaura Bonaccio was supported by a Fondazione Umberto Veronesi Fellowship. The present analyses were partially supported by the Italian Ministry of Health 2013 [Grant number GR-2013-02356060] and by the Italian Association for Cancer Research (A.I.R.C.) with grant AIRC"5x1000" Ref. n. 12237. Funders had no role in study design, collection, analysis, and interpretation of data; in the writing of the manuscript and in the decision to submit the article for publication. All authors were and are independent from funders. *Appendix 1 Moli-sani Study Investigators The enrolment phase of the Moli-sani study was conducted at the Research Laboratories of the Catholic University in Campobasso (Italy), the follow up of the Mol-sani cohort is being conducted at the IRCCS Neuromed, Pozzilli, Italy. Steering Committee: Licia Iacoviello (Neuromed, Pozzilli, Italy), Chairperson, Maria Benedetta Donati and Giovanni de Gaetano (Neuromed, Pozzilli, Italy). Safety and data monitoring Committee: Jos Vermylen (Catholic University, Leuven, Belgium), Chairman, Ignacio De Paula Carrasco (Accademia Pontificia Pro Vita, Roma, Italy), Simona Giampaoli (Istituto Superiore di Sanità, Roma, Italy), Antonio Spagnuolo (Catholic

References 1. Barbaresko J, Koch M, Schulze MB, Nöthlings U. Dietary pattern analysis and biomarkers of low-grade inflammation: a systematic literature review. Nutr Rev. 2013;71 (8):511-527. 2. Danesh J, Whincup P, Walker M, et al. Low grade inflammation and coronary heart disease: prospective study and updated metaanalyses. BMJ. 2000;321(7255):199-204. 3. Engstrom G, Hedblad B, Stavenow L, et al. Inflammation-Sensitive Plasma Proteins and

1440

4. 5. 6. 7.

University, Roma, Italy). Event adjudication Committee: Deodato Assanelli (Brescia, Italy), Vincenzo Centritto (Campobasso, Italy), Pasquale Spagnuolo and Dante Staniscia (Termoli, Italy). Scientific and organizing secretariat: Francesco Zito (Coordinator), Americo Bonanni, Chiara Cerletti, Amalia De Curtis, Augusto Di Castelnuovo, Licia Iacoviello, Roberto Lorenzet, Antonio Mascioli, Marco Olivieri and Domenico Rotilio. Data management and analysis: Augusto Di Castelnuovo (Coordinator), Marialaura Bonaccio, Simona Costanzo and Francesco Gianfagna. Informatics: Marco Olivieri (Coordinator), Maurizio Giacci, Antonella Padulo and Dario Petraroia. Biobank and biomedical analyses: Amalia De Curtis (Coordinator), Federico Marracino, Maria Spinelli, Christian Silvestri. Communication and Press Office: Americo Bonanni (Coordinator), Marialaura Bonaccio and Francesca De Lucia. Moli-family Project: Francesco Gianfagna, Branislav Vohnout. Recruitment staff: Franco Zito (General Coordinator), Secretariat: Mariarosaria Persichillo (Coordinator), Angelita Verna, Maura Di Lillo, Irene Di Stefano, Blood sample: Agostino Pannichella, Antonio Rinaldo Vizzarri, Branislav Vohnout, Agnieszka Pampuch; Spirometry: Antonella Arcari (Coordinator), Daniela Barbato, Francesca Bracone, Simona Costanzo, Carmine Di Giorgio, Sara Magnacca, Simona Panebianco, Antonello Chiovitti, Federico Marracino, Sergio Caccamo, Vanesa Caruso; Electrocardiograms: Livia Rago (Coordinator), Daniela Cugino, Francesco Zito, Francesco Gianfagna, Alessandra Ferri, Concetta Castaldi, Marcella Mignogna; Tomasz Guszcz; Questionnaires: Romina di Giuseppe (Coordinator), Paola Barisciano, Lorena Buonaccorsi, Floriana Centritto, Antonella Cutrone, Francesca De Lucia, Francesca Fanelli, Iolanda Santimone, Anna Sciarretta, Maura Di Lillo, Isabella Sorella, Irene Di Stefano, Emanuela Plescia, Alessandra Molinaro and Christiana Cavone. Call Center: Giovanna Galuppo, Maura Di Lillo, Concetta Castaldi, Dolores D'Angelo and Rosanna Ramacciato. Follow-up: Simona Costanzo (Coordinator); Data management: Simona Costanzo, Marco Olivieri; Event adjudication: Livia Rago (Coordinator), Simona Costanzo, Amalia De Curtis, Licia Iacoviello, Mariarosaria Persichillo. Regional Health Institutions: Azienda Sanitaria Regionale del Molise (ASReM, Campobasso, Italy), UOC Servizio Igiene e Sanità Pubblica Dipartimento di Prevenzione; Offices of vital statistics of the Molise region and Molise Dati Spa (Campobasso, Italy). Hospitals: Presidi Ospedalieri ASReM (Presidio Ospedaliero A. Cardarelli - Campobasso, Ospedale F. Veneziale - Isernia, Ospedale San Timoteo - Termoli (CB), Ospedale Ss. Rosario. Venafro (IS), Ospedale Vietri – Larino (CB), Ospedale San Francesco Caracciolo - Agnone (IS), Istituto di cura Villa Maria - Campobasso; Fondazione di Ricerca e Cura Giovanni Paolo II - Campobasso; IRRCS Neuromed - Pozzilli (IS), Italy.

Incidence of Myocardial Infarction in Men With Low Cardiovascular Risk. Arterioscler Thromb Vasc Biol. 2004;24(8):1498-1502. Lind L. Circulating markers of inflammation and atherosclerosis. Atherosclerosis. 2003;169(2):203-214. Coussens LM, Werb Z. Inflammation and cancer. Nature. 2002;420(6917):860-867. Samuels, Martin A. Inflammation and neurological disease. Curr Opin Neurol. 2004;17(3):307-309. Schnabel RB, Yin X, Larson MG, et al. Multiple inflammatory biomarkers in relation to cardiovascular events and mortality

in the community. Arterioscler Thromb Vasc Biol. 2013;33(7):1728-1733. 8. Mendall MA, Strachan DP, Butland BK, et al. C-reactive protein: relation to total mortality, cardiovascular mortality and cardiovascular risk factors in men. Eur Heart J. 2000;21(19):1584-1590. 9. Matsumoto K, Sera Y, Abe Y, Ueki Y, Tominaga T, Miyake S. Inflammation and insulin resistance are independently related to all-cause of death and cardiovascular events in Japanese patients with type 2 diabetes mellitus. Atherosclerosis. 2003;169 (2):317-321.

haematologica | 2016; 101(11)

Low-grade inflammation and mortality

10. Pierce BL, Ballard-Barbash R, Bernstein L, et al. Elevated biomarkers of inflammation are associated with reduced survival among breast cancer patients. J Clin Oncol. 2009;27(21):3437-3444. 11. Rathcke CN, Raymond I, Kistorp C, Hildebrandt P, Faber J, Vestergaard H. Low grade inflammation as measured by levels of YKL-40: association with an increased overall and cardiovascular mortality rate in an elderly population. Int J Cardiol. 2010;143 (1):35-42. 12. Candore G, Caruso C, Jirillo E, Magrone T, Vasto S. Low grade inflammation as a common pathogenetic denominator in age-related diseases: novel drug targets for anti-ageing strategies and successful ageing achievement. Curr Pharm Des. 2010;16 (6):584-596. 13. Rienstra M, Sun JX, Magnani JW, et al. White blood cell count and risk of incident atrial fibrillation (from the Framingham Heart Study). Am J Cardiol. 2012;109 (4):533-537. 14. Danesh J, Collins R, Appleby P, Peto R. Association of fibrinogen, C-reactive protein, albumin, or leukocyte count with coronary heart disease: meta-analyses of prospective studies. JAMA. 1998;279(18): 1477-1482. 15. Danesh J, Wheeler JG, Hirschfield GM, et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med. 2004;35 (14):1387-1397. 16. Arcari A, Zito F, Di Castelnuovo A, et al. C reactive protein and its determinants in healthy men and women from European regions at different risk of coronary disease: the IMMIDIET Project. J Thromb Haemost. 2008;6(3):436-443. 17. Bonaccio M, Di Castelnuovo A, De Curtis A, et al. Adherence to the Mediterranean diet is associated with lower platelet and leukocyte counts: results from the MOLI-SANI study. Blood. 2014;123(19):3037-3044. 18. Bhat T, Teli S, Rijal J, et al. Neutrophil to lymphocyte ratio and cardiovascular dis-

haematologica | 2016; 101(11)

19.

20.

21.

22.

23.

24. 25.

26. 27. 28.

29.

eases: a review. Expert Rev Cardiovasc Ther. 2013;11(1):55-59. Bonaccio M, Di Castelnuovo A, Rago L, et al. T-wave axis deviation is associated with biomarkers of low-grade inflammation. Findings from the MOLI-SANI study. Thromb Haemost. 2015;114(6):1199-1206. Yudkin JS, Juhan-Vague I, Hawe E, et al. Low-grade inflammation may play a role in the etiology of the metabolic syndrome in patients with coronary heart disease: the HIFMECH study. Metabolism. 2004;53(7): 852-857. van Bussel BC, Henry RM, Ferreira I, et al. A healthy diet is associated with less endothelial dysfunction and less low-grade inflammation over a 7-year period in adults at risk of cardiovascular disease. J Nutr. 2015;145(3):532-540. Pounis G, Bonaccio M, Di Castelnuovo A, et al. Polyphenol intake is associated with lowgrade inflammation, using a novel data analysis from the Moli-sani study. Thromb Haemost. 2015;115(2):344-352. Di Castelnuovo A, Costanzo S, Persichillo M, et al. Distribution of short and lifetime risks for cardiovascular disease in Italians. Eur J Prev Cardiol. 2012;19(4):723-730. Murtaugh PA. In defense of P values. Ecology. 2014;95(3):611-617. http://www.euro. who.int/__data/ assets/pdf_file/0004/185215/Leading-causes-of-death-in-Europe-Fact-Sheet.pdf. Accessed on December 2015 Ferrucci L, Corsi A, Lauretani F, et al. The origins of age-related proinflammatory state. Blood. 2005;105(6):2294-2299. Donati MB. The "common soil hypothesis": evidence from population studies? Thromb Res. 2010;125 Suppl 2:S92-95. Bonaccio M, Cerletti C, Iacoviello L, de Gaetano G. Mediterranean diet and lowgrade subclinical inflammation: the Molisani study. Endocr Metab Immune Disord Drug Targets. 2015;15(1):18-24. Engström G, Hedblad B, Rosvall M, Janzon L, Lindgärde F. Occupation, marital status,

30. 31. 32.

33.

34.

35.

36.

37.

38.

and low-grade inflammation: mutual confounding or independent cardiovascular risk factors? Arterioscler Thromb Vasc Biol. 2006;26(3):643-648. Ross R. Atherosclerosis-An inflammatory disease. N Engl J Med. 1999; 340(2):115–126. Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med. 2005;352(16):1685-1695. Ridker PM. Closing the loop on inflammation and atherothrombosis: why perform the CIRT and CANTOS trials? Trans Am Clin Climatol Assoc. 2013;124:174-190. Lawlor DA, Smith GD, Rumley A, Lowe GD, Ebrahim S. Associations of fibrinogen and C-reactive protein with prevalent and incident coronary heart disease are attenuated by adjustment for confounding factors. British Women’s Heart and Health Study. Thromb Haemost. 2005;93(5):955-963. Sanchez PL, Morinigo JL, Pabon P, et al. Prognostic relations between inflammatory markers and mortality in diabetic patients with non-ST elevation acute coronary syndrome. Heart. 2004;90(3):264-269. de Jager J, Dekker JM, Kooy A, et al. Endothelial Dysfunction and Low-Grade Inflammation Explain Much of the Excess Cardiovascular Mortality in Individuals With Type 2 Diabetes. Arterioscler Thromb Vasc Biol. 2006;26(5):1086-1093. Volpato S, Guralnik JM, Ferrucci L, et al. Cardiovascular disease, interleukin-6, and risk of mortality in older women: the women's health and aging study. Circulation. 2001;103(7):947-953. Badimon L, Padró T, Vilahur G. Atherosclerosis, platelets and thrombosis in acute ischaemic heart disease. Eur Heart J Acute Cardiovasc Care. 2012;1(1):60-74. van der Wal AC, Becker AE, van der Loos CM, Das PK. Site of intimal rupture or erosion of thrombosed coronary atherosclerotic plaques is characterized by an inflammatory process irrespective of the dominant plaque morphology. Circulation. 1994;89 (1):36-44.

1441

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

The origin of a name that reflects Europe’s cultural roots.

Ancient Greek

aÂma [haima] = blood a·matow [haimatos] = of blood lÒgow [logos]= reasoning

Scientific Latin

haematologicus (adjective) = related to blood

Scientific Latin

haematologica (adjective, plural and neuter, used as a noun) = hematological subjects

Modern English

The oldest hematology journal, publishing the newest research results. 2015 JCR impact factor = 6.671

Haematologica, as the journal of the European Hematology Association (EHA), aims not only to serve the scientific community, but also to promote European cultural identify.