Haematologica, volume 101, issue 12

Periodico – Sped. Abb. Post. – 45% art. 2, comma 20B, Legge 662/96 - Filiale di Pavia. Il mittente chiede la restituzione dei fascicoli non consegnati impegnandosi a pagare le tasse dovute

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

ISSN 0390-6078

Volume 101 1

DECEMBER

2016 12 2016|

www.haematologica.org

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

Editor-in-Chief Jan Cools (Leuven)

Deputy Editor Luca Malcovati (Pavia)

Managing Director Antonio Majocchi (Pavia)

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

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

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

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

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

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

Information for readers, authors and subscribers Haematologica (print edition, pISSN 0390-6078, eISSN 1592-8721) publishes peer-reviewed papers on all areas of experimental and clinical hematology. The journal is owned by a non-profit organization, the Ferrata Storti Foundation, and serves the scientific community following the recommendations of the World Association of Medical Editors (www.wame.org) and the International Committee of Medical Journal Editors (www.icmje.org).

Haematologica publishes editorials, research articles, review articles, guideline articles and letters. Manuscripts should be prepared according to our guidelines (www.haematologica.org/information-for-authors), and the Uniform Requirements for Manuscripts Submitted to Biomedical Journals, prepared by the International Committee of Medical Journal Editors (www.icmje.org).

Manuscripts should be submitted online at http://www.haematologica.org/.

Conflict of interests. According to the International Committee of Medical Journal Editors (http://www.icmje.org/#conflicts), “Public trust in the peer review process and the credibility of published articles depend in part on how well conflict of interest is handled during writing, peer review, and editorial decision making”. The ad hoc journal’s policy is reported in detail online (www.haematologica.org/content/policies).

Transfer of Copyright and Permission to Reproduce Parts of Published Papers. Authors will grant copyright of their articles to the Ferrata Storti Foundation. No formal permission will be required to reproduce parts (tables or illustrations) of published papers, provided the source is quoted appropriately and reproduction has no commercial intent. Reproductions with commercial intent will require written permission and payment of royalties.

Detailed information about subscriptions is available online at www.haematologica.org. Haematologica is an open access journal. Access to the online journal is free. Use of the Haematologica App (available on the App Store and on Google Play) is free.

For subscriptions to the printed issue of the journal, please contact: Haematologica Office, via Giuseppe Belli 4, 27100 Pavia, Italy (phone +39.0382.27129, fax +39.0382.394705, E-mail: info@haematologica.org).

Rates of the International edition for the year 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 November 2016.

haematologica calendar of events

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

The American Society of Hematology 58th ASH Annual Meeting and Exposition American Society of Hematology (ASH) Chairs: CS Abrams, J Di Paola, S Luger, R Brodsky, R Levine December 3-6, 2016 San Diego, US

EHA Scientific Meeting on Anemias 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

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 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 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

Calendar of Events updated on November 2, 2016

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

Table of Contents Volume 101, Issue 12: December 2016 Cover Figure Multiple Myeloma cells - image accompanying the review article at page 1451. (Image created by www.somersault1824.com)

Editorials 1443

Eltrombopag, a potent stimulator of megakaryopoiesis Hana Raslova et al.

1446

The Hippo-p53 pathway in megakaryopoiesis Praveen K Suraneni and John D. Crispino

1448

Ibrutinib in the real world patient: many lights and some shades Paolo Ghia and Antonio Cuneo

Review Articles 1451

The myeloma stem cell concept, revisited: from phenomenology to operational terms Hans Erik Johnsen et al.

1460

‘Trained immunity’: consequences for lymphoid malignancies Wendy B.C. Stevens et al.

Articles Hematopoiesis

1469

Uncoupling of the Hippo and Rho pathways allows megakaryocytes to escape the tetraploid checkpoint Anita Roy et al.

1479

Revealing eltrombopag’s promotion of human megakaryopoiesis through AKT/ERK-dependent pathway activation Christian A. Di Buduo et al.

Red Cell Biology & its Disorders

1489

Unexpected macrophage-independent dyserythropoiesis in Gaucher disease Nelly Reihani et al.

Iron Metabolism & its Disorders

1499

Wnt5a is a key target for the pro-osteogenic effects of iron chelation on osteoblast progenitors Ulrike Baschant et al.

Myelodysplastic Syndromes

1508

Response to treatment with azacitidine in children with advanced myelodysplastic syndrome prior to hematopoietic stem cell transplantation Nicolas Waespe et al.

Haematologica 2016; vol. 101 no. 12 - December 2016 http://www.haematologica.org/

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

1516

Chromosome abnormalities at onset of complete remission are associated with worse outcome in patients with acute myeloid leukemia and an abnormal karyotype at diagnosis: CALGB 8461 (Alliance) Christian Niederwieser et al.

Acute Lymphoblastic Leukemia

1524

International reference analysis of outcomes in adults with B-precursor Ph-negative relapsed/refractory acute lymphoblastic leukemia Nicola Gรถkbuget et al.

1534

CLIC5: a novel ETV6 target gene in childhood acute lymphoblastic leukemia Benjamin Neveu et al.

1544

A sequential approach with imatinib, chemotherapy and transplant for adult Ph+ acute lymphoblastic leukemia: final results of the GIMEMA LAL 0904 study Sabina Chiaretti et al.

Chronic Lymphocytic Leukemia

1553

SLP76 integrates into the B-cell receptor signaling cascade in chronic lymphocytic leukemia cells and is associated with an aggressive disease course Nili Dezorella et al.

1563

Ibrutinib for relapsed/refractory chronic lymphocytic leukemia: a UK and Ireland analysis of outcomes in 315 patients UK CLL Forum

1573

Real-world results of ibrutinib in patients with relapsed or refractory chronic lymphocytic leukemia: data from 95 consecutive patients treated in a compassionate use program. A study from the Swedish Chronic Lymphocytic Leukemia Group Maria Winqvist e tal.

Non-Hodgkin Lymphoma

1581

Non-Hodgkin lymphoma and pre-existing conditions: spectrum, clinical characteristics and outcome in 213 children and adolescents Andishe Attarbaschi et al.

Stem Cell Transplantation

1592

Comparison of graft-versus-host disease-free, relapse-free survival according to a variety of graft sources: antithymocyte globulin and single cord blood provide favorable outcomes in some subgroups Yoshihiro Inamoto et al.

Complications in Hematology

1603

Metabolic syndrome in long-term survivors of childhood acute leukemia treated without hematopoietic stem cell transplantation: an L.E.A. study Paul Saultier et al.

Letters to the Editor Letters are available online only at www.haematologica.org/content/101/12.toc

e469

Nocturnal enuresis and K+ transport in red blood cells from patients with sickle cell anemia Sanjay Tewari et al. http://www.haematologica.org/content/101/12/e469

Haematologica 2016; vol. 101 no. 12 - December 2016 http://www.haematologica.org/

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

e473

D-dimer measured at first venous thromboembolism is associated with future risk of cancer Olga V. Gran et al. http://www.haematologica.org/content/101/12/e473

e476

Remissions after long-term use of romiplostim for immune thrombocytopenia Ariela L. Marshall et al. http://www.haematologica.org/content/101/12/e476

e479

Lack of splice factor and cohesin complex mutations in pediatric myelodysplastic syndrome Julia C. Obenauer et al. http://www.haematologica.org/content/101/12/e479

e482

The impact of anemia on overall survival in patients with myelofibrosis treated with ruxolitinib in the COMFORT studies Vikas Gupta et al. http://www.haematologica.org/content/101/12/e482

e485

Peripheral neuropathy associated with subcutaneous or intravenous bortezomib in patients with newly diagnosed myeloma treated within the GMMG MM5 phase III trial Maximilian Merz et al. http://www.haematologica.org/content/101/12/e485

e488

Increased risk of axial fractures in patients with untreated chronic lymphocytic leukemia: a population-based analysis Adam J. Olszewski et al. http://www.haematologica.org/content/101/12/e488

Comment Letters are available online only at www.haematologica.org/content/101/12.toc

e492

Is ruxolitinib a potentially useful drug in hematological malignancies with RAS pathway hyperactivation?

Klaus Geissler et al. http://www.haematologica.org/content/101/12/e492

Haematologica 2016; vol. 101 no. 12 - December 2016 http://www.haematologica.org/

EDITORIALS Eltrombopag, a potent stimulator of megakaryopoiesis Hana Raslova,1 William Vainchenker1 and Isabelle Plo1 INSERM, UMR1170, Gustave Roussy, Villejuif, France E-mail: isabelle.plo@gustaveroussy.fr

doi:10.3324/haematol.2016.153668

I

n this issue of Haematologica, Di Buduo et al. show that eltrombopag induces human megakaryopoiesis in vitro and ex vivo through an activation of STAT, AKT and ERK pathways that is different from the other thrombomimetic, romiplostim.1 Megakaryopoiesis is the process leading to platelet production in the blood from the differentiation of bone marrow progenitors to platelet precursors called megakaryocytes (MKs).2 This phenomenon involves the commitment of a multipotent hematopoietic stem cell (HSC) towards a MK progenitor. This progenitor undergoes several divisions by classical mitosis, and further maturation involves two unique biological mechanisms: polyploidization through a process called endomitosis, and fragmentation of the MK cytoplasm to produce platelets. This process is highly regulated by numerous transcription factors, including GATA1/2, FOG-1, RUNX1, FLI1, SCL, GFI1b, NFE2 and MYB, and by many extrinsic factors such as cytokines, chemokines and extracellular matrix components. The major cytokine regulating megakaryopoiesis is the thrombopoietin (TPO). TPO binds to the extracellular domain of the type I homodimeric receptor MPL, mainly to two residues, D261 and L265, but also in a site around the residue F104, which is mutated in congenital amegakaryocytic thrombocytopenia.3,4 TPO binding results in conformational changes of the receptor, leading to the activation of pre-associated JAK2 molecules. Following the transphosphorylation of the receptor, downstream signaling molecules, including STAT, ERK and PI3K, become activated.

The expression of MPL and JAK2 gradually increases from the HSC to the MK throughout MK differentiation.5 Therefore, the TPO/MPL axis not only controls megakaryopoiesis, but also HSC homeostasis. Indeed, c-mpl-/- mice present both a defect in HSC function and thrombocytopenia.6 Defects in this TPO/MPL/JAK2 axis leads to hematological diseases such as thrombocytopenia or pancytopenia through the inhibition of the megakaryopoiesis process. Alternatively, thrombocytopenic states could occur following chemotherapies and immune or infectious diseases. To overcome thrombocytopenia, several TPO mimetics have been developed including romiplostim (Amgen) and eltrombopag (Novartis).

Romiplostim is a TPO mimetic that binds to MPL in a competitive manner, and consists of two identical single-chain subunits composed of human IgG1 Fc linked to a peptide containing two MPL-binding domains.7 In contrast, eltrombopag is a non-peptide TPO mimetic, which binds to the transmembrane domain of human MPL at critical residue H499, which is not conserved in the mouse Mpl.8 These two compounds have been approved by the US Food and Drug Administration (FDA) and by the European Medicines Agency (EMA) to raise platelet counts in immune thrombocytopenic purpura (ITP)9 and in infections caused by the hepatitis C virus.10 The stimulation of the TPO/MPL axis by TPO mimetics has also been successfully used in inherited MYH9haematologica | 2016; 101(12)

related thrombocytopenia and in aplastic anemia, in which eltrombopag showed benefits based on multilineage clinical responses.11,12 The effects of both romiplostim and eltrombopag in the treatment of both myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) have been investigated. The use of eltrombopag in their treatment had very limited clinical benefits; no significant improvements in platelet counts were noted, while a decrease in bleeding episodes was observed.13 Moreover, no difference in the proportions of malignant blasts was found in the bone marrow and peripheral blood of patients receiving eltrombopag versus placebo. Other studies showed that eltrombopag allowed the formation of normal MK colonies and a strong inhibition of leukemic cell proliferation.14,15 The binding of eltrombopag to MPL is specific to humans and other primates. However, it could affect murine and human leukemic cell proliferation by an MPL-independent pathway through the modulation of intracellular iron content. In aggregate, these studies have shown the importance of TPO mimetics in both the regulation of HSC homeostasis and megakaryopoiesis.

Eltrombopag is an oral drug and thus easy to administer, while romiplostim should be injected subcutaneously.

However, the mechanism of action of eltrombopag remains incompletely understood, since it could not be studied in preclinical mouse models.

In their work, Di Buduo et al. used a classical in vitro culture system with human primary cells from cord blood to demonstrate that doses of eltrombopag, ranging from 500-2000 ng/mL, stimulates the MK output at day 13 of culture as well as proplatelet formation by 4-fold, in comparison with 10 ng/mL TPO. The presence of MKs was confirmed by the expression of both surface markers (CD61 and CD42) and transcriptional factors (RUNX1, NFE2). No difference in the percentage of MKs or in their ploidy was found compared to TPO, similar to the findings of another group.16 In contrast, in another study, romiplostim was shown to display a different effect than TPO on megakaryopoiesis, stimulating MK proliferation but not maturation, with less polyploid MK and a drastic decrease in proplatelet formation. However, this effect was dose-related and linked to the over-activation of AKT but not of ERK with high romiplostim concentrations, while increased TPO doses over-activated both AKT and ERK pathways17 (Figure 1). These results indicate two different mechanisms of action for TPO and romiplostim, which bind to MPL at the same site and are expected to activate MPL in the same conformation.

Moreover, Di Buduo et al. have also developed a novel and original approach to studying megakaryopoiesis ex vivo, using a bioreactor technology based on a 3D silk-based bone marrow system, which reproduces the key steps of megakaryopoiesis and thrombopoiesis. Silk fibroin is a natural biocompatible material with non-thrombogenic features, perfectly adapted for the study of blood cell production ex vivo. Indeed, 1443

Editorials

Figure 1: Different effects of eltrombopag and romiplostim compared to thrombopoietin (TPO) during megakaryopoiesis. MPL is a homodimeric receptor important for the megakaryopoiesis process. TPO binding to the extracellular domain of MPL at residues F104, D261 and L265 induces changes in MPL conformation and the activation of STAT3/5, AKT and MAPK signaling pathways. Eltrombopag is a non-peptide TPO mimetic that binds to the intracellular domain of MPL mainly at H499, and greatly enhances the STAT3/5, AKT and MAPK activation at a higher extent than TPO. This leads to megakaryocyte (MK) amplification and overproduction of platelets. Romiplostim is a peptide TPO mimetic that binds to MPL in a similar manner to TPO, but induces a strong activation of AKT compared to STAT and ERK pathways. This leads to a strong MK amplification, which produces a lower platelet yield.

the authors had previously shown that this 3D system provides a tool not only for the study of fundamental biological mechanisms of hematopoiesis, but also for clinical applications.18 In the study by Di Buduo et al., human CD34+ progenitors from cord blood were seeded in a silk sponge for 13 days, after which MKs were observed based on specific surface markers. Alternatively, silk microtubes embedded in the silk sponge containing MKs was used for the analysis of platelet release, demonstrating that these MKs and the platelets were fully functional, even if the yield was lower than in in vitro 2D cultures. Thus, by using this technology, 1444

they have also demonstrated that eltrombopag mediates the differentiation of MKs and the release of platelets ex vivo.

To elucidate the mechanism of eltrombopag compared to TPO or to romiplostim, the authors investigated the signaling pathways activated in CD34+ progenitors and MKs. They found that eltrombopag induces the phosphorylation of STAT3/5, AKT and ERK pathways, in line with another study.16 These molecules are involved not only in proliferation and differentiation but also in survival and haematologica | 2016; 101(12)

Editorials

anti-apoptotic processes. Interestingly, they found an enhanced signaling in eltrombopag-restimulated MKs or progenitors after cytokine starvation compared to TPO. Moreover, and unexpectedly, they found that MKs or progenitors treated in vitro with various doses of eltrombopag also presented a greater dose-dependent activation of these signaling molecules than with TPO (Figure 1). These results may be explained knowing that MPL can assume different intermediate conformations between its inactive and ligand-bound states, in contrast to other receptors such as the erythropoietin receptor. These various conformations activate downstream signaling pathways differently.19 More particularly, it has recently been shown that eltrombopag activates MPL through the induction of an efficient dimerization of the transmembrane helix of MPL around the H499 residue.8 Moreover, AKT and ERK were activated by eltrombopag at a higher extent than by TPO, while romiplostim was previously shown to strongly activate AKT with a mild effect on ERK.17 While AKT phosphorylation seems to be activated during every megakaryopoiesis step, it has been shown that the ERK pathway is strongly induced at the beginning of megakaryopoiesis but must later be switched off for platelet production.20 Thus, the question of why the production of proplatelets was increased with eltrombopag, despite an enhanced and sustained activation of ERK, remains elusive. We could hypothesize that: i) there is a fine tuned cooperation between simultaneous activation of AKT and ERK pathways, ii) differential intensity and/or regulation of the signal mediated by eltrombopag (sustained signal over time), and iii) activation of new molecules and signaling pathways by eltrombopag.

The way in which eltrombopag activates MPL is analogous to the constitutive activation of MPL induced by MPLS505N and MPLW515K/L/R mutants associated with myeloproliferative neoplasms and hereditary thrombocytosis.8 Therefore, one can speculate that this non-physiological activation of signaling, which is similar to an oncogenic activation, may be very efficient, but must be carefully evaluated in patients treated long-term with eltrombopag in order to eliminate the risk of developing hematological malignancies.

In conclusion, the data obtained using eltrombopag in the study by Di Buduo et al. clearly supports a short period of use of this molecule for increasing platelet counts in thrombocytopenia patients. However, the use of this molecule for long-term treatments will require additional studies, in particular of the non-physiological activation of MPL signaling pathways.

haematologica | 2016; 101(12)

References 1. Di Buduo CA, Currao M, Pecci A, Kaplan DL, Balduini CL, Balduini A. Revealing Eltrombopag's promotion of human megakaryopoiesis through AKT/ERK-dependent pathway activation. Haematologica.

2016;101(12):1479-1488.

2. Chang Y, Bluteau D, Debili N, Vainchenker W. From hematopoietic stem cells to platelets. J Thromb Haemost. 2007 Jul;5 Suppl 1:318-27.

3. Chen WM, Yu B, Zhang Q, Xu P. Identification of the residues in the extracellular domain of thrombopoietin receptor involved in the binding of thrombopoietin and a nuclear distribution protein (human NUDC).

The Journal of biological chemistry. 2010;285(34):26697-26709.

4. Fox NE, Lim J, Chen R, Geddis AE. F104S c-Mpl responds to a transmembrane domain-binding thrombopoietin receptor agonist: proof of concept that selected receptor mutations in congenital amegakaryocytic thrombocytopenia can be stimulated with alternative thrombopoietic agents. Exp Hematol. 2010;38(5):384-391.

5. Besancenot R, Roos-Weil D, Tonetti C, et al. JAK2 and MPL protein levels determine TPO-induced megakaryocyte proliferation vs differentiation.

Blood. 2014;124(13):2104-2115.

6. Kimura S, Roberts AW, Metcalf D, Alexander WS. Hematopoietic stem cell deficiencies in mice lacking c-Mpl, the receptor for thrombopoietin.

Proc Natl Acad Sci U S A. 1998;95(3):1195-1200.

7. Cines DB, Yasothan U, Kirkpatrick P. Romiplostim. Nat Rev Drug Discov. 2008;7(11):887-888.

8. Leroy E, Defour JP, Sato T, et al. His499 regulates dimerization and prevents oncogenic activation by asparagine mutations of the human thrombopoietin teceptor. The Journal of biological chemistry.

2016;291(6):2974-2987.

9. Bussel JB. Update on eltrombopag for ITP. Oncology (Williston Park).

2009;23(13):1177-1178.

10. McHutchison JG, Dusheiko G, Shiffman ML, et al. Eltrombopag for thrombocytopenia in patients with cirrhosis associated with hepatitis C. N Engl J Med. 2007;357(22):2227-2236.

11. Favier R, Feriel J, Favier M, Denoyelle F, Martignetti JA. First successful use of eltrombopag before surgery in a child with MYH9-related thrombocytopenia. Pediatrics. 2013;132(3):e793-795.

12. Olnes MJ, Scheinberg P, Calvo KR, et al. Eltrombopag and improved hematopoiesis in refractory aplastic anemia. N Engl J Med. 2012;367(1):11-19.

13. Platzbecker U, Wong RS, Verma A, et al. Safety and tolerability of eltrombopag versus placebo for treatment of thrombocytopenia in patients with advanced myelodysplastic syndromes or acute myeloid leukaemia: a multicentre, randomised, placebo-controlled, double-blind, phase 1/2 trial. Lancet Haematol. 2015;2(10):e417-426.

14. Roth M, Will B, Simkin G, et al. Eltrombopag inhibits the proliferation of leukemia cells via reduction of intracellular iron and induction of differentiation. Blood. 2012;120(2):386-394.

15. Will B, Kawahara M, Luciano JP, et al. Effect of the nonpeptide thrombopoietin receptor agonist Eltrombopag on bone marrow cells from patients with acute myeloid leukemia and myelodysplastic syndrome.

Blood. 2009;114(18):3899-3908.

16. Jeong JY, Levine MS, Abayasekara N, Berliner N, Laubach J, Vanasse GJ. The non-peptide thrombopoietin receptor agonist eltrombopag stimulates megakaryopoiesis in bone marrow cells from patients with relapsed multiple myeloma. J Hematol Oncol. 2015;8:37.

17. Currao M, Balduini CL, Balduini A. High doses of romiplostim induce proliferation and reduce proplatelet formation by human megakaryocytes. PLoS One. 2013;8(1):e54723.

18. Di Buduo CA, Wray LS, Tozzi L, et al. Programmable 3D silk bone marrow niche for platelet generation ex vivo and modeling of megakaryopoiesis pathologies. Blood. 2015;125(14):2254-2264.

19. Staerk J, Defour JP, Pecquet C, et al. Orientation-specific signalling by thrombopoietin receptor dimers. EMBO J. 2011;30(21):4398-4413.

20. 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.

1445

Editorials

The Hippo-p53 pathway in megakaryopoiesis Praveen K Suraneni and John D. Crispino Division of Hematology/Oncology, Northwestern University, Chicago, IL, USA E-mail: j-crispino@northwestern.edu

doi:10.3324/haematol.2016.156125

egakaryocytes are among the largest and rarest cells in the body, accounting for approximately 0.01% of nucleated cells in the bone marrow.

Their differentiation involves a progression from hematopoietic stem cell to the megakaryocyte progenitor and finally to platelets.1 A key step in their maturation is the switch from a proliferating progenitor cell, which divides like any other cell, to a committed megakaryocyte that undergoes polyploidization through a modified form of the cell cycle termed endomitosis.2 During endomitosis, megakaryocytes proceed through successive cell cycles without cell division to reach DNA contents of 32N, 64N, and even 128N. After completion of G1, S and G2 phases, committed megakaryocytes enter mitosis, transition through anaphase, separate their chromosomes, and initiate cleavage furrow formation.3 However, the cleavage furrow regresses before cytokinesis is completed, resulting in formation of a single cell with a multi-lobulated, polyploid nucleus.4 Rho family small GTPases, including RhoA, Rac1, and Cdc42, are molecular switches that regulate various cellular processes including actin cytoskeleton reorganization, microtubule dynamics, cell cycle progression, cytokinesis, and platelet production.5,6 The furrow regression seen in

M

megakaryocytes appears to be due to a failure in either proper localization or activation of RhoA at the contractile ring.7 Evidence in support of a critical role of the RhoA and its effector Rho kinase (ROCK) in the regulation of the switch to polyploidy includes the finding that its inhibition or knockdown leads to increased polyploidy of megakaryocytes.7-10 Moreover, megakaryocyte-specific deletion of RhoA in mice resulted in macrothrombocytopenia due to premature release of platelets.11 Interestingly, the megakaryocytes in the animals were larger and more highly polyploidy, consistent with the inhibitor data.

Among other processes, Rho A regulates the Hippo-p53 tumor suppressor pathway, which controls proliferation, differentiation and apoptosis of cells from Drosophila to mammals.12 Decreased activity of Rho A contributes to the phosphorylation of the kinase LATS1/2 (Figure 1). Active LATS1/2 binds and inhibits MDM2, which releases p53 and allows for pathway activation.13 In addition, LATS1 phosphorylation of YAP/TAZ leads to its cytoplasmic sequestration and degradation, and impairs expression of its pro-proliferative and anti-apoptotic target genes.12 Recent studies by Ganem et al. demonstrated that the Hippo-p53 pathway controls the tetraploid checkpoint,14

A

B

Figure 1. Comparison of the Hippo-p53 pathway in erythroblasts versus megakaryocytes. (A) In diploid cells, such as erythroblasts, a reduction in RhoA GTPase activity promotes the Hippo pathway by increasing the phosphorylation of LATS1/2. The LATS1/2 kinases in turn phosphorylate and inhibit the transcriptional coactivators YAP and TAZ by subsequent degradation and/or cytoplasmic retention. In parallel, LATS1/2 stabilizes p53 through its direct association with MDM2, which disrupts the MDM2-p53 interaction. (B) In polyploid megakaryocytes, the Hippo-p53 pathway remains off as the reduction in RhoA activity fails to activate LATS1/2, allowing YAP/TAZ to translocate into the nucleus and promote target gene expression.

1446

haematologica | 2016; 101(12)

Editorials

which exists to prevent the continued growth of aneuploid cells. Moreover, the presence of extra centrosomes and actin filaments, a consequence of the increased number of chromosomes, was found to activate Hippo-p53 pathway by down-regulating RhoA activity.14 In proliferating cells, two guanine exchange factors, GEF-H1 and ECT2, play critical roles in cytokinesis by activating RhoA at the cleavage furrow.15 By contrast, GEFH1 and ECT2 are down-regulated during the 2N to 4N transition and during polyploidization beyond the 4N stage, respectively, resulting in suppression of RhoA signaling during endomitosis.10 Given that RhoA activity is low in megakaryocytes undergoing polyploidization, one might predict that the Hippo-p53 pathway would be activated and prevent the process. However, a new study by Roy et al. published in this issue of Haematologica sheds new light on Hippo-p53 pathway function in megakaryocytes.16 The critical observation is that, despite the presence of a functional Hippo-p53 pathway, low RhoA activation in megakaryocytes fails to activate the tetraploid checkpoint and instead allows for endomitosis. In addition, the sustained activation of YAP contributes to megakaryopoiesis by increasing expression of mitochondrial genes including PGC1Îą, which contributes to mitochondrial biogenesis.

To investigate how naturally polyploid cells such as megakaryocytes overcome the tetraploid checkpoint, the authors first validated the expression of Hippo-p53 pathway genes in human megakaryocytes at various developmental stages. Their results revealed that expression of LATS1, LATS2 and TAZ remain constant during MK maturation, but that there is a significant increase in the expression of YAP target genes, such as CTGF, CYR61, FSTL1 and INHBA. This increase was associated with reduced levels of phosphorylated YAP but an overall high level of total YAP protein, indicative of an inactive Hippop53 pathway.

Next, the authors examined the functionality of the Hippo-p53 pathway by exposing megakaryocytes to genotoxic stress induced by etoposide. With this treatment, they observed a strong increase in both LATS2 and p53 along with a spike in the phosphorylation of YAP. Furthermore, exposure to etoposide was associated with translocation of p53 from the cytoplasm to the nucleus.

These results reveal that there is an active Hippo-p53 surveillance pathway in human megakaryocytes.

To address whether polyploidy is interpreted as genotoxic stress in megakaryocytes, Roy et al. assessed the expression levels of p53 and Hippo pathway genes at different ploidy stages. However, there were no significant changes in the expression or the activity of p53 or level of YAP, and there was a steady rise in the expression of YAP target genes during polyploidization. This activation of YAP suggests that megakaryocytes fail to activate the Hippo pathway. To investigate this further, the Authors asked whether induced impairment of RhoA activity could force activation of the Hippo pathway in megakaryocytes. By treating cells with a ROCK inhibitor, they discovered that the Hippo pathway was, as expected, strongly induced in erythroid cells but that similar treatment did not drive Hippo signaling in megakaryocytes. This key result provides evidence that Hippo pathway activation is haematologica | 2016; 101(12)

uncoupled from Rho kinase activity in endomitotic megakaryocytes.

Roy et al. also investigated the contributions of p53, which normally eliminates aberrant tetraploid cells and prevents tumor formation to megakaryopoiesis. Previous studies have shown that loss of p53 in mice was associated with increased megakaryocyte ploidy levels; this effect was exacerbated in stress conditions.17 Furthermore, stabilization of p53 by MDM2 inhibition impaired polyploidization and proplatelet formation.18,19 In this new report, knockdown of p53 was shown to result in a modest, but significant, increase in MK polyploidization as well as increased numbers of pro-platelet forming cells and cytoplasmic maturation.16 Perhaps the most surprising finding of Roy et al. is the link between YAP and expression of the mitochondrial biogenesis regulator PGC1Îą during megakaryocyte differentiation and polyploidization. Although YAP regulated genes are generally thought to induce proliferation and survival of cells in the absence of Hippo-p53 activation, knockdown of YAP had no effect on polyploidization or apoptosis of megakaryocytes but rather the reduction of YAP did lead to decreased proplatelet formation and reduced mitochondrial mass, which the authors demonstrate is a notable feature of polyploid megakaryocytes.

These findings are consistent with previous studies that suggested that YAP signaling also has a role in mitochondrial regulation in Drosophila and human cell lines.20 In summary, the study by Roy et al. provides critical mechanistic insights into how naturally polyploid megakaryocytes overcome the tetraploid checkpoint that normally functions as a tumor suppressor pathway. Their findings clearly show that although the Hippo-p53 pathway is intact in megakaryocytes, it is not activated during polyploidization. Furthermore, they provide novel insights into the contributions of YAP to mitochondrial biogenesis, which is a notable feature of the larger, polyploid cells. However, a few important questions remain unanswered. What is the nature of the disconnect between low RhoA activity and Hippo-p53 activation, and how do megakaryocytes escape Hippo-p53 activation during polyploidization? Also, what is the link between Rac1 and RhoA in megakaryocytes? A previous study observed a 2-fold increase in Rac1 activity in tetraploid compared to diploid cells; this increased Rac1 activity was further found to suppress RhoA activity, which, in turn, typically activates the Hippo pathway.14 Comparing the level of Rac1 activity in megakaryocytes at different stages of polyploidization would be an interesting next step. Finally, a more comprehensive analysis of YAP target genes in endomitotic megakaryocytes would provide further insights into the way that this pathway contributes to increased mitochondrial mass and platelet production. In summary, the report by Roy et al. provides exciting and novel insights that improve our understanding of megakaryopoiesis and may lead to improved strategies to increase platelet production.

References 1. Geddis AE. Megakaryopoiesis. Semin Hemaol. 2010;47(3):212-219.

2. Ravid K, Lu J, Zimmet JM, Jones MR. Roads to polyploidy: the megakaryocyte example. J Cell Physiol. 2002;190(1):7-20.

1447

Editorials 3. Vitrat N, Cohen-Solal K, Pique C, et al. Endomitosis of human megakaryocytes are due to abortive mitosis. Blood. 1998;91(10):3711-3723.

4. Geddis AE, Fox NE, Tkachenko E, Kaushansky K. Endomitotic megakaryocytes that form a bipolar spindle exhibit cleavage furrow ingression followed by furrow regression. Cell Cycle. 2007;6(4):455-460.

5. Hall A. Rho family GTPases. Biochem Soc Trans. 2012;40(6):1378-1382.

6. Pleines I, DĂźtting S, Cherpokova D, et al. Defective tubulin organization and proplatelet formation in murine megakaryocytes lacking Rac1 and Cdc42. Blood. 2013;122(18):3178-3187.

7. Lordier L, Jalil A, Aurade F, et al. Megakaryocyte endomitosis is a failure of late cytokinesis related to defects in the contractile ring and Rho/Rock signaling. Blood. 2008;112(8):3164-3174.

8. Wen Q, Goldenson B, Silver SJ, et al. Identification of Regulators of Polyploidization Presents Therapeutic Targets for Treatment of AMKL. Cell. 2012;150(3):575-589.

9. Chang Y, Aurade F, Larbret F, et al. Proplatelet formation is regulated by the Rho/ROCK pathway. Blood. 2007;109(10):4229-4236.

10. Gao Y, Smith E, Ker E, et al. Role of RhoA-specific guanine exchange factors in regulation of endomitosis in megakaryocytes. Dev Cell.

2012;22(3):573-584.

11. Suzuki A, Shin JW, Wang Y, et al. RhoA is essential for maintaining normal megakaryocyte ploidy and platelet generation. Plos One. 2013;8(7):e69315.

12. Yu FX, Guan KL. The Hippo pathway: regulators and regulations. Genes

Dev. 2013;27(4):355-371.

13. Aylon Y, Michael D, Shmueli A, Yabuta N, Nojima H, Oren M. A positive feedback loop between the p53 and Lats2 tumor suppressors prevents tetraploidization. Genes Dev. 2006;20(19):2687-2700.

14. Ganem NJ, Cornils H, Chiu SY, et al. Cytokinesis failure triggers hippo tumor suppressor pathway activation. Cell. 2014;158(4):833-848.

15. Birkenfeld J, Nalbant P, Bohl BP, Pertz O, Hahn KM, Bokoch GM. GEFH1 modulates localized RhoA activation during cytokinesis under the control of mitotic kinases. Dev Cell. 2007;12(5):699-712.

16. Roy A, Lordier L, Pioche-Durieu C, et al. Uncoupling of the Hippo and Rho pathways allows megakaryocytes to escape the tetraploid checkpoint. Haematologica. 2016;101(12):1469-1478.

17. Apostolidis PA, Woulfe DS, Chavez M, Miller WM, Papoutsakis ET. Role of tumor suppressor p53 in megakaryopoiesis and platelet function. Exp Hematol. 2012;40(2):131-142.e4.

18. Iancu-Rubin C, Mosoyan G, Glenn K, Gordon RE, Nichols GL, Hoffman R. Activation of p53 by the MDM2 inhibitor RG7112 impairs thrombopoiesis. Exp Hematol. 2014;42(2):137-145.e5.

19. Mahfoudhi E, Lordier L, Marty C, et al. P53 activation inhibits all types of hematopoietic progenitors and all stages of megakaryopoiesis.

Oncotarget. 2016;7(22):31980-31992.

20. Nagaraj R, Gururaja-Rao S, Jones KT, et al. Control of mitochondrial structure and function by the Yorkie/YAP oncogenic pathway. Genes Dev. 2012;26(18):2027-2037.

Ibrutinib in the real world patient: many lights and some shades Paolo Ghia1 and Antonio Cuneo2 1 Strategic Research Program on CLL and B Cell Neoplasia Unit, Division of Experimental Oncology, Vita-Salute San Raffaele University and IRCCS San Raffaele Scientific Institute, Milan; and 2Department of Medical Sciences, Hematology Unit. University of Ferrara, Italy.

E-mail: ghia.paolo@hsr.it

W

doi:10.3324/haematol.2016.155986

With an estimated incidence of about 4.92 cases per 100,000/year in Europe1 and 14,620 new cases in 2015 in the USA,2 chronic lymphocytic leukemia (CLL) is the most frequent leukemia in Western countries. While a minority of patients may attain longlasting responses with chemoimmunotherapy,3,4,5 relapse and treatment-resistant diseases develop in the majority of cases; infections, progressive disease and second primary tumors being the most frequent causes of death.5 The firstin-class inhibitor of Bruton's tyrosine kinase (BTK), ibrutinib, was welcomed in 2013 as a new paradigm for the treatment of relapsed or refractory CLL, as it produced responses in 71% of the cases in a heavily pre-treated patient population who had few, if any, alternative treatment options.6,7 After a median observation of 3 years,8 exceptional overall survival (OS) and progression-free survival (PFS) rates were reported (79% and 69%, respectively), along with a low (12%) discontinuation rate due to adverse events. Following the publication of excellent efficacy data in patients with 17p deletion (del(17p)) or TP53 mutations,9,10 high expectations were generated in the belief that this drug was able to produce durable responses in the majority of patients, irrespective of the presence of unfavorable prognostic factors.11 However, the median age of the patients in the trials was 64 years8 and only 32% of them had a Cumulative Illness Rating Scale (CIRS) score of >6.12 This reflects the inclusion criteria in the clinical trials, which required that the patients had an Eastern Cooperative Oncology Group 1448

(ECOG) performance status of less than 2, with adequate liver and kidney function, no significant neutropenia or thrombocytopenia and who did not require warfarin or strong CYP3A4/5 inhibitors. Because CLL is diagnosed at a median age of 70-72 years and the majority of patients carry several comorbidities,13 the efficacy and safety data published in the literature were obtained in a patient population which did not reflect the typical patient found in everyday practice.

Two papers in this issue of Haematologica14,15 describe the efficacy and toxicity of ibrutinib in 315 and 95 realworld patients treated in the UK and in Sweden, respectively, within a named patient scheme or a compassionate use program. Both studies adopted rigorous methods, minimizing biases inherent in retrospective studies. The baseline characteristics of the enrolled patients are summarized in Table 1, along with the salient outcome measures.

Overall, these two studies are reassuring with regards to the excellent efficacy of this new class of inhibitors, even when utilized in routine clinical practice without the many constraints and controls typical of clinical trials.

Though the follow up is still short (around 1-1.5 years), objective outcome measures, i.e., median discontinuationfree survival and PFS, remain in the comfort zone, with PFS values of 77% after 10 months among the Swedish patients, and not yet reached in the UK series. These PFS values also include TP53 disrupted patients, and, in particular, those patients with del(17p), who make up one third haematologica | 2016; 101(12)

Editorials

Table 1. Salient results of published studies of ibrutinib in relapsed refractory chronic lymphocytic leukemia (CLL).

Patients’ characteristics and outcome measures N. of patients Median age Median follow-up (months) Progression free survival Overall survival Cases with dose reduction Cases with permanent discontinuation due to AEs

Patients enrolled in trials Byrd et al. (2015)8

UK CLL forum14

Swedish CLL group15

101 64 36 69% at 30 months 79% at 30 months 12% 12%

315 69 16 NR 83,8% at 12 months 44,4% 17,7% at 12 months

95 69 10.2 77% at 10 months 83% at 10 months 22% 11%

Real world patients Mayo Clinic series16 124 65 6,4 NR NR NR 10% at 6 months 22% at 12 months

Moffitt Cancer Center series17 54 60 9,1 NR NR NR 15%

NR: not reported; AEs: adverse events.

of all cases in the UK cohort and >50% in the Swedish group. Unfortunately, both real life studies were categorically able to demonstrate that even among the most advanced countries in Europe, testing for TP53 mutations in all CLL patients before starting any line of therapy remains a difficult goal to reach. Having said that, the efficacy of ibrutinib in CLL with TP53 disruption was also confirmed in these two studies, even though the affected patients experienced more discontinuations and earlier progression than the remaining individuals. It seems likely that a more advanced age and the number of previous treatments were the contributing factors to a less favorable prognosis.

This is probably also one of the possible explanations for the unexpectedly higher discontinuation rate, which ranged between 24-26% after 10-12 months, in contrast with the more reassuring percentage of patients (33% at 3 years) discontinuing ibrutinib treatment in published trials. Although disease progression was one of the reasons for discontinuing treatment, most patients, in particular in the UK study, stopped the drug because of adverse events. In general, the median age, which was 69 years in both studies, was higher than that reported in the registration trials, and 1/4 patients had poor performance status, thus providing, at least in part, an explanation for the discrepancies in drug tolerance. It is also worth noting that in the past, in the case of immunochemotherapy, the fludarabine, cyclophosphamide, and rituximab (FCR) regimen when used in routine clinical practice was associated with more dose reductions than previously reported.18 A worrisome possibility would be that, in contrast to the colleagues enrolling patients in clinical trials and who can be assisted by written guidelines or medical monitoring, hematologists in everyday life may not feel fully at ease in managing the typical non-hematological toxicity of the drug, resulting in earlier discontinuation. In line with this, a higher percentage of patients also reduced the dose, but this did not appear to be associated with poorer outcome, at least in the larger UK series. Long (>14 days) dose interruptions were associated with inferior outcome, and it is likely that this observation might be simply due to a selection bias for those patients with comorbidities or with more advanced disease.

Interestingly, the efficacy of ibrutinib was similar irrehaematologica | 2016; 101(12)

spective of the number of prior lines of therapy. This observation in everyday life is at variance with the ad hoc analysis within the RESONATE study,19 and requires longer follow-up and maybe sequential study in order to confirm the correct placement of the drug in the treatment history of our CLL patients.20,21 In conclusion, ibrutinib confirmed its efficacy and tolerability in over 400 patients treated outside clinical trials, without unexpected adverse events, but with infection being cited as a frequent cause of discontinuation. Dose reductions did not appear to influence outcome which also remained very good in patients with TP53 disruption, though less favorable in those with reduced performance status, more likely resulting in prolonged treatment breaks.

Along the same line, it becomes relevant to underscore that while initial reports suggested that most patients with relapsed or refractory CLL who discontinued ibrutinib had poor outcomes,22 more recent studies clearly show that switching to another kinase inhibitor23 or to venetoclax24 may rescue up to 70% of the cases, especially when discontinuation was prompted by an adverse event.

Therefore, there is life with ibrutinib, but also after it.

References 1. Sant M, Allemani T, Tereanu C et al. Incidence of hematologic malignancies in Europe by morphologic subtype: results of the HAEMACARE project. Blood. 2010; 116(19): 3724-3734.

2. Siegel R L, Miller K D, Jemal A. Cancer Statistics, 2015CA Cancer J Clin.

2015; 65 (1):5–29.

3. Thompson PA, Tam CS, O'Brien SM et al. Fludarabine, cyclophosphamide, and rituximab treatment achieves long-term disease-free survival in IGHV-mutated chronic lymphocytic leukemia. Blood. 2016; 127(3) :303-9 4. Rossi D, Terzi-di-Bergamo L, De Paoli L et al. Molecular prediction of durable remission after first-line fludarabine-cyclophosphamide-rituximab in chronic lymphocytic leukemia. Blood. 2015; 126(16): 1921-1924.

5. Fischer K, Bahlo J, Fink AM et al. Long-term remissions after FCR chemoimmunotherapy in previously untreated patients with CLL: updated results of the CLL8 trial. Blood. 2016;127(2): 208-215.

6. Byrd JC, Furman RR, Coutre SE et al.Targeting BTK with ibrutinib in relapsed chronic lymphocytic leukemia. N Engl J Med. 2013;369(1):3242.

7. Foà R, Guarini A. A mechanism-driven treatment for chronic lymphocytic leukemia? N Engl J Med. 2013; 369(1): 85-87.

8. Byrd JC, Furman RR, Coutre SE et al. Three-year follow-up of treatmentnaïve and previously treated patients with CLL and SLL receiving singleagent ibrutinib. Blood. 2015; 125(16): 2497-2506.

1449

Editorials 9. Farooqui MZ, Valdez J, Martyr S et al. Ibrutinib for previously untreated and relapsed or refractory chronic lymphocytic leukaemia with TP53 aberrations: a phase 2, single-arm trial. Lancet Oncol. 2015; 16(2): 169176.

10. O'Brien S, Jones JA, Coutre SE et al. Ibrutinib for patients with relapsed or refractory chronic lymphocytic leukaemia with 17p deletion (RESONATE-17): a phase 2, open-label, multicentre study. Lancet Oncol.

2016;17(10):1409-1418 11. Ghia P. Ibrutinib holds promise for patients with 17p deletion CLL. Lancet Oncol. 2016; 17(10): 1342-1343 12. Byrd JC, Brown JR, O'Brien S, et al. Ibrutinib versus ofatumumab in previously treated chronic lymphoid leukemia. N Engl J Med. 2014; 371(3):213-23 13. Yancik R. Cancer burden in the aged: an epidemiologic and demographic overview. Cancer. 1997; 80(7):1273-1283.

14. UK CLL Forum. Ibrutinib For relapsed/refractory chronic lymphocytic leukemia: A UK and Ireland analysis of outcomes in 315 Patients.

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

15. Winqvist M, Asklid A, Andersson PO, et al. Real-world results of ibrutinib in patients with relapsed or refractory chronic lymphocytic leukemia: data from 95 consecutive patients treated in a compassionate use program. Haematologica. 2016;101(12):1573-1580.

16. Parikh SA, Rabe Chaffee K, Call TG, et al. Ibrutinib therapy for chronic lymphocytic leukemia (CLL): an analysis of a large cohort of patients treated in routine clinical practice. Blood 2015 126:2935 17. Sandoval-Sus JD, Chavez JC, Dalia S, et al. Outcomes of patients with relapsed/refractory chronic lymphocytic leukemia after ibrutinib discon-

1450

18.

19.

20.

21.

22.

23.

24.

tinuation outside clinical trials: A single institution experience. Blood 2015 126:2945 Herishanu Y, Goldschmidt N, Bairey O et al. Israeli CLL Study Group.

Efficacy and safety of front-line therapy with fludarabine-cyclophosphamide-rituximab regimen for chronic lymphocytic leukemia outside clinical trials: the Israeli CLL Study Group experience. Haematologica.

2015; 100(5):662-669 Brown JR, Hillmen P, O'Brien S, et al Updated efficacy including genetic and clinical subgroup analysis and overall safety in the phase 3 RESONATETM trial of ibrutinib versus ofatumumab in previously treated chronic lymphocytic leukemia/small lymphocytic lymphoma. ASH 2014, Blood. 2014; 124:3331.

Eichhorst B, Robak T, Montserrat E, Ghia P, Hillmen P, Hallek M, Buske C; ESMO Guidelines Committee. Chronic lymphocytic leukaemia: ESMO Clinical Practice Guidelines for diagnosis, treatment and followup. Ann Oncol. 2015; 26 Suppl 5:78-84 Hallek M. Chronic lymphocytic leukemia: 2015 Update on diagnosis, risk stratification, and treatment. Am J Hematol. 2015; 90(5):446-460.

Jain P, Keating M, Wierda W et al. Outcomes of patients with chronic lymphocytic leukemia after discontinuing ibrutinib. Blood. 2015; 125(13):2062-2067.

Mato AR, Nabhan C, Barr PM et al. Outcomes of CLL patients treated with sequential kinase inhibitor therapy: a real world experience. Blood.

2016. [Epub ahead of print].

Jones JA, Wierda WG, Choi MY, et al. Venetoclax activity in CLL patients who have relapsed after or are refractory to ibrutinib or idelalisib. 2016 ASCO Annual Meeting. Abstract 7519.

haematologica | 2016; 101(12)

REVIEW ARTICLE

The myeloma stem cell concept, revisited: from phenomenology to operational terms

EUROPEAN HEMATOLOGY ASSOCIATION

Hans Erik Johnsen,1,8,9 Martin Bøgsted,1,8,9 Alexander Schmitz,1 Julie Støve Bødker,1 Tarec Christoffer El-Galaly,1,8,9 Preben Johansen,2 Peter Valent,3 Niklas Zojer,4 Els Van Valckenborgh,5 Karin Vanderkerken,5 Mark van Duin,6 Pieter Sonneveld,6 Martin Perez-Andres,7 Alberto Orfao7 and Karen Dybkær1,8,9 1 Department of Haematology Aalborg University Hospital, Denmark; 2Department of Hematopathology, Aalborg University Hospital, Denmark; 3The Department of Internal Medicine I, Division of Hematology Medical University of Vienna, Austria; 4Wilhelminen Cancer Research Institute and Ludwig Boltzmann Cluster Oncology, First Department of Medicine, Center for Oncology and Hematology, Vienna, Austria; 5Department of Hematology and Immunology-Myeloma Center, Vrije University Brussels, Belgium; 6 Department of Hematology, Erasmus Medical Center Cancer Institute, Rotterdam, The Netherlands; 7Department of Medicine and Cytometry Service (NUCLEUS), Cancer Research Center (IBMCC, USAL-CSIC), Institute for Biomedical Research of Salamanca (IBSAL), University of Salamanca (USAL), Spain; 8Clinical Cancer Research Center, Aalborg University Hospital, Denmark and 9The Department of Clinical Medicine, Aalborg University, Denmark

Ferrata Storti Foundation

Haematologica 2016 Volume 101(12):1451-1459

ABSTRACT

T

he concept of the myeloma stem cell may have important therapeutic implications, yet its demonstration has been hampered by a lack of consistency in terms and definitions. Here, we summarize the current documentation and propose single-cell in vitro studies for future translational studies. By the classical approach, a CD19–/CD45low/–/CD38high/CD138+ malignant plasma cell, but not the CD19+/CD38low/– memory B cell compartment, is enriched for tumorigenic cells that initiate myeloma in xenografted immunodeficient mice, supporting that myeloma stem cells are present in the malignant PC compartment. Using a new approach, analysis of c-DNA libraries from CD19+/CD27+/CD38– single cells has identified clonotypic memory B cell, suggested to be the cell of origin. This is consistent with multiple myeloma being a multistep hierarchical process before or during clinical presentation. We anticipate that further characterization will require single cell geno- and phenotyping combined with clonogenic assays. To implement such technologies, we propose a revision of the concept of a myeloma stem cell by including operational in vitro assays to describe the cellular components of origin, initiation, maintenance, and evolution of multiple myeloma. These terms are in accordance with recent (2012) consensus statements on the definitions, assays, and nomenclature of cancer stem cells, which is technically precise without completely abolishing established terminology. We expect that this operational model will be useful for future reporting of parameters used to identify and characterize the multiple myeloma stem cells. We strongly recommend that these parameters include validated standard technologies, reproducible assays, and, most importantly, supervised prospective sampling of selected biomaterial which reflects clinical stages, disease spectrum, and therapeutic outcome. This framework is key to the characterization of the cellular architecture of multiple myeloma and its use in precision medicine. haematologica | 2016; 101(12)

Correspondence: haej@rn.dk

Received: November 19, 2015.

Accepted: August 30, 2016.

Pre-published: November 10, 2016. doi:10.3324/haematol.2015.138826

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/100/12/1451

©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.

1451

H.E. Johnsen et al.

Introduction The multiple myeloma stem cell (MMSC) is defined as a cell within the malignant tissue that possesses the capacity to self-renew and to differentiate into the predominant lineages of myeloma plasma cells comprising the neoplasm. Self-renewal is cell division without the loss of differentiation potential, at least in some daughter cells. This concept is based on phenomenology, and MMSCs are defined experimentally by their ability to recapitulate the continuous growth of malignant tissue in vivo and/or in vitro. Experimental approaches underlie the terms “cancerinitiating cells” and “cell of origin”, which are used operationally to characterize cancer stem-cell (CSC) compartments. Several studies have analyzed the hematopoietic system by sorting a few or even single cells, tracking acquired genetic changes, and transplanting cells to determine whether subsets within the differentiating hierarchy act as CSCs.1 Hierarchical models have been described for hematological malignancies such as acute and chronic myeloid and lymphoblastic leukemia,2-6 and multiple myeloma (MM),712 suggesting that the malignant clone includes immature progenitors or stem cells present in the proliferating bonemarrow compartment. According to the “multistep-oncogenesis” theory, malignancies develop due to a series of molecular alterations that occur in such cells. Although this theory is appealing and, on the whole, accepted by the scientific community, many aspects of this theory require rigorous scientific questions and answers.

The practical usefulness of the hierarchical model relies on emerging insights into the pathogenesis and identification of new biomarkers. Progress in this area depends on recent technological advances that have enabled researchers to study single-cell gene expression13 and to address changes in cellular programming in a global fashion by analyzing gene mutations, expression and deregulation.14-16 Lagging slightly behind this approach is the rapidly developing technology for examining the protein compartment of the cell.17-19 In the present review, we have updated the latest scientific reports and propose a revision of the MMSC concept to include operational terms (Table 1), in accordance with recent (2012) consensus statements on the definitions, assays, and nomenclature of cancer stem cells, which is technically precise without completely abolishing established terminology.20 We expect this will empower the design of translational research activities based on prospectively sampled biomaterial which reflects the spectrum of clinical disease. We expect such a revision to acquire data that indirectly document the existence of multiple MMSCs and validate its impact via targeted therapy.

The stem-cell concept and the B-cell hierarchy A normal stem cell is a unique cell type present at low frequency which can renew itself and produce progenitors of one or more specialized cell types. Beginning with the fertilized oocyte, cellular differentiation into specialized cell types, tissues, and organs follows a strict pattern. The classical view is that during such processes, cells gradually lose their capacity for self-renewal, their plasticity, and their ability to develop into different lineages.21 Stem cells may be classified into two major classes: 1) pluripotent stem cells derived from early embryos that are 1452

able to replenish all cell types in the human body, and 2) multipotent stem cells located in various organs that are dedicated to the replenishment of specific tissues such as blood. Embryonic stem cells, which are derived from the inner cell mass of the blastocyst, can be cultured in vitro nearly indefinitely. Unlike embryonic stem cells, multipotent organ restricted stem cells, which may be isolated from a variety of tissues in fetal and adult humans, are lineage specific; hematopoietic stem cells, neuronal stem cells, and hepatic stem cells are all multipotent.

In this review, we consider hematopoietic stem cells and putative CSCs as prototypes of multipotent stem cells.

However, not all are multipotent; for example, ´endstage´ effector B cells may regain self-renewing mechanisms in order to expand and maintain immunity.22,23 In normal B-cell lymphopoiesis, a number of well-characterized subpopulations have been defined by membrane marker phenotyping, as reviewed and illustrated in the upper part of Figure 1. The very early B-cell precursors develop into pro- and pre-B cells before they migrate as immature B cells into the blood to reach peripheral lymphoid organs as naive B cells.24-31 Germinal and post-germinal-center centrocytes, centroblasts, memory cells, plasmablasts, and end-stage plasma cells (PCs) are included in the later stages of the mature B-cell differentiation hierarchy. Most malignant B-cell lymphomas, chronic lymphoblastic leukemias, and MMs are considered to originate from these cells following analyses of the somatic hypermutation and class switch-recombination status of the gene encoding the immunoglobulin heavy chain (IgH) which defines the hierarchical status of any clonotypic cell.32-36 Further understanding of the molecular mechanisms that regulate the malignant B-cell hierarchy requires investigations of purified subpopulations or even single cells.

The phenomenon and its markers The MMSC concept is based on phenomenology: the outcome of studies in animal and/or humans that rely on in vivo and in vitro assays. However, these assays address the future potential of the stem cell, while study outcomes address the expression of this potential.37 Therefore, identifying a stem cell by allowing it to differentiate loses the original cell; at the same time, only a limited range of responses may be evident based on the model used. All stem cell assays reveal an outcome after cells are perturbed, and it is still an open question how the stem cell phenomena should be identified at the single-cell level.

At the cellular level, MM is characterized by uncontrolled expansion of PCs in the bone marrow. In addition, cells belonging to the myeloma IgH-defined clone (clonotypic cells) which precede the PC stage have been detected in peripheral blood, lymph nodes, and bone marrow.10,3336,38,39 The earliest clonotypic cells were exclusively identified in the CD38– B-cell compartment, suggesting a potential precursor and a myeloma hierarchy. The main scientific question was whether these clonal cells were intrinsic to the maintenance of the malignant PC clone defined by membrane markers, as illustrated in the lower part of Figure 1. Functional studies supported this idea,10,11,38,39 but attempts to identify clonogenic potential in this compartment were unsuccessful,8,9 and it remains a controversial issue.12 haematologica | 2016; 101(12)

Operational terms for study of the myeloma stem cell

Table 1. Terms that define MMSCs and associated information.

Type of cell Pre-malignant MMSC Malignant MMSC

Cell of origin Myeloma-initiating cell (in vivo) Myeloma long-term culture-initiating cell (in vitro) Neoplastic sphere-forming cell (in vitro)

Definition Conceptual context Member of a subpopulation of neoplastic stem cells that can propagate clones that may or may not develop into MMSCs over time, but that have no immediate cancer-initiating potential Member of a subpopulation of neoplastic stem cells within the tumor that indefinitely propagates malignant clones and produces overt myeloma Operational context A normal cell that acquires the first myeloma-promoting mutation(s); not necessarily linearly related to the MMSC and myeloma populations A cell that regenerates detectable myelomaa populations in xenografted immunodeficient mice that are sustainedb; usually measured via limiting dilution A cell that can initiate the sustainedc production of neoplastica populations when cultured in supportive conditions with or without stromal cells; usually measured via limiting dilution A cell that initiates non-adherent clusters or colonies of neoplastic progeny in in vitro cultures; usually measured by counting clusters/colonies that generate secondary units when re-plated.

a Defined by obviously abnormal biological features exhibited by cells in the primary sample, for example, the formation of a palpable growth or tumor, the production of myeloma plasma cells, abnormal growth properties, and clonal karyotype or genotype. bBased on our experience with normal stem cells, in this context “sustained” usually means ≥16 weeks, with (ideally) demonstrable activity on serial transplantation into secondary mice. cBased on our experience with normal hematopoietic stem cells, in this context “sustained” usually means ≥6 weeks with stromal cells ± stimulatory growth factors.

As illustrated in Figure 2, MM is the clinical outcome of a multistep transformation process that includes a premalignant state, the monoclonal gammopathy of undetermined significance (MGUS).40 Two early pathways in MM oncogenesis have been identified: a nonhyperdiploid pathway characterized by translocations involving the IgH locus (14q32), and a hyperdiploid pathway.41,42 IgH translocations are introduced at the MGUS state; the majority of breakpoints fall within the switch regions of the gene encoding IgH. Thus, the mechanisms and timing of translocation are those of normal IgH class switchrecombination, and define an early oncogenic event or targeting of a gene with oncogenic potential during initiation.

In the hyperdiploid pathway, recurrent changes in chromosome number are considered to constitute an early event in MM oncogenesis.

MM cells display stable VDJ joining sequences, an accumulation of somatic mutations and absence of ongoing somatic hypermutation.43,44 These findings support the idea that some MM cells are derived from a germinal center or post-germinal-center B cell that differentiates into a clonotypic memory B cell or a PC as illustrated in Figure 1.

In search of the MMSC Although the malignant cells that represent terminally differentiated PCs are readily identifiable via morphological criteria, the phenotype of the MMSC is not yet known with certainty due to several observations that suggest a less-mature clonal precursor. Critically, this idea was supported by data from single-cell analyses.33,36 More specifically, we identified, sorted and studied CD19+/CD27+/CD38– single-cell libraries of documented clonotypic cells that met all the criteria for a memory B cell, which were present in all 10 patients studied at a median frequency of 0.3% of CD19+ B cells (range 0.001-1 cell per 1000 circulating cells).45,46 This observation, together with contradictory MMSC studies8-11,33-36 and technological progress, motivated detailed multiparametric characterization of MMSCs and related B-cell subsets. To this end, members of the European Myeloma Network initiated a collaborative nethaematologica | 2016; 101(12)

work of laboratories and scientists (MSCNET) in 2007.

As reviewed above, in 2007, the state of the art of MMSC suggested that post-germinal CD19+/CD20+/CD138–/CD38– pre-PC B cells constituted the putative MMSC.10,11,38,39 However, some evidence pointed to a CD19–/CD20– /CD138+/CD38+ plasmablast/PC.8,9 It was conceivable that both or additional stem-cell compartments could be the cellular basis for the acquired oncogenetic changes that underlie myeloma initiation, maintenance, and evolution.

During the last decade, MSCNET has performed detailed studies to confirm or refute previous studies47-60 and established protocols17,61-70 to delineate the phenotypes of subpopulations of cells in randomly selected primary tumor samples and in preclinical disease models. While our investigations did not confirm that pre-PC B cells are myeloma initiating,50,51,52 several observations suggested that further study of plasmablasts/PCs at the single cell level will be key to elucidating MMSC functions.

This suggestion is in accordance with recent studies documenting that “CD19–/CD45low/–/CD38high/CD138+ [PCs] are enriched for human tumorigenic myeloma cells” that regenerate detectable myeloma populations in xenografted immunodeficient mice.12 In a parallel investigation, CD19+/CD38low/– memory B cells engrafted into human bone grafts, resulting in the repopulation of polyclonal B cells, which supports the hypothesis that memory B cells have the ability to self-renew.23 However, since few clonotypic B cells were present in these grafts, the specific questions about pre-PC B cells have not been definitively answered; these cells may be present at quantities below the limit of detection. The same limitation may pertain to the first cells of origin that depart from normal B lymphopoiesis and harbor early, but not late,50 changes in genetic, epigenetic, or other regulatory events that underlie the generation of malignant myeloma-initiating cells.

Another controversy surrounds the CD138– myeloma PCs studied in our preclinical models. These PCs exhibit engraftment and clonogenic potential in vitro (unpublished data),51,52 in accordance with findings from others.53,54 The nature of the CD138– and CD138+ PCs in the lineage of 1453

H.E. Johnsen et al. myeloma cells remains to be determined.55 Of particular importance, MSCNET has identified a new CD19–/CD45–/CD138+/CD38+ subpopulation associated to the CD19+/CD45–/CD138+/CD38+ normal PC compartment with different gene expression in normal bone marrow, suggesting a differentiation pathway that has not yet been studied in terms of MM pathogenesis.56 Our next step will be to study the function of these compartments and patients in more detail, guided by novel technological progress in single-cell analysis, a common prospective biobank strategy, and well-characterized preclinical functional in vitro and in vivo models.

In our search, we propose a revision of the conceptual context of MMSCs to use more operational in vitro func-

tions that will enable studies of the origin, initiation, maintenance, and evolution driven by deregulated genetic events. We anticipate that our understanding of MMSC will encompass the multiple dynamic cell compartments that are present from the initial to the final steps of the differentiation, evolution and selection of clonal PCs in myeloma.

Hallmarks: self-renewal, plasticity, and drug resistance CSC function, a hallmark of cancer in general, is defined to include normal, specific functions such as self-renewal,

Figure 1. Membrane marker defined subpopulations of the normal B-cell differentiation and the myeloma hierarchy. Upper panel: Cytomic phenotyping of the normal, lineage-specific pro- and pre-B cells in the bone marrow that develops from hematopoietic stem cells and migrates into the blood as immature B cells to reach peripheral tissue as naive B cells. Here, the B-cell receptor is activated and cells develop into short-term PCs during the primary response or enter the germinal center. Germinal-center B cells differentiate from centroblasts and centrocytes into long-term end-stage circulating memory cells or PCs that migrate to tissue survival niches and differentiate into immobile mature PCs. Lower panel: The earliest clonotypic cells were exclusively identified in the CD38- memory B-cell compartment, suggesting a precursor and a myeloma hierarchy that includes circulating memory cells or PCs that migrate to tissue survival niches and differentiate into mature premalignant PCs, giving rise to MGUS. Within this neoplasia, later genetic changes yield a range of myeloma-initiating cells that drives the propagation of a medullary neoplasia at multiple sites that is clinically known as MM. Ultimately, evolution continues to select niche-independent PCs that circulate, resulting in the extramedullary growth of myeloma subclones and advanced disease stages clinically known as extramedullary MM, PC leukemia, and HMCL.

1454

haematologica | 2016; 101(12)

Operational terms for study of the myeloma stem cell

plasticity, and drug resistance.71,72,68,69 Since most end-stage myeloma cells are short-lived,73 MMSCs are thought to continuously generate myeloma daughter cells by precisely balancing self-renewal and differentiation. In theory, a CSC can accomplish this balance via asymmetric cell division in which it divides to generate one progeny pool with a stem-cell phenotype (self-renewal) and another progeny pool that differentiates and gives rise to end-stage tumor cells.74,75 However, stem cells can also use symmetric divisions, defined as the generation of daughter cells that acquire identical phenotypes, to self-renew or to generate differentiated progeny.75 Stem cells are thus defined by their capacity to generate more stem cells and differentiated daughter cells, rather than by symmetric production of a stem cell and a differentiated daughter at each division.

The picture is even more complex in MM as some patients harbor heterogeneous populations of PCs that were most likely initiated from different populations of MMSCs within the myeloma hierarchy after many genetic lesions.

Such heterogeneity may be explained by an alternative stochastic model in which all myeloma cells have the potential to self-renew, yet experience a varying probability of entering the cell cycle and finding an environment that supports subclonal evolution and heterogeneity.

Future studies should therefore be based on genome-wide fingerprinting of clonal heterogeneity in order to identify factors that drive stepwise malignant transformation from normal PCs to MGUS, medullary MM, extramedullary MM, PC leukemia, and HMCL. Data from analyses of selected tissues and samples76-79 will qualify biobank material for future studies of the deregulation of normal selfrenewal pathways.

Novel technologies and potential strategies include single-cell analyses,79,80,36,49,55 transgenic mouse models54,81-82 in addition to the xenogene immuno compromised SCID-hu models,53 syngeneic BALB/c plasmacytoma or the 5T serie,83 and investigations of oncogene transformation in primary organoid miniature tissue culture.84-87 Bmi-1, Notch, Hedgehog, and Wnt, which were initially identified based on their roles in tumor formation, have been shown to be involved in the regulation of self-renewal in normal stem cells in many tissues.88-90 Since myeloma cells are influenced by the host, the microenvironment may play a key role in the initiation of myeloma and associated phenotypic changes—a phenomenon called “plasticity”, defined as altered cellular phenotype and function during deregulated differentiation. Of interest, this refers to malignant mature B cells that share features of different maturation steps, including precursors. Plasticity in MM is perhaps best illustrated by the subtyping of clinical tumor samples based on B-cell subset-associated gene signatures; tumors previously assigned to PreB-II and memory-cell subtypes of malignant PCs were associated with inferior prognoses.57-60 This observation provides a new tool for generating insight into the stages of clonal plasticity associated with oncogenesis and deregulated differentiation. The mechanisms of myelomacell plasticity should be exploited, and their significance for the concept of MMSCs assessed.

Since a major group of patients suffer from disease recurrence or clinical relapse after chemotherapy, MM is thought to be a consequence of molecular resistance mechanisms that protect the MMSC compartments. The idea that resistant MMSCs are the source of post-therapeutic recurrence is not a new one; it was first described in haematologica | 2016; 101(12)

studies of the stem-cell hierarchy and the self-renewal gene expression signature in leukemia with poor clinical outcome, which was also suggested to be associated with drug resistance.91-94 In myeloma, it has recently been documented that the level of drug resistance is a function related to the cellular hierarchy95 and its active or dormant stage96 that we need to identify and target to overcome it.

These findings highlight the potential to develop predictive, drug-specific sensitivity assays. We have taken the first step toward defining gene signatures of drug-specific resistance in pre-clinical models of HMCL.57-60 Self-renewal, plasticity, and resistance have been studied in the preclinical model of HMCL, which is considered to be the most advanced and homogenous myeloma tissue available. It is important to recognize that each HMCL reflects the end stage of an individual patient’s genetic evolution and selection, but each HMCL also reflects the aggregation of stepwise oncogenic events over time, some of which may have deregulated the hallmarks of MMSCs. Although we acknowledge that this model may be irrelevant for studies of the hierarchical model, it is a tool for identifying potential markers for MMSC. Such findings should be traced back through the myeloma hierarchy in prospective qualified clinical myeloma cases, as exemplified by the MSCNET single-cell approach.49

Biological assays of MMSCs In a conceptual context, classical stem-cell assays capture the phenomenon of a subpopulation that can propagate malignant clones indefinitely, and produce overt myeloma in vivo — the MMSCs. In an operational context, these assays indirectly seek to detect MMSCs as engrafting myeloma-initiating cells (in vivo), long-term culture-initiating cells (in vitro), or short-term sphere-forming cells (in vitro) as described in Table 1.

To date, immunodeficient mice have served as the most sensitive recipients for the growth, detection, and quantification of MMSC. Several xenografted mouse models enabled successful detection of the malignant regeneration of myeloma, usually measured via limiting dilution to identify the low frequent cells. However, the ability or failure of a cell compartment to produce myeloma in a transplanted mouse may not directly reflect the function of this compartment in patients. In this regard, the absence of the MMSC niche is a major limiting factor, as most samples include myeloma cells that have not yet completely acquired the ability to grow autonomously. To address this limitation, humanized and genetically modified mice have been designed as the state of the art fundament for functional studies of MMCS in myelomagenesis.8,12,39,81-83 Some studies have also evaluated MMSCs on the basis of their presumed self-renewal activity in vitro by investigating cells that initiate the sustained production of clonal myeloma cells when cultured in supportive conditions with or without stromal cells; these cells are usually subjected to analysis via limiting dilution. The generation of cellular spheres (as clusters or colonies in non-adherent liquid culture) constitutes a simple yet indirect strategy for identifying MMSCs in cell suspensions from myeloma tissue. It is unlikely that either of these in vitro systems fully replicates the three-dimensional structure and environment of myeloma in patients. On the contrary, it is likely that variables important for in vivo growth and self-renewal may not be present in in vitro investigations; the cells 1455

H.E. Johnsen et al.

under study may show no or selected growth, and/or may be anomalously and rapidly induced to differentiate to an end state without prior expansion. Therefore, in vitro assays may be useless in terms of clinically meaningful predictions.

Using HMCLs as a source of MMSCs for assays of growth in vitro or in transplanted mice is also problematic.

Although it is unlikely that these cells reflect the original genotype or origin of MMSCs, HMCLs contain cells that display the functions of MMSCs. Accordingly, considerable caution needs to be taken when formulating questions that will be addressed through the analysis of cells passaged in vitro. However, meta-analyses of HMCL responses to known anticancer drugs illustrate how these cells can be used to reveal associations between drug sensitivity and gene-expression profiling in cell lines from individual patients; these gene signatures of resistance have documented prognostic value.57,58

Single cells, single genes, single clones By the time MM is diagnosed, it consists of millions of

myeloma cells carrying genetic abnormalities that initiate malignant proliferation, and other mutations are acquired during disease evolution. Some of these secondary mutations emerge due to selective pressure and act as “drivers”; others may be “passengers” resulting from random mutational exposures or genomic instability during many cell divisions. In theory, this instability may yield one MMSC per driver lesion. It is likely that individual MM tumors have multiple MMSCs with different phenotypes that are closely linked to the deregulated functions of self-renewal, plasticity, and drug resistance.

Initial DNA sequencing studies76-78 have provided insight into mutational profiles in MM and have identified recurrent genes and potential molecular mechanisms responsible for MM initiation, maintenance, and progression. It is becoming clear that complexity beyond the landscape of mutations exists at the level of intraclonal heterogeneity, which directly affects disease progression and treatment resistance during clonal evolution and selection.69,72,76-78 Understanding these processes and characterizing these subclones will require investigations, of single or few plas-

Figure 2 Multi-steps of origin, initiation and the clinical spectrum of multiple myeloma.40 Left panel: Illustrates the normal B-cell subpopulations of interest in defining the cell of origin or the myeloma initiating cells. Current results support that a detectable, but rare, subpopulation of early oncogene-positive memory-like B cells in lymph node, blood, and bone marrow is descended from the cell of origin in the germinal center. These cells differentiate into premalignant PCs in the bone marrow, propagate through peripheral blood, and give rise to a benign neoplasia, clinically known as MGUS. Right panel: Within this benign neoplasia, later oncogenic events give rise to a range of myeloma-initiating cells that drives the propagation of medullary neoplasia at multiple sites, clinically known as MM. The ultimate selection of niche-independent cells results in extramedullary growth of myeloma subclones and advanced disease stages clinically known as extramedullary MM and PC leukemia. These advanced diseases are the origin of most HMCL. It is thought that transition through this clinical disease spectrum requires an inherited genetic background and the acquisition of genetic events that lead to expression of biological hallmarks that can be used for novel molecular disease classification systems, including drug resistance and plasticity.

1456

haematologica | 2016; 101(12)

Operational terms for study of the myeloma stem cell

ma cells,97-98 into the functional impact of specific genetic variants.

It will be challenging to combine gene identification and in vitro functional approaches in order to separate the true genetic drivers from the many passengers along the tracks of myelomagenesis. There is an increasing interest in defining the exact phylogeny of individual subclonal populations;80 patient-specific single-cell genetic profiling provides a potential resolution to this problem. Our original work in this area, which included the estimation of subset frequencies,33 36,45,46 also encompassed studies of differential oncogene expression in cDNA libraries from normal PCs and cases of MGUS, MM, extramedullary MM, and HMCL, as well as the prognostic impact of these differences. In brief, tissue samples were phenotyped via multiparametric flow cytometry, potential subclones were identified, and single cells (or a few cells) were sorted into individual wells containing lysis medium and followed via the global amplification of cDNA from each well. The quality of each library was confirmed via a set of highly targeted QRT-PCRs for chimeric gene fusions, deregulated genes, and singlenucleotide variants (for example, in the IgH locus). These patient-specific libraries were associated with parameters such as tissue type, disease stage, sample site, therapy, and outcome.97,98 This strategy, which has been automated and optimized to have a low error rate, has been integrated into high-throughput platforms that are useful for DNA and RNA sequencing; this may yield insight into subclonal genetic architectures and phylogenies in MM.13,75-78 We propose to combine this state of the art technique with an operational context as summarized in Table 1, in order to better define the phylogenetic relationships among clonal populations in myeloma at clinical presentation, during follow-up, and at relapse. This strategy will enable us to generate quantitative measures of stem-cell activities and functions at the level of single subclones by assessing self-renewal via the analysis of gene-expression signatures,3,4 plasticity via subtyping by B-cell subset-associated gene signatures,63,64 and drug resistance via the assignment of gene signatures of drug resistance.57-60

Summary and perspective The term “CSC” captures the idea that a stable, minor, quiescent, and phenotypically definable subpopulation exists within the malignant tissue. This subpopulation has the potential to self-renew and to enhance tumor resistance to toxic stress, thereby propagating the cancer for prolonged or even unlimited periods of time.

This idea is supported by phenomenology, and it plays a major role in our understanding of the pathogenesis of cancer. Although this concept is also widely appreciated for MMSCs, controversy persists regarding the identification and their origin, selection, plasticity, phenotype(s), and het-

References 1. Clarke MF, Dick JE, Dirks PB, et al. Cancer stem cells - perspectives on current status and future directions: AACR Workshop on cancer stem cells.

Cancer Res. 2006;66(19):9339-9344.

2. Dick JE. Stem cell concepts renew cancer

haematologica | 2016; 101(12)

erogeneity in various stages of MM. The existence of MMSCs, and whether they can be documented as a welldefined and characterized entity in MM remains to be demonstrated — and it may never be. However, introducing a more operational context (as suggested in Table 1) into our descriptions may enable the acquisition of data that indirectly support the existence of MMSCs and allow clinical validation of their impact; for example, via targeted therapy. Here we have summarized results from recent experimental work within and outside MSCNET, with a focus on the identification, isolation, and characterization of MMSCs. To date, these results support the hypothesis that myeloma-initiating cells are present in the malignant PC compartment, but the cell of origin is a normal counterpart of a germinal-center B cell that differentiates into a premalignant PC compartment identified in MGUS as indicated in Figure 1. This is consistent with our current understanding of the pathogenesis of MM as a multistep, cellular, hierarchical and linear process of disease initiation, evolution, selection and clinical presentation as illustrated in Figure 2.

A revision of the MMSC concept should include operational terms as described in Table 1 that enable the design of research plans based on prospectively sampled biomaterial that reflects the clinical disease spectrum. We anticipate that such a revision will lead to the acquisition of data that indirectly documents the existence of multiple MMSCs. The ultimate validation of their existence may then be achieved via targeted therapy in clinical trials.

The revision will allow us to identify a range of specific genetic events in various B-cell and PC subsets, and to design research activities focused on targeted therapy,99,100 based on prospectively sampled biomaterial from all myeloma subtypes classified at diagnosis and during follow-up. This strategy is in accordance with recent consensus statements on the definitions, assays, and nomenclature of CSCs, including a more operational nomenclature that achieves technical precision without completely abolishing established terminology.1,20 Funding This work on MMSC and myeloma pathogenesis and classification was supported by research funding from the EU 6th FP to MSCNET (LSHC-CT-2006-037602), the Danish Research Agency (grants no. 99 00 771, 271-05-0286, 271-05-0537, and 22-00-0314; to CHEPRE, #2101-07-0007), the Danish Cancer Society (grants no. DP 07014, DR 07017, and DP 9810009), Novo Nordic Foundation (Senior Fellowship 2001-4), KE Jensen Foundation (2005-2013), the Multiple Myeloma Research Foundation (senior grant 2003-4, contract no. 14).

The Myeloma Stem Cell Network is a consortium of nine partners within the European Myeloma Network collaborating on studies of cancer stem cells in multiple myeloma. The present report is prepared by active partners in work package 1. The partners in the European Myeloma Network and MSCNET are listed at http://www.myeloma-europe.org/.

research. Blood. 2008;112(13):4793-807.

3. Eppert K, Takenaka K, Lechman ER et al.

Stem cell gene expression programs influence clinical outcome in human leukemia.

Nat Med. 2011;17(9):1086-1093.

4. Greaves M. Cancer stem cells renew their impact. Nat Med. 2011;17(9):1046-1048.

5. Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that

originates from a primitive hematopoietic cell. Nat Med. 1997;3:730−737.

6. Anderson K, Lutz C, van Delft FW et al.

Genetic variegation of clonal architecture and propagating cells in leukaemia. Nature.

2011;469(7330):356-361.

7. Cobaleda C, Gutiêrrez-Cianca N, P erezLosada J, et al. A primitive hematopoietic cell is the target for the leukemic transfor-

1457

H.E. Johnsen et al.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

1458

mation in human philadelphia-positive acute lymphoblastic leukemia. Blood.

2000;95(3):1007−1013.

Yaccoby S, Epstein J. The proliferative potential of myeloma plasma cells manifest in the SCID-hu host.

Blood.

1999;94(10):3576-3582.

Guikema JE, Vellenga E, Bakkus MH, Bos NA. Myeloma clonotypic B cells are hampered in their ability to undergo B-cell differentiation in vitro. Br J Haematol. 2002; 119(1):54-61.

Matsui W, Huff CA, Wang Q et al.

Characterization of clonogenic multiple myeloma cells. Blood. 2004;103(6): 2332– 2336.

Huff CA, Matsui W. Multiple myeloma cancer stem cells. J Clin Oncol. 2008;26(17) :2895-2900.

Kim D, Park CY, Medeiros BC, Weissman IL. CD19-CD45 low/- CD38 high/CD138+ plasma cells enrich for human tumorigenic myeloma cells. Leukemia. 2012;26(12): 2530-2537.

Dey SS, Kester L, Spanjaard B, Bienko M, van Oudenaarden A. Integrated genome and transcriptome sequencing of the same cell. Nat Biotechnol. 2015;33(3):285-289.

Zong C, Lu S, Chapman AR, Xie XS. Genome-wide detection of singlenucleotide and copy-number variations of a single human cell. Science. 2012;338 (6114):1622-1626.

Streets AM, Zhang X, Cao C et al.

Microfluidic single-cell whole-transcriptome sequencing. Proc Natl Acad Sci USA. 2014;111(19):7048-7053.

Baslan T, Kendall J, Rodgers L et al.

Genome-wide copy number analysis of single cells. Nat Protoc. 2012;7(6):10241041.

Johnsen HE, Bøgsted M, Klausen TW et al.

Multiparametric flow cytometry profiling of neoplastic plasma cells in multiple myeloma. Cytometry B Clin Cytom.

2010;78(5):338-347.

Paiva B, Almeida J, Pérez-Andrés M et al.

Utility of flow cytometry immunophenotyping in multiple myeloma and other clonal plasma cell-related disorders. Cytometry B Clin Cytom. 2010;78(4):239-252.

Salonen J, Rönnholm G, Kalkkinen N, Vihinen M. Proteomic changes during B cell maturation: 2D-DIGE approach. PLoS One. 2013;8(10):e77894.

Valent P, Bonnet D, De Maria R et al.

Cancer stem cell definitions and terminology: the devil is in the details. Nat Rev Cancer. 2012;12(11):767-775.

Clements WK, Traver D. Signalling pathways that control vertebrate haematopoietic stem cell specification. Nat Rev Immunol. 2013;13(5):336-348.

Cooper MD. The early history of B cells.

Nat Rev Immunol. 2015;15(3):191-197.

Luckey CJ, Bhattacharya D, Goldrath AW, Weissman IL, Benoist C, Mathis D. Memory T and memory B cells share a transcriptional program of self-renewal with long-term hematopoietic stem cells.

Proc Natl Acad Sci U S A. 2006;103(9):33043309.

Shlomchik MJ, Weisel F. Germinal center selection and the development of memory B and plasma cells. Immunol Rev. 2012;247(1):52-63.

Matthias P, Rolink AG. Transcriptional networks in developing and mature B-cells.

Nat Rev Immunol. 2005;5(6):497–508.

Klein U, Dalla-Favera R. Germinal centres: role in B-cell physiology and malignancy.

Nat Rev Immunol. 2008;8(1):22–33.

27. McHeyzer-Williams M, Okitsu S, Wang N, et al. Molecular programming of B-cell memory. Nat Rev Immunol. 2011;12(1): 24–34.

28. Allen CD, Okada T, Cyster JG. Germinalcenter organization and cellular dynamics.

Immunity. 2007;27(2):190–202.

29. Muramatsu M, Kinoshita K, Fagarasan S, et al. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell. 2000;102(5):553–563.

30. Jackson SM, Harp N, Patel D et al. Key developmental transitions in human germinal center B cells are revealed by differential CD45RB expression.

Blood.

2009;113(17):3999-4007.

31. Tooze RM. A replicative self-renewal model for long-lived plasma cells: questioning irreversible cell cycle exit. Front Immunol. 2013;4:460.

32. González D, van der Burg M, García-Sanz R et al. Immunoglobulin gene rearrangements and the pathogenesis of multiple myeloma. Blood. 2007;110(9):3112-3121.

33. Rasmussen T, Lodahl M, Hancke S, Johnsen HE. In Multiple Myeloma Clonotypic CD38-/CD19+/CD27+ Memory B Cells Recirculate Through Bone Marrow, Peripheral Blood and Lymph Nodes. Leuk Lymphoma. 2004;45(7):1413-1417.

34. Billadeau D, Ahmann G, Greipp P, Van NB. The bone marrow of multiple myeloma patients contains B cell populations at different stages of differentiation that are clonally related to the malignant plasma cell. J Exp Med. 1993;178(3):1023-1031.

35. Corradini P, Boccadoro M, Voena C, Pileri A. Evidence for a bone marrow B cell transcribing malignant plasma cell VDJ joined to C mu sequence in immunoglobulin (IgG)- and IgA-secreting multiple myelomas. J Exp Med. 1993;178(3):1091-1096.

36. Rasmussen T, Jensen L, Johnsen HE. The Clonal Hierachy in Multiple Myeloma.

Acta Oncol. 2001;39(7):765-770.

37. Potten CS, Loeffler M. Stem cells: attributes, cycles, spirals, pitfalls and uncertainties. Lessons for and from the crypt.

Development. 1990;110(4):1001−1020.

38. Pilarski LM, Giannakopoulos NV, Szczepek AJ, Masellis AM, Mant MJ, Belch AR. In multiple myeloma, circulating hyperdiploid B cells have clonotypic immunoglobulin heavy chain rearrangements and may mediate spread of disease. Clin Cancer Res. 2000;6(2):585-596.

39. Pilarski LM, Hipperson G, Seeberger K et al.

Myeloma progenitors in the blood of patients with aggressive or minimal disease: engraftment and self-renewal of primary human myeloma in the bone marrow of NOD SCID mice.

Blood.

2000;95(3):1056-1065.

40. Kuehl WM, Bergsagel PL. Multiple myeloma: Envolving genetic events and host interactions. Nat Rev Cancer. 2002;2(3): 175-187.

41. Bergsagel PL, Kuehl WM, Zhan F et al.

Cyclin D dysregulation: an early and unifying pathogenic event in multiple myeloma.

Blood. 2005;106(1):296-303.

42. Bergsagel PL, Kuehl M. Chromosome translocation in multiple myeloma.

Oncogene. 2001;20(40):5611-5622.

43. Vescio RA, Cao J, Hong CH et al. Myeloma Ig heavy chain V region sequences reveal prior antigenic selection and marked somatic mutation but no intraclonal diversity. J Immunol. 1995;155(5):2487-2497.

44. Bakkus MH, Heirman C, Van RI, Van CB, Thielemans K. Evidence that multiple myeloma Ig heavy chain VDJ genes contain somatic mutations but show no intraclonal variation. Blood. 1992;80(9):2326-2335.

45. Rasmussen T, Jensen L, Honoré L, Johnsen HE. Frequency and kinetics of polyclonal and clonal B-cells in the peripheral blood of patients being treated for multiple myeloma. Blood. 2000;96(13):4357-4359.

46. Rasmussen T, Jensen L, Johnsen HE. The CD19 compartment in myeloma includes a population of clonal cells persistent after high-dose treatment. Leuk Lymphoma.

2002;43(5):1075-1077.

47. Pfeifer S, Perez-Andres M, Ludwig H, Sahota SS, Zojer N. Evaluating the clonal hierarchy in light-chain multiple myeloma: implications against the myeloma stem cell hypothesis. Leukemia. 2011;25(7):12131216.

48. Thiago LS, Perez-Andres M, Balanzategui A et al. Circulating clonotypic B cells in multiple myeloma and monoclonal gammopathy of undetermined significance.

Haematologica. 2014;99(1):155-162.

49. Rasmussen T, Haaber J, Dahl IM, et al.

Identification of translocation products but not K-RAS mutations in memory B cells from patients with multiple myeloma.

Haematologica. 2010;95(10):1730-1737.

50. Paino T, Ocio EM, Paiva B, et al. CD20 positive cells are undetectable in the majority of multiple myeloma cell lines and are not associated with a cancer stem cell phenotype. Haematologica. 2012;97(7):11101114.

51. Paíno T, Sarasquete ME, Paiva B, et al.

Phenotypic, genomic and functional characterization reveals no differences between CD138++ and CD138low subpopulations in multiple myeloma cell lines. PLoS One. 2014 ;9(3):e92378.

52. Van Valckenborgh E, Matsui W, Agarwal P, et al. Tumor-initiating capacity of CD138and CD138+ tumor cells in the 5T33 multiple myeloma model.

Leukemia.

2012;26(6):1436-1439.

53. Hosen N. Multiple myeloma-initiating cells. Int J Hematol. 2013;97(3):306-312.

54. Tanno T, Lim Y, Wang Q, et al. Growth differentiating factor 15 enhances the tumorinitiating and self-renewal potential of multiple myeloma cells. Blood. 2014;123(5): 725-733.

55. Christensen JH, Jensen PV, Kristensen IB, et al. Characterization of potential CD138 negative myeloma "stem cells".

Haematologica. 2012;97(6):e18-20.

56. Schmidt-Hieber M, Paivia B, Perez-Andres M, et al. CD56+ Clonal Plasma Cells In Multiple Myeloma Are Associated With Unique Disease Characteristics and Have a Counterpart Of CD56+ Normal Plasma Cells With Increased Maturity. Blood.2013; 122(21):751.

57. Bøgsted M, Holst JM, Fogd K, et al.

Generation of a predictive melphalan resistance index by drug screen of B-cell cancer cell lines. PLoS One. 2011;6(4): e19322.

58. Bøgsted M, Bilgrau AE, Wardell CP, et al.

Proof of the concept to use a malignant B cell line drug screen strategy for identification and weight of melphalan resistance genes in multiple myeloma. PLoS One. 2013;8(12):e83252.

59. Falgreen S, Laursen MB, Bødker JS, et al.

Exposure time independent summary statistics for assessment of drug dependent cell line growth inhibition. BMC

haematologica | 2016; 101(12)

Operational terms for study of the myeloma stem cell Bioinformatics. 2014;15:168.

60. Laursen MB, Falgreen S, Bødker JS, et al.

Human B-cell cancer cell lines as a preclinical model for studies of drug effect in diffuse large B-cell lymphoma and multiple myeloma. Exp Hematol. 2014;42(11):927938.

61. Kjeldsen MK, Perez-Andres M, Schmitz A, et al. Multiparametric flow cytometry for identification and fluorescence activated cell sorting of five distinct B-cell subpopulations in normal tonsil tissue. Am J Clin Pathol. 2011;136(6):960-969.

62. Bergkvist KS, Nyegaard M, Bøgsted M, et al. Validation and implementation of a method for microarray gene expression profiling of minor B-cell subpopulations in man. BMC Immunol. 2014;15:3.

63. Johnsen HE, Bergkvist KS, Schmitz A, et al.

Cell of origin associated classification of Bcell malignancies by gene signatures of the normal B-cell hierarchy. Leuk Lymphoma.

2014;55(6):1251-1260.

64. Johnsen HE, Bødker JS, Schmitz A, et al. A Multiple Myeloma Classification System That Associates Normal Bone Marrow BCell Subset Phenotypes with Disease Stage and Prognosis. Blood. 2014;124(21):3352.

65. Paiva B, Pérez-Andrés M, Vídriales MB, et al. Competition between clonal plasma cells and normal cells for potentially overlapping bone marrow niches is associated with a progressively altered cellular distribution in MGUS vs myeloma. Leukemia.

2011;25(4):697-706.

66. Perez-Andres M, Paiva B, Nieto WG, Caraux A, et al. Human peripheral blood Bcell compartments: a crossroad in B-cell traffic. Cytometry B Clin Cytom. 2010;78 Suppl 1:S47-60.

67. Ross FM, Avet-Loiseau H, Ameye G, et al.

Report from the European Myeloma Network on interphase FISH in multiple myeloma and related disorders.

Haematologica. 2012;97(8):1272-1277.

68. Caraux A, Klein B, Paiva B, et al. Circulating human B and plasma cells. Age-associated changes in counts and detailed characterization of circulating normal CD138- and CD138+ plasma cells. Haematologica.

2010;95(6):1016-1020.

69. Johnsen HE, Kjeldsen MK, Urup T, et al.

Cancer stem cells and the cellular hierarchy in haematological malignancies. Eur J Cancer. 2009;45 Suppl 1:194-201.

70. Rawstron AC, Orfao A, Beksac M, et al.

Report of the European Myeloma Network on multiparametric flow cytometry in multiple myeloma and related disorders.

Haematologica. 2008;93(3):431-438.

71. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell.

2011;144(5):646-674.

72. O'Brien CA, Kreso A, Jamieson CH. Cancer stem cells and self-renewal. Clin Cancer Res. 2010;16(12):3113-3120.

73. Cenci S, van Anken E, Sitia R.

haematologica | 2016; 101(12)

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

Proteostenosis and plasma cell pathophysiology. Curr Opin Cell Biol. 2011;23(2):216222.

Morrison SJ, Kimble J. Asymmetric and symmetric stem-cell divisions in development and cancer.

Nature.

2006;441(7097):1068-1074.

Keats JJ, Chesi M, Egan JB, et al. Clonal competition with alternating dominance in multiple myeloma.

Blood.

2012;120(5):1067-1076.

Egan JB, Shi CX, Tembe W, et al. Wholegenome sequencing of multiple myeloma from diagnosis to plasma cell leukemia reveals genomic initiating events, evolution, and clonal tides.

Blood.

2012;120(5):1060-1066.

Walker BA, Wardell CP, Melchor L, et al.

Intraclonal heterogeneity and distinct molecular mechanisms characterize the development of t(4;14) and t(11;14) myeloma. Blood. 2012;120(5):1077-1086.

Walker BA, Wardell CP, Melchor L, et al.

Intraclonal heterogeneity is a critical early event in the development of myeloma and precedes the development of clinical symptoms. Leukemia. 2014;28(2):384-390.

Moignard V, Macaulay IC, Swiers G, et al.

Characterization of transcriptional networks in blood stem and progenitor cells using high-throughput single-cell gene expression analysis. Nat Cell Biol.

2013;15(4):363-372.

Melchor L, Brioli A, Wardell CP, et al.

Single-cell genetic analysis reveals the composition of initiating clones and phylogenetic patterns of branching and parallel evolution in myeloma. Leukemia.

2014;28(8):1705-1715.

Vicente-Dueñas C, Romero-Camarero I, González-Herrero I, et al. A novel molecular mechanism involved in multiple myeloma development revealed by targeting MafB to haematopoietic progenitors.

EMBO J. 2012;31(18):3704-3717.

Asai T, Hatlen MA, Lossos C, et al.

Generation of a novel, multi-stage, progressive, and transplantable model of plasma cell neoplasms. Sci Rep. 2016;6:22760.

Mitsiades CS, Anderson KC, Carrasco DR. Mouse models of human myeloma.

Hematol Oncol Clin North Am. 2007;21(6):1051-1069, viii.

McMillin DW, Negri JM, Mitsiades CS. The role of tumour-stromal interactions in modifying drug response: challenges and opportunities.

Nat Rev Drug Discov.

2013;12(3):217-228.

Li X, Nadauld L, Ootani A, et al. Oncogenic transformation of diverse gastrointestinal tissues in primary organoid culture. Nat Med. 2014;20(7):769-777.

Cantrell MA, Kuo CJ. Organoid modeling for cancer precision medicine. Genome Med. 2015;7(1):32.

Gracz AD, Williamson IA, Roche KC, et al.

A high-throughput platform for stem cell

niche co-cultures and downstream gene expression analysis. Nat Cell Biol.

2015;17(3):340-349.

88. Reya T, Morrison S, Clarke M, Weissman I. Stem cells, cancer, and cancer stem cells.

Nature. 2001;414(6859):105−111.

89. Takebe N, Harris PJ, Warren RQ, Ivy SP. Targeting cancer stem cells by inhibiting Wnt, Notch, and Hedgehog pathways. Nat Rev Clin Oncol. 2011;8(2):97-106.

90. Takebe N, Miele L, Harris PJ, et al.

Targeting Notch, Hedgehog, and Wnt pathways in cancer stem cells: clinical update.

Nat Rev Clin Oncol. 2015;12(8):445-464.

91. Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med. 1997;3(7):730-7.

92. Shlush LI, Zandi S, Mitchell A, et al.

Identification of pre-leukaemic haematopoietic stem cells in acute leukaemia. Nature. 2014;506(7488):328333.

93. Eppert K, Takenaka K, Lechman ER, et al.

Stem cell gene expression programs influence clinical outcome in human leukemia.

Nat Med. 201;17(9):1086-1093.

94. Potter NE, Ermini L, Papaemmanuil E, et al.

Single-cell mutational profiling and clonal phylogeny in cancer. Genome Res. 2013;23(12):2115-2125.

95. Leung-Hagesteijn C, Erdmann N, Cheung G, et al. Xbp1s-negative tumor B cells and pre-plasmablasts mediate therapeutic proteasome inhibitor resistance in multiple myeloma. Cancer Cell. 2013;24(3):289-304.

96. Lawson MA, McDonald MM, Kovacic N, et al. Osteoclasts control reactivation of dormant myeloma cells by remodelling the endosteal niche. Nat Commun. 2015; 6:8983.

97. Rasmussen T, Kuehl M, Lodahl M, Johnsen HE, Dahl IM. Possible roles for activating RAS mutations in the MGUS to MM transition and in the intramedullary to extramedullary transition in some plasma cell tumors. Blood. 2005;105(1):317323.

98. Rasmussen T, Theilgaard-Mönch K, Hudlebusch HR, Lodahl M, Johnsen HE, Dahl IM. Occurrence of dysregulated oncogenes in primary plasma cells representing consecutive stages of myeloma pathogenesis: indications for different disease entities. Br J Haematol. 2003;123(2): 253-262.

99. Walker BA, Boyle EM, Wardell CP, et al.

Mutational Spectrum, Copy Number Changes, and Outcome: Results of a Sequencing Study of Patients With Newly Diagnosed Myeloma. J Clin Oncol.

2015;33(33):3911-3920.

100.Paíno T, Paiva B, Sayagués JM, et al.

Phenotypic identification of subclones in multiple myeloma with different chemoresistant, cytogenetic and clonogenic potential. Leukemia. 2015;29(5):1186-1194.

1459

REVIEW ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

‘Trained immunity’: consequences for lymphoid malignancies Wendy B.C. Stevens,1 Mihai G. Netea,2,3 Arnon P. Kater4 and Walter J.F.M. van der Velden1,3 Department of Hematology, Radboud University Medical Centre, Nijmegen; Department of Internal Medicine, Radboud University Medical Centre, and Radboud Center for Infectious Diseases, Nijmegen; 3Radboud Institute of Health Sciences, Radboud University Medical Center, Nijmegen and 4Department of Hematology, Lymphoma and Myeloma Center Amsterdam (LYMMCARE) Academic Medical Center, University of Amsterdam, The Netherlands

1 2

Haematologica 2016 Volume 101(12):1460-1468

ABSTRACT

I

Correspondence: walter.vandervelden@radboudumc.nl

Received: May 15, 2016.

Accepted: June 29, 2016.

Pre-published: November 10, 2016. doi:10.3324/haematol.2016.149252

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/100/12/1460

©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.

1460

n hematological malignancies complex interactions exist between the immune system, microorganisms and malignant cells. On one hand, microorganisms can induce cancer, as illustrated by specific infection-induced lymphoproliferative diseases such as Helicobacter pylori-associated gastric mucosa-associated lymphoid tissue lymphoma.

On the other hand, malignant cells create an immunosuppressive environment for their own benefit, but this also results in an increased risk of infections. Disrupted innate immunity contributes to the neoplastic transformation of blood cells by several mechanisms, including the uncontrolled clearance of microbial and autoantigens resulting in chronic immune stimulation and proliferation, chronic inflammation, and defective immune surveillance and anti-cancer immunity. Restoring dysfunction or enhancing responsiveness of the innate immune system might therefore represent a new angle for the prevention and treatment of hematological malignancies, in particular lymphoid malignancies and associated infections. Recently, it has been shown that cells of the innate immune system, such as monocytes/macrophages and natural killer cells, harbor features of immunological memory and display enhanced functionality long-term after stimulation with certain microorganisms and vaccines. These functional changes rely on epigenetic reprogramming and have been termed ‘trained immunity’. In this review the concept of ‘trained immunity’ is discussed in the setting of lymphoid malignancies. Amelioration of infectious complications and hematological disease progression can be envisioned to result from the induction of trained immunity, but future studies are required to prove this exciting new hypothesis.

Introduction In order to combat infections and cancer the human body is equipped with an innate and adaptive immune system. The two systems are highly intertwined and closely collaborate with a bridging role for antigen-presenting cells e.g., dendritic cells. Their differences and commonalities are depicted in Table 1 (for more detailed reviews see references1-3). After birth, the immune system evolves from and is shaped by exposure to foreign antigens, which for the most part come from microbes belonging to the gut microbiota.4 Gradually, an immune armamentarium is build which, by virtue of memory properties, acts increasingly specifically, rapidly and efficiently. Immune memory has in the past been attributed solely to the adaptive immune system, but recently the paradigm has shifted with evidence showing that the innate immune system possesses the capacity for immunological memory, designated ‘trained immunity’.5 This feature is crucial for organisms that haematologica | 2016; 101(12)

Trained immunity in hematology

have no adaptive immune response, e.g., plants and invertebrates, but it also exists in humans, where it might be of special benefit in neonates that have yet to develop a mature adaptive immune repertoire.6 Importantly, 'training' of the innate immune system by the use of vaccination, resulting in increased immune activation, has been shown feasible. Hopefully these insights can be exploited in the near future for the design of new treatments for immunodeficiencies, infections and cancer.7 Currently, immunotherapies used for the treatment of hematological malignancies have focused on the adaptive immune system, mostly T and B lymphocyte responses.

Examples are numerous and include the use of allogeneic stem cell transplantation (SCT), monoclonal antibodies, immune checkpoint inhibitors and cellular therapies (e.g., adoptive T cell transfer).8-10 However, accumulating evidence has confirmed the significant impact of deregulated interactions between host innate immune cells (e.g., monocytes, dendritic cells, and NK cells) and microbes (e.g., chronic infections and dysbiosis) in the pathogenesis of cancer.2,11-13 This seems especially true for lymphoid malignancies, where antigenic stimulation by microbes and chronic inflammation drive lymphoproliferation and hence tumor progression.14 Moreover, the immunosuppressive environment that develops in many hematological malignancies, consists of a considerable part of functionally altered innate immune cells, including myeloidderived suppressor cells (MDSC), that cause immune evasion, progression and dissemination of neoplastic cells.15 Therefore, considering the pivotal role of the innate immune system in the initiation and progression of hematological malignancies, it might prove a valuable target for prevention and treatment. One option would be exploiting the recently discovered feature of innate immune memory which can be induced, for instance, by a Bacillus Calmette-Guérin (BCG) vaccination.5 By enhancing and restoring the function of innate immune cells, clearance of microbial and neoantigens can be achieved and the immunosuppressive tumor microenvironment reversed.

In addition, by innate immune training, the infection incidence might be reduced in the high-risk setting of cancer therapy.

In the review herein we summarize the current knowledge on the concept of ‘trained immunity’, and hypothesize ways of extending this concept to the field of lymphoid malignancies.

The concept of trained immunity: a novel type of immune memory For decades the prevailing assumption has been that immunological memory was a feature characterizing only the acquired immune system. However, recent studies have shown that the mammalian innate immune system also exhibits adaptive properties compatible with immunological memory, for which the term ‘trained immunity’ has been proposed.5 On reinfection or rechallenge with microbial ligands, prototypical innate immune cells such as monocytes/macrophages exhibit enhanced functionality with the release of pro-inflammatory cytokines and effector functions e.g., phagocytosis (Figure 1).16,17 The training of monocytes/macrophages is mediated by the activation of pattern-recognition receptors by microbe-associated molecular patterns (MAMPs) from bacteria and fungi; for example dectin-1 by β-glucan, and nucleotide-binding oligomerization domain-containing haematologica | 2016; 101(12)

protein 2 (NOD2) by components of the BCG vaccine.16-19 Several mechanisms are involved in the development of innate immune memory, among which epigenetic histone modifications (e.g., histone methylation and acetylation), and autophagy play a central role. Monocytes are functionally reprogrammed for either enhanced (training) or decreased (tolerance) cytokine production, depending on the type and concentration of MAMPs they encountered17,20-22 (Figure 1). The epigenetic reprogramming of monocytes after MAMP stimulation results in functional as well as morphological changes, including cell surface marker modifications such as upregulation of TLR expression.17,20 Autophagy contributes to the process of ‘trained immunity’ induced by a BCG vaccination, as pharmacological or genetic (autophagy gene polymorphisms) inhibition blocks the epigenetic programming of monocytes from occuring.23 Natural killer (NK) cells belonging to the innate immune system also display memory functions, mainly during viral infections. Expression of the Ly49H receptor and chemokine receptor CXCR6 seem to be of importance for the memory characteristics of NK cells.24,25 In addition, epigenetic modifications with changes in promoter methylation status are a hallmark of adaptive NK cells.26,27 The adaptive immune features of NK cells result in their longterm activation and protection from infection or reinfection with both herpesviruses and also influenza.24-26,28 NK cell training can also be achieved by BCG vaccination with the induction of non-specific immune memory, as has been shown in healthy volunteers.29 ‘Trained immunity’ of innate immune cells provides protection against reinfection in a direct T/B-cell-independent manner. In addition, a more or less indirect effect on the acquired immune system may also occur as monocytes/macrophages and NK cells influence cells of the adaptive immune system.1 For instance, after BCG vaccination complex cytokine profiles are induced that include elevated T-helper (Th) 1 cytokines like IFNγ and also interleukin (IL)-4 (Th2), IL-17 (Th17), and IL-10.30-33 These responses belong to the adaptive immune responses that facilitate and contribute to infection control and vaccination efficacy.

Important hallmarks of ‘trained immunity’ are the nonspecific nature and long duration of the enhanced immune responses. Trained monocytes have been shown to circulate in the peripheral blood for up to 3 months after BCG vaccination, and NK responses, e.g., release of IFNγ, are also enhanced for months.16,29 Immune training of both monocytes and NK cells induces protection from infectious agents other than the primary stimulus. In vitro studies demonstrated an increased innate immune response against a plethora of pathogens after BCG vaccination, including Mycobacterium tuberculosis, Candida albicans, and Staphylococcus aureus.16,29 Moreover, in vivo data exist from large vaccination trials. In low birth weight children from Guinea-Bissau, vaccination with BCG resulted in reduced neonatal mortality, which was the result of a composite effect consisting of reduced neonatal sepsis, respiratory infections and fever.34 However, this protective effect did not result in overall reduced infant mortality, suggesting that it was most pronounced in neonates.6 An alternative explanation, however, might be the fact that the longevity of the training effect is in the order of months and probably not years, although data on the exact duration are lacking. In Danish cohorts similar results were also achieved 1461

W.B.C. Stevens et al.

with BCG and smallpox vaccinations, significantly reducing the risk of hospitalization for most subgroups of infectious diseases, especially respiratory tract infections.35,36 Intriguingly, non-specific protection might go beyond infections, as the BCG vaccination has been shown to be effective in treating carcinomas (e.g., bladder cancer or melanoma).37 Considering the protective effects of ‘trained immunity’ against infections, the concept might be exploited to improve care for cancer patients who suffer from infections. The concept might even be more important in malignancies whose pathogenesis entails exposure to microorganisms (chronic antigenic stimulation) and impaired innate and adaptive immune responses, hence hematological, but in particular, lymphoid malignancies.

Role for microbes in lymphoid malignancies Complex interactions between microorganisms, immune cells and cells exhibiting features of neoplastic transformation eventually determine the development of overt malignant disease. This intriguing interrelatedness of biological events is especially encountered in those circumstances where immune cells themselves are the cells at risk for malignant transformation, thus in the setting of lymphoid malignancies including multiple myeloma (MM), chronic lymphocytic leukemia (CLL) and malignant lymphomas. These diseases constitute ‘exemplary models’ in which the role of ‘trained immunity’ can be explored.

Infections in patients with lymphoid malignancies Patients with cancer suffer from an increased risk for common and opportunistic infections that result from disease intrinsic and anti-cancer therapy-induced immune deficiencies. The severity of these deficiencies and the specific nature of the defects determine which infections can be expected to occur in individual patients. An elaborate review of this topic is beyond the scope of this article and an overview can be found in a recent publication.38 Malignant lymphomas and CLL significantly increase the risk for serious infections.39,40 The cause of this increased risk for infections is multifactorial. Early on, the humoral immune abnormalities, resulting from B-lymphocyte dysfunction often accompanied by the emergence of hypogammaglobulinemia, dominate, and infections are caused mainly by viruses and encapsulated bacteria. With the progression of the lymphoid malignancies T cell dysfunction occurs, which is often therapy-related, but not exclusively, as early after diagnosis T cell defects/alterations are already present.41,42 Malignant B lymphocytes produce anti-inflammatory cytokines, like IL-10 and TGF-β, and indoleamine 2,3-dioxygenase that skew the balance from a Th1/Th17 towards a regulatory (FOXP3) T cell phenotype.43,44 Moreover, tumor cells express immunosuppressive ligands like programmed death-ligand 1 (PDL1) and PD-L2 that inhibit proliferation and activation of effector memory T cells.41,45 In addition, the recruitment of tumor supporting innate myeloid cells; including nurselike cells, tolerogenic DCs, tumor-associated macrophages (M2-polarized) and MDSCs; to the tumor micro-environment occurs, which enhance the state of tolerance.43,46-49 Through these mechanisms, malignant B cells, besides inhibiting anti-tumor immunity, also impair immune responses that are crucial for the control of both common and opportunistic pathogens. Treatment with anti-neo1462

Table 1. Characteristics of innate and adaptive immune responses.

Cellular Polymorphonuclear components cells, monocytes, macrophages, dendritic cells, NK cells, innate lymphoid cells Humoral Complement, host defense components peptides e.g., defensins, natural antibodies Time Early: minutes-hours-days to activation Recognition Non-specific or semi-specific (pattern-recognition receptors) Immune Epigenetic memory reprogramming: trained immunity Lasts for weeks to months

T-lymphocytes, B-lymphocytes Specific antibodies Late: days-weeks Highly specific (T-cell receptors and antibodies) Gene recombination and clonal expansion Lasts for years to decades

plastic drugs, including purine analogs, and monoclonal antibodies, further impair specific cellular immunity with an increased risk of opportunistic infections.

Infection-induced lymphoid malignancies The role of microorganisms in the pathogenesis of cancer is increasingly appreciated. Several classical pathogens have been implicated in the origin of specific hematologic malignancies. The Epstein-Barr virus is exemplary. It is responsible for the disease infectious mononucleosis, which has also been associated with a diverse range of malignant lymphomas in immunocompetent and immunocompromised patients e.g., Burkitt lymphoma, and post-transplant lymphoproliferative disorder. Other well-known examples are Helicobacter pylori, Borelia burgdorferi, Coxiella burnetii and the hepatitis C virus (HCV), which have been associated with marginal zone lymphoma (MZL) and diffuse large B-cell lymphomas (DLBCL).14,50 Some of these diseases can initially be treated with antimicrobial agents with a considerable chance for cure, despite the fact that they can be considered monoclonal lymphoproliferative disorders.51,52 Beyond specific pathogens, it has been shown that experiencing common community-acquired viral and bacterial infections early in life increases the risk for developing lymphoid malignancies later in life. Large epidemiological studies have shown clear associations between infections and monoclonal B-cell lymphocytosis (MBL) and CLL, non-Hodgkin lymphoma (NHL), and MM.53-55 More recently, the commensal bacterial flora of the gut (microbiota) has also been implicated in lymphomagenesis.12 The expression of restricted immunoglobulin gene repertoires/B-cell receptors (BCR) in CLL and malignant lymphomas underscores a key role of an antigenic drive in the initiation and perpetuation of lymphoproliferation; mainly microbial antigens but also self-antigens released on cell apoptosis.56-59 In CLL, over 30% of cases can be grouped together based on the expression of stereotypic BCRs with characteristic complementarity determining region 3 (CDR3) amino acid sequences. As CDR3s are most decisive for the antigen specificity of immunoglobulins (Igs), this strongly suggests that distinctive antigens haematologica | 2016; 101(12)

Trained immunity in hematology

Figure 1. Epigenetic regulation of memory-like activity of innate immune cells. The biological phenomena of endotoxin tolerance and trained immunity are depicted in the illustration. Stimulation of monocytes and/or macrophages via pattern recognition receptors (PRRs), such as Toll-like receptor 4 (TLR4) or dectin-1, leads to increased expression of pro-inflammatory genes. Decreased or increased responsiveness of these cells to subsequent PRR stimulation may then occur, depending on the initial type of stimulus that the cell received. TLR4 stimulation with lipopolysaccharide (LPS) can induce a state of endotoxin tolerance (represented by the red line), whereas the stimulation of dectin-1 with β glucan from Candida albicans leads to a state of trained immunity (represented by the blue line). Initial PRR stimulation leads to trimethylation of histone H3K4 (H3K4me3) and global acetylation of histone H4 on promoters of pro-inflammatory genes. In the case of endotoxin tolerance, the removal of the stimulus results in the loss of activating marks and gene expression returns to basal levels. Following a second encounter with the stimulus, ‘tolerized’ genes will not regain the H3K4me3 mark or acetylation, and will remain silent to stimulation. In the case of 'trained immunity', pro-inflammatory genes will retain enhancers marked with monomethylation of histone H3K4. A second stimulus will induce transcription factors to bind to the enhancers and promoters of these genes, thereby promoting an increase in H3K4me3 and thus the expression of ‘trained’ genes.22

are involved in the development of subsets of CLL. It was recently shown that a newly identified subset of CLL with mutated IgVH heavy chain, expresses stereotypic BCRs highly specific for β-(1,6)-glucan, a major antigenic determinant of yeast and molds.60 The clonal B-cells of these patients were shown to proliferate in response to β-(1,6)glucan, suggesting that the fungal microbiota can deliver functional ligands in the process of CLL. Mechanisms involved in the neoplastic transformation of lymphocytes by microorganisms include: the release of genotoxic metabolites, antigen-driven lymphoproliferation, induction of chronic inflammation, impaired apoptosis, inactivation of the tumor suppressor gene p53, disrupted DNA repair, mitochondrial dysfunction, and oxidative stress.12,14,61-63 Antigen-driven lymphoproliferation is implicated in the pathogenesis of CLL and NHL, including MZL, mantle cell lymphoma (MCL), follicular lymphoma, and DLBCL (Figure 2).64 However, B-cells express many other (surface) receptors involved in microbial antigen recognition in addition to the BCR, including CD5 and CD6 (CLL) and pattern recognition receptors like TLR1, TLR2, TLR6, TLR7, TLR9, NOD1 and NOD2 (CLL, NHL).65,66 Signaling through these receptors is exploited by the malignant cells for their own survival, precisely as shown for BCR activation. The TLRs expressed by CLL cells, for instance, are functional, as upon stimulation, the nuclear factor-κB signaling pathway becomes activated, protecting CLL cells from spontaneous apoptosis.65 haematologica | 2016; 101(12)

Cancer-induced immune-suppression; cancer progression and infections Many cancers facilitate their preservation and progression by protecting themselves against the host’s immune surveillance armamentarium by inducing an immunosuppressive environment. The acquired immune system seems most affected with pronounced deficits occurring in anti-tumor T cell responses resulting from several interrelated mechanisms, as described above, with an important role for the accumulation of immunosuppressive innate myeloid cells (e.g., MDSC).15 T cells from CLL patients exhibit deviant T cell subset distributions, and have functional defects, including impaired ability to form immunological synapses, decreased proliferative capacity and an impaired effector function.67,68 These functional defects coincide with an increased expression of CD244, CD160, and PD-1 on CLL-derived T cells, a phenotype that is similar to the phenotype of exhausted T cells in chronic viral infections.69 Targeting these immune checkpoints has been a new approach in the treatment of malignant lymphomas, and is now being explored in large clinical studies.70 Also relevant to anti-tumor immune responses are immune cells belonging to the innate immune systems, such as NK cells and monocytes/macrophages. Defects in these immune effectors that contribute to cancer initiation and progression have been increasingly described in lymphoid malignancies. For instance, NK cells found in the cir1463

W.B.C. Stevens et al.

Figure 2. Antigen-driven B cell receptor (BCR) signaling in lymphoproliferative diseases. Chronic stimulation of the BCR by microbial antigens and autoantigens results in the activation of several intracellular signaling pathways resulting in lymphoproliferation, reduced apoptosis, prolonged cell survival and disrupted cell migration that can contribute to the emergence of lymphoid malignancies. HCV: Hepatitis C virus; H. pylori: helicobacter pylori; S. pneumoniae: streptococcus pneumoniae; mTOR: mechanistic target of rapamycin; BLNK: B cell linker protein; BTK: Bruton's tyrosine kinase; PLCγ2: phospholipase C Gamma 2; PI3K: phosphatidylinositol 3-kinase; PIP3: phosphatidylinositol (3,4,5)-trisphosphate; IP3: inositol 1,4,5-trisphosphate; DAG: diacylglycerol; PDK: protein kinase D; Ca: calcium; JNK: Jun N-Terminal kinases; IKK: IκB kinase; PKC: protein kinase C; NFAT: nuclear factor of activated T-cells; NF-κB: nuclear factor kappa B; ATF2: activating transcription factor 2.

culation of CLL patients appear to have several functional defects, including impaired cytotoxic activity, possibly because of the defective expression of the NKG2D coreceptor.71 Moreover, the monocytic population of CLL patients has an altered composition and monocytes exhibit deregulation of genes that are involved in phagocytosis and inflammation.72 Higher numbers of non-classical CD14+CD16++ monocytes are present that are known to have immune suppressive features opposed to the classical CD14++CD16– counterparts. Intriguingly, monocytes from CLL patients exhibit the primary features of ‘endotoxin tolerance’, including low cytokine production, high phagocytic activity and impaired antigen presentation.73 The refractory state of these cells prohibits sufficient inflammatory responses to occur after pathogens and cancer cells (‘tumor tolerance’). This introduces a new mechanism of innate immune failure that contributes to the susceptibility of CLL patients to infections and tumor progression.73

Trained immunity in hematological malignancies Evidence for ‘trained immunity’ in cancer therapy Training the innate immune system seems feasible and 1464

effective, as for decades BCG vaccination has been successfully applied in the treatment of urothelial cell carcinomas and melanomas.37 Data from large epidemiological studies and vaccination trials reveal circumstantial evidence for a similar potential in hematological malignancies. In a previous Danish case-cohort study it was shown that BCG vaccination during infancy significantly reduces the risk of developing lymphoma’s (HR 0.49 (95% CI: 0.26–0.93)).74 Mechanisms proposed, attempting to explain this beneficial result, included the stimulating effect of vaccination on immune surveillance of cancer and the decreased incidence of infectious diseases involved in lymphoma pathogenesis. Older studies have tested the concept of the induction of anti-tumour immunity in patients with AML by vaccination.75,76 However, these studies were small, and although some small benefits were shown, these data preclude drawing definite conclusions. At least BCG vaccination in these patients did not result in untoward complications.

The use of β-glucan as an immune adjuvant in the treatment of solid and hematological malignancies has evoked considerable interest for many years, as the MAMP shows promising activity both in vitro and in vivo. The anti-tumor haematologica | 2016; 101(12)

Trained immunity in hematology

Figure 3. Concept of ‘trained immunity’ of the innate immune system in lymphoid malignancies. Training of innate immune cells, including monocytes/macrophages and NK cells, with microbial ligands results in enhanced effector functions of these cells. By facilitating antigen eradication and restoration of tumor surveillance and anti-tumor immunity, the succeeding events that would ultimately result in the development of lymphoid malignancies might be interrupted or at least delayed. In addition, enhanced immunity will also prevent and reduce infectious complications. Ig: immunoglobin; BCR: B cell receptor; TLR: toll-like receptor; MDP: muramyl dipeptide; BCG: Bacillus Calmette-Guérin; CMV: cytomegalovirus; NK: natural killer; PD-1L: programmed death-ligand 1; MDSC: myeloid-derived suppressor cells; Th: T helper; TREG: regulatory T cell; IL-10: interleukin 10; TGF-β: transforming growth factor-beta; NOD2: nucleotide-oligomerization domain-containing protein 2.

activity directly relates to signaling through dectin-1,77,78 suggesting that the innate and acquired immunity elicited by β-glucan could be of therapeutic value. In a small phase I/II study, twenty patients with advanced malignancies who were receiving chemotherapy were additionally treated with a β-(1,3)/(1,6)-D-glucan preparation.

Preliminary results showed that β-glucan was well tolerated in these patients and suggested a beneficial effect on hematopoiesis. Moreover, one patient with a chemotherapy-refractory malignant lymphoma achieved a partial response.79

Mechanisms of action The concept of ‘trained immunity’ can be explored as a new modality of cancer prevention and therapy, as well as in infection control. The concept seems appropriate to consider in the setting of lymphoid malignancies, and three major mechanisms can be envisioned to contribute to the effects of innate immune enhancement in this setting (Figure 3).

1. Decreasing antigen-driven lymphoproliferation: Many data support the hypothesis that malignant B cells, found in CLL and NHL e.g., MZL and MCL, resemble antigen-activated B cells and that ongoing antigeninduced modulation of cell responses occur in the context haematologica | 2016; 101(12)

of antigen recognition by the BCR and affiliated receptors (Figure 2). Inhibiting this signaling cascade has already proven to be of clinical use with the introduction of potent BCR inhibitors, e.g., ibrutinib, and therefore eradicating B-cell stimulating antigens and infectious antigens and autoantigens may also alter the course of lymphoproliferative diseases. This might be achieved by 'trained immunity', as monocytes and macrophages can be activated to prevent many infections that have been implicated in the pathogenesis of lymphoma. Since monocytes and macrophages are also pivotal in apoptotic cell clearance, training of these cells might contribute to autoantigen eradication.80,81 2. Reversing immune tolerance: Innate immunity Considering the contribution of monocytes that exhibit ‘endotoxin and tumor tolerance’ to tumor progression and infection risk, the reversal of their refractory state by ‘trained immunity’ might prove clinically beneficial. An increased tolerogenic state of monocytes and macrophages has clearly been implicated in lymphoid malignancies.82 Nevertheless, whether vaccination with BCG or β-glucan can restore the balance and skew tumor associated macrophages from the M2 towards a more M1 phenotype with beneficial anti-tumor characteristics, 1465

W.B.C. Stevens et al.

remains to be proven. However, recently data have shown that another type of immunosuppressive myeloid cell, i.e., MDSC, can be successfully reprogrammed by exposure to β-glucan, resulting in the loss of its immunosuppressive phenotype and gain of enhanced antigen presenting capacities.83 In addition, a therapeutic approach in which NK cell immunity can be enhanced, for instance by BCG vaccination, may lead to an improved function of NK cells, thereupon overcoming cancer-induced NK cell function defects.29,84

Acquired immunity General T cell dysfunction and exhaustion contribute to tumor tolerance and defective immune surveillance.

Training the innate immune system also influences T cell activity and skewing of acquired immune responses towards Th1 and Th17 phenotypes that contribute to antitumor immunity.

Generation of adaptive NK cells NK cells play a pivotal role during the treatment of hematological malignancies, and their activity might be enhanced by ‘training’. The adoptive transfer of ex vivo generated autologous and allogeneic NK cells is being explored for immunotherapy in the setting of lymphoid malignancies.85,86 Interestingly, considerable evidence exists for a close relationship between CMV infection post SCT and relapse of AML, in which NK cells are deemed to be the effectors. Moreover, CMV infection results in a mature phenotype of NK cells which also have memorylike features.87 Hence, the training of NK cells by exposure to BCG or CMV antigens and/or cytokines ex vivo, might be a novel approach for the optimization of NK cell based immunotherapies.26

How to apply “trained immunity” in hematology patients?

The prevention of lymphoid, and maybe, myeloid malignancies on the population might be achieved with immune training. Vaccination programs have underlined the effect of immune training and the impact on cancer development.74 However, additional studies are needed before standard vaccination can be introduced on a larger scale. But what about the individual patient? Can innate immune training alter the course of manifest hematologi-

References 1. Iwasaki A, Medzhitov R. Control of adaptive immunity by the innate immune system. Nat Immunol. 2015;16(4):343-353.

2. Vesely MD, Kershaw MH, Schreiber RD, Smyth MJ. Natural innate and adaptive immunity to cancer. Ann Rev Immunol.

2011;29(235-271.

3. Rossi M, Young JW. Human dendritic cells: potent antigen-presenting cells at the crossroads of innate and adaptive immunity. J Immunol. 2005;175(3):1373-1381.

4. Levy O. Innate immunity of the newborn: basic mechanisms and clinical correlates.

Nat Rev Immunol. 2007;7(5):379-390.

5. Netea MG, Quintin J, van der Meer JW.

1466

6.

7.

8.

9.

cal malignancies? Malignancies that rely on antigen-driven proliferation, hence lymphoid malignancies, seem, at least conceptually, the best candidates for innate immune training. The timing of immune training might, however, prove crucial. The efficacy of this approach is probably at the highest level early on in the disease process, when the tumor burden is low and before tumor evolution has resulted in BCR signaling, independent proliferation and profound immune exhaustion. Moreover, applying a live attenuated vaccine such as BCG, which is normally considered safe, might prove to be less safe in patients with severely impaired T cell immunity. Alternatives, such as gamma-irradiated BCG, may prove more desirable.88 Since these severe immune deficits occur in late stage disease, an early initiation of vaccinations must be pursued. Two conditions therefore seem ideal for testing the concept of innate immune training: MBL/early stage CLL and MZL. Training the immune system might be most feasible by applying the BCG vaccination. Considerable experience exists with this old vaccination strategy and safety is hardly an issue, even in vulnerable patients, including low birth weight children. Currently no standard immune adjuvant, which is based on β-glucan or MDP, exists, and considerable questions remain to be answered about the route of administration, dosage and safety of these antigens.

Conclusion Epigenetic reprogramming of innate immune cells results in enhanced non-specific immunity and immune memory.

‘Training’ of the innate immune system is a novel concept in immunology and infectious disease that may be exploited in hematological malignancies, especially lymphoid malignancies, where antigen-driven lymphoproliferation and immune impairments are the hallmarks of the disease.

Amelioration of infectious complications and cancer progression can be envisioned as goals of 'trained immunity' in cancer therapy. Acknowledging the current limited data on the effects of ‘trained immunity’, and the hypothetical status, thus far, of the concept in the setting of clinical hematology, further studies are mandatory.

Funding MGN was supported by an ERC Consolidator Grant (#310372).

Trained immunity: a memory for innate host defense. Cell Host Microbe. 2011; 9(5):355-361.

Levy O, Wynn JL. A prime time for trained immunity: innate immune memory in newborns and infants. Neonatology. 2014; 105(2):136-141.

Netea MG, Joosten LA, Latz E, et al.

Trained immunity: A program of innate immune memory in health and disease.

Science. 2016;352(6284):aaf1098.

Pardoll DM. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer. 2012;12(4):252-264.

Vanneman M, Dranoff G. Combining immunotherapy and targeted therapies in cancer treatment. Nat Rev Cancer.

2012;12(4):237-251.

10. Bachireddy P, Burkhardt UE, Rajasagi M, Wu CJ. Haematological malignancies: at the forefront of immunotherapeutic innovation. Nat Rev Cancer. 2015;15(4):201215.

11. de Martel C, Ferlay J, Franceschi S, et al.

Global burden of cancers attributable to infections in 2008: a review and synthetic analysis. Lancet Oncol. 2012;13(6):607-615.

12. Yamamoto ML, Maier I, Dang AT, et al.

Intestinal bacteria modify lymphoma incidence and latency by affecting systemic inflammatory state, oxidative stress, and leukocyte genotoxicity. Cancer Res. 2013;73(14):4222-4232.

13. Zitvogel L, Ayyoub M, Routy B, Kroemer

haematologica | 2016; 101(12)

Trained immunity in hematology

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

G. Microbiome and anticancer immunosurveillance. Cell. 2016;165(2):276-287.

Suarez F, Lortholary O, Hermine O, Lecuit M. Infection-associated lymphomas derived from marginal zone B cells: a model of antigen-driven lymphoproliferation.

Blood. 2006;107(8):3034-3044.

Senovilla L, Aranda F, Galluzzi L, Kroemer G. Impact of myeloid cells on the efficacy of anticancer chemotherapy. Curr Opin Immunol. 2014;30:24-31.

Kleinnijenhuis J, Quintin J, Preijers F, et al.

Bacille Calmette-Guerin induces NOD2dependent nonspecific protection from reinfection via epigenetic reprogramming of monocytes. Proc Natl Acad Sci USA. 2012;109(43):17537-17542.

Quintin J, Saeed S, Martens JH, et al.

Candida albicans infection affords protection against reinfection via functional reprogramming of monocytes. Cell Host Microbe. 2012;12(2):223-232.

Klinman DM, Conover J, Coban C. Repeated administration of synthetic oligodeoxynucleotides expressing CpG motifs provides long-term protection against bacterial infection. Infect Immun.

1999;67(11):5658-5663.

Ifrim DC, Joosten LA, Kullberg BJ, et al.

Candida albicans primes TLR cytokine responses through a Dectin-1/Raf-1-mediated pathway. J Immunol. 2013; 190(8):4129-4135.

Ifrim DC, Quintin J, Joosten LA, et al.

Trained immunity or tolerance: opposing functional programs induced in human monocytes after engagement of various pattern recognition receptors. Clinical Vaccine Immunol. 2014;21(4):534-545.

Saeed S, Quintin J, Kerstens HH, et al.

Epigenetic programming of monocyte-tomacrophage differentiation and trained innate immunity.

Science.

2014; 345(6204):1251086.

Alvarez-Errico D, Vento-Tormo R, Sieweke M, Ballestar E. Epigenetic control of myeloid cell differentiation, identity and function. Nat Rev Immunol. 2015;15(1):717.

Buffen K, Oosting M, Quintin J, et al.

Autophagy controls BCG-induced traine immunity and the response to intravesical BCG therapy for bladder cancer. PLoS Pathog. 2014;10(10):e1004485.

Paust S, Gill HS, Wang BZ, et al. Critical role for the chemokine receptor CXCR6 in NK cell-mediated antigen-specific memory of haptens and viruses. Nat Immunol. 2010; 11(12):1127-1135.

Sun JC, Beilke JN, Lanier LL. Adaptive immune features of natural killer cells.

Nature. 2009;457(7229):557-561.

Lee J, Zhang T, Hwang I, et al. Epigenetic modification and antibody-dependent expansion of memory-like NK cells in human cytomegalovirus-infected individuals. Immunity. 2015;42(3):431-442.

Schlums H, Cichocki F, Tesi B, et al.

Cytomegalovirus infection drives adaptive epigenetic diversification of NK cells with altered signaling and effector function.

Immunity. 2015;42(3):443-456.

Sun JC, Madera S, Bezman NA, Beilke JN, Kaplan MH, Lanier LL. Proinflammatory cytokine signaling required for the generation of natural killer cell memory. J Exp Med. 2012;209(5):947-954.

Kleinnijenhuis J, Quintin J, Preijers F, et al.

BCG-induced trained immunity in NK cells: Role for non-specific protection to infection. Clinical Immunol. 2014; 155(2):213-219.

haematologica | 2016; 101(12)

30. Lalor MK, Smith SG, Floyd S, et al.

Complex cytokine profiles induced by BCG vaccination in UK infants. Vaccine.

2010;28(6):1635-1641.

31. Gopal R, Lin Y, Obermajer N, et al. IL-23dependent IL-17 drives Th1-cell responses following Mycobacterium bovis BCG vaccination. Eur J Immunol. 2012;42(2):364373.

32. Pitt JM, Stavropoulos E, Redford PS, et al.

Blockade of IL-10 signaling during bacillus Calmette-Guerin vaccination enhances and sustains Th1, Th17, and innate lymphoid IFN-gamma and IL-17 responses and increases protection to Mycobacterium tuberculosis infection. J Immunol. 2012; 189(8):4079-4087.

33. Kleinnijenhuis J, Quintin J, Preijers F, et al.

Long-lasting effects of BCG vaccination on both heterologous Th1/Th17 responses and innate trained immunity. J innate Immun. 2014;6(2):152-158.

34. Aaby P, Roth A, Ravn H, et al. Randomized trial of BCG vaccination at birth to lowbirth-weight children: beneficial nonspecific effects in the neonatal period? J Infect Dis. 2011;204(2):245-252.

35. Sorup S, Villumsen M, Ravn H, et al.

Smallpox vaccination and all-cause infectious disease hospitalization: a Danish register-based cohort study. Int J Epidemiol.

2011;40(4):955-963.

36. de Castro MJ, Pardo-Seco J, MartinonTorres F. Nonspecific (heterologous) protection of neonatal BCG vaccination against hospitalization due to respiratory infection and sepsis. Clin Infect Dis. 2015; 60(11):1611-1619.

37. Hersh EM, Gutterman JU, Mavligit GM. BCG as adjuvant immunotherapy for neoplasia. Ann Rev Med. 1977;28:489-515.

38. Donnelly JP, Blijlevens NM, van der Velden WJ. Host impairments in patients with neoplastic diseases. Cancer Treat Res. 2014; 161:1-41.

39. Morrison VA. Infectious complications in patients with chronic lymphocytic leukemia: pathogenesis, spectrum of infection, and approaches to prophylaxis. Clin Lymphoma Myeloma. 2009;9(5):365-370.

40. Morrison VA. Infections in patients with leukemia and lymphoma. Cancer Treat Res. 2014;161:319-349.

41. Forconi F, Moss P. Perturbation of the normal immune system in patients with CLL. Blood. 2015;126(5):573-581.

42. Riches JC, Gribben JG. Immunomodulation and immune reconstitution in chronic lymphocytic leukemia. Semin Haematol. 2014; 51(3):228-234.

43. Jitschin R, Braun M, Buttner M, et al. CLLcells induce IDOhi CD14+HLA-DRlo myeloid-derived suppressor cells that inhibit T-cell responses and promote TRegs. Blood. 2014;124(5):750-760.

44. Yang ZZ, Novak AJ, Ziesmer SC, Witzig TE, Ansell SM. Malignant B cells skew the balance of regulatory T cells and TH17 cells in B-cell non-Hodgkin's lymphoma. Cancer Res. 2009;69(13):5522-5530.

45. McClanahan F, Riches JC, Miller S, et al.

Mechanisms of PD-L1/PD-1-mediated CD8 T-cell dysfunction in the context of aging-related immune defects in the Emicro-TCL1 CLL mouse model. Blood.

2015;126(2):212-221.

46. Tsukada N, Burger JA, Zvaifler NJ, Kipps TJ. Distinctive features of "nurselike" cells that differentiate in the context of chronic lymphocytic leukemia. Blood. 2002;99(3):10301037.

47. Nam SJ, Go H, Paik JH, et al. An increase of

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

M2 macrophages predicts poor prognosis in patients with diffuse large B-cell lymphoma treated with rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone.

Leuk Lymphoma.

2014;55(11):2466-2476.

Serafini P, Mgebroff S, Noonan K, Borrello I. Myeloid-derived suppressor cells promote cross-tolerance in B-cell lymphoma by expanding regulatory T cells. Cancer Res. 2008;68(13):5439-5449.

Ysebaert L, Fournie JJ. Genomic and phenotypic characterization of nurse-like cells that promote drug resistance in chronic lymphocytic leukemia. Leuk Lymphoma.

2011;52(7):1404-1406.

Melenotte C, Million M, Audoly G, et al. Bcell non-Hodgkin lymphoma linked to Coxiella burnetii. Blood. 2016;127(1):113121.

Ferreri AJ, Sassone M, Kiesewetter B, et al.

High-dose clarithromycin is an active monotherapy for patients with relapsed/refractory extranodal marginal zone lymphoma of mucosa-associated lymphoid tissue (MALT): the HD-K phase II trial. Ann Oncol. 2015;26(8):1760-1765.

Ferreri AJ, Govi S, Raderer M, et al.

Helicobacter pylori eradication as exclusive treatment for limited-stage gastric diffuse large B-cell lymphoma: results of a multicenter phase 2 trial.

Blood.

2012;120(18):3858-3860.

Anderson LA, Landgren O, Engels EA. Common community acquired infections and subsequent risk of chronic lymphocytic leukaemia. Br J Haematol. 2009;147(4):444449.

McShane CM, Murray LJ, Engels EA, Anderson LA. Community-acquired infections associated with increased risk of lymphoplasmacytic lymphoma/Waldenstrom macroglobulinaemia. Br J Haematol.

2014;164(5):653-658.

McShane CM, Murray LJ, Engels EA, Landgren O, Anderson LA. Common community-acquired infections and subsequent risk of multiple myeloma: a populationbased study. Int J Cancer. 2014;134(7):17341740.

Agathangelidis A, Darzentas N, Hadzidimitriou A, et al. Stereotyped B-cell receptors in one-third of chronic lymphocytic leukemia: a molecular classification with implications for targeted therapies.

Blood. 2012;119(19):4467-4475.

Lanemo Myhrinder A, Hellqvist E, Sidorova E, et al. A new perspective: molecular motifs on oxidized LDL, apoptotic cells, and bacteria are targets for chronic lymphocytic leukemia antibodies. Blood.

2008;111(7):3838-3848.

Hadzidimitriou A, Agathangelidis A, Darzentas N, et al. Is there a role for antigen selection in mantle cell lymphoma?

Immunogenetic support from a series of 807 cases. Blood. 2011;118(11):3088-3095.

Davis RE, Ngo VN, Lenz G, et al. Chronic active B-cell-receptor signalling in diffuse large B-cell lymphoma. Nature. 2010; 463(7277):88-92.

Hoogeboom R, van Kessel KP, Hochstenbach F, et al. A mutated B cell chronic lymphocytic leukemia subset that recognizes and responds to fungi. J Exp Med. 2013;210(1):59-70.

Machado AM, Desler C, Boggild S, et al.

Helicobacter pylori infection affects mitochondrial function and DNA repair, thus, mediating genetic instability in gastric cells.

Mech Ageing Dev. 2013;134(10):460-466.

Strickertsson JA, Desler C, Martin-

1467

W.B.C. Stevens et al.

63.

64.

65.

66.

67.

68.

69.

70.

71.

1468

Bertelsen T, et al. Enterococcus faecalis infection causes inflammation, intracellular oxphos-independent ROS production, and DNA damage in human gastric cancer cells.

PloS one. 2013;8(4):e63147.

Siegl C, Rudel T. Modulation of p53 during bacterial infections. Nat Rev Microbiol.

2015;13(12):741-748.

Niemann CU, Wiestner A. B-cell receptor signaling as a driver of lymphoma development and evolution. Semin Cancer Biol.

2013;23(6):410-421.

Muzio M, Scielzo C, Bertilaccio MT, Frenquelli M, Ghia P, Caligaris-Cappio F. Expression and function of toll like receptors in chronic lymphocytic leukaemia cells. Br J Haematol. 2009;144(4):507-516.

Fonte E, Agathangelidis A, Reverberi D, et al. Toll-like receptor stimulation in splenic marginal zone lymphoma can modulate cell signaling, activation and proliferation.

Haematologica. 2015;100(11):1460-1468.

Riches JC, Davies JK, McClanahan F, et al.

T cells from CLL patients exhibit features of T-cell exhaustion but retain capacity for cytokine production. Blood. 2013; 121(9):1612-1621.

Ramsay AG, Johnson AJ, Lee AM, et al.

Chronic lymphocytic leukemia T cells show impaired immunological synapse formation that can be reversed with an immunomodulating drug. J Clin Invest.

2008;118(7):2427-2437.

Yang ZZ, Grote DM, Xiu B, et al. TGF-beta upregulates CD70 expression and induces exhaustion of effector memory T cells in Bcell non-Hodgkin's lymphoma. Leukemia.

2014;28(9):1872-1884.

Kater AP, van der Windt GJ. PD-L1 blockade: rejuvenating T cells in CLL. Blood.

2015;126(2):126-128.

Huergo-Zapico L, Acebes-Huerta A, Gonzalez-Rodriguez AP, et al. Expansion of NK cells and reduction of NKG2D expression in chronic lymphocytic leukemia.

72.

73.

74.

75.

76.

77.

78.

79.

80.

Correlation with progressive disease. PloS one. 2014;9(10):e108326.

Maffei R, Bulgarelli J, Fiorcari S, et al. The monocytic population in chronic lymphocytic leukemia shows altered composition and deregulation of genes involved in phagocytosis and inflammation.

Haematologica. 2013;98(7):1115-1123.

Jurado-Camino T, Cordoba R, EstebanBurgos L, et al. Chronic lymphocytic leukemia: a paradigm of innate immune cross-tolerance.

J Immunol.

2015; 194(2):719-727.

Villumsen M, Sorup S, Jess T, et al. Risk of lymphoma and leukaemia after bacille Calmette-Guerin and smallpox vaccination: a Danish case-cohort study. Vaccine. 2009; 27(49):6950-6958.

Omura GA, Vogler WR, Lefante J, et al.

Treatment of acute myelogenous leukemia: influence of three induction regimens and maintenance with chemotherapy or BCG immunotherapy. Cancer. 1982;49(8):15301536.

Powles RL, Russell JA, Selby PJ, et al.

Maintenance of remission in acute myelogenous leukaemia by a mixture of B.C.G. and irradiated leukaemia cells. Lancet.

1977;2(8048):1107-1110.

Ikeda Y, Adachi Y, Ishii T, et al. Blocking effect of anti-Dectin-1 antibodies on the anti-tumor activity of 1,3-beta-glucan and the binding of Dectin-1 to 1,3-beta-glucan.

Biol Pharm Bull 2007;30(8):1384-1389.

Leibundgut-Landmann S, Osorio F, Brown GD, Reis e S. Stimulation of dendritic cells via the dectin-1/Syk pathway allows priming of cytotoxic T-cell responses. Blood.

2008;112(13):4971-4980.

Weitberg AB. A phase I/II trial of beta(1,3)/(1,6) D-glucan in the treatment of patients with advanced malignancies receiving chemotherapy. J Exp Clin Cancer Res 2008;27(40.

Poon IK, Lucas CD, Rossi AG,

81.

82.

83.

84.

85.

86.

87.

88.

Ravichandran KS. Apoptotic cell clearance: basic biology and therapeutic potential. Nat Rev Immunol. 2014;14(3):166-180.

Arandjelovic S, Ravichandran KS. Phagocytosis of apoptotic cells in homeostasis. Nat Immunol. 2015;16(9):907-917.

Hanna BS, McClanahan F, Yazdanparast H, et al. Depletion of CLL-associated patrolling monocytes and macrophages controls disease development and repairs immune dysfunction in vivo. Leukemia.

2015; Albeituni SH, Ding C, Liu M, et al. Yeastderived particulate beta-gluten treatment subverts the suppression of myeloidderived suppressor cells (MDSC) by inducing polymorphonuclear MDSC apoptosis and monocytic MDSC differentiation to APC in Cancer. J Immunol. 2016; 196(5):2167-2180.

Shatnyeva OM, Hansen HP, Reiners KS, Sauer M, Vyas M, von Strandmann EP. DNA damage response and evasion from immunosurveillance in CLL: new options for NK cell-based immunotherapies. Front Genet. 2015;6(11)eCollection 2015.

Cruz CR, Bollard CM. T-cell and natural killer cell therapies for hematologic malignancies after hematopoietic stem cell transplantation: enhancing the graft-versusleukemia effect.

Haematologica.

2015;100(6):709-719.

Cheng M, Chen Y, Xiao W, Sun R, Tian Z. NK cell-based immunotherapy for malignant diseases. Cell Mol Immunol.

2013;10(3):230-252.

Della Chiesa M, Falco M, Muccio L, Bertaina A, Locatelli F, Moretta A. Impact of HCMV Infection on NK Cell Development and Function after HSCT. Front Immunol.

2013;4(458.

Arts RJ, Blok BA, Aaby P, et al. Long-term in vitro and in vivo effects of gamma-irradiated BCG on innate and adaptive immunity.

J Leukoc Biol. 2015;98(6):995-1001.

haematologica | 2016; 101(12)

ARTICLE

Hematopoiesis

Uncoupling of the Hippo and Rho pathways allows megakaryocytes to escape the tetraploid checkpoint

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Anita Roy,1,2,3 Larissa Lordier,1,2,3 Catherine Pioche-Durieu,2,3,4 Sylvie Souquere,2,3,5 Lydia Roy,1,6 Philippe Rameau,3 Valérie Lapierre,7 Eric Le Cam,2,3,4 Isabelle Plo,1,2,3 Najet Debili,1,2,3 Hana Raslova1,2,3 and William Vainchenker1,2,3 Institut National de la Santé et la Recherche Médicale (INSERM) UMR1170, Equipe Labellisée par la Ligue Nationale Contre le Cancer, Villejuif; 2Université Paris-Saclay, Villejuif; 3Gustave Roussy, Villejuif; 4Centre Nationale de la Recherche Scientifique (CNRS), UMR 8126, Gustave Roussy, Villejuif; 5CNRS UMR 8122, Gustave Roussy, Villejuif; 6Assistance Publique des Hôpitaux de Paris (AP-HP), Service d’Hématologie Clinique, Hôpital Henri Mondor, Créteil and 7Gustave Roussy, Unité de Thérapie Cellulaire, Villejuif, France

1

Haematologica 2016 Volume 101(12):1469-1478

AR and LL, and HR and WV contributed equally to this work.

ABSTRACT

M

egakaryocytes are naturally polyploid cells that increase their ploidy by endomitosis. However, very little is known regarding the mechanism by which they escape the tetraploid checkpoint to become polyploid. Recently, it has been shown that the tetraploid checkpoint was regulated by the Hippo-p53 pathway in response to a downregulation of Rho activity. We therefore analyzed the role of Hippo-p53 pathway in the regulation of human megakaryocyte polyploidy. Our results revealed that Hippo-p53 signaling pathway proteins are present and are functional in megakaryocytes.

Although this pathway responds to the genotoxic stress agent etoposide, it is not activated in tetraploid or polyploid megakaryocytes.

Furthermore, Hippo pathway was observed to be uncoupled from Rho activity. Additionally, polyploid megakaryocytes showed increased expression of YAP target genes when compared to diploid and tetraploid megakaryocytes. Although p53 knockdown increased both modal ploidy and proplatelet formation in megakaryocytes, YAP knockdown caused no significant change in ploidy while moderately affecting proplatelet formation. Interestingly, YAP knockdown reduced the mitochondrial mass in polyploid megakaryocytes and decreased expression of PGC1α, an important mitochondrial biogenesis regulator. Thus, the Hippo pathway is functional in megakaryocytes, but is not induced by tetraploidy. Additionally, YAP regulates the mitochondrial mass in polyploid megakaryocytes.

Correspondence: William.Vainchenker@gustaveroussy.fr

Received: May 23, 2016.

Accepted: August 8, 2016.

Pre-published: August 8, 2016. doi:10.3324/haematol.2016.149914

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/100/12/1469

Introduction Megakaryopoiesis is a unique model of differentiation characterized by a physiological polyploidization and a maturation that leads to platelet production.1,2 Polyploidy in megakaryocytes (MKs) is achieved by endomitosis, which corresponds to defective cytokinesis and karyokinesis.3-5 An increase in ploidy is believed to augment cell size, a crucial parameter for increasing platelet production.6,7 While the modal ploidy of MKs in the bone marrow is 16N, the ploidy of individual MKs can reach 64N or more.7,8 This indicates that MKs are able to escape the 4N control called the tetraploid checkpoint either because they are devoid of the tetraploid checkpoint machinery or can overcome this checkpoint that normally exists in haematologica | 2016; 101(12)

©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.

1469

A. Roy et al.

most cell types except embryonic stem cells and undifferentiated embryos.9 Key checkpoints exist in cells and include the tetraploid checkpoint that ensures cessation of proliferation and apoptosis of tetraploid cells. This is essential as tetraploidy not only promotes genetic instability, but also contributes to tumorigenesis.10,11 Reports have identified the Hippop53 pathway as an important component of the tetraploid checkpoint.12,13 The conserved Hippo tumor suppressor pathway consists of the STE (yeast Sterile20 kinase) family protein kinases (MST1/2) which when activated can phosphorylate LATS1/2.14 This activates the kinase activity of LATS1/2 leading to direct interaction and phosphorylation of the transcription co-activators YAP/TAZ. Phosphorylated YAP/TAZ is sequestered in the cytoplasm resulting in the inhibition of target gene transcription.15 In contrast when upstream kinases are inactive, YAP/TAZ can translocate into the nucleus and activate transcription of their target genes. More recent studies have described the tumor suppressor LATS2 as a key link between p53 and tetraploid arrest. p53 has long been known to play a central role in this checkpoint since p53 knockout cells were found to be prone to accumulate tetraploid cells that divide subsequently.16,17 The Hippo pathway is triggered in response to tetraploidy.12 This stabilizes p53 through direct interaction of LATS2 with MDM2 leading to p53 stabilization and proliferation arrest of tetraploid cells.13 The role of p53 during MK differentiation has been previously studied. It has been shown that p53–/– mice showed increased numbers of MKs with higher ploidy, more particularly in stress conditions.18-20 Further, p53 stabilization by MDM2 inhibitors was found to impair all stages of megakaryopoiesis including polyploidy and proplatelet formation.20,21 However, no information exists about the Hippo pathway in MKs. In this study we analyzed the role of the Hippo-p53 pathway in the regulation of human MK polyploidy. MKs were observed to harbor a functional Hippo-p53 pathway that responds to genotoxic stress, but not to polyploidy by decoupling the Hippo pathway from Rho activity.

Elmer Applied Biosystems) using the Power SYBR-Green PCR Master Mix (ABI) containing the specific primers (1.2 mM). The expression levels of all genes were calculated relatively to HPRT and PPIA1. Details of primer sequences are provided in the Online Supplementary Appendix.

Immunofluorescence The cells were plated on poly-L-lysine-coated slides (O. Kindler GmbH&Co, Freiburg, Germany) for 1 h at 37°C. Immunofluorescence staining was performed using mouse antip53, anti-β1-tubulin (Sigma-Aldrich) or anti-YAP antibodies and appropriate secondary antibodies conjugated with Alexa-488 or Alexa-546 (Molecular Probes Life Technology). TOTO-3 iodide or DAPI (Molecular Probes, Life Technology) was applied for nuclear staining. Cells were examined under a Zeiss LSM 510 laser scanning microscope (Carl Zeiss, Le Pecq, France) or Leica TCS SP8 MP (Leica Microsystems, Wetzlar, Germany) with a 63X oil immersion objective.

Western blot analysis Western blots were performed as described previously.22 Details of the primary antibodies are given in the Online Supplementary Appendix.

Proplatelet formation assay CD41+GFP+ or CD41+mCherry+ MKs were sorted at day 8 of culture and proplatelet formation was evaluated as previously described.22 A total of 200 cells per well were counted during four days. Images were obtained using an inverted microscope (Carl Zeiss, Göttingen, Germany) at a magnification of 40X using the Axio v.4.6 software.

Transmission electron microscopy Details of transmission electron microscopy may be found in the Online Supplementary Appendix.

Statistical analysis Student’s t-test and one-way Anova test were used to determine the significance of the data.

Results Methods Cultures of megakaryocytes and erythroblasts derived from human CD34+ cells in serum-free liquid medium Leukapheresis and cord blood samples were obtained after approval from the Assistance Publique des Hopitaux de Paris. All adult participants in this study gave their informed written consent in accordance with the Declaration of Helsinki. The study was approved by the Local Research Ethics Committee of Hospital Saint Louis, Paris, France, for the cord blood samples and the Local Research Ethics Committee of Institut Gustave Roussy, Villejuif, France. Details of culture conditions are given in the Online Supplementary Appendix.

Cell sorting and flow cytometry Cell sorting and analysis of ploidy level were previously described.4

Real-time quantitative PCR Primers for qRT-PCR were designed using Primer Express Software (Perkin-Elmer Applied Biosystems, Foster City, CA, USA) and were synthesized by Eurogentec (Angers, France). qRTPCR was carried out in the ASI Prism GeneAmp 5700 (Perkin1470

Hippo-p53 pathway proteins are expressed in megakaryocytes The Hippo-p53 pathway constitutes the tetraploid checkpoint.12 We investigated the status of this pathway in human MKs at different ontogenic stages. Analysis of previously reported global microarray expression data revealed that key genes of the pathway were expressed in MKs derived from human cord blood and adult cytapheresis23 (Online Supplementary Figure S1). This was confirmed by real-time analysis of mRNA expression in in vitro cultured mature MKs (defined as CD41+CD42+ cells) derived from cord blood and adult cytapheresis (Online Supplementary Figure S1). Furthermore, adult MKs were sorted on day 6 of culture on the expression of CD41 and further cultured to the end of MK maturation.

Representative data of the ploidy distribution across the days of culture are shown in Online Supplementary Figure S2. We observed an initial increase followed by a marked decrease in the p53 transcript level at the end of MK maturation with a corresponding increase in the expression of p21. No statistically significant change was observed in the mRNA expression of BCL2L1, BAX and MDM2 haematologica | 2016; 101(12)

Hippo pathway and megakaryocyte polyploidization

A

B

C

D

E

F

Figure 1. Expression of Hippo-p53 pathway genes in megakaryocytes (MKs). (A) qRT-PCR data indicating relative expression of p53 and related genes on days 10 and 13 with respect to day 8 of in vitro cultured CD41+CD42+ MKs normalized against HPRT. Data represent meanÂąSEM (n=3; **P<0.01). (B) Protein expression of p53, p21, BAX, MDM2 and BCL-XL (BCL2L1) in the CD41+ sorted cells during MK differentiation and investigated by Western blot analysis. HSC70 indicates the loading in each lane. (C) The CD41+ sorted cells was treated with proteasome inhibitors (ALLN or MG132) for three hours, then p53 level was investigated by Western blot. HSC70 indicates the loading in each lane. (D) qRT-PCR data indicating relative expression of Hippo pathway genes on days 10 and 13 with respect to day 8 of in vitro cultured CD41+CD42+ MKs normalized against HPRT. Data represent meanÂąSEM (n=3; *P<0.05, **P<0.01) Similar results were obtained using PPIA. (E) Western blot analysis of Hippo pathway proteins and their expression in CD41+ MKs on different days of culture. HSC70 indicates the loading in each lane. (F) Confocal microscopic image of CD41+ MKs on different days of culture and stained for YAP and showing its localization in the cytosol and nucleus. DAPI was used to stain the nucleus. 150 cells were counted and categorized according to the distribution of YAP between nucleus (N) and cytosol (C). Day 8: N>C(6%), N=C(21%), N<C(73); Day 13: N>C(74%), N=C(14%), N<C(12%). Scale bar=100 mm.

(Figure 1A). In agreement with the mRNA expression profile, p53 protein expression remained constant and decreased at the end (day 13) of in vitro MK differentiation (Figure 1B). Moreover, two negative regulators of p53, MDM2 and MDMX, were also present in MKs (Figure 1B and C). Treatment of sorted CD41+ MKs with the proteasome inhibitors ALLN and MG132 dramatically increased the expression of p53, demonstrating that p53 was mainly regulated by proteasomal degradation during MK differentiation (Figure 1C). The mRNA expression of Hippo pathway genes LATS1, LATS2 and TAZ remained invariant during the course of MK maturation. However, a consistent and significant increase in the expression of the transcriptional targets of YAP (CTGF, CYR61, FSTL1 and INHBA) was observed indicating that YAP activity increased in mature MKs (Figure 1D). This was associated with an unchanged YAP protein level (Figure 1E), but with an increased nuclear localization of the protein (Figure 1F).

Reminiscent of p53, the LATS2 protein expression remained fairly constant and decreased only at the very end of MK maturation (Figure 1E). Interestingly and in contrast to the mRNA levels, protein expression of LATS2 and p21 decreased at day 13 of in vitro culture. This may be due to the fact that day 13 MKs are at the very end of their maturation with a heterogeneous population of MKs in haematologica | 2016; 101(12)

terms of ploidy and proplatelet production. Together, our results reveal that genes of the Hippo-p53 pathway are expressed throughout the various stages of MK maturation.

Hippo-p53 pathway is functional in megakaryocytes To understand whether Hippo-p53 signaling pathway was functional in MKs, cells were treated with a genotoxic agent (etoposide). Staining for p53-BP1 confirmed the genotoxicity of a 3-h treatment with 10 mM etoposide in MKs (Online Supplementary Figure S3). MKs were exposed to etoposide at various days of culture. Etoposide induced a drastic increase in the expression of both LATS2 and p53 (Figure 2A). Consistent with the canonical Hippo pathway, increased phosphorylation of YAP on ser127 was observed (Figure 2A). Enhanced LATS2 expression leading to increased p53 stability was reflected by increased p21 expression (Figure 2B). Under basal conditions, p53 protein was mostly cytoplasmic (>95%) in MKs. Upon etoposide exposure, p53 trafficked from the cytoplasm to the cell nucleus (Figure 2C). Moreover, consistent with its increased phosphorylation, YAP was completely sequestered in the cytoplasm of mature MKs (Figure 2D).

Taken together, etoposide-induced genotoxic stress was found to activate the Hippo-p53 axis in MKs. 1471

A. Roy et al.

A

C

B

D

Figure 2. Hippo-p53 axis is activated by genotoxic stress. (A) Western blot analysis of Hippo-p53 pathway proteins in CD41+ megakaryocytes treated with/without 10 mM etoposide for five hours. β-actin was used as loading control. (B) Western blot analysis of p53 and its target genes performed after 12 hours of etoposide treatment at different days (D) of culture (D5, D7 and D9). HSC70 was used as loading control. (C) p53 localization in untreated and 12-h etoposide treated CD41+ MK cells. Cells were stained with anti-p53 antibody. TOTO-3 was used to stain the nucleus. Scale bar=20 mm. (D) YAP localization in untreated CD41+ MK cells and CD41+ MK cells treated by 10 mM etoposide for five hours on day 10 of culture. Cells were stained with anti-YAP antibody. DAPI was used to stain the nucleus. Cells (150) were counted and categorized according to the distribution of YAP between nucleus (N) and cytosol (C). Whereas for control cells, nearly 70% had YAP staining the nucleus, none of the cells in etoposide treated samples showed nuclear localization of YAP. Scale bar=30 mm.

Hippo-p53 pathway is not activated in polyploid megakaryocytes We next checked whether polyploidy could induce the activation of Hippo and p53 pathway genes. MKs were sorted based on their ploidy level into diploid (2N), tetraploid (4N) and polyploid (â&#x2030;Ľ8N) cell populations. We did not detect any significant link between ploidization and the mRNA expression of p53, BAX, p21, MDM2 and MDMX genes (Figure 3A). Similarly, we did not detect any significant link between p53 protein expression and ploidy (Figure 3B). The expression of LATS2 also remained invariant at the different ploidy levels (Figure 3C and densitometric quantification in Online Supplementary Figure S4).

YAP expression remained fairly constant in the three ploidy states with a modest but not significant decrease in the phosphorylation of YAP in 4N ploidy stage (Figure 3C and Online Supplementary Figure S4). However, a pronounced increase in the expression of YAP target genes was observed in 4N and polyploid MKs indicating an increase in YAP transcriptional activity and, therefore, an inactivation of the Hippo pathway (Figure 3D). Thus, our data indicate that, in MKs, the Hippo-p53 pathway failed to sense polyploidy as a genotoxic stress.

RhoA/ROCK pathway has been widely studied in relation to MK differentiation. RhoA and ROCK proteins are 1472

well expressed in MKs and have been shown to regulate ploidy and proplatelet formation.4,22,24 Because, reduced RhoA activity in tetraploid cells induced Hippo-p53 signaling, and as MK differentiation and ploidization are associated with a decrease in RhoA activity,24 we checked if a further decrease in RhoA activity through ROCK inhibition (Y27632) could induce Hippo-p53 signaling in MKs. CD41+CD42+ MKs and CD71+ erythroblasts sorted on day 5 of in vitro culture were treated with Y27632. ROCK inhibition induced p53 expression in erythroblasts and reduced YAP expression increasing the ratio of phosphorylated YAP-S127 to total YAP (Figure 3E and densitometric quantification in Online Supplementary Figure S4).

However, Y27632 treatment did not induce p53 expression in MKs. In addition, there was no accompanying change in the expression of LATS2 or YAP and in the phosphorylation of YAP on ser127 (Figure 3E). Furthermore, increased MK ploidy was observed upon prolonged exposure to Y27632, a result that we have previously reported.4 Lastly, treatment of erythroblasts with ROCK inhibitor decreased the expression of YAP downstream target genes (Online Supplementary Figure S5) without any effect on their expression in MKs. This indicated that, in MKs in contrast to erythroblasts, RhoA activity is uncoupled from the Hippo pathway. haematologica | 2016; 101(12)

Hippo pathway and megakaryocyte polyploidization

A

B

C

D

E

Figure 3. Hippo-p53 axis is not activated by polyploidy. (A) qRT-PCR data indicating relative expression of p53 and related genes in cultured megakaryocytes (MKs) at different ploidy levels: 4N and ≥8N in comparison to 2N CD41+CD42+ MKs and normalized against HPRT. Data represent mean±SEM of 3 independent experiments. No significant difference was observed. (B) p53 expression in the CD41+ MKs sorted on ploidy and investigated by Western blot analysis. HSC70 indicates the loading in each lane. The mean density of p53 normalized against HSC70 is indicated in the histogram plot below. (C) Protein expression of Hippo pathway genes in the CD41+ MKs sorted on ploidy and investigated by Western blot analysis. HSC70 indicates the loading in each lane. (D) qRT-PCR data indicating relative expression of YAP downstream target genes in MK cells of ploidy 4N and ≥8N with respect to 2N of in vitro cultured CD41+CD42+ MKs and normalized against HPRT. Data represent mean±SEM (n=4; *P<0.05, **P<0.01). (E) Western blot analysis of the expression of Hippo-p53 pathway proteins in CD41+ MKs and CD71+ erythroblasts (ER) treated for five hours with 10 mM Y27632. GAPDH was used as a loading control. The corresponding DNA ploidy/cell cycle analysis analyzed after 72 hours of culture with/without 10 mM Y27632 for each sample with Hoechst 33342 staining is provided. Data are representative of 3 independent experiments. Densitometric analysis of the western blots are provided in Online Supplementary Figure S3. A significant increase in the percentage of polyploid MKs was observed with Y27632 (Control MKs = 3.50 ± 0.2%, MK + Y227632= 18.15 ± 2.5%, n=3; P<0.004)

p53 knockdown does not affect ploidy, but increases proplatelet formation in megakaryocytes To determine whether p53 knockdown facilitates human MK differentiation, CD34+ cells were induced into MK differentiation and transduced at day 3 or 4 of culture with GFP+ lentivirus encoding scrambled sequence or shp53 constructs (shp53-0, shp53-2, shp53-4). Transcript analysis of sorted GFP+CD41+ cells after 72-h transduction demonstrated that shp53-0 alone almost completely depleted p53, whereas a combination of shp53-2 and shp53-4 (shp53-2/4) reduced the mRNA level by approximately 60% (Online Supplementary Figure S6i). A decrease in p53 expression was also seen at the protein level (Online Supplementary Figure S6ii). Reduced p53 expression was accompanied by a decrease in the expression of p53 targets such as p21, BAX, MDM2, DR5 and PUMA (Online Supplementary Figure S6i). We also checked whether p53 expression was similarly down-regulated at the different ploidy states (Online Supplementary Figure S7). In subsequent experiments, we used shp53-0 and confirmed our results with shp53-2/4. p53 knockdown had a modest effect on 2N and 4N ploidy MKs while significantly increasing the percentage of MKs with ploidy 8N and 16N (Figure 4A). It had no effect on MK differentiation as the percentage of mature CD41+CD42+ MKs was not signifihaematologica | 2016; 101(12)

cantly modified in culture (n=5, 54%-61%) (Figure 4B).

Furthermore, p53 downregulation had limited effects on the MK cell cycle as attested by the incorporation of BrdU in the control and p53 knockdown samples at different ploidy states (Figure 4C). Lastly, as changes in p53 expression are frequently associated with genomic instability, we examined the separation of chromosomes and the number of centromeres during mitosis and endomitosis.

We observed that centrosomes were paired and localized correctly, and that the segregation and separation of these chromosomes was normal (Online Supplementary Figure S8). At the same time, p53 downregulation decreased MK response to apoptotic stimuli as observed in MK cells treated with 2 mM staurosporin, 4 mM etoposide and 1 mg/mL mitomycin C (Figure 4D). Next, we analyzed the proplatelet formation in p53 knockdown MKs. No effect was observed on proplatelet branching (Online Supplementary Figure S9i). However, p53 knockdown increased the number of proplatelet-forming MKs. At day 14 of culture, a 4-fold increase was observed with shp530 and 3.5-fold increase with shp53-2/4 (P=0.015) (Figure 4E). p53 knockdown also caused a marked increase in cytoplasmic maturation as observed by an increased development of the demarcation membrane system (Online Supplementary Figure S9ii). At the same time, p53 1473

A. Roy et al.

A

B

D

C

E

Figure 4. p53 knockdown increases proplatelet formation. Cells were transduced at day 4 of culture with a control lentivirus encoding scrambled shRNA (Cnt) or lentiviruses encoding either shRNA p53-0 or shRNAs p53-2/4. (A) Representative image of the ploidy level of GFP+ CD41+CD42+ cell population as analyzed by Hoechst staining. The percentage of cells at each ploidy level was calculated. Data represent mean±SEM of 4 independent experiments (*P<0.006, **P<0.002).

(B) Flow cytometric analysis of mature MKs expressing CD41 and CD42 in the GFP+ cells at day 9 of culture (n=5; P=0.03). (C) Flow cytometric analysis showing percentage of BrdU positive cells in control and p53 knockdown MKs (CD41+ GFP+ cells) at day 7 of culture. Data represent the mean±SEM of 3 independent experiments. (D) Annexin V binding assay on p53 knockdown or control cells at day 7 of culture exposed to 2 mM staurosporine (STS), 4 mM etoposide (Et), 1 mg/mL mitomycine C (Mit) and DMSO for 24 h was performed by flow cytometry. Data represent mean±SEM (n=3; **P≤0.01). (A) GFP+ CD41+ sorted cells were seeded at 2x103 cells/well in a 96-well plate. The percentage of proplatelet-forming megakaryocytes (MKs) was estimated by counting MKs exhibiting one or more cytoplasmic processes with areas of constriction. A total of 200 cells per well were counted during four days. Error bars in histograms represent the standard deviation (SD) of one representative experiment performed in triplicate wells. Similar results were obtained in 4 repeated experiments (n=4; **P=0.015).

knockdown did not affect the mRNA level of Hippo pathway genes (data not shown). Our results clearly indicate that p53 knockdown has a minor effect on ploidy level, further demonstrating that, in basal conditions, p53 is not a major determinant of MK polyploidization, but markedly increases proplatelet formation.

Knockdown of YAP moderately decreases proplatelet formation and does not affect megakaryocyte ploidy To determine the effects of YAP on human MK differentiation, CD34+ cells were induced into MK differentiation and transduced at day 3 or 4 of culture with mCherry+ lentivirus encoding scrambled or shYAP. Western blot and real-time PCR analysis on the sorted mCherry+CD41+ cells after 72 h of transduction demonstrated the efficacy of shYAP with more than a 70% decrease in mRNA expression and a corresponding decrease in protein expression (Online Supplementary Figure S10i and ii). Reduction in YAP expression was also accompanied by a decrease in the expression of YAP target genes (Online Supplementary Figure S10ii). shRNA mediated knockdown of YAP did not significantly affect MK ploidy (Figure 5A), although YAP knockdown by the shRNA was identical in all ploidy classes (Online Supplementary Figure S11). YAP knockdown did not affect MK differentiation as the percentage of mature CD41+CD42+ MKs was not modified in culture in comparison to control cells (Figure 5B). At the same time, YAP knockdown had no significant effect on apoptosis as observed in MKs treated with etoposide (Figure 5C). In addition, we analyzed the proplatelet formation in YAP knockdown MKs. A moderate but significant decrease in 1474

the number of cells bearing proplatelets was consistently observed in YAP knockdown MKs (Figure 5D). Taken together, the results clearly indicate that YAP knockdown does not significantly affect ploidy, but decreases proplatelet formation. Given the increased expression of YAP target genes with the ploidy level and MK maturation, it was plausible that YAP had pleotropic functions in MK biology.

YAP sustains the expression of PGC1α in megakaryocytes

Therefore, we tried to understand the implications of increased YAP activity in polyploid MKs. We hypothesized that the increased cell size associated with polyploidy would require increased energy production. We therefore analyzed the mitochondria in MKs at different ploidy levels. Live cell imaging of mitochondria revealed that diploid and tetraploid (2N-4N) nuclei containing MKs and polyploid MKs have small, punctate mitochondria.

Mitochondria were perinuclear in 2N-4N ploidy MKs while they were dispersed throughout the cytoplasm in polyploid MKs (Figure 6A). Electron micrographs also confirmed the perinuclear and dispersed localizations of mitochondria in the two ploidy classes. Furthermore, mitochondria were observed to have clearly defined inner membrane with well-defined cristae (Figure 6B). Next, we analyzed the mitochondrial mass in the various ploidy states of MKs. Increase in MK ploidy was accompanied by an increase in mitochondrial mass (Figure 6C and D). An increase in the number of mitochondria was also evident in the electron micrographs of polyploid MKs. haematologica | 2016; 101(12)

Hippo pathway and megakaryocyte polyploidization

A

B

C

D

Figure 5. YAP knockdown does not affect ploidy. Cells were transduced at day 4 or day 5 of culture with a control mCherry+ lentivirus encoding either scrambled (Cnt) or shYAP. (A) Representative image of ploidy distribution in mCherry+CD41+CD42+ cells analyzed by Hoechst staining. The percentage of cells at each ploidy level was calculated. Corresponding histogram plot represents the mean±SEM of mean ploidy of 5 independent experiments. (P>0.05, indicates that differences between the 2 samples are not significant). (C) Flow cytometric analysis of mature MKs expressing CD41 and CD42 in the mCherry+ cells at day 9 of culture. Data represent mean±SEM of 4 independent experiments. (N=4; P>0.05 indicates that the difference between the samples is not significant.) (D) Annexin V binding assay in scrambled and YAP knockdown megakaryocyte (MK) cells exposed to 10 mM etoposide (Et) for five hours was performed by flow cytometry. Data represent mean ± SEM of 4 independent experiments. (n=4; P>0.05 indicates that the difference between the samples is not significant.) (D) mCherry+CD41+ sorted cells were seeded at 2x103 cells/well in a 96-well plate. The percentage of PPT-forming MKs was estimated by counting MKs on day 11 or day 12, exhibiting one or more cytoplasmic processes with areas of constriction. shYAP slightly but significantly decreased proplatelet formation. Data represent mean±SEM of 4 independent experiments (n=4; * P<0.014).

Furthermore, mitochondria in polyploid MKs were observed to be smaller in comparison to diploid-tetraploid MKs (Figure 6C and Online Supplementary Figure S12). We also analyzed the expression of key genes involved in mitochondrial biogenesis and mitochondrial fission-fusion kinetics in the three ploidy states. The mRNA expression of these genes remained unaltered with ploidy (Online Supplementary Figure S13i and ii). Taken together, this indicated that polyploidization was accompanied by an increase in mitochondrial mass.

We next analyzed the effects of YAP knockdown on mitochondrial mass of MKs. Staining of mitochondria using mitotracker green followed by live cell imaging revealed that mitochondrial morphology remained unaltered in YAP knock-down MKs (Online Supplementary Figure S14). YAP knockdown did not alter the mitochondrial mass in 2N-4N MKs. However, a significant decrease in mitochondrial mass was observed in cells with ploidy of greater than 8N (Figure 6D). Remarkably, YAP knockdown substantially decreased the expression of PGC1α, a key regulator of mitochondrial biogenesis (Figure 6E and F).

PGC1α expression was also found to be constant between diploid, tetraploid and polyploid MKs (Online Supplementary Figure S13i). However, neither ploidy nor YAP knockdown had any significant effect on the expression of key genes of the oxidative phosphorylation system or other genes regulating mitochondrial biogenesis (Figure 6E and Online Supplementary Figure S4). To check if the reduced mitochondrial mass affected cellular energetics, we assessed the ATP content along with cellular NAD+/NADH ratio (Online Supplementary Figure S15i and ii). No change in the ATP content was observed in diploid and tetraploid cells. However, a decrease was observed in the ATP content of polyploid cells demonstrating again that the regulation of YAP is important in polyploid MK haematologica | 2016; 101(12)

cells. However, the NAD+/NADH ratio remained unchanged in the three ploidy classes analyzed. Thus, YAP was found to regulate mitochondrial mass along with the expression of a key mitochondrial biogenesis regulator PGC1α.

Discussion Megakaryocytes are endowed with unconventional properties that make them unique systems to study various biological processes, especially the switch from mitotic mode of cell division to endomitosis whereby the cells become polyploid. Thus, MKs provide an ideal platform to study how naturally polyploid cells overcome the tetraploid checkpoint that normally arrests cell cycle. It was previously reported that the Hippo-p53 pathway maintains the tetraploid checkpoint and reduction of RhoA activity induced by extra centromeres was found to activate Hippo-p53 pathway.12 Here, we provide evidence for a functional Hippo-p53 axis in MKs. Figure 7 shows a schematic representation that highlights the key players of the Hippo-p53 pathway.

Previous reports have demonstrated the importance of the Hippo pathway in fly hematopoiesis.25,26 It was also reported that key proteins of the Hippo pathway are expressed in mantle cell lymphoma.27 Furthermore YAP, a key component of the Hippo pathway, is highly expressed in stem cells and decreases progressively during differentiation.28 Therefore, we first confirmed the expression of the Hippo and p53 pathway genes in MKs derived from cord blood or adult blood.

While various mechanisms induce the Hippo pathway activity, tetraploidy and induction of apoptosis activate both p53 and the Hippo pathway.12,29 We employed a 1475

A. Roy et al.

A

B

E

C

D

F

Figure 6. YAP knockdown decreases mitochondrial mass. (A) Fluorescence microscopic image of CD41+ megakaryocytes (MKs) stained with mitotracker red. DAPI was used to stain the nucleus. Scale bar=20 μm. (B) Electron micrographs of sections of MKs of 2N-4N ploidy (a,b) and ≥8N (c,d) ploidy states. Representative image shown at 4400X magnification (a,c) and mitochondria indicated at higher magnification of 21600X (b,d) of a field within marked by a box. Scale bars indicate 1 mm (a,c) and 200 nm (b,d). (C) The corresponding graph indicates the number of mitochondria versus the median mitochondrial diameter in 2N-4N and ≥8N ploidy MKs. The parameters are measured on ultrathin sections by TEM using a photo series taken at the same magnification. Further statistical analysis is provided in Online Supplementary Figure S12. (D) MK cells were transduced at day 4 or day 5 of culture with a control mCherry+ lentivirus encoding either scrambled (Cnt) or shYAP. Cells were stained with mitotracker green, Hoechst and anti-CD41 antibody. The fluorescence intensity of mitotracker green in each individual ploidy state (2N, 4N, etc.) was plotted against ploidy number and fitted against a straight line passing through origin by least square fit method (n=4; *P<0.07). (E) qRT-PCR data indicating relative expression of YAP and genes involved in mitochondrial biogenesis and oxidative phosphorylation in MK cells transduced with the indicated vectors.

Relative expression of shYAP samples with respect to control (Cnt) was calculated. Data were normalized against HPRT. Data represent mean±SEM of 3 independent experiments (n=3; P<0.003). (F) Protein expression of PGC1α and YAP in the CD41+ MKs sorted on mCherry+ Cnt or mCherry+ shYAP and investigated by Western blot analysis. HSC70 indicates the loading in each lane. Densitometric analysis showing the normalized expression of PGC1alpha (mean±SEM) of 3 independent experiments (P<0.006).

genotoxic agent etoposide to decipher the functionality of the Hippo-p53 pathway in MKs and found that it activates this pathway. Next, we analyzed the expression of the components of this pathway in diploid, tetraploid and polyploid MKs. No appreciable change in their expression was detected between the various ploidy states. However, a consistent increase in the expression of YAP and its target genes was observed during polyploidization, as well as during the course of MK differentiation. This indicates that the Hippo-p53 pathway does not sense polyploidy as a stress in MKs. This is in agreement with previous reports that indicated that tetraploid cells re-entering the cell cycle harbored an inactive Hippo pathway.12 Although ploidization has minor effects on p53 expression and its activity, p21 expression markedly increased during late stages of megakaryopoiesis. This is related to regulation by other signaling pathways, such as the MAPK/ERK pathway.30,31 The activation of the Hippo-p53 pathway in tetraploid cells was reported to be acutely dependent upon RhoA activity. As RhoA activity was reported to decrease during the first endomitotic division of MKs, we assessed the effects of decreased RhoA activity on Hippo-p53 pathway.12,24 Our data demonstrate that, in contrast to erythroblasts, the Hippo pathway is decoupled from RhoA activity in MKs. Importantly, decreased RhoA activ1476

ity had previously been linked to increased YAP phosphorylation in HEK293A cells.32 Our results show that decreased Rho/ROCK signaling did not induce YAP phosphorylation in MKs in contrast to erythroblasts. Moreover, in erythroblasts, inhibition of ROCK was also found to decrease the expression of total YAP protein while increasing the ratio of phosphorylated YAP to total YAP. This was consistent with previous reports that demonstrate that phosphorylation of YAP on S127 induced its degradation.33 p53 knockdown induced a moderate, but significant increase in MK polyploidization, without significant increase in DNA replication. It is likely that the modest increase in polyploidization could be the consequence of a decreased basal apoptosis in p53 knockdown MKs. This is in agreement with previous results on p53–/– mice that show that p53 knockout only increased MK ploidy under stress conditions such as during induced thrombocytopenia, but not in basal conditions in the bone marrow.16,17 However, p53 knockdown significantly increased proplatelet formation in accordance with our previous reports.21,34 Furthermore, YAP knockdown caused no significant change in ploidy, but a consistent and significant decrease in the percentage of proplatelet bearing MKs was observed.

Given the observed increase in YAP target gene expreshaematologica | 2016; 101(12)

Hippo pathway and megakaryocyte polyploidization

NUCLEUS

Figure 7. Illustration of Hippo-p53 pathway in megakaryocytes (MKs). Low RhoA activity leads to increased Hippo pathway activity by increasing the expression and phosphorylation of LATS2 leading to increased interaction with MDM2. The interaction between LATS2 and MDM2 inhibits MDM2-p53 interaction thereby releasing p53, which can now enter the nucleus and act as a transcription factor blocking cell cycle progression. At the same time, increased expression and phosphorylation of LATS2 increases phosphorylation of YAP. YAP is thus sequestered away from the nucleus, thereby blocking its transcriptional activity. In MKs, the cells fail to respond to decreased RhoA activity and do not increase the expression of LATS2 or phosphorylation of YAP.

sion both during the course of MK differentiation as well as polyploidization, we tried to identify its possible effects. Previous works had suggested links between mitochondria and the Hippo-p53 pathway.35,36 We hypothesized that polyploid MKs would have a large energy demand to sustain their size and functionality, and it is plausible that this could be regulated by the Hippo-p53 pathways. YAP knockdown was found to decrease mitochondrial mass in polyploid MKs with a ploidy greater than 8N while keeping the mitochondrial mass unchanged in diploid and tetraploid MKs. Although mitochondria in polyploid MKs were on average smaller when compared to mitochondria in diploid MKs, no significant differences could be observed in the expression of genes regulating mitochondrial fission-fusion dynamics. Instead, YAP knockdown substantially decreased the expression of PGC1α, a key mitochondrial biogenesis factor37 which was associated with a decrease in ATP content in polyploid MKs. However, no significant change was observed in the NAD+/NADH ratio. Neither was any change observed in the expression of factors regulating biogenesis such as TFAM, which are themselves regulated by PGC1α. Thus, it appears that the haematologica | 2016; 101(12)

expression level of PGC1α in YAP knockdown MKs is sufficient to drive biogenesis, albeit at a slower rate.

Interestingly, activation of Yorkie (the fly homolog of YAP) did not alter the expression of PGC1α in Drosophila.

Instead, activation of Yorkie increased mitochondrial fusion without altering the cellular ATP content.35 YAPmediated regulation of mitochondrial mass may explain its effects on proplatelet formation because it has recently been shown that mitochondria play a key role in platelet formation and function, particularly in stress conditions by controlling ROS production.38,39 Interestingly, mitochondria in diploid and tetraploid MKs were found clustered around the nucleus whereas in polyploid MKs they were distributed in the cytoplasm.

The perinuclear localization of mitochondria has been previously reported to be associated with high nuclear ROS levels and increased expression of hypoxia-associated genes.40 Therefore, it is plausible that mitochondrial biogenesis and its direct impact on the energetics of the cell could affect proplatelet formation. Further research is required to fully understand the impact of cellular energetics on demarcation membrane formation, proplatelet formation and, ultimately, platelet function.

1477

A. Roy et al.

In summary, our results clearly show the existence of functional Hippo-p53 machinery in MKs that is not activated during MK polyploidization, suggesting that the decrease in RhoA activity is not sensored in MKs. Finally, our study revealed the unexpected role of YAP in the regulation of mitochondrial biogenesis in polyploid MKs and proplatelet formation.

References 1. Bluteau D, Lordier L, Di Stefano A, et al.

Regulation of megakaryocyte maturation and platelet formation. J Thromb Haemost.

2009;7(Suppl 1):227-234.

2. Deutsch VR, Tomer A. Advances in megakaryocytopoiesis and thrombopoiesis: from bench to bedside. Br J Haematol. 2013;161(6):778-793.

3. Geddis AE, Kaushansky K. Endomitotic megakaryocytes form a midzone in anaphase but have a deficiency in cleavage furrow formation. Cell Cycle. 2006;5(5): 538-545.

4. Lordier L, Jalil A, Aurade F, et al.

Megakaryocyte endomitosis is a failure of late cytokinesis related to defects in the contractile ring and Rho/Rock signaling.

Blood. 2008;112(8):3164-3174.

5. Lordier L, Pan J, Naim V, et al. Presence of a defect in karyokinesis during megakaryocyte endomitosis. Cell Cycle. 2012;11(23): 4385-4389.

6. Sher N, Von Stetina JR, Bell GW, Matsuura S, Ravid K, Orr-Weaver TL. Fundamental differences in endoreplication in mammals and Drosophila revealed by analysis of endocycling and endomitotic cells. Proc Natl Acad Sci USA. 2013;110(23):9368-9373.

7. Ravid K, Lu J, Zimmet JM, Jones MR. Roads to polyploidy: the megakaryocyte example.

J Cell Physiol. 2002;190(1):7-20.

8. Zimmet J, Ravid K. Polyploidy: occurrence in nature, mechanisms, and significance for the megakaryocyte-platelet system. Exp Hematol. 2000;28(1):3-16.

9. Horii T, Yamamoto M, Morita S, Kimura M, Nagao Y, Hatada I. p53 suppresses tetraploid development in mice. Sci Rep. 2015;5:8907.

10. Davoli T, de Lange T. The causes and consequences of polyploidy in normal development and cancer. Annu Rev Cell Dev Biol.

2011;27:585-610.

11. Dewhurst SM, McGranahan N, Burrell RA, et al. Tolerance of whole-genome doubling propagates chromosomal instability and accelerates cancer genome evolution.

Cancer Discov. 2014;4(2):175-185.

12. Ganem NJ, Cornils H, Chiu SY, et al.

Cytokinesis failure triggers hippo tumor suppressor pathway activation. Cell.

2014;158(4):833-848.

13. Aylon Y, Michael D, Shmueli A, Yabuta N, Nojima H, Oren M. A positive feedback loop between the p53 and Lats2 tumor suppressors prevents tetraploidization. Genes

1478

Funding This work was supported by grants from Ligue Nationale Contre le Cancer (“Equipe labellisée HR 2013 and 2016”: AR, IP, ND, WV, HR) and Institut National de la Santé et de la Recherche Médicale (INSERM). AR was funded by a grant from FRM (SPF20140129106). IP and WV are partners of the Laboratory of Excellence Globule Rouge-Excellence funded by the program “Investissements d’avenir”.

Dev. 2006;20(19):2687-2700.

14. Yu FX, Guan KL. The Hippo pathway: regulators and regulations. Genes Dev. 2013;27(4):355-371.

15. Zhao B, Wei X, Li W, et al. Inactivation of YAP oncoprotein by the Hippo pathway is involved in cell contact inhibition and tissue growth control. Genes Dev. 2007;21 (21):2747-2761.

16. Margolis RL. Tetraploidy and tumor development. Cancer Cell. 2005;8(5):353-354.

17. Fujiwara T, Bandi M, Nitta M, Ivanova EV, Bronson RT, Pellman D. Cytokinesis failure generating tetraploids promotes tumorigenesis in p53-null cells.

Nature.

2005;437(7061):1043-1047.

18. Apostolidis PA, Woulfe DS, Chavez M, Miller WM, Papoutsakis ET. Role of tumor suppressor p53 in megakaryopoiesis and platelet function.

Exp Hematol.

2012;40(2):131-142.e4.

19. Fuhrken PG, Apostolidis PA, Lindsey S, Miller WM, Papoutsakis ET. Tumor suppressor protein p53 regulates megakaryocytic polyploidization and apoptosis. J Biol Chem. 2008;283(23):15589-15600.

20. Iancu-Rubin C, Mosoyan G, Glenn K, Gordon RE, Nichols GL, Hoffman R. Activation of p53 by the MDM2 inhibitor RG7112 impairs thrombopoiesis. Exp Hematol. 2014;42(2):137-145.e5.

21. Mahfoudhi E, Lordier L, Marty C, et al. P53 activation inhibits all types of hematopoietic progenitors and all stages of megakaryopoiesis. Oncotarget. 2016;7(22):3198031992.

22. Chang Y, Aurade F, Larbret F, et al.

Proplatelet formation is regulated by the Rho/ROCK pathway. Blood. 2007; 109(10):4229-4236.

23. Bluteau O, Langlois T, Rivera-Munoz P, et al. Developmental changes in human megakaryopoiesis. J Thromb Haemost.

2013;11(9):1730-1741.

24. Gao Y, Smith E, Ker E, et al. Role of RhoAspecific guanine exchange factors in regulation of endomitosis in megakaryocytes.

Dev Cell. 2012;22(3):573-584.

25. Ferguson GB, Martinez-Agosto JA. Yorkie and Scalloped signaling regulates Notchdependent lineage specification during Drosophila hematopoiesis. Curr Biol.

2014;24(22):2665-2672.

26. Milton CC, Grusche FA, Degoutin JL, et al.

The Hippo pathway regulates hematopoiesis in Drosophila melanogaster.

Curr Biol. 2014;24(22):2673-2680.

27. Hartmann EM, Campo E, Wright G, et al.

Pathway discovery in mantle cell lym-

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

phoma by integrated analysis of high-resolution gene expression and copy number profiling. Blood. 2010;116(6):953-961.

Lian I, Kim J, Okazawa H, et al. The role of YAP transcription coactivator in regulating stem cell self-renewal and differentiation.

Genes Dev. 2010;24(11):1106-1118.

Pan D. Hippo signaling in organ size control. Genes Dev. 2007;21(8):886-897.

Baccini V, Roy L, Vitrat N, et al. Role of p21(Cip1/Waf1) in cell-cycle exit of endomitotic megakaryocytes. Blood.

2001;98(12):3274-3282.

Besancenot R, Chaligne R, Tonetti C, et al.

A senescence-like cell-cycle arrest occurs during megakaryocytic maturation: implications for physiological and pathological megakaryocytic proliferation. PLoS Biol.

2010;8(9).

Yu FX, Zhao B, Panupinthu N, et al.

Regulation of the Hippo-YAP pathway by G-protein-coupled receptor signaling. Cell.

2012;150(4):780-791.

Zhao B, Li L, Lei Q, Guan KL. The HippoYAP pathway in organ size control and tumorigenesis: an updated version. Genes Dev. 2010;24(9):862-874.

Ali A, Bluteau O, Messaoudi K, et al.

Thrombocytopenia induced by the histone deacetylase inhibitor abexinostat involves p53-dependent and -independent mechanisms. Cell Death Dis. 2013;4:e738.

Nagaraj R, Gururaja-Rao S, Jones KT, et al.

Control of mitochondrial structure and function by the Yorkie/YAP oncogenic pathway. Genes Dev. 2012;26(18):20272037.

Sankaran VG, Orkin SH, Walkley CR. Rb intrinsically promotes erythropoiesis by coupling cell cycle exit with mitochondrial biogenesis. Genes Dev. 2008;22(4):463-475.

LeBleu VS, O'Connell JT, Gonzalez Herrera KN, et al. PGC-1alpha mediates mitochondrial biogenesis and oxidative phosphorylation in cancer cells to promote metastasis.

Nat Cell Biol. 2014;16(10):992-1003.

Garcia-Souza LF, Oliveira MF. Mitochondria: biological roles in platelet physiology and pathology. Int J Biochem Cell Biol. 2014;50:156-160.

Ramsey H, Zhang Q, Wu MX. Mitoquinone restores platelet production in irradiation-induced thrombocytopenia.

Platelets. 2015;26(5):459-466.

Al-Mehdi AB, Pastukh VM, Swiger BM, et al. Perinuclear mitochondrial clustering creates an oxidant-rich nuclear domain required for hypoxia-induced transcription.

Sci Signal. 2012;5(231):ra47.

haematologica | 2016; 101(12)

ARTICLE

Hematopoiesis

Revealing eltrombopag's promotion of human megakaryopoiesis through AKT/ERK-dependent pathway activation

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Christian A. Di Buduo,1,2 Manuela Currao,1,2 Alessandro Pecci,3 David L. Kaplan,4 Carlo L. Balduini,3 and Alessandra Balduini1,2,4

1 Department of Molecular Medicine, University of Pavia, Italy; 2Biotechnology Research Laboratories, IRCCS San Matteo Foundation, Pavia, Italy; 3Department of Internal Medicine, IRCCS Policlinico San Matteo Foundation and University of Pavia, Italy; and 4Department of Biomedical Engineering, Tufts University, Medford, MA, USA.

Haematologica 2016 Volume 101(12):1479-1488

ABSTRACT

E

ltrombopag is a small, non-peptide thrombopoietin mimetic that has been approved for increasing platelet count not only in immune thrombocytopenia and Hepatitis C virus-related thrombocytopenia, but also in aplastic anemia. Moreover, this drug is under investigation for increasing platelet counts in myelodysplastic syndromes.

Despite current clinical practice, the mechanisms governing eltrombopag’s impact on human hematopoiesis are largely unknown, in part due to the impossibility of using traditional in vivo models. To investigate eltrombopag’s impact on megakaryocyte functions, we employed our established in vitro model for studying hematopoietic stem cell differentiation combined with our latest 3-dimensional silk-based bone marrow tissue model. Results demonstrated that eltrombopag favors human megakaryocyte differentiation and platelet production in a dose-dependent manner. These effects are accompanied by increased phosphorylation of AKT and ERK1/2 signaling molecules, which have been proven to be crucial in regulating physiologic thrombopoiesis. These data further clarify the different mechanisms of action of eltrombopag when compared to romiplostim, which, as we have shown, induces the proliferation of immature megakaryocytes rather than platelet production, due to the unbalanced activation of AKT and ERK1/2 signaling molecules. In conclusion, our research clarifies the underlying mechanisms that govern the action of eltrombopag on megakaryocyte functions and its relevance in clinical practice.

Correspondence: alessandra.balduini@unipv.it

Received: March 25, 2016.

Accepted: August 4, 2016.

Pre-published: August 11, 2016. doi:10.3324/haematol.2016.146746

Introduction Hematopoiesis occurs in a complex microenvironment within the bone marrow, which provides an ideal habitat for the production of mature blood cells from the multipotent, self-renewing hematopoietic stem cells (HSCs).1,2 The failure of HSCs to guarantee the physiologic homeostasis of one or more progenitors for circulating blood cells leads to pathologic conditions, such as aplastic anemia (AA) or myelodysplastic syndromes (MDS), characterized by peripheral pancytopenia of heterogeneous severity.3,4 A selective cytopenia of blood platelets, namely thrombocytopenia, may occur because of mutations in genes relevant for the functions of maturing megakaryocytic progenitors, as in inherited thrombocytopenias (IT).5 However, thrombocytopenia may also be secondary to viral infections or autoimmune diseases.6,7 It is well known that thrombopoietin, through binding to its receptor (c-Mpl), which is expressed by HSCs and megakaryocytes, is a critical regulator of both HSC homeostasis and megakaryopoiesis.8 The opportunity to synthesize in vitro molecules able to mimic the physiologic effect of thrombopoietin haematologica | 2016; 101(12)

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/100/12/1479

©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.

1479

C.A. Di Buduo et al.

on platelet production recently opened new perspectives for the treatment of thrombocytopenic states.9 Among the small, non-peptide c-Mpl agonists,10,11 eltrombopag has been successfully employed to stimulate platelet production in patients suffering from IT caused by mutations of the MYH9 gene, immune thrombocytopenia (ITP), and thrombocytopenia due to hepatitis C infection.12-14 Moreover, improvement of thrombocytopenia has been obtained in patients with acute myeloid leukemia (AML) and MDS.15,16 Interestingly, promising clinical results have also demonstrated multi-lineage responses and the maintenance of normalized hematopoiesis in several patients suffering from AA,17-19 suggesting that eltrombopag may exert a beneficial effect on HSCs recovery by yet unexplored ways.

Unfortunately, the use of traditional animal models for studies on the mechanisms of action of eltrombopag is not possible due to its selective activity only in humans and chimpanzees.20 The management of both thrombocytopenias and hematologic malignancies is challenging.

Therefore, further improvement of the current knowledge strongly depends on the possibility to create laboratory assays in order to understand the effects of eltrombopag on human cells, both in physiologic and pathologic conditions.21 To this end, we have established a translational study made up of two complementary approaches. The first, based on the development of a culture system for the study of the basic mechanisms of the action of eltrombopag on human HSCs, focuses on the evaluation of the ability to promote megakaryopoiesis and platelet formation. The second, dedicated to the implementation of this knowledge into our recently established silk-based bone marrow model, integrates important physical and physiological elements characterizing the hematopoietic niche, conducive to evaluating platelet production.22 Together our studies demonstrate that eltrombopag significantly increases the activation of all the major c-Mpl downstream signaling pathways in a dose-dependent manner, and that this is paralleled by the differentiation of human HSCs, which results in an increased output of mature megakaryocytes which show an improved ability to form proplatelets and release platelets. Furthermore, we propose a novel mechanism explaining such effects on thrombopoiesis through the activation of AKT and ERK1/2 signaling molecules.

Methods Cell culture Human cord blood was collected following normal pregnancies and deliveries upon the informed consent of the parents; peripheral blood samples were collected from healthy volunteers after written informed consent. All samples were processed in accordance with the ethical committee of the IRCCS Policlinico San Matteo Foundation and the principles of the Declaration of Helsinki. Hematopoietic progenitor cells from cord and peripheral blood were separated by immunomagnetic bead selection, as previously described,23-25 and cultured in StemSpan medium (Stem Cell Technologies, Vancouver, Canada) supplemented with 1% L-glutamine, 1% penicillinstreptomycin, 10 ng/ml interleukin (IL)-6 and IL-11 (PeproTech, London, UK) and 50, 100, 200, 500 or 2000 ng/ml of eltrombopag (kindly supplied by GlaxoSmithKline), at 37°C in a 5%

1480

CO2 fully humidified atmosphere for 13 days. Instead of eltrombopag, 10 ng/ml recombinant human thrombopoietin (rHuTPO, PeproTech) was always used as the standard control to ensure normal megakaryocyte differentiation and functions.

Eltrombopag and all cytokines were dissolved in sterile distilled deionized water.

Immunofluorescence analysis At the end of the culture, a sample of 105 cells was cytospun on glass coverslips, fixed with 4% paraformaldehyde (PFA) for 20 minutes at room temperature (RT), blocked with 5% bovine serum albumin (BSA, Sigma, Milan, Italy) for 30 minutes, at RT, and then stained with the mouse monoclonal (clone SZ21) antibody against CD61 (1:100, Immunotech, Marseille, France), for 1 hour at RT, and the Alexa Fluor secondary antibody (1:500, Invitrogen, Milan, Italy) for 2 hours at RT. Nuclear counterstaining was performed using Hoechst 33258 (100 ng/ml, Sigma Aldrich, Milan, Italy) for 3 minutes at RT. Specimens were mounted in ProLong Gold Antifade Reagent (Invitrogen).

Negative controls were routinely performed by omitting the primary antibody. Immunofluorescence images were acquired by an Olympus BX51 microscope (Olympus, Deutschland GmbH, Hamburg, Germany).

For immunofluorescence imaging of the silk-based bone marrow system, samples were fixed in 4% PFA for 20 minutes and blocked with 5% BSA for 30 minutes, at RT. Samples were probed with anti-CD61 (1:100) overnight at 4°C, and then immersed in Alexa Fluor secondary antibody (1:500) for 2 hours at RT. Nuclei were stained with Hoechst. Samples were imaged by an Olympus FluoView FV10i confocal laser-scanning microscope (Olympus). Silk fibroin scaffolds were stained with Hoechst in all experiments.26

Analysis of proplatelet formation For the analysis of proplatelet yields by mature megakaryocytes, 12 mm glass coverslips were coated with 100 µg/ml fibrinogen (Merck Millipore, Milan, Italy) for 2 hours at RT and subsequently blocked with 1% BSA (Sigma) for 1 hour at RT. Megakaryocytes at day 13 of differentiation, either from liquid or 3D cultures, were harvested and plated onto fibrinogencoated coverslips in 24-well plates, and allowed to adhere for 16 hours at 37°C in a 5% CO2 fully humidified atmosphere.

After incubation, cells were fixed in 4% PFA, permeabilized with 0.5% Triton X-100, blocked with 5% BSA and stained with a rabbit anti-β1-tubulin antibody (kindly supplied by Prof.

J. Italiano Jr). Megakaryocytes extending proplatelets were identified as cells extending β1-tubulin positive long filamentous structures ending with platelet-sized tips.

Immunofluorescence images were acquired as reported above.

Phase contrast images were obtained using an Olympus IX53 microscope (Olympus).

Statistics Values are expressed as the mean plus or minus the standard deviation (mean±SD). A Student’s t-test was performed for paired observations. ANOVA, followed by the Bonferroni post hoc t-test, was performed for grouped observations. Spearman’s rank-order correlation coefficient (rs) was used to measure the association between two variables. A value of at least P<0.05 was considered statistically significant. All experiments were independently replicated at least three times.

A detailed description of silk bone marrow model preparation, western blotting and flow cytometry analysis have been reported in the Online Supplementary Methods.

haematologica | 2016; 101(12)

Ex vivo inquiry on Eltrombopag mechanism of action

Results

utilized, a concentration able to induce optimal megakaryocytic maturation and function, as shown in our previous research.23-25 After 13 days of culture, 50 and 100 ng/ml of eltrombopag failed to promote megakaryocyte differentiation (data not shown). Conversely, when concentrations of eltrombopag, in the same range as those measured in the serum of healthy subjects during oral eltrombopag administration (Cmax 7-10 mg/ml),27,28 were used in culture, megakaryocytes efficiently differentiated from their progenitors. Specifically, starting from 200 to 500 and 2000

Eltrombopag supports full megakaryocyte differentiation and maturation To investigate the effect of eltrombopag in promoting megakaryocyte differentiation, human cord blood-derived HSCs were cultured in the presence of 50, 100, 200, 500 or 2000 ng/ml eltrombopag. As a positive control of normal megakaryocyte differentiation, 10 ng/ml of rHuTPO was

A

B

C

D

E

Figure 1. Eltrombopag promotes megakaryocyte differentiation in vitro. Megakaryocytes (MKs) were differentiated from human umbilical cord blood progenitors and cultured for 13 days in the presence of 200, 500 and 2000 ng/mL eltrombopag or 10 ng/mL recombinant human thrombopoietin (rHuTPO) as positive control.

(A) Representative immunofluorescence staining of CD61 (green=CD61; blue=nuclei; scale bar=25 mm) and flow cytometry analysis of ploidy levels (PI=propidium iodide) in mature megakaryocytes at the end of the culture. (B) Percentage of CD61+CD42b+ megakaryocytes at the end of the culture. Data are presented as mean±SD (P= not significant (NS)). (C) Statistical analysis of ploidy levels. Data are presented as mean±SD (P=NS). (D) Western blot analysis of megakaryocytic transcription factors RUNX-1 and NF-E2. Samples were probed with 'anti-β-actin' antibody to ensure equal loading. (E) Output was calculated as the number of CD61+CD42b+ cells at day 13 of culture with respect to the total starting number of hematopoietic progenitor cells. Histograms show the fold increases of megakaryocyte output in the presence of 200, 500 or 2000 ng/mL eltrombopag with respect to 10 ng/mL rHuTPO. Data are expressed as mean ± SD (*P<0.05).

haematologica | 2016; 101(12)

1481

C.A. Di Buduo et al.

ng/ml of eltrombopag, megakaryocyte maturation was observed with more than 90% of mature megakaryocytes displaying increased DNA content, as assessed by both immunofluorescence and flow cytometry analysis (Figure 1A-C). Furthermore, Western blot of protein lysates revealed similar expression levels of RUNX1 and NF-E2 transcription factors with respect to rHuTPO, suggesting comparable induction of late megakaryocyte differentiation (Figure 1D). Despite the similar efficacy with respect to 200 ng/ml eltrombopag in promoting megakaryocyte differentiation, 500 ng/ml and 2000 ng/ml eltrombopag also resulted in an increased absolute number of megakaryocytes. Specifically, megakaryocyte output in the presence of 500 ng/ml and 2000 ng/ml eltrombopag prompted a significant 2- and 3-fold increase, respectively, compared to 200 ng/ml eltrombopag. The data were normalized with respect to 10 ng/ml rHuTPO as the positive control (Figure 1E). Similar results were obtained by culturing human adult peripheral blood progenitor cells with 500 ng/ml eltrombopag, which determined an increase of megakaryocyte output with respect to 10 ng/ml rHuTPO (Online Supplementary Figures S1A and S1B).

500 or 2000 ng/ml eltrombopag, or 10 ng/ml rHuTPO, used as the positive control condition (Figure 2Ai-iv).

Importantly, the beneficial effect of eltrombopag on megakaryocyte function was further sustained by the evidence that the increased output observed in response to 500 or 2000 ng/ml was paralleled by a significant increase in the number of cells displaying proplatelet branching, with respect to both 200 ng/ml eltrombopag or 10 ng/ml rHuTPO (Figure 2Av-viii). Specifically, a comparison of the percentage of proplatelet forming megakaryocytes among the different tested conditions highlighted that 500 ng/ml eltrombopag determined a significant 2-fold increase in the percentage of proplatelet forming megakaryocytes, while a 4-fold increase was reached with 2000 ng/ml eltrombopag (Figure 2B).

Accordingly, megakaryocytes likewise differentiated from human adult peripheral blood progenitor cells in the presence of 500 ng/ml eltrombopag, showing a significantly increased proplatelet formation compared to those differentiated in the presence of 10 ng/ml rHuTPO (Online Supplementary Figures S1C and S1D).

Eltrombopag promotes proplatelet formation

Eltrombopag promotes megakaryocyte differentiation in the silk-based 3D bone marrow model

To investigate the ability of megakaryocyte to form platelets, at the end of the culture cells were harvested and plated on fibrinogen, an extracellular matrix component known to support proplatelet formation and branching.23 Proplatelet formation did not occur in the presence of 50 and 100 ng/ml eltrombopag (data not shown), consistent with the lack of megakaryocyte differentiation in these conditions. Conversely, immunofluorescence analysis showed a normal architecture of proplatelet forming megakaryocytes, with cells displaying β1-tubulin positive long filamentous structures and bulbous tips at their terminal ends, in the presence of 200,

We have recently engineered a silk protein scaffoldbased tissue system to mimic the 3D spongy architecture that surrounds the marrow vasculature, in order to support mature megakaryocyte function ex vivo.22 We took advantage of this system to assess HSC differentiation towards the megakaryocyte lineage in the presence of eltrombopag (Figure 3A). Human cord blood-derived HSCs were seeded inside the sponge immediately after isolation and differentiated for two weeks, in the same conditions used for liquid cultures. The silk scaffolds, in the presence of 200 ng/ml of eltrombopag, supported megakaryocyte differentiation, which appeared fully

A

B

Figure 2. Eltrombopag sustains proplatelet formation in vitro. (Ai-iv) Analysis of proplatelet structure was performed by immunofluorescence staining of the megakaryocyte-specific cytoskeleton component β1-tubulin (green=β1-tubulin; blue=nuclei; scale bar=25 mm). In all tested conditions, the representative pictures show similar elongation of proplatelet shafts, with the presence of bulbous tips at the terminal ends of each branch, resembling mature platelets. (Av-viii) Representative light microscopy images of proplatelet formation by human megakaryocytes cultured in the presence of recombinant human thrombopoietin (rHuTPO) or increasing concentrations of eltrombopag (scale bar=50 mm). Arrows indicate proplatelet-forming megakaryocytes. (B) The percentage of proplatelet forming megakaryocytes was calculated as the number of cells displaying long filamentous pseudopods with respect to the total number of round megakaryocytes per analyzed field. Histograms show the fold increase of proplatelet formation in the presence of 200, 500 or 2000 ng/mL eltrombopag with respect to 10 ng/mL rHuTPO, used as optimum standard condition to test proplatelet formation. Data are expressed as mean ± SD (*P<0.05).

1482

haematologica | 2016; 101(12)

Ex vivo inquiry on Eltrombopag mechanism of action

mature as compared to 10 ng/ml of rHuTPO (Figure 3B,C).

To confirm this feature, megakaryocytes were recovered from the silk sponge and seeded onto fibrinogen coated coverslips, where they formed normal branched proplatelets displaying β1-tubulin staining (Figure 3D). The same results were obtained with 500 ng/ml and 2000 ng/ml eltrombopag (data not shown).

Eltrombopag sustains human platelet generation ex vivo In parallel, silk sponges, previously assembled around a silk microtube functionalized via the entrapment of extracellular matrix components and SDF-1α as a chemoattractant, were perfused at a shear rate of 60 s-1 to allow ex vivo platelet collection (Figure 4A). Confocal microscopy revealed that human mature CD61+ megakaryocytes, contained in the silk sponges, were able to elongate proplatelets through the silk microtube wall (Figure 4B). In addition, media flow mimicking the bloodstream into the bone marrow sinusoids, helped platelet collection into gaspermeable bags containing acid citrate dextrose (ACD) as an anticoagulant. Ex vivo collected platelets were double stained with anti-CD61 and anti-CD42b antibodies, and their number evaluated by a counting bead standard (Figure 4C). The mean number of CD61+CD42b+ platelets collected per 3D tissue perfusion system increased with increasing concentrations of eltrombopag (Figure 4D). We then calculated the Spearman’s rank-order correlation coefficient (rs) to assess the statistical correlation between the percentage of proplatelet formation in vitro and the number of platelets collected in the 3D ex vivo system obtaining a value of rs =0.9002, with a P value of <0.0001. Importantly, analysis of PAC-1 binding demonstrated comparable platelet activation in all tested conditions (Figure 4E).

Eltrombopag activates c-Mpl downstream pathways and keeps a balanced activation of AKT and ERK1/2 signaling molecules Upon its binding to the c-Mpl receptor, thrombopoietin stimulates multiple biochemical signals, including the Janus Kinase (JAK)/signal transducers and activators of transcription 3 (STAT3) and STAT5, the phosphoinositide 3-kinase/AKT (PI3K/AKT), and the extracellular signalregulated Kinase1/2 (ERK1/2).29 Here, we demonstrated that all these signals are activated upon stimulation of megakaryocytes with eltrombopag (Online Supplementary Figure S2). Among these, the JAK/STAT3 and JAK/STAT5 pathways are known to promote megakaryocyte differentiation, survival, and expansion.30-32 Eltrombopag determined an increase of STAT3 and STAT5 phosphorylation as compared to rHuTPO (Figure 5A-C), consistent with our data demonstrating normal megakaryocyte differentiation, but increased output. Of note, the balance between the activation of AKT and ERK1/2 has been found to be crucial to regulate proplatelet formation. In particular, inhibition of AKT phosphorylation resulted in impaired proplatelet formation,33,34 while over-activation of AKT not associated with a parallel ERK1/2 over-activation led to hyperproliferation of immature megakaryocytes with a defective capacity to form proplatelets.25 In contrast, pathologic over-activation of ERK1/2 only, not paralleled

by AKT over-activation, inhibited proplatelet formation.35 Here we demonstrate that eltrombopag sustains parallel AKT and ERK1/2 phosphorylation in a dose-dependent manner (Figure 5A), with a significantly increased activation of both pathways in the presence of 500 and 2000 ng/ml of eltrombopag with respect to 10 ng/ml of rHuTPO (Figure 5D,E). Similar results were obtained by culturing

A

B

C

D

Figure 3. Eltrombopag stimulates ex vivo megakaryocyte differentiation within the 3D silk bone marrow model. (A) Aqueous silk solution was mixed with salt particles and dried at room temperature overnight. After leaching out the salt, the resulting porous silk sponge was trimmed and sterilized. CD34+ hematopoietic stem cells from human cord blood were then seeded into the sponges and cultured for 13 days in the presence of 10 ng/mL recombinant human thrombopoietin (rHuTPO) or 200 ng/mL eltrombopag. (B) Confocal microscopy analysis of CD61+ megakaryocytes after 13 days of differentiation within the silk sponges under the different tested conditions (red=CD61; blue=silk; scale bar=100 mm). (C) Flow cytometry analysis of samples from the silk-sponges demonstrated an almost comparable percentage of CD61+CD42b+ cells between rHuTPO and eltrombopag. (D) Representative β1-tubulin staining of proplatelet formation by megakaryocytes collected from the silk sponge scaffold and seeded on fibrinogen. Both rHuTPO and eltrombopag supported normal proplatelet extension (green=β1-tubulin; blue=nuclei; scale bar=20 mm).

haematologica | 2016; 101(12)

1483

C.A. Di Buduo et al.

A

B

C

D

E

Figure 4. Eltrombopag sustains ex vivo platelet release. (A) Silk microtubes were prepared by gel spinning aqueous silk solutions, containing polyethylene oxide porogen, around a wire and functionalized via entrapment of SDF-1α and extracellular matrix components. The resulting microtubes were fitted into the bioreactor chamber and a silk sponge was fabricated around them. Megakaryocytes (MKs) cultured within the system could migrate toward the microtube, adhere and extend proplatelets through the microtube wall to release platelets into the microtube lumen. (Bi) Bioreactor chamber containing the silk microtube-sponge system (scale bar=5 mm). (Bii) Confocal microscopy analysis of the silk microtube-sponge system before megakaryocyte seeding (blue=silk; scale bar=100 mm). (Biii) Mature megakaryocytes cultured into the silk sponge scaffold (red=CD61; blue=silk; scale bar=100 mm). (Biv) Representative megakaryocyte extending proplatelets through the silk microtube wall (red=CD61; blue=silk; scale bar=50 mm). (C) Ex vivo produced platelets (PLTs) were stained with CD61 and CD42b antibodies, and their number calculated by mixing samples with counting beads before flow cytometry analysis. (D) The graph shows the absolute number of CD61+CD42b+ platelets released in the presence of rHuTPO or increasing doses of eltrombopag (mean±SD, *P<0.05). (E) Ex vivo collected platelets revealed increased PAC-1 binding after stimulation with ADP or thrombin (mean±SD). rHuTPO: recombinant human thrombopoietin; SSC: side scatter; FSC: forward scatter; ADP: adenosine diphosphate; SDF-1α: stromal cell-derived factor 1α.

1484

haematologica | 2016; 101(12)

Ex vivo inquiry on Eltrombopag mechanism of action

CD34+ hematopoietic progenitors for 1 day with 2000 ng/ml eltrombopag (Figure 6A-E). Consistently, even within 1 hour upon stimulation of both megakaryocytes and hematopoietic progenitors, 2000 ng/ml eltrombopag sustained a marked increase of STAT3, STAT5, AKT and ERK1/2 phosphorylation, compared to 10 ng/ml rHuTPO (Online Supplementary Figures S3A and S3B).

Discussion Eltrombopag is a non-peptide synthetic thrombopoietin-mimetic that binds to the transmembrane domain of c-Mpl, leading to its activation.10,11 This compound has been approved by the Food and Drug Administration

A

(FDA) for the treatment of ITP and severe AA, and also for increasing platelet counts in subjects with thrombocytopenia due to hepatitis C infection who are undergoing antiviral treatment.12,13,17-19 Moreover, a pivotal study on IT demonstrated that eltrombopag can ameliorate thrombocytopenia in subjects affected by mutations of the MYH9 gene.14 Preliminary data are available concerning the treatment of AML and MDS.15,16,36,37 Despite the wide spectrum of conditions where eltrombopag has been used, little is known about the mechanisms of action of this drug on human HSCs. The results of our study demonstrated that eltrombopag promoted successful in vitro human cord and peripheral blood-derived HSCs differentiation towards the megakaryocytic lineage, as assessed by the maturation of high ploidy megakaryocytes expressing the lineage-spe-

B

C

D

E

Figure 5. Eltrombopag activates all main biochemical pathways downstream of c-Mpl in in vitro differentiated megakaryocytes. (A) Lysates from megakaryocyte cultures in the presence of 10 ng/mL recombinant human thrombopoietin (rHuTPO), or 200, 500 and 2000 ng/mL eltrombopag, were obtained at day 13 of differentiation. Samples were probed for phosphorylated STAT3 (pSTAT3), phosphorylated STAT5 (pSTA5), phosphorylated AKT (pAKT) and phosphorylated ERK1/2 (pERK1/2). Total STAT3, STAT5, AKT, ERK1/2 and β-actin were revealed to ensure equal loading. (B-D) Densitometric analysis demonstrated sustained activation by eltrombopag of all signaling molecules in a dose-dependent manner, with a significant difference with respect to 10 mg/mL rHuTPO, as standard control condition (*P<0.05).

haematologica | 2016; 101(12)

1485

C.A. Di Buduo et al.

cific markers CD61 and CD42b and the transcription factors RUNX-1 and NF-E2. Our data were consistent with recent findings by Jeong et al., demonstrating that eltrombopag can stimulate megakaryocyte differentiation from hematopoietic progenitors isolated from patients with relapsed multiple myeloma and normal controls.38 Interestingly, concentrations of eltrombopag, similar to those observed in patients receiving this drug in vivo,27,28 stimulated the activation of all the main biochemical pathways downstream of the c-Mpl receptor. This activation resulted in the expansion of fully differentiated megakaryocytes with an enhanced ability to extend long branched proplatelets. Taking advantage of our recently developed 3D bone marrow tissue model,22 we also demonstrated that eltrombopag promoted the release of functional platelets under flow conditions mimicking the bloodstream into marrow sinusoids.

The effect of eltrombopag appears to be markedly different from that recently observed in the same experimental conditions with the thrombopoietin mimetic romiplostim,25 a peptibody that binds to the c-Mpl competing

A

with endogenous thrombopoietin for the same binding site.39 Romiplostim promoted the significant proliferation of immature megakaryocytes, displaying an impaired ability to form proplatelets.25 Of note, these different effects reflected different patterns of activation of AKT and ERK1/2 kinases, which are known to be crucial for the regulation of megakaryocyte maturation and the process of platelet production which occurs in healthy conditions.3335,40-42 In the present study we observed that eltrombopag sustained a concomitant phosphorylation of both AKT and ERK1/2 both in hematopoietic progenitors and mature megakaryocytes, while in our previous work romiplostim was shown to stimulate mainly megakaryocyte proliferation by increasing AKT phosphorylation only.25 Overall, our studies on the effects of eltrombopag and romiplostim on in vitro human megakaryopoiesis suggest that these two drugs increase platelet output by different mechanisms: the former by stimulating increased megakaryocyte maturation and proplatelet formation, the latter by increasing mainly megakaryocyte proliferation without a correspondent parallel stimulation of megakary-

B

C

D

E

1486

Figure 6. c-Mpl downstream signaling molecules are highly activated in hematopoietic progenitors in the presence of eltrombopag. (A) Lysates from hematopoietic progenitors in the presence of 10 ng/mL recombinant human thrombopoietin (rHuTPO) or 2000 ng/mL eltrombopag (Eltromb.) were obtained after 1 day of culture. Samples were probed for phosphorylated STAT3 (pSTAT3), phosphorylated STAT5 (pSTA5), phosphorylated AKT (pAKT) and phosphorylated ERK1/2 (pERK1/2). Total STAT3, STAT5, AKT, ERK1/2 and β-actin were revealed to ensure equal loading. (B-D) Densitometric analysis demonstrated significant over-activation of all signaling molecules in the presence of 2000 ng/mL eltrombopag with respect to 10 mg/mL rHuTPO, as standard control condition (*P<0.05).

haematologica | 2016; 101(12)

Ex vivo inquiry on Eltrombopag mechanism of action Figure 7. Representative model of different mechanisms of action of eltrombopag and romiplostim. Bone marrow, contained in spongy bones, is a tridimensional network of branching sinusoids surrounding islets of hematopoietic cells.

Within this environment hematopoietic stem cells (HSCs) undergo self-renewal as well as differentiation into committed lineages in order to support the physiological homeostasis of all blood cells. Megakaryopoiesis takes place due to the activation of c-Mpl, the thrombopoietin receptor, which promotes HSC commitment and differentiation. Upon c-Mpl binding, eltrombopag promotes higher phosphorylation of both AKT and ERK than recombinant human thrombopoietin (rHuTPO), thus ensuring proper proplatelet formation (PPF). At variance with eltrombopag, romiplostim promotes increased AKT but not ERK activation with respect to rHuTPO, thus promoting megakaryocyte proliferation rather than PPF. PLTs: platelets; RBCs: red blood cells.

ocyte maturation (Figure 7). Further studies are required to ascertain whether these differences in vitro also occur in vivo in patients receiving these drugs. Nevertheless, these data provide a rational basis to the observation that ITP patients who do not respond to romiplostim may respond to eltrombopag, and vice versa.43 The peculiarities of the mechanisms of action of eltrombopag could also provide an explanation for the observation that its effect is additive to that of thrombopoietin or romiplostim both in vitro and in vivo.20,44 It is well known that c-Mpl is expressed not only by hematopoietic progenitors, but also by many malignant hematopoietic cell types, and different in vitro experiments demonstrated that thrombopoietin may sustain the growth of both leukemic cell lines and human leukemic cells.45,46 The concern that thrombopoietin mimetics might favor leukemogenesis has been extensively debated. A rise in blast counts has been seen in patients with MDS receiving romiplostim,47 and a recent retrospective study of a large series of ITP patients indicated an association between the administration of thrombopoietin mimetics and the development of AML.48 On the other hand, recent studies concluded that eltrombopag allowed the formation of normal megakaryocytic colonies without the stimulation of malignant blasts from AML and MDS patients, and also had a strong anti-leukemic effect.37,49 Interestingly, this effect was independent of c-Mpl, but was instead mediated through the modulation of intracellular iron content.49 Finally, Kalota et al. demonstrated that eltrombopag dramatically decreases ROS levels, leading to a disruption of AML intracellular

References 1. Morrison SJ, Scadden DT. The bone marrow niche for haematopoietic stem cells. Nature.

2014;505(7483):327-334.

2. Malara A, Abbonante V, Di Buduo CA,

haematologica | 2016; 101(12)

metabolism and rapid cell death.50 Thus, conflicting results were obtained concerning the potential leukemogenic effect of thrombopoietin mimetics, and further studies are required to clarify this important issue.

In conclusion, this study defined the mechanisms of action of the thrombopoietin mimetic eltrombopag on hematopoietic progenitors and human megakaryocytes, and identified relevant differences with respect to those of romiplostim. Importantly, this observation warrants further investigation on outcomes from the HSCs of patients affected by thrombocytopenias and overall bone marrow pathologies. Understanding the efficacy and safety of thrombopoietin mimetics on pathologic samples before treating patients in vivo would represent an important step towards creating more personalized and effective therapies.

Acknowledgements The authors would like to thank Dr. Gianluca Viarengo and Prof. Federica Meloni for technical assistance with the flow cytometry analysis; Dr. Cesare Perotti for supplying human cord blood; Dr. Lorenzo Tozzi and Daniel Smoot for technical assistance in the preparation of gel-spun silk microtubes; Prof. Joseph Italiano for providing β1-tubulin antibody. This paper was supported by the Cariplo Foundation (2012-0529, 2013-0717), ERA-Net for Research Programmes on Rare Diseases (EUPLANE), the Italian Ministry of Health (RF-20102310098) and US National Institutes of Health (R01 EB016041-01). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Tozzi L, Currao M, Balduini A. The secret life of a megakaryocyte: emerging roles in bone marrow homeostasis control. Cell Mol Life Sci. 2015;72(8):1517-1536.

3. Marsh JC. Bone marrow failure syndromes.

Clin Med. 2005;5(4):332-336.

4. Tiu R, Gondek L, O'Keefe C, Maciejewski

JP. Clonality of the stem cell compartment during evolution of myelodysplastic syndromes and other bone marrow failure syndromes. Leukemia. 2007;21(8):1648-1657.

5. Balduini CL, Pecci A, Noris P. Diagnosis and management of inherited thrombocytopenias. Semin Thromb Hemost. 2013;39(2):

1487

C.A. Di Buduo et al.

161-171.

6. Nagamine T, Ohtuka T, Takehara K, Arai T, Takagi H, Mori M. Thrombocytopenia associated with hepatitis C viral infection. J Hepatol. 1996;24(2):135-140.

7. Rodeghiero F, Stasi R, Gernsheimer T, et al.

Standardization of terminology, definitions and outcome criteria in immune thrombocytopenic purpura of adults and children: report from an international working group.

Blood. 2009;113(11):2386-2393.

8. Hitchcock IS, Kaushansky K. Thrombopoietin from beginning to end. Br J Haematol. 2014;165(2):259-268.

9. Kuter DJ. The biology of thrombopoietin and thrombopoietin receptor agonists. Int J Hematol. 2013;98(1):10-23.

10. Erickson-Miller CL, DeLorme E, Tian SS, et al. Discovery and characterization of a selective, nonpeptidyl thrombopoietin receptor agonist. Exp Hematol. 2005;33(1):85-93.

11. Safonov IG, Heerding DA, Keenan RM, et al.

New benzimidazoles as thrombopoietin receptor agonists. Bioorg Med Chem Lett.

2006;16(5):1212-1216.

12. Bussel JB. Update on eltrombopag for ITP. Oncology (Williston Park). 2009;23(13): 1177-1178.

13. McHutchison JG, Dusheiko G, Shiffman ML, et al. Eltrombopag for thrombocytopenia in patients with cirrhosis associated with hepatitis C. N Engl J Med. 2007;357 (22):2227-2236.

14. Pecci A, Gresele P, Klersy C, et al.

Eltrombopag for the treatment of the inherited thrombocytopenia deriving from MYH9 mutations. Blood. 2010;116(26): 5832-5837.

15. Platzbecker U, Wong RS, Verma A, et al.

Safety and tolerability of eltrombopag versus placebo for treatment of thrombocytopenia in patients with advanced myelodysplastic syndromes or acute myeloid leukaemia: a multicentre, randomised, placebo-controlled, double-blind, phase 1/2 trial. Lancet Haematol.

2015;2(10):e417-e426.

16. Santini V, Fenaux P. Treatment of myelodysplastic syndrome with thrombomimetic drugs. Semin Hematol. 2015;52(1):38-45.

17. Olnes MJ, Scheinberg P, Calvo KR, et al.

Eltrombopag and improved hematopoiesis in refractory aplastic anemia. N Engl J Med. 2012;367(1):11-19.

18. Desmond R, Townsley DM, Dumitriu B, et al. Eltrombopag restores trilineage hematopoiesis in refractory severe aplastic anemia that can be sustained on discontinuation of drug. Blood. 2014;123(12):18181825.

19. McCormack PL. Eltrombopag: a review of its use in patients with severe aplastic anaemia. Drugs. 2015;75(5):525-531.

20. Erickson-Miller CL, Delorme E, Tian SS, et al. Preclinical activity of eltrombopag (SB497115), an oral, nonpeptide thrombopoietin receptor agonist. Stem Cells. 2009;27 (2):424-430.

21. Balduini A, Di Buduo CA, Kaplan DL. Translational approaches to functional platelet production ex vivo. Thromb Haemost. 2016;115(2):250-256.

22. Di Buduo CA, Wray LS, Tozzi L, et al.

Programmable 3D silk bone marrow niche

1488

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

for platelet generation ex vivo and modeling of megakaryopoiesis pathologies. Blood.

2015;125(14):2254-64.

Balduini A, Pallotta I, Malara A, et al.

Adhesive receptors, extracellular proteins and myosin IIA orchestrate proplatelet formation by human megakaryocytes. J Thromb Haemost. 2008;6(11):1900-1907.

Balduini A, Badalucco S, Pugliano MT, et al.

In vitro megakaryocyte differentiation and proplatelet formation in Ph-negative classical myeloproliferative neoplasms: distinct patterns in the different clinical phenotypes.

PLoS One. 2011;6(6):e21015.

Currao M, Balduini CL, Balduini A. High doses of romiplostim induce proliferation and reduce proplatelet formation by human megakaryocytes. PLoS One. 2013;8(1): e54723.

Talukdar S, Nguyen QT, Chen AC, Sah RL, Kundu SC. Effect of initial cell seeding density on 3D-engineered silk fibroin scaffolds for articular cartilage tissue engineering.

Biomaterials. 2011;32(34):8927-8937.

Deng Y, Madatian A, Wire MB, et al.

Metabolism and disposition of eltrombopag, an oral, nonpeptide thrombopoietin receptor agonist, in healthy human subjects. Drug Metab Dispos. 2011;39(9):1734-1746.

Stasi R, Rhodes E, Benjamin R, et al. The Emergence of Thrombopoietin Receptor Agonists as a Novel Treatment for Immune Thrombocytopenia. Eur Oncol Haematol.

2011;7(1):63-70.

Kaushansky K. The molecular mechanisms that control thrombopoiesis. J Clin Invest.

2005;115(12):3339-3347.

Matsumura I, Ishikawa J, Nakajima K, et al.

Thrombopoietin-induced differentiation of a human megakaryoblastic leukemia cell line, CMK, involves transcriptional activation of p21(WAF1/Cip1) by STAT5. Mol Cell Biol. 1997;17(5):2933-2943.

Kirito K, Watanabe T, Sawada K, Endo H, Ozawa K, Komatsu N. Thrombopoietin regulates Bcl-xL gene expression through Stat5 and phosphatidylinositol 3-kinase activation pathways. J Biol Chem. 2002;277(10):83298337.

Kirito K, Osawa M, Morita H, et al. A functional role of Stat3 in in vivo megakaryopoiesis. Blood. 2002;99(9):3220-3227.

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.

Badalucco S, Di Buduo CA, Campanelli R, et al. Involvement of TGFβ1 in autocrine regulation of proplatelet formation in healthy subjects and patients with primary myelofibrosis. Haematologica. 2013;98(4):514-517.

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.

Li W, Morrone K, Kambhampati S, Will B, Steidl U, Verma A. Thrombocytopenia in MDS: epidemiology, mechanisms, clinical consequences and novel therapeutic strategies. Leukemia. 2016;30(3):536-544.

Will B, Kawahara M, Luciano JP, et al. Effect of the nonpeptide thrombopoietin receptor

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

agonist Eltrombopag on bone marrow cells from patients with acute myeloid leukemia and myelodysplastic syndrome. Blood.

2009;114(18):3899-3908.

Jeong JY, Levine MS, Abayasekara N, Berliner N, Laubach J, Vanasse GJ. The nonpeptide thrombopoietin receptor agonist eltrombopag stimulates megakaryopoiesis in bone marrow cells from patients with relapsed multiple myeloma. J Hematol Oncol. 2015;8:37.

Kuter DJ. Milestones in understanding platelet production: a historical overview. Br J Haematol. 2014;165(2):248-258.

Mazharian A, Watson SP, Séverin S. Critical role for ERK1/2 in bone marrow and fetal liver-derived primary megakaryocyte differentiation, motility, and proplatelet formation. Exp Hematol. 2009;37(10):12381249.e5.

Séverin S, Ghevaert C, Mazharian A. The mitogen-activated protein kinase signaling pathways: role in megakaryocyte differentiation. J Thromb Haemost. 2010;8(1):17-26.

Besancenot R, Chaligné R, Tonetti C, et al. A senescence-like cell-cycle arrest occurs during megakaryocytic maturation: implications for physiological and pathological megakaryocytic proliferation. PLoS Biol.

2010;8(9). pii: e1000476.

D'Arena G, Guariglia R, Mansueto G, et al.

No cross-resistance after sequential use of romiplostim and eltrombopag in chronic immune thrombocytopenic purpura. Blood.

2013;121(7):1240-1242.

Ruiz-Arguelles GJ, Ruiz-Delgado GJ, Velázquez-Sánchez-de-Cima S, ZamoraOrtiz G. Simultaneous romiplostin, eltrombopag, and prednisone were successful in severe thrombocytopenia of Evans syndrome refractory to hydrocortisone, splenectomy, intravenous IgG, and rituximab. Hematology. 2013;18(3):175-177.

Drexler HG, Zaborski M, Quentmeier H. Thrombopoietin supports the continuous growth of cytokine-dependent human leukemia cell lines. Leukemia. 1997;11(4): 541-551.

Corazza F, Hermans C, D'Hondt S, et al.

Circulating thrombopoietin as an in vivo growth factor for blast cells in acute myeloid leukemia. Blood. 2006;107(6):2525-2530.

Prica A, Sholzberg M, Buckstein R. Safety and efficacy of thrombopoietin-receptor agonists in myelodysplastic syndromes: a systematic review and meta-analysis of randomized controlled trials. Br J Haematol.

2014;167(5):626-638.

Oshima Y, Yuji K, Tanimoto T, Hinomura Y, Tojo A. Association between acute myelogenous leukemia and thrombopoietin receptor agonists in patients with immune thrombocytopenia. Intern Med. 2013;52(19): 2193-2201.

Roth M, Will B, Simkin G, et al.

Eltrombopag inhibits the proliferation of leukemia cells via reduction of intracellular iron and induction of differentiation. Blood.

2012;120(2):386-394.

Kalota A, Selak MA, Garcia-Cid LA, Carroll M. Eltrombopag modulates reactive oxygen species and decreases acute myeloid leukemia cell survival. PLoS One. 2015; 10(4):e0126691.

haematologica | 2016; 101(12)

ARTICLE

Red Cell Biology & its Disorders

Unexpected macrophage-independent dyserythropoiesis in Gaucher disease

Nelly Reihani,1 Jean-Benoit Arlet,2 Michael Dussiot,3 Thierry Billette de Villemeur,4 Nadia Belmatoug,5 Christian Rose,6 Yves Colin-Aronovicz,1 Olivier Hermine,7 Caroline Le Van Kim1* and Melanie Franco1*

Université Sorbonne Paris Cité, Université Paris Diderot, Inserm, INTS, Unité Biologie Intégrée du Globule Rouge, Laboratoire d’Excellence GR-Ex, Paris; 2 Sorbonne Paris-Cité, Université Paris Descartes, Service de Médecine Interne, Assistance publique-Hôpitaux de Paris, Hôpital Européen Georges Pompidou, Inserm UMR 1163, CNRS ERL 8254, Hôpital Necker, Institut Imagine, Laboratoire d’excellence GR-Ex, Paris; 3Sorbonne Paris-Cité, Université Paris Descartes, Inserm UMR 1163, CNRS ERL 8254, Institut Imagine, Hôpital Necker, Laboratoire d’excellence GR-Ex, Paris; 4 Sorbonne Université, Université Pierre et Marie Curie, Service de Neuropédiatrie Hôpital Trousseau, Assistance publique-Hôpitaux de Paris, Hôpital et GRC ConCer-LD, Paris; 5Hôpitaux universitaires Paris Nord Val de Seine, Assistance publique-Hôpitaux de Paris, Hôpital Beaujon, Service de Médecine Interne, Centre de Référence des Maladies Lysosomales, Clichy; 6Université Catholique de Lille, Hôpital Saint Vincent de Paul, Service d’Hématologie, Lille and 8Sorbonne Paris-Cité, Université Paris Descartes, Assistance publique-Hôpitaux de Paris, Hôpital Necker, Service d'Hématologie, Inserm UMR 1163, CNRS ERL 8254, Institut Imagine, Laboratoire d’excellence GR-Ex, Paris, Inserm UMR 1163, CNRS, France

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

1

*MF and CLVK contributed equally to this work

Haematologica 2016 Volume 101(11):1489-1498

ABSTRACT

G

aucher disease is a rare inherited disease caused by a deficiency in glucocerebrosidase leading to lipid accumulation in cells of mononuclear-macrophage lineage known as Gaucher cells.

Visceral enlargement, bone involvement, mild anemia and thrombocytopenia are the major manifestations of Gaucher disease. We have previously demonstrated that the red blood cells from patients exhibit abnormal properties, which indicates a new role in Gaucher disease pathophysiology. To investigate whether erythroid progenitors are affected, we examined the in vitro erythropoiesis from the peripheral CD34+ cells of patients and controls. CD34- cells were differentiated into macrophages and co-cultivated with erythroblasts. We showed an accelerated differentiation of erythroid progenitors without maturation arrest from patients compared to controls. This abnormal differentiation persisted in the patients when the same experiments were performed without macrophages, which strongly suggested that dyserythropoiesis in Gaucher disease is secondary to an inherent defect in the erythroid progenitors. The accelerated differentiation was associated with reduced cell proliferation. As a result, less mature erythroid cells were generated in vitro in the Gaucher disease cultures compared to the control. We then compared the biological characteristics of untreated patients according to their anemic status. Compared to the non-anemic group, the anemic patients exhibit higher plasma levels of growth differentiation factor-15, a marker of ineffective erythropoiesis, but they had no indicators of hemolysis and similar reticulocyte counts. Taken together, these results demonstrated an unsuspected dyserythropoiesis that was independent of the macrophages and could participate, at least in part, to the basis of anemia in Gaucher disease. haematologica | 2016; 101(12)

Correspondence: melanie.franco@inserm.fr

Received: May 2, 2016.

Accepted: July 26, 2016.

Pre-published: July 28, 2016. doi:10.3324/haematol.2016.147546

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/100/12/1489

©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.

1489

N. Reihani et al.

Introduction Gaucher disease (GD), the most common lysosomal storage disorder, is caused by autosomal recessive mutations in the gene encoding glucocerebrosidase (GCerase).

This enzyme is required for the degradation of glycosphingolipids, and its reduced activity results in the accumulation of the immediate substrates glucosylceramide (GlcCer) and its deacylated products, the glucosylsphingosines, within the macrophages.1 GD has been classified into three types depending on the absence (type 1, the most frequent form comprising 90% of patients) or the presence (type 2 and type 3) of neurological features. Type 2 GD is the most severe form, and the afflicted subjects have a very short life expectancy (less than 2 years). In GD types 1 and 3, visceral enlargement (splenomegaly and hepatomegaly), bone involvement and hematological manifestations (anemia, thrombocytopenia) represent the major symptoms of the disease. GD type 1 is treated efficiently by enzyme replacement therapy (ERT) that leads to a specific uptake of recombinant GCerase by the deficient macrophages.2 Substrate reduction therapy represents an alternative oral approach based on the reduced synthesis of glucosylceramide. These treatments have been demonstrated to have beneficial effects on anemia, thrombocytopenia, hepatosplenomegaly and bone crisis.1 Complications associated with GD have been attributed to the accumulation of GCerase substrates in cells of the monocytic lineage.3,4 Indeed, lipid-laden macrophages transformed into Gaucher cells that infiltrate the spleen, the liver and the bone marrow are considered the basis for the major GD symptoms. However, a growing number of studies performed on murine and in vitro models as well as those using cells from GD patients, indicate that the pathophysiology of GD may involve a wider array of cell types, including hematopoietic and mesenchymal cells, thymic T cells and progenitors, dendritic cells and osteoblasts.5-10 We have previously shown that the red blood cells (RBCs) of GD patients exhibit abnormal morphological, rheological and functional properties, and could be considered additional factors in the GD pathophysiology on the basis that they can trigger ischemic events.11 Importantly, a study of a large cohort revealed that anemia was the only risk factor for avascular osteonecrosis, the most debilitating skeletal complication for GD type 1 patients.12 Anemia affects 36% of GD patients.13 The underlying mechanism has not been fully described, although it has been generally attributed to hypersplenism and splenic sequestration.14 However, the severity of the anemia is not directly associated with the degree of splenomegaly, and the anemia may sometimes persist after splenectomy.15,16 Recent data have shown both bone marrow and hematopoietic abnormalities in GD, and bone marrow infiltration by Gaucher cells was observed in GD patients.17-20 It has also been reported that GD may affect the growth of erythroid progenitors and myelopoiesis of induced pluripotent stem cells (iPSCs).21 Moreover, an important extramedullary hematopoiesis with the presence of erythroid cells was observed in the spleen of a murine model of GD.9,10 We previously detected GCerase activity in normal erythroblasts but not in circulating RBCs.11 High lipid levels have been frequently reported in the plasma and RBCs of GD patients,22-25 which raises the possibility that the RBCs are overloaded with lipids due to the passive incorporation of GlcCer and/or 1490

that the erythroid progenitors are primarily affected. To evaluate the alterations in the erythroid progenitors, we performed in vitro experiments to study the erythroid differentiation from peripheral blood samples from GD type 1 patients. The circulating CD34+ and CD34â&#x20AC;&#x201C; cells from these patients were differentiated into erythroblasts and macrophages, respectively. By performing co-culture experiments of the erythroblasts with the macrophages, we showed accelerated differentiation that was associated with reduced cell proliferation of erythroid progenitors in the cells from the GD subjects compared to controls. This dyserythropoiesis was independent of the macrophage defects because the accelerated differentiation was also observed in the absence of macrophages. Although no evidence of peripheral hemolysis was observed in the anemic GD patients from our cohort, we observed an increased level of growth differentiation factor-15 (GDF-15), a marker of ineffective erythropoiesis, in their plasma. We proposed that dyserythropoiesis per se is the basis of the anemia observed in GD.

Methods Patients Patients were followed in the French Reference Center for Lysosomal Diseases. A total of 35 type 1 GD patients were recruited prospectively between June 2012 and July 2015. Patients were considered anemic when the hemoglobin (Hb) levels were strictly below 11.5 g/dL for children between 2 and 12 years old, 12 g/dL for women and 13 g/dL for men. No patient was splenectomised.

In this cohort, GD was moderate and 10 patients exhibited anemia. Hemograms and biochemical data related to anemia for the 20 untreated GD patients who had never received ERT were compared according to their anemic status. For the in vitro erythropoiesis study, blood samples were collected from 24 GD patients, including 9 untreated patients and 15 patients treated with ERT. Detailed clinical and biological information is provided in Table 1.

Blood samples were obtained after informed consent according to approved institutional guidelines (Assistance Publique-HĂ´pitaux de Paris, France).

Colony forming unit assay These experiments are described in the Online Supplementary Methods.

In vitro differentiation of human erythroid cells by two-phase liquid culture Peripheral blood mononuclear cells (PBMCs) were obtained from the blood samples of the GD patients (GD) and from healthy donors used as control. These donors were treated with granulocyte colony-stimulating factor to induce hematopoietic stem cell mobilization. The PBMCs were subjected to Ficoll density gradient separation. The CD34+ cells were isolated by magnetic sorting (Miltenyi Biotec), and an in vitro two-phase liquid culture to allow erythroid differentiation was performed, as described by Freyssinier et al.26 During the first phase, the non-adherent cells were expanded for 7 days in a medium containing 100 ng/mL human recombinant (hr) interleukin (IL)-6, 10 ng/mL hr IL-3 and 50 ng/mL hr stem cell factor (SCF). On day 7, the cells were harvested and cultured for 8 days with the second phase medium (10 ng/mL hr IL-3, 50 ng/mL hr SCF, and 2 U/mL hr erythropoietin (EPO). The cells were counted every day, and their concentration was maintained at 0.5Ă&#x2014;106 cells/mL. The SCF, IL-6 and IL-3 cytokines were obtained from Miltenyi Biotec. haematologica | 2016; 101(12)

Dyserythropoiesis in Gaucher disease

In vitro differentiation of human macrophages and co-culture with erythroblasts

Results

CD34– PBMCs from patients and control were differentiated into macrophages as described by Ramos et al.27 Erythroblast differentiation was carried out on day 8 with the second phase medium in the presence of differentiated macrophages. The protocol is provided in the Online Supplementary Methods.

Erythroid differentiation is accelerated in GD

Flow cytometry The antibodies and the methods used for the erythroid differentiation and measurements of GCerase activity are provided in the Online Supplementary Methods.

ELISA Assay The GDF-15 levels were measured in the plasma of 8 healthy controls and 15 untreated GD patients using ELISA according to the manufacturer’s protocols (R&D Systems).

Statistical analysis For the in vitro studies, Wilcoxon paired tests were used to compare control and GD cultures. For the biochemical data analyses, the results are presented as the median (extremes). Mann-Whitney tests were used to compare control and GD patients and anemic and non-anemic patients.

To explore the impact of GCerase deficiency on erythropoiesis, we studied the erythroid precursors produced by the peripheral CD34+ cells of the GD patients and control.

We first explored the clonogenic capacity of GD and control CD34+ cells using a colony-forming unit assay in semi-solid culture. The progenitor cells of different lineages and stages of maturation produce colonies that differ in their size, morphology and cellular composition. The total colonies (including erythroid and non-erythroid colonies) were counted after 14 days of culture. We observed that the CD34+ cell progenitors from GD patients and control exhibited similar total (Figure 1A) and erythroid (Figure 1B) clonogenic capacities. To identify the erythroid-associated CD34+ stages, we scored the BFU-E colonies based on the number of early and late BFU-E colonies. The data showed similar and homogeneous progenitor populations from the control and GD subjects (Figure 1C).

To more precisely study the erythropoiesis in the cells from the GD patients and controls, we then performed a two-phase liquid culture over a period of 15 days. In the first phase, we differentiated the CD34– cells into

Table 1. Biological parameters of the Gaucher disease (GD) patients whose cells were used to study in vitro erythropoiesis.

Pat Age n. (y)

Sex

Genotype

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

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

N370S/?

N370S/?

N370S/1265-1317DEL55 D409H/D409H N370S/L444P+E326K N370S/?

N370S/L324P N370S/L444P ND N370S/R120W N370S/?

N370S/N370S N370S/?

N370S/?

N370S/L324P N370S/1263 DEL 55 N370S/?

N370S/?

ND N370S/?

ND D409H/D409H N370S/?

N370S/?

18 18 8 3 50 14 20 60 6 12 20 35 15 15 80 11 19 19 14 20 18 4 20 20

Time under ERT (year) no no no no no no no no no 7 1 3 11 11 3 3 0.5 0.5 9 1 0.5 1 2 2

SMG

HMG

+ + + + + + ND + + + -

+ + + + ND + + + -

Hb MCV Reticulocytes Ferritin Platelets CCL-18 chitotriosidase (g/dL) (fL) (103/mm3) (mg/L) (103/mm3) (pg/mol) activity 11.9 11 10.5 11.7 11.6 14 12 13.6 12.5 11.9 11.2 16.2 13.9 14.7 13.6 ND 13.4 13.2 13.6 14.1 14,9 12.5 14.1 13.1

96 97 78 69 87 84 92.7 89.8 74 79 104 81 84 84 89 78 96 96 82 99 90 70 91 96

82.6 81.8 ND 33.6 62 ND 54 85 68.5 78.3 85.1 ND 38.9 38.6 ND ND ND ND 22.7 29.4 124 42.6 30.5 37

369 346 276 228 526 638 1316 126 215 172 130 545 111 99 1341 103 237 152 142 192 451 97 132 72

80 93 87 119 65 74 126 109 153 210 123 176 159 182 157 128 140 182 271 167 214 318 216 206

765 663 153 754 689 139 289 149 320 ND 347 82 233 476 518 ND 257 354 163 205 ND 263 175 193

1994 1904 ND 14630 ND ND Deficiency 517 ND ND 726 ND 12750 6750 9400 ND ND ND 1180 361 149 888 ND ND

SMG: splenomegaly. HMG: hepatomegaly. Hb: hemoglobin. MCV: mean corpuscular volume. ND: not determined. Time under ERT (year): "no" means that the patient did not meet the criteria for ERT and did not receive this treatment. The clinical and biological parameters, i.e., SMG, HMG, Hb level, MCV, reticulocytes and platelet counts, ferritin level, CCL-18 (C-C motif chemokine ligand 18) level and chitotriosidase activity, reported here are the results of the last recorded data, usually obtained on the day of blood sampling for the study or within the previous few weeks. Genotype: "?" means that the mutation was not identified despite screening of the most frequent mutations of the GBA gene. Patient 7 had a constitutive chitotriosidase deficiency.

haematologica | 2016; 101(12)

1491

N. Reihani et al.

A

B P=0.27

C P=0.19

P=0.54

P=0.73

Figure 1. Clonogenic potential of CD34+ cells. A. Number of total colony counts including erythroid (BFU-E: burst forming unit), non-erythroid (granulocyte/macrophage colony forming units: CFU-GM, CFU-G and CFU-M) and multilineage progenitors (CFU-GEMM) on day 14 of culture. B. Number of BFU-E burst-forming units (BFU-E). C. Number of early and late erythroid BFU-E. The colonies were scored by direct microscopic visualization on day 14 of culture for the CTL (n=13) and GD (n=13) subjects, which are shown as gray and black boxes, respectively. CTL: control; GD: Gaucher disease.

macrophages and expanded the CD34+ cells in parallel for 7 days. In a second phase, EPO was added to the expanded CD34+ cells to trigger the erythroid differentiation, and the macrophages were cultivated with erythroblasts to mimic the bone marrow microenvironment for 8 subsequent days. We measured the GCerase activity in the CD34+ early progenitors from the control and GD patients. We showed that GCerase was active in the early erythroid progenitors from both GD and control, although at a much lower level in the GD progenitors (Online Supplementary Figure S1A, left panel). The macrophages derived from GD CD34– adherent cells also exhibited reduced GCerase activity compared to those from the control (Online Supplementary Figure S1A, right panel). The monocyte- macrophage profile was confirmed by CD14 (more than 90%), CD68 and CD163 markers (Online Supplementary Figure S1B) as well as by their morphological appearance (Online Supplementary Figure S1C). The expression of CD169, the marker of the central macrophages,28 was also demonstrated as well as the one of CD49e (α5 integrin) since it is involved in the interaction between macrophages and the maturing erythroblasts in the erythroblastic island (Online Supplementary Figure S1B).29 We then compared the erythroid differentiation of the GD erythroblasts cultivated with GD macrophages, and the control erythroblasts cultivated with control macrophages. We observed accelerated differentiation of the GD erythroblasts compared to controls (Figure 2A) on day 15 of the erythroid culture, as shown by the significant increase in the percentage of the differentiated GPA+ CD117– cell subpopulation at day 15 (89.4 vs. 73.5%, P=0.01, Figure 2B). Moreover, we observed an increased percentage of mature cells (acidophilic erythroblasts and reticulocytes) that highly expressed the band 3 marker (20.1 vs. 16.6%, P=0.02, Figure 2C), referred to as band 3hi cells on the basis of the gating strategy (see Online Supplementary Figure S2). The accelerated erythroid differentiation and maturation in the cells from the GD patients was also confirmed by the MGG staining, which showed a lower percentage of polychromatophilic cells and a higher percentage of mature acidophilic cells and reticulocytes (Figure 2D). On day 15, the terminal maturation index was not reduced in the GD vs. control cultures (Figure 2E), which suggested the absence of maturation arrest during the erythroid differentiation. Finally, there was no differ1492

ence in the rate of cell apoptosis between the GD and control cells during the differentiation phase, as indicated by annexin V staining (data not shown). Taken together, these experiments showed accelerated erythroid differentiation without any maturation arrest or cell death in the GD erythroblasts.

The accelerated erythroid differentiation in GD is independent of the macrophages Because macrophages are the key players in GD, we next determined whether the dyserythropoiesis was mediated by the GD macrophages by performing cross coculture experiments. The control erythroblasts were cultivated with GD or control macrophages. The percentage of differentiated erythroblasts was the same in both conditions (Figure 2F). This result strongly suggested that the GD macrophages are not per se the cause of the dyserythropoiesis observed in GD. These observations were also confirmed by the cross-culture of GD EBs cultivated with control or GD macrophages (data not shown). It is noteworthy that erythroid differentiation was more efficient when the EBs were cultivated in the presence of either type of MP (GD or control), as shown by the higher percentage of well-differentiated cells in the co-culture experiments (Online Supplementary Figure S3). This latter result confirms the important role of MPs for efficient erythroid differentiation, as previously demonstrated by others.29 To confirm that the EB differentiation in GD was primarily affected independently of the presence of abnormal macrophages, we performed an in vitro erythropoiesis experiment using CD34+ cells cultivated without macrophages. Proliferation as well as the late differentiation of GD and control erythroblasts were compared. Flow cytometry analysis revealed that before adding EPO, CD117 (cKit) expression levels are equivalent in control and GD erythroblasts indicating the same stage of differentiation (data not shown). We then observed an increased percentage of differentiated cells from the GD subjects, as shown by the increased percentage of the GPA+ CD117- cell subpopulation on day 12 (GD=42 vs. control =21.6%, P=0.0003) and day 15 (GD=60 vs. control =41%, P=0.0001, Figure 3A,B) and the band 3hi erythroblasts (GD=11.3 vs. control=9.1%, P=0.03, Figure 3C). These results indicate that GD erythroid differentiation was significantly accelerated compared to control. Moreover, the MGG analysis showed a lower percentage of basophilic cells and a higher haematologica | 2016; 101(12)

Dyserythropoiesis in Gaucher disease

Table 2. Biological parameters of untreated Gaucher disease (GD) patients according to their hemoglobin level.

Age (year) Sex (F), n (%) Hb (g/dL) MCV (fL) MCH (pg) MCHC (g/L) Reticulocytes (103/mm3) Platelets (103/mm3) Haptoglobin (g/L) LDH (IU/L) Total bilirubin (mmol/L) Conjugated bilirubin CRP (mg/L) Ferritin (Âľg/L)

Anemic n= 10

Non-Anemic n= 10

P

17 [3 - 61] 8 (66.7%) 11.0 [7.6 - 12.9] 81 [74 - 94] 28 [22 - 33.5] 34.6 [29.6 - 35.9] 62 [49.4 - 114] 88 [43 - 191] 0.67 [0.2 - 1.9] 288 [201 - 404] 17 [12 - 22] Absence 1 [1 - 24] 372 [128 - 1316]

10 [3 - 50] 6 (54.5%) 12.5 [11.6 - 14] 77 [69 - 89] 26.6 [24.4 - 29.5] 34.2 [32.5 - 36.5] 59.1 [36.5 - 93] 105 [79 - 208] 0.71 [0.2 -1.52] 389 [261 - 573] 17 [6 - 18.2] Absence 1 [1 - 7] 221.5 [53 - 594]

0.56 0.003 0.25 0.6 0.73 0.62 0.08 0.86 0.27 0.4 0.99 0.06

The hematological and biochemical data for 20 untreated patients who had never received ERT were compared according to their anemic status. Patients were considered anemic when their hemoglobin (Hb) levels were strictly below 11.5 g/dL for children between 2 and 12 years old, 12 g/dL for women and 13 g/dL for men. MCV: mean corpuscular volume.

MCH: Mean corpuscular hemoglobin. MCHC: mean corpuscular hemoglobin concentration. LDH: lactate dehydrogenase. CRP: C-reactive protein. The values are expressed as the median (extremes). The P values were determined using the Mann-Whitney test to compare continuous variables between the anemic and non-anemic patients.

percentage of acidophilic cells and reticulocytes in the GD erythroblasts vs. those from the control (Figure 3D,F). On day 15, the terminal maturation index was significantly increased in the GD compared to the control progenitor cells, which indicated accelerated differentiation without any arrest in the maturation (Figure 3E). These results showed an abnormal erythroid differentiation in GD that was independent of the macrophage defects.

GD erythroid progenitors exhibit decreased proliferative capacity Cell expansion and viability were measured during the two culture phases using trypan blue staining. The cell proliferation capacity is reported as the absolute number of viable cells at days 0 and 15 of culture. Our results showed a decrease of the proliferation capacity of 7.55-fold in GD cells compared to those from the control (GD=121.9 vs. control=920.7-fold, P<0.0001, Figure 4A).

This reduced proliferation resulted in a dramatically smaller absolute number of band 3hi erythroblasts produced at the end of the differentiation (GD=0.6x106 vs. control=14x106, P=0.0001, Figure 4B), despite a higher percentage of differentiated cell population.

Taken together, these data showed that acceleration of erythroid differentiation and maturation occurred to the detriment of cell proliferation and provided evidence for dyserythropoiesis in GD. We then explored whether this dyserythropoiesis could explain the mild anemia observed in GD patients.

were observed for hemolytic markers (lactate dehydrogenase (LDH), haptoglobin and bilirubin levels), and none of the patients had an elevated absolute reticulocyte count (<120.103/mm3 in all patients). These data ruled out the hypothesis that GD patients have peripheral anemia as a result of hemolytic processes. Moreover, the C-reactive protein (CRP) and ferritin levels indicated no evidence of either inflammation or iron deficiency. Additionally, the plasma level of EPO was increased in untreated GD patients vs. the healthy controls (15.24 vs. 3.46 IU/L, P<0.0001). This result indicated that the anemia in GD is not due to an EPO deficiency and instead could represent a compensatory mechanism preventing a more pronounced anemia in GD patients. The fact that the anemic patients did not exhibit higher reticulocyte counts than those without anemia indicated that the bone marrow response was not appropriately responding to the anemia, which suggested a central defect.

To further test this hypothesis, we analyzed markers of ineffective erythropoiesis in plasma from GD patients.

Growth differentiation factor-15, a marker of ineffective erythropoiesis in several hemoglobinopathies,30,31 was significantly increased in the plasma from the untreated GD patients compared to the healthy controls (Figure 5A).

Among our untreated GD cohort, the anemic patients exhibited higher GDF-15 levels than the non-anemic patients (4311 vs. 2298, P=0.006, Figure 5B). Taken together, these data suggest that the anemia observed in GD has a central origin.

An anemia of central origin in GD To investigate the mechanism for the anemia in GD, hematological and biochemical parameters were examined in 20 GD untreated patients. Of these 20 subjects, 10 patients exhibited mild anemia. These anemic GD patients were compared to the non-anemic subjects (Table 2). The ages and sex ratios were similar. No differences haematologica | 2016; 101(12)

Discussion In spite of the progress made toward the molecular characterization of GD, much still remains to be investigated about the mechanisms involved in the pathophysiology of the clinical complications.

1493

N. Reihani et al.

In the study herein, we demonstrated for the first time a dyserythropoiesis in GD characterized by an accelerated differentiation accompanied by decreased cell proliferation. The clinical data from the GD patients strongly suggested that the anemia observed in GD is of central origin.

Moreover, we observed high levels of GDF-15, a marker of dyserythropoiesis, in the anemic GD patients.

Previous studies have described infiltration of the bone marrow as well as some erythrophagocytosis by Gaucher cells in GD patients.17-19,32,33 These occurrences could be considered partly responsible for the bone marrow erythroid insufficiency. Lee et al. first described the ery-

throphagocytic events in Gaucher cells in subjects with slight erythroid hyperplasia.17 In the study herein, ferrokinetic studies already suggested a mild dyserythropoiesis.17 In our study, it would have been useful to obtain marrow from the anemic patients for examination, but bone marrow aspiration is not ethically justified in GD. Consistent with the results of our in vitro study in semisolid culture, Lecourt et al. reported that CD34+ bone marrow progenitors from control and GD patients exhibit similar basic characteristics.7 These cells gave rise to a comparable number of total erythroid and non-erythroid colonies and exhibited similar erythroid potential. Our

A

B P=0.06

E

%Band 3hi

C

D

1494

F

Figure 2. In vitro erythroid differentiation of Gaucher disease (GD) erythroblasts cultivated with macrophages. Erythroblasts (EBs) derived from the CD34+ peripheral blood cells of GD patients (GD) or healthy controls (CTL) were cultivated with macrophages (MPs) derived from CD34- cells from the GD or CTL subjects. The GD EBs cultivated with GD MPs were compared with the CTL EBs cultivated with CTL MPs. Expression of erythroid surface markers was measured by flow cytometry on day 12 and 15 of erythroid differentiation. A. Representative flow cytometry plots of glycophorin A (GPA) and c-Kit (CD117) surface expression on days 12 and 15 in the GD and CTL cultures. The GPA+ CD117– cell population represents the differentiating EBs. B. Percentage of GPA+ CD117– cells derived from CTL or GD patients on days 12 and 15 of erythroid differentiation. C. Percentage of band 3hi cells on day 15 of erythroid differentiation. These cells represent the mature EBs. D. The boxes represent the percentage of GD or CTL progenitors on day 15 of erythroid differentiation. Morphological analysis after May– Grünwald–Giemsa (MGG) staining was used. ProEB: proerythroblasts; Baso: basophilic cells; Polych: polychromatophilic cells; Acido+Retic: acidophilic cells and reticulocytes E. T h e boxes represent the terminal maturation index (on day 15 of differentiation) as defined in the Online Supplementary Methods section. F. CTL or GD MPs were co-cultivated with CTL EBs. The percentages of GPA+ CD117– cells on days 12 and 15 of erythroid differentiation were compared. The results are presented as box-and-whisker plots. Gray boxes, culture of CTL EBs with CTL MPs; black boxes, culture of GD EBs with GD MPs; tiled boxes, culture of CTL EBs with GD MPs. For B, C, D, E and F, n=8 for each condition. The medians are represented as horizontal bars (-); the upper and lower quartiles are represented as the top and the bottom of the box, respectively; and the maximum and minimum data values are shown by dashes (-) at the top and the bottom, respectively, of the whiskers. The P values were determined using the Wilcoxon signed-rank test to compare the parameters of erythroid differentiation between the CTL and GD cultures on days 12 and 15 (*P<0.05). ns= non significant.

haematologica | 2016; 101(12)

Dyserythropoiesis in Gaucher disease

results also showed that CD34+ peripheral hematopoietic stem cells in controls and GD patients are basically at the same stage of differentiation. However, starting from the same point, the GD progenitors demonstrated moderately accelerated differentiation in liquid culture conditions with a prolonged differentiating step that permitted a detailed study for each stage of the maturation. We observed clear dyserythropoiesis in the GD type 1 patients (n=24). This dyserythropoiesis was characterized by

accelerated erythroid differentiation and diminished cellular proliferation without maturation arrest. Sgambato et al. have recently performed an in vitro hematopoiesis study using iPSCs from a limited number of GD patients (n=4).

They showed an enhanced myeloid differentiation and decreased erythroid differentiation and maturation in the severe types of GD, but they did not report any clear differences in the single type 1 GD patient in their study.21 It should be noted that since the starting materials were dif-

A

B C

E D

F

haematologica | 2016; 101(12)

Figure 3. In vitro erythroid differentiation of Gaucher disease (GD) erythroblasts cultivated without macrophages. CD34+ cells derived from the peripheral blood cells from GD patients (GD) or healthy controls (CTL) were cultivated without macrophages (MPs) and differentiated into the erythroid lineage until day 15 as described in methods. The surface expression of the erythroid differentiation markers was measured using flow cytometry during erythroblast (EB) differentiation.

A. Representative flow cytometry plots of the cell surface expression of glycophorin A (GPA) and cKit (CD117) on days 12 and 15 of the erythroid differentiation culture. The GPA+ CD117–cells represent the differentiating EBs. B. Percentage of GPA+ CD117– cells derived from CTL or GD patients on days 12 and 15 of the erythroid differentiation performed without MPs (CTL n=18; GD n=24). C. Percentage of band 3hi cells on day 15 of erythroid differentiation (CTL n=13; GD n=13). These cells represent the mature EBs. D. The boxes represent the percentage of GD or CTL erythroid cells on day 15 of the erythroid differentiation performed without MPs (CTL n=12; GD n=17). Morphological analysis after MGG (May-Grünwald-Giemsa) staining was used. ProEB, proerythroblasts; Baso: basophilic cells; Polych: polychromatophilic cells; Acido+Retic: acidophilic cells and reticulocytes. E. The boxes represent the terminal maturation index as defined in the Online Supplementary Methods section (CTL n=12; GD n=17). F. Representative morphological analysis of erythroid differentiation as indicated by MGG staining on day 15 of cell culture (magnification 60x). The results are presented as box-and-whisker plots.

Gray boxes (CTL EBs); black boxes (GD EBs). The medians are represented as horizontal bars (-); the upper and lower quartiles are represented as the top and the bottom of the box, respectively; and the maximum and minimum data values are shown as dashes (-) at the top and the bottom, respectively, of the whiskers. The P values were determined using the Wilcoxon signed-rank test to compare the parameters of erythroid differentiation between the CTL and GD cultures on days 12 and 15 (*P<0.05; *** P<0.001).

1495

N. Reihani et al.

ferent (fibroblasts vs. peripheral blood from patients), our study and that of Sgambato et al. are not comparable.

Dyserythropoiesis is a hallmark of β-thalassemia, a major inherited hemoglobinopathy that is caused by a quantitative defect in the synthesis of the β−globin chains.

However, the dyserythropoiesis in this disease is characterized not only by accelerated erythroid differentiation, but also by maturation arrest and apoptosis at the polychromatophilic stage.34-36 In our study of GD subjects, the increased erythroid differentiation was associated with a decrease in cell expansion, but we did not observe any premature death of the erythroid precursors or maturation

Number of cells (Log)

A

arrest. This may explain the mild anemia in GD compared to the much more severe anemia observed in patients afflicted with β-thalassemia.

Erythropoiesis requires fine regulation of cell survival, proliferation and differentiation that is controlled primarily by regulatory signals provided by cytokines.

Transforming growth factor β1 (TGF- β1) is known to accelerate the erythroid differentiation, allowing full terminal differentiation toward enucleated red cells.37,38 TGF- β1 also induces a massive inhibition of cell proliferation that mainly involves cell cycle arrest rather than apoptosis.38 We observed similar effects, albeit to a lesser extent, in GD during in vitro erythropoiesis. Because TGF-β1, or cytokines belonging to the same family (such as GDF-11 and GDF-15), represent autocrine factors released by erythroblasts, they could induce or impair the signaling pathways responsible for the dyserythropoiesis in GD. However, due to the limited number of circulating progenitor cells, the cell cycle could not be investigated, and the question of the cell cycle kinetics remains open.

Currently, Gaucher macrophages are considered the main cause of the GD complications. In our experiments, in vitro dyserythropoiesis was observed even in the absence of macrophages. Although we could not exclude a local paracrine action of the GD macrophages in the

A

Number of Band 3hi erythroblasts (x106)

B

Figure 4. Cell proliferation during in vitro erythropoiesis. A. Absolute numbers of cells derived from CD34+ cells from the GD and CTL subjects on day 0 and day 15 of erythroid differentiation without MPs (macrophages) (n=24 for each GD and control group). B. Absolute numbers of mature cells derived from the CD34+ cells from the GD and CTL subjects on day 15 of erythroid differentiation without MPs (n=17 for each GD and control group). This value has been calculated from the percentage of band 3hi cells multiplied by the total number of cells at day 15.

Gray boxes and curve show the results for the CTL erythoblasts (EBs), and black boxes and curve show the results for the GD EBs. The P values were determined using the Wilcoxon signed-rank test to compare the expansion of cells between CTL and GD cultures. For figure A, the ratios of the cell numbers on day 15 to those on day 0 were compared by a Wilcoxon signed-rank test (**** P<0.0001). For figure B, the P value was determined using the Mann-Whitney test to compare the number of Band 3hi erythroblasts (***P<0.001). For figure B, the P value was determined using the Mann-Whitney test to compare the number of Band 3hi erythroblasts (***P<0.001). CTL: control; GD: Gaucher disease.

1496

B

Figure 5. Plasma growth differentiation factor-15 (GDF-15) levels in GD patients. A. Increased level of GDF-15 in the plasma from untreated GD patients (UT GD, n=15) compared to CTL (CTL, n=8). B. GDF-15 level measured in GD patients according to their anemic status (n=7 anemic patients vs. n=8 nonanemic patients). The P values were determined using the Mann-Whitney test to compare the GDF-15 levels between control and GD subjects. (**P<0.01, ****P<0.0001). The bars represent the medians. CTL: control; GD: Gaucher disease.

haematologica | 2016; 101(12)

Dyserythropoiesis in Gaucher disease

bone marrow, these data strongly suggest that the primary defects leading to the dyserythropoiesis in GD may intrinsically affect the erythroid lineage. Because the macrophages are essential components of the erythroblast islands that play an important role in enucleation processes,29 we cannot rule out any effect of macrophages during the last step of erythroid terminal maturation, which could not be efficiently analyzed in our culture system.

These late stages could also be affected by autophagy, a process that has been linked to GD pathophysiology4,39 and is known to play a critical role during erythroid differentiation. Mesenchymal stem cells also exhibit GCerase activity and are important components of the bone marrow microenvironment.7 The effects of these cells in GD hematopoiesis were not investigated in our study.

Anemia in GD patients is efficiently corrected by a few months of ERT. Unfortunately, information on the longitudinal effects of ERT on the indirect biological characteristics of erythropoiesis (Hb, reticulocyte count, GDF15) in anemic GD patients was not available in our cohort.

Moreover, the effect of ERT therapy on in vitro erythropoiesis was not investigated in our study. Indeed, the blood samples from treated patients were collected two weeks after their last ERT infusion, and the progenitor cells were washed and then cultivated in vitro during two additional weeks without any ERT drugs.

Key sphingolipids are known to accumulate in GD and are proposed to be responsible for numerous aspects of GD pathophysiology.10,24 The results of the study herein do not permit a conclusion as to whether the dyserythropoiesis is directly due to the reduced enzyme activity and

References 1. Linari S, Castaman G. Hematological manifestations and complications of Gaucher disease. Expert Rev Hematol. 2016;9(1):51-58.

2. Barton NW, Brady RO, Dambrosia JM, et al.

Replacement therapy for inherited enzyme deficiency--macrophage-targeted glucocerebrosidase for Gaucher's disease. N Engl J Med. 1991;324(21):1464-1470.

3. Beutler E. Gaucher disease. Blood Rev. 1988;2(1):59-70.

4. Cox TM, Cachon-Gonzalez MB. The cellular pathology of lysosomal diseases. J Pathol.

2012;226(2):241-254.

5. Berger J, Lecourt S, Vanneaux V, et al.

Glucocerebrosidase deficiency dramatically impairs human bone marrow haematopoiesis in an in vitro model of Gaucher disease. Br J Haematol. 2010;150 (1):93-101.

6. Campeau PM, Rafei M, Boivin MN, Sun Y, Grabowski GA, Galipeau J. Characterization of Gaucher disease bone marrow mesenchymal stromal cells reveals an altered inflammatory secretome.

Blood.

2009;114(15):3181-3190.

7. Lecourt S, Mouly E, Freida D, et al. A prospective study of bone marrow hematopoietic and mesenchymal stem cells in type 1 Gaucher disease patients. PLoS One. 2013;8(7):e69293.

8. Lecourt S, Vanneaux V, Cras A, et al. Bone marrow microenvironment in an in vitro

haematologica | 2016; 101(12)

9.

10.

11.

12.

13.

14.

15.

lipid accumulation, or to an alteration in the mechanisms involved in enzyme folding. Alternatively, since sphingosine 1-phosphate (S1P), as a bioactive lipid, can affect the hematopoiesis and RBC signaling in addition to its multiple cellular and systemic functions,40-42 the possibility that the abnormal sphingolipid contents in the erythroid progenitors could affect erythropoiesis could be an interesting hypothesis to investigate.

In conclusion, we have shown that anemia in GD has a central origin and is associated with an unexpected dyserythropoiesis. This dyserythropoiesis is characterized by accelerated erythroid differentiation and reduced proliferation capacities that are independent of the macrophage defects. Our data shed new light on the mechanism for the anemia in GD and highlight the role of the erythroid cells as important contributors to the pathophysiology of GD. Acknowledgments The authors thank Monia Bengherbia and Karima Yousfi for their assistance in collecting GD patient data, Emmanuel Collec for his contribution to the collection of patient samples and Geneviève Courtois and Cyril Mignot for their helpful discussions and advice.

Funding This study was supported by a grant from Shire. N.R. was supported by a Labex GR-Ex fellowship. The Labex GR-Ex, reference ANR-11-LABX-0051, is funded by the “Investissements d’avenir” program of the French National Research Agency, reference ANR-11-IDEX-0005-02.

model of Gaucher disease: consequences of glucocerebrosidase deficiency. Stem Cells Dev. 2012;21(2):239-248.

Liu J, Halene S, Yang M, et al. Gaucher disease gene GBA functions in immune regulation. Proc Natl Acad Sci USA. 2012;109(25):10018-10023.

Mistry PK, Liu J, Yang M, et al.

Glucocerebrosidase gene-deficient mouse recapitulates Gaucher disease displaying cellular and molecular dysregulation beyond the macrophage. Proc Natl Acad Sci USA. 2010;107(45):19473-19478.

Franco M, Collec E, Connes P, et al. Abnormal properties of red blood cells suggest a role in the pathophysiology of Gaucher disease.

Blood. 2013;121(3):546-555.

Khan A, Hangartner T, Weinreb NJ, Taylor JS, Mistry PK. Risk factors for fractures and avascular osteonecrosis in type 1 Gaucher disease: a study from the International Collaborative Gaucher Group (ICGG) Gaucher Registry. J Bone Miner Res. 2012;27(8):1839-1848.

Hughes D, Cappellini MD, Berger M, et al.

Recommendations for the management of the haematological and onco-haematological aspects of Gaucher disease. Br J Haematol. 2007;138(6):676-686.

Zimran A, Altarescu G, Rudensky B, Abrahamov A, Elstein D. Survey of hematological aspects of Gaucher disease.

Hematology. 2005;10(2):151-156.

Weinreb NJ. Introduction. Advances in Gaucher Disease: therapeutic goals and eval-

16.

17.

18.

19.

20.

21.

uation and monitoring guidelines. Semin Hematol. 2004;41(4 Suppl 5):1-3.

Weinreb NJ, Charrow J, Andersson HC, et al. Effectiveness of enzyme replacement therapy in 1028 patients with type 1 Gaucher disease after 2 to 5 years of treatment: a report from the Gaucher Registry.

Am J Med. 2002;113(2):112-119.

Lee RE, Balcerzak SP, Westerman MP. Gaucher's disease. A morphologic study and measurements of iron metabolism. Am J Med. 1967;42(6):891-898.

Machaczka M, Klimkowska M, Regenthal S, Hagglund H. Gaucher disease with foamy transformed macrophages and erythrophagocytic activity. J Inherit Metab Dis. 2011;34(1):233-235.

Markuszewska-Kuczynska A, Klimkowska M, Regenthal S, Bulanda A, Kampe Bjorkvall C, Machaczka M. Atypical cytomorphology of Gaucher cells is frequently seen in bone marrow smears from untreated patients with Gaucher disease type 1. Folia Histochem Cytobiol. 2015;53(1):62-69.

Rudzki Z, Okon K, Machaczka M, Rucinska M, Papla B, Skotnicki AB. Enzyme replacement therapy reduces Gaucher cell burden but may accelerate osteopenia in patients with type I disease - a histological study. Eur J Haematol. 2003;70(5):273-281.

Sgambato JA, Park TS, Miller D, et al.

Gaucher Disease-Induced Pluripotent Stem Cells Display Decreased Erythroid Potential and Aberrant Myelopoiesis. Stem Cells Transl Med. 2015;4(8):878-886.

1497

N. Reihani et al.

22. Dekker N, van Dussen L, Hollak CE, et al.

Elevated plasma glucosylsphingosine in Gaucher disease: relation to phenotype, storage cell markers, and therapeutic response. Blood. 2011;118(16):e118-127.

23. Meikle PJ, Whitfield PD, Rozaklis T, et al.

Plasma lipids are altered in Gaucher disease: biochemical markers to evaluate therapeutic intervention. Blood Cells Mol Dis. 2008;40 (3):420-427.

24. Mistry PK, Liu J, Sun L, et al.

Glucocerebrosidase 2 gene deletion rescues type 1 Gaucher disease. Proc Natl Acad Sci USA. 2014;111(13):4934-4939.

25. Nilsson O, Hakansson G, Dreborg S, Groth CG, Svennerholm L. Increased cerebroside concentration in plasma and erythrocytes in Gaucher disease: significant differences between type I and type III. Clin Genet.

1982;22(5):274-279.

26. Freyssinier JM, Lecoq-Lafon C, Amsellem S, et al. Purification, amplification and characterization of a population of human erythroid progenitors. Br J Haematol. 1999;106 (4):912-922.

27. Ramos P, Casu C, Gardenghi S, et al.

Macrophages support pathological erythropoiesis in polycythemia vera and beta-thalassemia. Nat Med. 2013;19(4):437-445.

28. Chow A, Huggins M, Ahmed J, et al.

CD169(+) macrophages provide a niche pro-

1498

29.

30.

31.

32.

33.

34.

35.

moting erythropoiesis under homeostasis and stress. Nat Med. 2013;19(4):429-436.

Chasis JA, Mohandas N. Erythroblastic islands: niches for erythropoiesis. Blood.

2008;112(3):470-478.

Ramirez JM, Schaad O, Durual S, et al.

Growth differentiation factor 15 production is necessary for normal erythroid differentiation and is increased in refractory anaemia with ring-sideroblasts. Br J Haematol.

2009;144(2):251-262.

Tanno T, Noel P, Miller JL. Growth differentiation factor 15 in erythroid health and disease. Curr Opin Hematol. 2010;17(3):184190.

Bitton A, Etzell J, Grenert JP, Wang E. Erythrophagocytosis in Gaucher cells. Arch Pathol Lab Med. 2004;128(10):1191-1192.

Hibbs RG, Ferrans VJ, Cipriano PR, Tardiff KJ. A histochemical and electron microscopic study of Gaucher cells. Arch Pathol.

1970;89(2):137-153.

Arlet JB, Ribeil JA, Guillem F, et al. HSP70 sequestration by free alpha-globin promotes ineffective erythropoiesis in beta-thalassaemia. Nature. 2014;514(7521):242-246.

Mathias LA, Fisher TC, Zeng L, et al.

Ineffective erythropoiesis in beta-thalassemia major is due to apoptosis at the polychromatophilic normoblast stage. Exp Hematol. 2000;28(12):1343-1353.

36. Ribeil JA, Arlet JB, Dussiot M, Moura IC, Courtois G, Hermine O. Ineffective erythropoiesis in beta-thalassemia.

ScientificWorldJournal. 2013;2013(394295.

37. Krystal G, Lam V, Dragowska W, et al.

Transforming growth factor beta 1 is an inducer of erythroid differentiation. J Exp Med. 1994;180(3):851-860.

38. Zermati Y, Fichelson S, Valensi F, et al.

Transforming growth factor inhibits erythropoiesis by blocking proliferation and accelerating differentiation of erythroid progenitors. Exp Hematol. 2000;28(8):885-894.

39. Vitner EB, Dekel H, Zigdon H, et al. Altered expression and distribution of cathepsins in neuronopathic forms of Gaucher disease and in other sphingolipidoses. Hum Mol Genet. 2010;19(18):3583-3590.

40. Bendall LJ, Basnett J. Role of sphingosine 1phosphate in trafficking and mobilization of hematopoietic stem cells. Curr Opin Hematol. 2013;20(4):281-288.

41. Zhang L, Urtz N, Gaertner F, et al.

Sphingosine kinase 2 (Sphk2) regulates platelet biogenesis by providing intracellular sphingosine 1-phosphate (S1P). Blood.

2013;122(5):791-802.

42. Zhang Y, Berka V, Song A, et al. Elevated sphingosine-1-phosphate promotes sickling and sickle cell disease progression. J Clin Invest. 2014;124(6):2750-2761.

haematologica | 2016; 101(12)

ARTICLE

Iron Metabolism & its Disorders

Wnt5a is a key target for the pro-osteogenic effects of iron chelation on osteoblast progenitors

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Ulrike Baschant,1* Martina Rauner,1* Ekaterina Balaian,2 Heike Weidner,2 Antonella Roetto,3 Uwe Platzbecker2 and Lorenz C. Hofbauer1,4,5

Department of Medicine III, Technische Universität Dresden, Saxony, Germany; Department of Medicine I, Technische Universität Dresden, Saxony, Germany; 3 Department of Clinical and Biological Science, University of Torino, Italy; 4Center for Regenerative Therapies Dresden, Saxony, Germany and 5Center for Healthy Aging, Technische Universität Dresden, Saxony, Germany

1

2

*UB and MR contributed equally to this work.

Haematologica 2016 Volume 101(12):1499-1507

ABSTRACT

I

ron overload due to hemochromatosis or chronic blood transfusions has been associated with the development of osteoporosis.

However, the impact of changes in iron homeostasis on osteoblast functions and the underlying mechanisms are poorly defined. Since Wnt signaling is a critical regulator of bone remodeling, we aimed to analyze the effects of iron overload and iron deficiency on osteoblast function, and further define the role of Wnt signaling in these processes.

Therefore, bone marrow stromal cells were isolated from wild-type mice and differentiated towards osteoblasts. Exposure of the cells to iron dose-dependently attenuated osteoblast differentiation in terms of mineralization and osteogenic gene expression, whereas iron chelation with deferoxamine promoted osteogenic differentiation in a time- and dose-dependent manner up to 3-fold. Similar results were obtained for human bone marrow stromal cells. To elucidate whether the proosteogenic effect of deferoxamine is mediated via Wnt signaling, we performed a Wnt profiler array of deferoxamine-treated osteoblasts.

Wnt5a was amongst the most highly induced genes. Further analysis revealed a time- and dose-dependent induction of Wnt5a being up-regulated 2-fold after 48 h at 50 mM deferoxamine. Pathway analysis using specific inhibitors revealed that deferoxamine utilized the phosphatidylinositol-3-kinase and nuclear factor of activated T cell pathways to induce Wnt5a expression. Finally, we confirmed the requirement of Wnt5a in the deferoxamine-mediated osteoblast-promoting effects by analyzing the matrix mineralization of Wnt5a-deficient cells. The promoting effect of deferoxamine on matrix mineralization in wild-type cells was completely abolished in Wnt5a–/– cells. Thus, these data demonstrate that Wnt5a is critical for the pro-osteogenic effects of iron chelation using deferoxamine.

Correspondence: lorenz.hofbauer@uniklinikum-dresden.de

Received: February 23, 2016.

Accepted: August 10, 2016.

Pre-published: August 18, 2016. doi:10.3324/haematol.2016.144808

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/100/12/1499

Introduction Iron is an essential nutrient for life as it plays a key role in several physiological processes, including the transport of oxygen in erythrocytes, generating ATP as a source of energy, and controlling innate immune responses to combat bacteria.1,2 However, the otherwise desirable redox potential of iron can also generate cellular toxicity in conditions of iron overload, as it produces reactive oxygen species intermediates that damage lipids, DNA, and proteins.3 Therefore, iron concentrations within the body are tightly controlled to maintain an optimal range.

Bone remodeling is susceptible to changes in iron homeostasis, and both iron haematologica | 2016; 101(12)

©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.

1499

U. Baschant et al.

deficiency and iron excess affect bone health in humans and rodents. Dietary iron is positively associated with bone mineral density in healthy postmenopausal women.4,5 Accordingly, rats rendered iron deficient display low bone mass due to a reduced bone turnover.6,7 On the other hand, elderly women with low bone mass and ovariectomized rats have increased iron levels in the bone, and by chelating iron in ovariectomized rats, bone loss is partially prevented.8,9 Consistent with this, patients with iron overload due to hemochromatosis, β-thalassemia and sickle-cell anemia have been shown to have a higher incidence of osteoporosis.10-13 Importantly, bone mineral density was shown to increase again in patients with β-thalassemia after iron chelation with deferasirox.14 Similar to humans, rats fed with a high iron diet or injected with colloidal iron also experience bone loss, mainly due to an increased osteoclast activation and bone resorption.15-17 Collectively, these data show that certain iron levels are required to maintain bone homeostasis.

Despite the clear evidence that bone homeostasis is tightly regulated by iron, the underlying cellular and molecular mechanisms are not fully understood. In osteoclasts, iron excess stimulates osteoclastogenesis by supporting mitochondrial respiration,18-20 while sequestering iron from osteoclasts inhibits their maturation and function.21-22 In contrast, osteoblast function is suppressed by iron excess in osteoblast cell lines,23-25 while iron chelation using deferoxamine (DFO) seems to exert positive effects on osteoblast maturation.26-28 In addition, bone regeneration during distraction osteogenesis and after radiation therapy has been shown to be promoted in the presence of DFO.29-30 However, some studies have reported an inhibitory effect of DFO on osteoblasts and a suppression of alkaline phosphatase (ALP).31 Thus, further investigations are required to conclusively address the effects of iron excess and chelation on osteoblast function and to assess the underlying mechanisms.

One of the most important pathways for maintaining bone homeostasis is the Wnt signaling pathway.32 Both canonical (β-catenin-dependent) and non-canonical (β-catenin-independent) pathways regulate bone turnover, mainly by promoting osteoblast differentiation and indirectly controlling osteoclastogenesis.33-34 So far, the modulation of Wnt signaling by iron has not been investigated in vivo or in vitro.

Herein, we set out to examine the effects of iron excess and iron chelation via DFO on osteoblast differentiation and function using human and murine primary cells.

Moreover, we investigated the iron-dependent regulation of Wnt signaling to assess whether Wnt signaling plays a role in mediating the effects of iron and/or DFO on osteoblasts.

Methods Mice Primary murine bone marrow stromal cells (BMSCs) were collected from 10-12 week old C57BL/6J mice of both genders. Also, Wnt5afl/fl mice that were crossed with ROSA26-Cre-ERT2 mice (B6.129-Gt(ROSA)26Sortm1(cre/ERT2)Tyj/J, Jax Mice, Stock number: 8463) to produce mice in which Wnt5a can be globally deleted upon the administration of tamoxifen.35 Wnt5afl/fl mice without the cre transgene were used as littermate controls. Transferrin receptor 2 knockout mice on a 129XI/SvJ background (male, 10-12 weeks of 1500

age) were used as a model of iron overload.36 Wild-type 129XI/SvJ mice served as controls. The sacrifice of the animals for the collection of organs was approved by the institutional animal care committee and the Landesdirektion Sachsen. All mice were fed a standard diet with water ad libitum and were kept in groups of 5 animals per cage. Mice were exposed to a 12 h light/dark cycle and an air-conditioned room at 23 °C.

DFO treatment in mice and mCT analysis

Twelve week old male 129/Sv wild-type mice were treated for three weeks with 250 mg/kg body weight deferoxamine (daily i.p. injections). Micro-CT of the vertebrae was performed ex vivo using the vivaCT 40 (Scanco Medical AG, Brüttisellen, Switzerland) with an X-ray energy of 70 kVp, 114 mA, 200 msec integration time, and an isotropic voxel size of 20 mm. Pre-defined scripts from Scanco were used for the reconstruction and evaluation of the trabecular bone.

For mRNA expression analysis, the mice were sacrificed and primary murine bone BMSCs were yielded by flushing the femora and tibiae with PBS. After centrifugation at 300 x g for 5 min, 1 ml TriFast (Peqlab, Erlangen, Germany) was added to the cell pellet.

For the isolation of RNA from the cortical bone tissue of iron-overloaded mice (transferrin receptor 2-deficient mice and wild-type controls), bone marrow was flushed and the cortical bone was crushed in liquid nitrogen. The bone powder was recovered in TriFast. Total RNA from the cells was isolated according to the manufacturer`s protocol. Five-hundred ng RNA were reverse transcribed using Superscript II (Invitrogen, Darmstadt, Germany), and subsequently used for SYBR green-based real-time PCRs using the standard protocol (Applied Biosystems Inc., Carlsbad, CA, USA).

Culture of murine bone marrow stromal cells Femurs and tibias were dissected from the mice and cells were obtained by flushing the bone marrow with DMEM (supplemented with 10% FCS and 1% penicillin/streptomycin, all from Invitrogen).37,38 Cells were maintained in growth medium (DMEM with 10% FCS and 1% penicillin/streptomycin) until 70% confluence before they were switched to differentiation media (growth medium supplemented with 100 mM ascorbate phosphate and 5 mM β-glycerol phosphate, both from Sigma-Aldrich) for up to 21 days. To delete Wnt5a expression in vitro, cells were treated with 1 mM tamoxifen (Sigma-Aldrich) for 3 days prior to starting the differentiation process.

For some experiments, iron (III) chloride (FeCl3) or deferoxamine (DFO, both from Sigma-Aldrich) was added at different concentrations either for the entire differentiation period or for 48 h.

Where indicated, ferric ammonium citrate or holo-transferrin were used as alternative iron sources (both from Sigma-Aldrich). During the 21 day differentiation period, the medium as well as the supplemented FeCl3 and DFO were replaced three times per week.

For the signaling studies, cells were pre-treated for 1 h with the specific pathway inhibitors BAY-11-7082 (I-κB phosphorylation inhibitor, 1 mM, Calbiochem), UO126 (MEK1 inhibitor, 10 mM, Calbiochem), LY-294002 (PI3K inhibitor, 10 µM, Cell Signaling), and VIVIT (NFAT inhibitor, 1 mM, Calbiochem). Gene and protein expression of Wnt5a were analyzed after another 48 h of treatment with DFO and the inhibitors.

Culture of human BMSCs Primary human BMSCs were collected from healthy donors (aged 22-49 years, mixed sex); following Institutional Review Board approval and having obtaining written informed consent; and cultured according to previously reported methods.39 Briefly, cells at passages 2-4 were used. Cells were grown in DMEM with 10% FCS and 1% penicillin/streptomycin. Osteogenic differentiahaematologica | 2016; 101(12)

Wnt5a mediates DFO effects on osteoblasts

tion was induced using growth medium supplemented with 100 µM ascorbate phosphate, 5 mM β-glycerol phosphate, and 10 nM dexamethasone (all from Sigma-Aldrich) for 21 days. Cells were treated with 25 mM FeCl3 or 25 mM DFO for 21 days or for 48 h.

Cell viability assay Osteoblast cultures were differentiated for 7 days before starting the 48 h treatment with FeCl3 and DFO in a medium with a reduced FCS content (1%) at varying concentrations. Cell viability was measured using the CellTiter-Blue® assay according to the manufacturer’s protocol (Promega, Mannheim, Germany). The fluorescent signal was measured using the FLUOstar Omega (BMG labtech, Jena, Germany).

Alizarin red staining Osteoblast cultures were fixed in 70% ethanol for 30 min and stained with 1% alizarin red S (pH 5.5, Sigma-Aldrich) for 10 min at room temperature. Excess dye was removed by repeatedly washing the plates with distilled water. The amount of incorporated calcium was eluted with 100 mM cetylpyridinium chloride (Sigma-Aldrich) for 10 min at room temperature. Aliquots were taken and measured photometrically at 540 nm in duplicates.

Alkaline phosphatase activity assay

Cells were lysed in 100 ml lysis buffer (1.5 mM Tris-HCl pH 7, 1

A

B

C

D

E

F

haematologica | 2016; 101(12)

G

mM ZnCl2, 1 mM MgCl2, 1% triton X-100) and centrifuged for 20 min at 6,000 rpm at 4 °C. Aliquots of each sample were incubated with 100 ml alkaline phosphatase (ALP) substrate buffer (100 mM diethanolamine, 0.1% triton X-100 supplemented with 1:10 37 mM p-nitrophenyl phosphate) for 30 min at 37 °C. The enzymatic reaction was stopped with 40 mM NaOH, measured at a wavelength of 405 nm, and normalized to the total protein content determined by the BCA method from the same protein extracts.

RNA isolation, RT, and real-time RT-PCR Total RNA from cell culture was isolated using the High Pure RNA Isolation kit (Roche, Mannheim, Germany) according to the manufacturer`s protocol. Five-hundred ng RNA were reverse transcribed using Superscript II (Invitrogen, Darmstadt, Germany) and subsequently used for SYBR green-based real-time PCRs using standard protocol (Applied Biosystems Inc., Carlsbad, CA, USA).

The primer sequences for mice were: β-actin: ATCTGGCACCACACCTTCT, β-actin as: GGGGTGTTGAAGGTCTCAAA; Runx2: CCCAGCCACCTTTACCTACA, Runx2 as: TATGGAGTGCTGCTGCTGGTCTG; Alp as: CTACTTGTGTGGCGTGAAGG, Alp: CTGGTGGCATCTCGTTATCC; Ocn: GCGCTCTGTCTCTCTGACCT, Ocn as: ACCTTATTGCCCTCCTGCTT; Wnt5a as: CCAACTGGCAGGACTTTCTC, Wnt5a: GCATTCCTTGATGCCTGTCT. The primer sequences for the human genes were: β-ACTIN: CCAACCGCGAGAAGAT-

H

Figure 1. Iron excess inhibits and iron chelation supports osteogenic differentiation. (A-D) Bone marrow stromal cells were differentiated towards osteoblasts in the presence of (A) various concentrations of iron chloride (FeCl3), (B) 25 mM FeCl3, 1 mM ferric ammonium citrate (FAC), and 250 mM holo-transferrin (Holo-Tf), (C) the iron-chelator deferoxamine (DFO) as well as (D) combinations of FeCl3 and DFO for 21 days. The mineralization was visualized with alizarin red S staining and quantified after elution with cetylpyridinium chloride. (E-G) Gene expression analysis of Runx2, alkaline phosphatase (Alp) and osteocalcin (Bglap) using real-time polymerase chain reaction (PCR) after treating day 10 differentiated cells with 5 mM FeCl3 (Fe), 25 µM DFO or Fe+DFO for an additional 48 hours. (H) Enzyme activity of ALP in day 10 osteoblasts after 48 hours of Fe (5 µM) and DFO (25 mM) treatment. N=4-6. *P<0.05, **P<0.01, ***P<0.001 vs. control (CO).

1501

U. Baschant et al.

GA, β-ACTIN as: CCAGAGGCGTACAGGGATAG; ALP: CAACCCTGGGGAGGAGAC, ALP as: GCATTGGTGTTGTACGTCTTG; RUNX2: CACCATGTCAGCAAAACTTCTT, RUNX2 as: TCACGTCGCTCATTTTGC, WNT11: CAAGTTTTCCGATGCTCCTATGAA, WNT11 as: TTGTGTAGACGCATCAGTTTATTGG; WNT5a: CCTGCCAAAAACAGAGGTGT. WNT5a as: CCTGCCAAAAACAGAGGTGT. The results were calculated using the DDCT method and are presented in x-fold increase relative to β-actin mRNA levels.

Wnt profiler PCR array Cells were differentiated with osteogenic medium for 7 days and treated with 25 mM DFO for 48 h. Afterwards, RNA was isolated as described above, reverse transcribed using the RT2 First Strand Kit (SABiosciences) and 500 ng cDNA, and were subjected to the Wnt profiler PCR array, containing 84 Wnt-related genes, according to the manufacturer´s protocol (SABiosciences). Genes were normalized to the mean of five housekeeping genes (β-actin, Gapdh, Hsp90ab1, Gusb, b2m).

Western blot analysis For the Western Blot analysis, cells were either stimulated with 25 mM DFO for 30 min (signaling studies) or 48 h (Wnt5a expression). Cells were lysed in lysis buffer containing 20 mM Tris/HCl pH 7.4, 1% SDS and a protease inhibitor (complete mini, Roche).

Lysates were processed through a 24-gauge needle and centrifuged at 20,000 x g for 20 min. The protein content was measured using the BCA method. For electrophoresis, 20 mg protein was loaded on a 10% SDS-PAGE and transferred onto a 0.2 mm nitrocellulose membrane (Whatman). After blocking for 1 h with 5% nonfat dry milk (Wnt5a) or 1% BSA (signaling antibodies) in Tris-buffered saline with 1% Tween-20 (TBS-T), membranes were incubated with an anti-WNT5a antibody (Santa Cruz Biotech, 1:1,000) or the respective signaling antibodies (AKT, pAKT, ERK, pERK, p38, pp38, NF-κB, pNF-κB, all from Cell Signaling, 1:1,000; NFATc1: Thermo Fischer Scientific, 1:500) overnight. GAPDH was used as a loading control (1:2,000, Santa Cruz Biotech). Membranes were washed and incubated with the appropriate HRP conjugated secondary antibodies for 1 h at RT. After washing again, membranes were incubated with an ECL substrate (Pierce, Thermo Fisher Scientific) to visualize the proteins using the MF-ChemiBIS 3.2 bioimaging system (Biostep, Jahnsdorf, Germany).

A

Mineralization - hBMSC

B

Statistical analysis Data are presented as mean ± standard deviation (SD).

Statistical evaluations of two group comparisons were performed using a two-sided Student’s t-test. One-way analysis of variance (ANOVA) was used for experiments with more than two groups or time/dose experiments.

Results Excess iron and iron chelation exhibit opposing effects on osteogenic differentiation To determine the effects of iron excess and iron deficiency on the osteogenic differentiation, we treated mBMSC with increasing concentrations of iron (III) chloride or DFO for the entire differentiation period. Iron excess led to a marked decrease in osteoblast differentiation as measured by the amount of mineralized matrix reaching a maximum suppression of 67% at a concentration of 50 mM (Figure 1A). Similar results were also seen when ferric ammonium citrate or holo-transferrin were used as different iron sources (Figure 1B). In contrast, DFO stimulated osteogenic differentiation showing the most potent effect at a concentration of 25 mM (3.7-fold) (Figure 1C). To address whether the osteoblast-promoting effect of DFO is specifically due to chelating iron, DFO-treated cells were incubated with increasing concentrations of iron chloride. Indeed, higher doses of iron dose-dependently suppressed the DFO-induced mineralization (Figure 1D).

Cell viability was not altered by iron concentrations up to 50 mM (50606 ± 11205 vs. 48175 ± 9875 fluorescence units, not significant), while DFO started to inhibit osteoblast viability at concentrations of 50 mM (50606 ± 11205 vs.

42671 ± 14690 fluorescence units, P<0.05). The inhibitory effects of iron excess and stimulatory effects of iron chelation were further demonstrated by analyzing the gene expression levels of the osteoblast markers Runx2, Alp and osteocalcin (Bglap) as well as by measuring the enzyme activity of ALP (Figures 1E-H). In all cases, iron excess suppressed osteoblast markers, while DFO increased their expression. The DFO-mediated induction of the osteoblastic genes was reversed when cells were concomitantly exposed to iron (Figures 1E-G).

Runx2 - hBMSC

C

ALP- hBMSC

P=0.06

Figure 2. Osteogenic differentiation of human bone marrow stromal cells is impaired by iron excess and increased by iron chelation. (A) Bone marrow stromal cells were obtained from healthy donors (hBMSCs) and differentiated towards osteoblasts in the presence of 5 mM iron chloride (Fe) and 25 mM deferoxamine (DFO).

Mineralization was determined using alizarin red S staining and subsequent elution and quantification. (B-C) Quantitative gene expression analysis of Runx2 and alkaline phosphatase (ALP) after treating day 10 differentiated cells with Fe (5 mM) or DFO (25 mM) for an additional 48 hours. N=4. *P<0.05 vs. control (CO).

1502

haematologica | 2016; 101(12)

Wnt5a mediates DFO effects on osteoblasts

To assess whether altering the iron balance also affects osteogenic differentiation in human cells, BMSCs isolated from human donors were treated with iron (III) chloride and DFO. Similar to murine cells, iron excess inhibited osteoblast function by 42%, while DFO stimulated matrix mineralization by 2.6-fold (Figure 2A). Also, corresponding effects were seen on the gene expression of RUNX2 and ALP (Figure 2B,C). Thus, iron excess suppresses osteogenic differentiation while iron chelation has proosteogenic effects in human and murine BMSC.

Wnt5a expression is induced by deferoxamine As DFO exerted potent pro-osteogenic effects and Wnt signaling is required for osteoblast differentiation, we analyzed the Wnt profile of osteoblasts after treatment with 25 mM DFO for 48 h. As indicated in Figure 3A, three distinct Wnt inhibitors were significantly down-regulated after DFO treatment (Sfrp1, Sfrp2 and Sfpr4) along with the canonical Wnt ligands Wnt2b, Wnt4, Wnt7b, Wnt8 and Wnt16 and the transcription factor Tcf7 and Lef1 (Figure

3A). In addition, the canonical Wnt target genes axin 2, cyclin D1 (Ccnd1) and cyclin D2 (Ccnd2) were significantly down-regulated by DFO suggesting an overall down-regulation of canonical Wnt signaling (Figure 3A). With regard to non-canonical Wnt signaling, Wnt5a and Wnt11 as well as the downstream mediators Jun and Fosl1 were up-regulated 2.5- 6-fold (Figure 3A).

The induction of Wnt5a was further validated using timeand dose-response curves. Wnt5a expression was dosedependently increased after DFO treatment reaching a maximum at 100 mM (2.3-fold) after 48 h (Figure 3B,C). The induction of Wnt5a expression by DFO was finally also confirmed by Western blot analysis showing a 3.7-fold induction (Figure 3D,E). In addition, WNT5a expression was also increased after 48 h of DFO treatment in human BMSC (Figure 3F). Iron stimulation did not alter Wnt5a expression (Figure 3G). Hence, DFO targets several components of the Wnt signaling pathway, with the robust induction of Wnt5a representing a pro-osteogenic Wnt ligand that may contribute to the osteoblast-supporting effects of DFO.

A

B

C

E

F

G D

Figure 3. Deferoxamine (DFO) modulates the expression of components of the Wnt pathway. (A) Bone marrow stromal cells were differentiated towards osteoblasts for 10 days and treated with 25 mM DFO for 48 hours. A Wnt profiler polymerase chain reaction (PCR) array was performed using real-time PCR. Genes are separated into three categories: Wnt ligands/receptors, Wnt target genes and Wnt inhibitors. Genes in green indicate down-regulated genes (lighter green means more intensively down-regulated); red ones are up-regulated. All genes presented in the bar graph were significantly regulated (P<0.05). N=3. (B-C) Day 10 differentiated cells were stimulated with various concentrations and durations of DFO. Afterwards, Wnt5a expression was determined using real-time PCR. N=4. (D-E) Western blot analysis of WNT5a in osteoblasts after DFO treatment (25 mM) for 48 hours. Three independent experiments are shown. Graph indicates quantification experiments using ImageJ. (F) Human bone marrow stromal cells (hBMSCs) were obtained from healthy donors and differentiated towards osteoblasts for 10 days. Afterwards, cells were treated with Fe (iron chloride) (5 mM) or DFO (25 mM) for an additional 48 hours to assess WNT5a expression using real-time PCR. N=4. (G) Murine bone marrow stromal cells were differentiated towards osteoblasts for 10 days and treated with 5-50 mM FeCl3 for 48 hours. Afterwards, Wnt5a expression was determined using real-time PCR. N=3. *P<0.05, **P<0.01 vs. control (CO).

haematologica | 2016; 101(12)

1503

U. Baschant et al.

The PI3K and NFAT pathways mediate the induction of WNT5a expression by DFO To examine the pathways involved in the induction of Wnt5a levels after DFO treatment, we analyzed the activation of different pathways that have previously been reported to be activated by DFO.40-43 After 30 min stimulation, DFO (50 mM) activated several pathways, most notably NFATc1, AKT, and NF-κB. The MAPK downstream pathways ERK and p38 were not activated by DFO (Figure 4A). To verify the necessity of the NFATc1, AKT, and NF-κB pathways, we used specific pathway inhibitors prior to stimulation with DFO. Only the PI3K/AKT inhibitor (LY-294002) and the NFAT inhibitor (VIVIT) were able to block the DFO-induced Wnt5a expression, as shown in Figure 4B,C. Inhibition of the NF-κB (BAY-117082) and ERK pathways (UO126) did not interfere with the induction of Wnt5a expression after 48 h of DFO treatment (Figure 4B,C). Thus, DFO signals through the PI3K/AKT as well as the NFATc1 pathways to up-regulate Wnt5a expression in osteoblasts.

Wnt5a mediates the pro-osteogenic effect of DFO As Wnt5a is a known pro-osteogenic Wnt ligand and its expression is highly induced by DFO, we next assessed whether the pro-osteogenic effect of DFO is mediated via Wnt5a. Therefore, we isolated BMSC from

A

Wnt5afl/fl:ROSA26-Cre-ERT2 or their littermate controls (Wnt5afl/fl), and deleted Wnt5a expression in osteogenic cells in vitro using tamoxifen for 3 consecutive days. After 7 days, the recombination efficiency was analyzed using real-time PCR analysis and revealed a 75% deletion of Wnt5a expression levels (Wnt5afl/fl: 1.17 ± 0.22; Wnt5afl/fl:ROSA26-Cre-ERT2: 0.28 ± 0.08 relative expression, P<0.001). Next, we treated Wnt5a proficient and Wnt5a deficient cells with increasing concentrations of DFO throughout the entire differentiation period of 21 days. Alizarin red staining showed that the pro-osteogenic effect of DFO at the concentrations of 25 and 50 mM were completely abrogated in Wnt5a deficient cells (Figure 5A,B). Therefore, WNT5a mediates the osteoinductive effect of DFO. Finally, we treated healthy 12-week-old male mice with DFO for three weeks to assess whether Wnt5a is also regulated in vivo and determine whether DFO alters bone mass. As shown in Figure 5C, Wnt5a expression was induced 2-fold in bone marrow cells isolated from DFOtreated animals. Moreover, DFO tended to increase the bone volume fraction in the fourth lumbar vertebrae (Figure 5D). In addition, the trabecular thickness and the trabecular number also tended to be increased, whereas trabecular separation was decreased, suggesting bone-promoting effects (Figure 5E-G). Similar to the in vitro experi-

B

C

Figure 4. Deferoxamine (DFO) modulates the expression of Wnt5a via the PI3K and NFAT pathways. (A) Day 10 differentiated bone marrow stromal cells (BMSCs) were treated with 50 mM DFO for 30 min. Western blot analysis was performed for various signaling pathways. Protein amounts were normalized to GAPDH. Three independent experiments were performed. Two are shown. (B-C) BMSCs were treated with the following inhibitors prior to the addition of 50 mM DFO for 48 hours: BAY-11-7082 (I-κB phosphorylation inhibitor, 1 mM), UO126 (MEK1 inhibitor, 10 mM), LY-294002 (PI3K inhibitor, 10 mM), and VIVIT (NFAT inhibitor, 1 mM). Afterwards, Wnt5a expression was determined using either real-time polymerase chain reaction (PCR) (B) or Western Blot analysis (C). All three Western blots were quantified using ImageJ. N=3-4. *P<0.05 vs. control (CO).

1504

haematologica | 2016; 101(12)

Wnt5a mediates DFO effects on osteoblasts

ments, Wnt5a expression was not altered in the bones derived from iron-overloaded mice (wild-type: 37.3 Âą 10.0 transferrin receptor 2 knockout: 27.3 Âą 11.9, relative expression, P=0.14). Thus, DFO also modulates Wnt5a expression in a complex in vivo system.

Discussion Herein we show that altering the exogenous iron concentration in osteoblast cultures critically influences their differentiation capacity, with iron excess inhibiting and iron chelation promoting osteoblast differentiation.

Moreover, we have identified Wnt5a as a target gene of DFO that contributes to its pro-osteogenic effects.

Even though there is strong evidence that bone remodeling is affected by iron concentrations which are too high or too low,6-7,10-13 the cellular effects, in particular on osteoblast biology, are not fully understood. The results of our study confirm the inhibitory effect of iron excess on osteoblast function as indicated by a decreased matrix production and a reduced expression of osteoblast markers.2325,43 Wnt5a expression was not reduced by the iron treatment, suggesting that the suppression of Wnt5a expression may not be an underlying mechanism. Our study is in line with a number of other studies that have shown an

A

C

osteoblast-supporting effect of DFO.26-28,44,45 Interestingly, osteoblasts treated with 50 mM DFO showed a somewhat decreased mineralization potential as compared to 25 mM DFO, indicating that this concentration might already exert toxic effects. In fact, we found a reduced cell viability in murine cells treated with 50 mM DFO, while cell vitality was not affected at 25 mM. Depending on the cell type, other studies have also shown toxic effects of DFO in concentrations above 30 mM.27,28 In contrast, human periodontal ligament cells do not seem to be particularly sensitive to DFO, as even concentrations of up to 200 mM still exerted pro-osteogenic effects.26 In our hands, DFO reduced human BMSC viability at similar concentrations as murine BMSC (i.e., 50 mM, data not shown). However, it should be noted that the impact on cell viability did not seem to be of biological significance, as osteoblasts treated with 50 mM still differentiate better than untreated cells.

Therefore, at non-toxic concentrations, the iron-chelating effect of DFO seems to promote osteoblast differentiation.

To gain more insights into the molecular mechanisms that contribute to the osteoblast-promoting effects of DFO, we investigated whether Wnt signaling, a critical pathway for osteoblast differentiation, is involved.

Canonical and non-canonical Wnt pathways both positively affect bone mass. While canonical signaling increases osteoblast function and suppresses osteoclast function,

B

D

E

F

G

Figure 5. Wnt5a mediates the pro-osteogenic effect of deferoxamine (DFO). (A-B) Bone marrow stromal cells were isolated from wild-type (WT) or Wnt5a:ROSA26ERT2Cre mice. Deletion of Wnt5a was induced by treating the cells with 1 mM tamoxifen for 3 consecutive days. After deletion, cells were differentiated for 21 days with or without the addition of increasing concentrations of DFO. (A) Mineralization was assessed using alizarin red S. (B) Quantification of alizarin red S after elution with cetylpyridinium chloride. N=4. (C-G) Twelve-week-old male 129Sv mice were treated with 250 mg/kg DFO or phosphate buffered saline (PBS) daily for three weeks. Thereafter, mice were sacrificed to assess the expression of Wnt5a in the bone marrow using real-time polymerase chain reaction (PCR). (C) (D) Bone volume/total volume (BV/TV), (E) trabecular thickness (Tb.Th), (F) trabecular number (Tb.N), and (G) trabecular separation (Tb.Sp) at the fourth lumbar vertebrae were determined using mCT. N=5. *P<0.05 vs. control (CO).

haematologica | 2016; 101(12)

1505

U. Baschant et al.

non-canonical WNT5a signaling increases both osteoblast and osteoclast function. This mechanism nonetheless also results in an increased bone mass.32 Using a Wnt profiler array we found that several genes involved in canonical Wnt signaling were down-regulated, including Wnt2b, Wnt7b, Tcf7, Fzd5, Fzd6 and axin-2. Previously, Qu et al. demonstrated that DFO promotes the stabilization of β-catenin through phosphorylating glycogen synthase kinase-3β, and that the pro-osteogenic effect of DFO is dependent on β-catenin.27 Other studies, however, report an inhibition of Wnt target genes by DFO in osteoblasts46 and a suppression of canonical Wnt signaling in neuronal cells.47,48 As Wnt signaling is strongly context- and cell type-dependent, it would not be surprising if DFO would exert opposing effects on Wnt signaling in different cell types. Nevertheless, more research is required to address the discrepancies on the regulation of canonical Wnt signaling by DFO. In contrast to the canonical Wnt genes, the non-canonical ligands Wnt5a and Wnt11 were up-regulated after DFO treatment. In particular Wnt5a was robustly induced by DFO in a time- and dose-dependent manner and required the activation of the PI3K/AKT and NFAT pathways. Both the PI3K/Akt and NFAT pathways have also previously been shown to be activated upon stimulation with DFO in epidermal and pheochromocytoma cells.40,41 Even though the ERK and p38 pathways have also been reported to be activated by DFO, these pathways were not necessary to induce Wnt5a expression in osteoblasts.42,43 Besides directly activating specific signaling pathways, DFO is also known to inhibit prolyl hydroxylases (PHDs) by sequestering the iron that is required for their activity. Inhibition of PHDs leads subsequently to the stabilization of hypoxiainducible factors (HIFs).49 Although the exact role of HIF proteins on osteoblast differentiation is not fully understood, the majority of studies show that HIF activation reduces osteoblast differentiation and bone mass.49-52 While we cannot exclude that hypoxia-dependent pathways contribute to the effects of DFO on osteoblasts, it seems unlikely to mediate its osteoblast supportive effects.

As Wnt5a promotes osteoblast differentiation,34 we further examined whether the induction of Wnt5a is necessary for the stimulation of osteoblast differentiation by DFO. Even though the ablation of Wnt5a expression in osteoblasts did not affect the differentiation potential in vitro, it completely abolished the pro-osteogenic effect of

References 1. Cassat JE & Skaar EP. Iron in infection and immunity.

Cell Host Microbe.

2013;13(5):509-520.

2. Hentze MW, Muckenthaler MU, Andrews NC. Balancing acts: molecular control of mammalian iron metabolism. Cell.

2004;117(3):285-297.

3. Koskenkorva-Frank TS, Weiss G, Koppenol WH, Burckhardt S. The complex interplay of iron metabolism, reactive oxygen species, and reactive nitrogen species: insights into the potential of various iron therapies to induce oxidative and nitrosative stress. Free Radic Biol Med. 2013;65:1174-1194.

4. Abraham R, Walton J, Russell L, et al.

1506

5.

6.

7.

8.

DFO, indicating that this is a crucial pathway of DFO to stimulate osteoblast differentiation in vitro. Likewise, DFO treatment in mice led to an increased expression of Wnt5a in the bone marrow, underlining its significance in vivo.

Finally, bone volume fraction was almost significantly increased, suggesting bone-promoting effects of DFO in vivo.

Our study is potentially limited by the use of only one iron chelator (DFO) and one type of iron source (FeCl3) in most experiments. In other studies, different sources of iron are found, for example ferric ammonium citrate.28 However, all studies have shown suppressive effects of iron on osteoblast differentiation, suggesting that the exact chemical composition may not be of relevance for in vitro studies. This is further supported by our study showing that ferric ammonium citrate and holo-transferrin also suppressed osteoblast differentiation. Furthermore, considering that DFO also binds to aluminum, copper and zinc, although with a lower affinity,53 we could verify that the osteoblast-promoting effects of DFO are specific for iron, as the addition of iron to DFO-treated cultures inhibited the mineralization capacity of the osteoblasts.

In conclusion, our results demonstrate that certain iron concentrations are necessary for proper osteoblast differentiation and that Wnt5a is required for the pro-osteogenic effect of DFO. Thus, in addition to its potent effect in reducing the iron load in patients with hemochromatosis or sickle cell anemia, treatment with DFO and perhaps also novel iron chelating agents may also improve bone health in those patients. However, this assumption needs to be verified in further studies.

Funding This work was supported by Start-up grants of the Medical Faculty of the Technische Universität Dresden (Meddrive) to UB and MR, the German Research Foundation (SFB655/P13) to UP and LCH, the Josè Carreras foundation to MR and UP, and funding from the Excellence Initiative by the German Federal and State Governments" (Institutional Strategy, measure "Support the best") to MR and LCH. DFG funding SFB-655, project B13 to UP and LCH; Startup grants from the Medical Faculty of the TUD to UB and MR; Jose Carreras foundation to MR and UP (DJCLS R 13/15); Excellence Initiative by the German Federal and State Governments" (Institutional Strategy, measure "Support the best") to MR and LCH.

Dietary determinants of post-menopausal bone loss at the lumbar spine: a possible beneficial effect of iron. Osteoporos Int. 2006;17(8):1165-1173.

Harris MM, Houtkooper LB, Stanford VA, et al. Dietary iron is associated with bone mineral density in healthy postmenopausal women. J Nutr. 2003;133(11):3598-3602.

Katsumata S, Katsumata-Tsuboi R, Uehara M, Suzuki K. Severe iron deficiency decreases both bone formation and bone resorption in rats. J Nutr. 2009;139(2):238243.

Medeiros DM, Plattner A, Jennings D, Stoecker B. Bone morphology, strength and density are compromised in iron-deficient rats and exacerbated by calcium restriction.

J Nutr. 2002;132(10):3135-3141.

Kenner GH, Brik AB, Liu G, et al. Variation

of long-lived free radicals responsible for the EPR native signal in bone of aged or diseased human females and ovariectomized adult rats. Radiat Meas. 2005; 39(3):255262.

9. Liu G, Men P, Kenner GH, Miller SC. Ageassociated iron accumulation in bone: implications for postmenopausal osteoporosis and a new target for prevention and treatment by chelation. Biometals.

2006;19(3):245-251.

10. Guggenbuhl P, Deugnier Y, Boisdet JF, et al.

Bone mineral density in men with genetic hemochromatosis and HFE gene mutation.

Osteoporos Int. 2005;16(12):1809-1814.

11. Skordis N, Toumba M. Bone disease in thalassaemia major: recent advances in pathogenesis and clinical aspects. Pediatr Endocrinol Rev. 2011;Suppl2:300-306.

haematologica | 2016; 101(12)

Wnt5a mediates DFO effects on osteoblasts

12. Sarrai M, Duroseau H, D'Augustine J, Moktan S, Bellevue R. Bone mass density in adults with sickle cell disease. Br J Haematol. 2007;136(4):666-672.

13. Valenti L, Varenna M, Fracanzani AL, et al.

Association between iron overload and osteoporosis in patients with hereditary hemochromatosis. Osteoporos Int. 2009; 20(4):549-555.

14. Casale M, Citarella S, Filosa A, et al.

Endocrine function and bone disease during long-term chelation therapy with deferasirox in patients with -thalassemia major. Am J Hematol. 2014;89(12):11021106.

15. Isomura H, Fujie K, Shibata K, et al. Bone metabolism and oxidative stress in postmenopausal rats with iron overload.

Toxicology. 2004;15:197(2):93-100.

16. Kudo H, Suzuki S, Watanabe A, et al.

Effects of colloidal iron overload on renal and hepatic siderosis and the femur in male rats. Toxicology. 2008;246(2-3):143-147.

17. Tsay J, Yang Z, Ross FP, et al. Bone loss caused by iron overload in a murine model: importance of oxidative stress. Blood.

2010;116(14):2582-2589.

18. Ishii KA, Fumoto T, Iwai K, et al.

Coordination of PGC-1beta and iron uptake in mitochondrial biogenesis and osteoclast activation. Nat Med. 2009; 15(3):259-266.

19. Jia P, Xu YJ, Zhang ZL, et al. Ferric ion could facilitate osteoclast differentiation and bone resorption through the production of reactive oxygen species. J Orthop Res. 2012;30(11):1843-1852.

20. Xiao W, Beibei F, Guangsi S, et al. Iron overload increases osteoclastogenesis and aggravates the effects of ovariectomy on bone mass. J Endocrinol. 2015;226(3):121-134.

21. Xie W, Lorenz S, Dolder S, Hofstetter W. Extracellular iron is a modulator of the differentiation of osteoclast lineage cells.

Calcif Tissue Int. 2016;98(3):275-283.

22. Guo JP, Pan JX, Xiong L, et al. Iron Chelation Inhibits Osteoclastic Differentiation In Vitro and in Tg2576 Mouse Model of Alzheimer's Disease.

PLoS One. 2015;10(11):e0139395.

23. Doyard M, Fatih N, Monnier A, et al. Iron excess limits HHIPL-2 gene expression and decreases osteoblastic activity in human MG-63 cells.

Osteoporos Int. 2012;23(10):2435-2445.

24. Yamasaki K, Hagiwara H. Excess iron inhibits osteoblast metabolism. Toxicol Lett. 2009;191(2-3):211-215.

25. Messer JG, Kilbarger AK, Erikson KM, Kipp DE. Iron overload alters iron-regulatory genes and proteins, down-regulates osteoblastic phenotype, and is associated with apoptosis in fetal rat calvaria cultures.

Bone. 2009;45(5):972-979.

26. Chung JH, Kim YS, Noh K, et al.

Deferoxamine promotes osteoblastic differentiation in human periodontal ligament cells via the nuclear factor erythroid 2-related factor-mediated antioxidant signaling pathway. J Periodontal Res. 2014;49(5):563573.

haematologica | 2016; 101(12)

27. Qu ZH, Zhang XL, Tang TT, Dai KR. Promotion of osteogenesis through betacatenin signaling by desferrioxamine.

Biochem Biophys Res Commun.

2008;370(2):332-337.

28. Zhao GY, Zhao LP, He YF, et al. A comparison of the biological activities of human osteoblast hFOB1.19 between iron excess and iron deficiency. Biol Trace Elem Res. 2012;150(1-3):487-495.

29. Farberg AS, Jing XL, Monson LA, et al.

Deferoxamine reverses radiation induced hypovascularity during bone regeneration and repair in the murine mandible. Bone.

2012;50(5):1184-1187.

30. Wan C, Gilbert SR, Wang Y, et al. Role of hypoxia inducible factor-1 alpha pathway in bone regeneration. J Musculoskelet Neuronal Interact. 2008;8(4):323-324.

31. Messer JG, Cooney PT, Kipp DE. Iron chelator deferoxamine alters iron-regulatory genes and proteins and suppresses osteoblast phenotype in fetal rat calvaria cells. Bone. 2010;46(5):1408-1415.

32. Baron R & Kneissel M. WNT signaling in bone homeostasis and disease: from human mutations to treatments. Nat Med. 2013;19(2):179-192.

33. Gong Y, Slee RB, Fukai N, et al. LDL receptor-related protein 5 (LRP5) affects bone accrual and eye development. Cell. 2001; 107(4):513-523.

34. Maeda K, Kobayashi Y, Udagawa N, et al.

Wnt5a-Ror2 signaling between osteoblastlineage cells and osteoclast precursors enhances osteoclastogenesis. Nat Med. 2012;18(3):405-412.

35. Miyoshi H, Ajima R, Luo CT, Yamaguchi TP, and Stappenbeck TS. Wnt5a potentiates TGF-beta signaling to promote colonic crypt regeneration after tissue injury.

Science. 2012;338(6103):108-113.

36. Roetto A, Di Cunto F, Pellegrino RM, et al.

Comparison of 3 Tfr2-deficient murine models suggests distinct functions for Tfr2alpha and Tfr2-beta isoforms in different tissues. Blood. 2010;115(16):3382-3389.

37. Rauner M, Föger-Samwald U, Kurz MF, et al. Cathepsin S controls adipocytic and osteoblastic differentiation, bone turnover, and bone microarchitecture. Bone.

2014;64:281-287.

38. Sinningen K, Albus E, Thiele S, et al. Loss of milk fat globule-epidermal growth factor 8 (MFG-E8) in mice leads to low bone mass and accelerates ovariectomy-associated bone loss by increasing osteoclastogenesis.

Bone. 2015;76(25):107-114.

39. Rauner M, Stein N, Winzer M, et al.

WNT5A is induced by inflammatory mediators in bone marrow stromal cells and regulates cytokine and chemokine production.

J Bone Miner Res. 2012;27(3):575-585.

40. Alvarez-Tejado M, Naranjo-Suarez S, Jiménez C, et al. Hypoxia induces the activation of the phosphatidylinositol 3kinase/Akt cell survival pathway in PC12 cells: protective role in apoptosis. J Biol Chem. 2001;276(25):22368-22374.

41. Huang C, Li J, Zhang Q, Huang X. Role of bioavailable iron in coal dust-induced acti-

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

vation of activator protein-1 and nuclear factor of activated T cells: difference between Pennsylvania and Utah coal dusts.

Am J Respir Cell Mol Biol. 2002;27(5):568574.

Klettner A, Koinzer S, Waetzig V, Herdegen T, Roider J. Deferoxamine mesylate is toxic for retinal pigment epithelium cells in vitro, and its toxicity is mediated by p38. Cutan Ocul Toxicol 2010;29(2):122-129.

Markel TA, Crisostomo PR, Wang M, et al.

Iron chelation acutely stimulates fetal human intestinal cell production of IL-6 and VEGF while decreasing HGF: the roles of p38, ERK, and JNK MAPK signaling. Am J Physiol Gastrointest Liver Physiol. 2007; 292(4):958-963.

Chen B, Yan YL, Liu C, et al. Therapeutic effect of deferoxamine on iron overloadinduced inhibition of osteogenesis in a zebrafish model. Calcif Tissue Int. 2014; 94(3):353-360.

Naves Díaz ML, Elorriaga R, Canteros A, Cannata Andía JB. Effect of desferrioxamine and deferiprone (L1) on the proliferation of MG-63 bone cells and on phosphatase alkaline activity. Nephrol Dial Transplant. 1998;Suppl3:23-28.

Chen D, Li Y, Zhou Z, et al. Synergistic inhibition of Wnt pathway by HIF-1 and osteoblast-specific transcription factor osterix (Osx) in osteoblasts. PLoS One. 2012;7(12):e52948.

Meng H, Li F, Hu R, et al. Deferoxamine alleviates chronic hydrocephalus after intraventricular hemorrhage through iron chelation and Wnt1/Wnt3a inhibition.

Brain Res. 2015;1602:44-52.

Ziaei A, Ardakani MR, Hashemi MS, et al.

Acute course of deferoxamine promoted neuronal differentiation of neural progenitor cells through suppression of Wnt/ catenin pathway: a novel efficient protocol for neuronal differentiation. Neurosci Lett.

2015;590:138-144.

Rauner M, Franke K, Murray M, et al.

Increased EPO Levels Are Associated with Bone Loss in Mice Lacking PHD2 in EPOProducing Cells. J Bone Miner Res. 2016 Apr 15 (Epub ahead of print).

Hsu SH, Chen CT, Wei YH. Inhibitory effects of hypoxia on metabolic switch and osteogenic differentiation of human mesenchymal stem cells. Stem Cells.

2013;31(12):2779-2788.

Liu T, Zou W, Shi G, Xu J, Zhang F, Xiao J, Wang Y. Hypoxia-induced MTA1 promotes MC3T3 osteoblast growth but suppresses MC3T3 osteoblast differentiation. Eur J Med Res. 2015;20:10.

Xu Y, Wang S, Tang C, Chen W. Upregulation of long non-coding RNA HIF 1 -anti-sense 1 induced by transforming growth factor- -mediated targeting of sirtuin 1 promotes osteoblastic differentiation of human bone marrow stromal cells. Mol Med Rep. 2015;12(5):7233-7238.

Cuajungco MP, Faget KY, Huang X, Tanzi RE, Bush AI. Metal chelation as a potential therapy for Alzheimer's disease. Ann N Y Acad Sci. 2000;920:292–304.

1507

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Myelodysplastic Syndromes

Ferrata Storti Foundation

Haematologica 2016 Volume 101(12):1508-1515

Response to treatment with azacitidine in children with advanced myelodysplastic syndrome prior to hematopoietic stem cell transplantation

Nicolas Waespe,1,2 Machiel Van Den Akker,1,3 Robert J. Klaassen,4 Lani Lieberman,5 Meredith S. Irwin,6 Salah S. Ali,7 Mohamed Abdelhaleem,8 Bozana Zlateska,1,2 Mira Liebman,1 Michaela Cada,1 Tal Schechter7 and Yigal Dror1,2,9

Marrow Failure and Myelodysplasia Program, Division of Hematology/Oncology, Department of Pediatrics, The Hospital for Sick Children, Toronto, Canada; 2Genetics and Genome Biology Program, The Hospital for Sick Children, Toronto, Canada; 3 Pediatric Hematology/Oncology, UZ Brussel, Jette, Belgium; 4Department of Pediatrics, Children's Hospital of Eastern Ontario, Ottawa, Canada; 5Department of Laboratory Medicine, University Health Network, Toronto, Canada; 6Division of Hematology/ Oncology, Department of Pediatrics, The Hospital for Sick Children, Toronto, Canada; 7 Bone Marrow Transplantation Program, Division of Hematology/Oncology, Department of Pediatrics, The Hospital for Sick Children, Toronto, Canada; 8Department of Pediatric Laboratory Medicine, Division of Hematopathology, The Hospital for Sick Children, Toronto, Canada and 9Institute of Medical Science, University of Toronto, Canada

1

ABSTRACT

A

Correspondence: yigal.dror@sickkids.ca

Received: March 16, 2016.

Accepted: August 18, 2016.

Pre-published: August 18, 2016. doi:10.3324/haematol.2016.145821

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/100/12/1508

Š2016 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.

1508

dvanced myelodysplastic syndrome harbors a high risk of progression to acute myeloid leukemia and poor prognosis. In children, there is no established treatment to prevent or delay progression to leukemia prior to hematopoietic stem cell transplantation.

Azacitidine is a hypomethylating agent, which was shown to slow progression to leukemia in adults with myelodysplastic syndrome. There is little data on the efficacy of azacitidine in children. We reviewed 22 pediatric patients with advanced myelodysplastic syndrome from a single center, diagnosed between January 2000 and December 2015. Of those, eight patients received off-label azacitidine before hematopoietic stem cell transplantation. A total of 31 cycles were administered and modification or delay occurred in four of them due to cytopenias, infection, nausea/vomiting, and transient renal impairment. Bone marrow blast percentages in azacitidine-treated patients decreased significantly from a median of 15% (range 9-31%) at the start of treatment to 5.5% (0-12%, P=0.02) before hematopoietic stem cell transplantation.

Following azacitidine treatment, four patients (50%) achieved marrow remission, and none progressed. In contrast, three untreated patients (21.4%) had progressive disease characterized by >50% increase in blast counts or progression to leukemia. Azacitidine-treated patients had significantly increased 4-year event-free survival (P=0.04); predicted 4-year overall survival was 100% versus 69.3% in untreated patients (P=0.1). In summary, azacitidine treatment prior to hematopoietic stem cell transplantation was well tolerated in pediatric patients with advanced myelodysplastic syndrome, led to partial or complete bone marrow response in seven of eight patients (87.5%), and correlated with superior event-free survival in this cohort.

Introduction Myelodysplastic Syndrome (MDS) is a clonal disorder with cytopenias, cytogenetic aberrations, varying degrees of bone marrow dysplasia, and leukemic blasts.

Pediatric MDS is rare and comprises about 3% of childhood cancers.1 Advanced MDS in children with an increase in leukemic blasts in the bone marrow (5-29%) haematologica | 2016; 101(12)

Azacitidine in advanced pediatric MDS

and/or peripheral blood (2-29%), is termed either refractory cytopenia with excess blasts (RCEB),2,3 or divided into refractory anemia with excess blasts (RAEB, bone marrow blasts 5-19% and/or peripheral blasts 2-19%) and refractory anemia with excess blasts in transformation (RAEB-T, bone marrow and/or peripheral blasts 20-29%).4 RCEB harbors a high propensity of transformation to leukemia, mainly MDS-related acute myeloid leukemia (MDRAML).1,5 The outcome of RCEB in childhood is poor, with a 5-year overall survival (OS) of about 35-63%,6–9 which is inferior to survival in MDS without excess blasts10,11 and de novo AML.12 Allogeneic hematopoietic stem cell transplantation (HSCT) is currently the only curative treatment for RCEB.8 Earlier HSCT correlates with better outcome,13 but preparations for HSCT are often lengthy and appropriate stem cell donors may not be readily available. The main causes of poor outcome include progression to leukemia, high treatment-related toxicity and high relapse incidence of MDS or occurrence of MDR-AML after HSCT.8 The administration of AML-type induction chemotherapy to children with RCEB prior to HSCT did not improve outcome.13 However, the occurrence of progression to MDR-AML leads to reduction in survival by about 50%, and intensive pre-transplant chemotherapy is typically given, resulting in significant toxicity.8 There is currently no pre-HSCT treatment that has been established in children with RCEB to prevent progression to MDR-AML, and patients are closely monitored and typically managed without any cytoreductive treatment prior to HSCT. Aberrant methylation of critical genes was seen in adult and pediatric patients with advanced MDS, and is believed to be one of the driving alterations of the disease.14,15 Implicated genes act as cell differentiation, cell cycle and cell growth regulators, or play roles in the stress response and apoptosis pathways. Hypermethylation leads to the silencing of regulatory genes and aberrant cell behavior. Furthermore, hypermethylation of the promoters of various genes, such as p15,15,16 DLX4,17 p73,18 and VTRNA1-3,19 has been associated with unfavorable prognosis in MDS. Azacitidine (5-azacytidine, AZA) is a hypomethylating agent that has been approved for the treatment of MDS in adults. The mechanism of action of AZA is related to interference with DNA methyltransferases leading to DNA hypomethylation and subsequent cell cycle exit and differentiation of blast cells. High doses of AZA result in direct cytotoxicity.20 Response to AZA was seen in adults with high-risk MDS after a median of 2-3 cycles with a maximum efficacy after 4-6 cycles.21 AZA reduced the percentage of leukemic blasts in the bone marrow and the rate of transformation to MDR-AML, it also prolonged survival and improved quality of life.22 In adult patients eligible for HSCT, treatment with AZA prior to HSCT improved survival23-25 or resulted in similar outcomes,26 compared to standard AML-type induction chemotherapy prior to HSCT. The short- and long-term side effect profile of AZA is better than that of AML-type chemotherapy.25,27 AZA prior to HSCT was shown to improve outcome in a smaller MDS cohort, including pediatric patients without separate evaluation of this subgroup.25 In a single case report, AZA was shown to induce complete remission in a child with treatment-related MDS and signs of early relapse after HSCT, however, long-term follow-up was not haematologica | 2016; 101(12)

reported.28 The European Working Group of Myelodysplastic Syndromes in Childhood recently reported their experience with AZA in MDS and MDR-AML.29 Their retrospective analysis comprised a heterogeneous group of children with low-grade MDS (refractory cytopenia of childhood), advanced MDS, secondary MDS, MDRAML, and relapsed disease after HSCT. Among the seven patients with RAEB/-T in this report, only one was declared alive in remission at 24 months of follow-up; the remaining 6/7 patients died after 1-6 months.

Notwithstanding this, AZA was given in 4/7 patients after relapse only, and half of those received AZA with palliative intent. Furthermore, there was no description of what the inclusion criteria were in order to offer AZA treatment.

Finally, this study did not report details of the response criteria, such as changes in leukemic blast counts, and did not have a comparative analysis to a non-treated group.

Herein we report on the use of AZA in pediatric patients with RCEB prior to HSCT and compare their outcomes to patients who were not treated with AZA.

Methods Patient population All consecutive patients from 1 to 18 years of age who fulfilled the diagnostic criteria of pediatric refractory cytopenia with excess blasts (RCEB) according to the Category Cytology Cytogenetics (CCC) classification of childhood MDS,3 and were treated, at least partially, at the Hospital for Sick Children, Toronto, Canada between January 1st 2000 and December 31st 2015, were identified.

The definition of RCEB included the categories of “refractory anemia with excess blasts (RAEB)” and “refractory anemia with excess blasts in transformation (RAEB-T)” as defined by Hasle et al.4 In summary, patients were included if they had bone marrow blast counts of 5-29% in addition to one or more of the following: (i) sustained unexplained cytopenia, (ii) prominent multilineage dysplasia, and (iii) acquired clonal cytogenetic abnormality in hematopoietic cells. Throughout the article the term RCEB is used for the patient population.

Patients with Down syndrome (DS)-related MDS were excluded since DS-related MDS is associated with different genetic alterations, and leads to more favorable outcomes after AML-type chemotherapy alone without proceeding to HSCT.30 Patients deemed not eligible for HSCT at the time of evaluation for MDS treatment were also excluded.

The charts of all identified patients were retrospectively reviewed with a collection of demographic and clinical data, laboratory test results, and outcome. Bone marrow blast percentages as determined by morphological examination of bone marrow aspirate samples were included. Bone marrow morphology was assessed by expert hematopathologists at the Hospital for Sick Children, Toronto (n=18), or by expert hematopathologists of other Canadian centers for pediatric hematology and oncology (n=4), not blinded for diagnosis or treatment. Most patients also had flow cytometric evaluation. Response to treatment was assessed using the 2006 revision of the International Working Group (IWG) response criteria in myelodysplasia.31 Bone marrow blast percentages at presentation, after AZA cycles, at progression if present, or before HSCT were used to assign remission status.

Adverse events to administered treatment were graded using the common terminology criteria of adverse events score (CTCAE v4.0).32 The study was approved by the institution’s research ethics board and abided by the principles of the Declaration of Helsinki.

1509

N. Waespe et al.

Statistical analysis Results were presented using descriptive statistics, including calculation of the median and range for continuous variables, and the percentage for categorical variables. Overall survival (OS) was defined as the time from diagnosis of RCEB until death from any cause. Event-free survival (EFS) was defined as the time from diagnosis until evidence of progression to leukemia, relapse of MDS or leukemia after HSCT, or death from any cause. OS and EFS data were compared using the Kaplan-Meier estimate and Mantel-Cox log-rank test to assess statistical difference. Fisher’s exact test was used to compare dichotomous variables, Mann-Whitney U test to

compare continuous variables, and Wilcoxon signed-rank test to compare repeated measurements. A P-value of <0.05 was considered statistically significant. All statistical analyses were performed using GraphPad Prism 6.0h.

Results A total of 29 patients with a diagnosis of RCEB were identified during the 16-year period. Six patients with Down syndrome-associated MDS were excluded. One

Table 1. Characteristics and outcome of the patients included in the present study.

Patient Predisposition number or preceding disease

Age (y) /sex

Cytopenias

BM cytogenetics

1

Idiopathic

9.2/ M

A/T/N

t(3;5)(q25;q34)

AZA: 3 cycles

10/ 14/ 6 Cycle 1: SD; cycle 3: PR

2

Idiopathic

10.3/ F

A/T/N

Normal

AZA: 3 cycles

13/ 9/ 12

Cycle 3: SD

3

Idiopathic

11.7/ F

A/N

Normal

AZA: 3 cycles

13/ 10/ 0

Cycle 1 and 2: CR

4

Idiopathic

17.2/ F

A/T

28/ 28/ 5

Cycle 1: CR

5

Prior chemotherapy

6.6/ F

-7,+8 at diagnosis; AZA: 1 cycle then complex A/T t(1;7)(p10;q10), AZA: 13 cycles t(11;17)(p15;q21); then additional clone: t(X;1)(q21;q12) A/T/N Complex AZA: 2 cycles

15/ 17/ 0

Cycle 1: SD; cycle 3-13: CR

40/ alive in remission

13/ 14/ 3

25/ alive in remission

6 7 8

Germline 11.1/ M RUNX1 mutation Unclassified syndrome 12.3/ F HLH 11.5/ M

15.5/ alive in remission 54.5/ alive in remission 10.5/ alive in remission 50/ alive in remission

Normal Normal

AZA: 2 cycles AZA: 4 cycles

15/ 16/ 8 15/ 31/ 6

AML-I: 2 cycles

7/ 12/ 0

None None

18/ -/ 18 10/ -/ 6

NA NA

None None

11/ -/ 8 13/ -/ NA

NA NA

5/ -/ 5 8/ -/ 18 16/ 22/ 21 14/ 93/ 4 7/ -/ 10 9/ -/ 9

NA NA SD CR NA NA

81/ alive in remission 108/ relapse after first HSCT, salvaged 7.5/ died, DRM 111/ relapse after first HSCT, salvaged 38.5/ died, DRM 98/ alive in remission 58/ alive in remission 14/ died, DRM 26.5/ alive in remission 13.5/ alive in remission

None None

22/ -/ 14 9/ -/ 3

NA NA

53.5/ died, DRM 7.5/ died, TRM post-HSCT

None

5/ -/ 5

NA

53/ alive in remission

1.3/ M

A/T

10 11

Idiopathic Idiopathic

2/ F 6.8/ F

A/T A/T

-7/ i(11)(q10)/ del(5)(q15q31) Complex -7

12 13

Idiopathic Idiopathic

7/ M 8.8/ M

A/T A/N

Normal -7

14 15 16 17 18 19

Idiopathic Idiopathic Idiopathic Idiopathic Prior chemotherapy Germline RUNX1 mutation Constitutional trisomy 8 Hepatitisassociated SAA Behçet disease

9.8/ F 10.6/ F 13.7/ M 15.1/ M 16.6/ F 16.5/ F

A/T A/T/N T/N A/T/N A/N A/T/N

10.3/ F 13/ F

A/N A/T

13.3/ F

T

22

Last follow-up from presentation (months)/ status

A/T T/N

Idiopathic

21

BM treatment response

Cycle 1: SD; cycle 2: CR Cycle 2: PR Before cycle 1: PD; cycles 1 and 3: PR CR

9

20

PreBM blasts HSCT at presentation/ treatment before treatment (if present)/ prior to HSCT (%)

-7 None Normal None t(3;12)(q26.2;p13) ARA-C: 2 weeks del9(q13q22) AML-I: 2 cycles -7,r(6)(p22-23q22) None -7 None +8c,+8 -7 at diagnosis; then complex +8

50/ alive in remission 7/ alive in remission

66/ alive in remission

A: anemia; AML-I: acute myeloid leukemia induction chemotherapy; ARA-C: low-dose cytarabine SC; AZA: azacitidine IV/ SC; BM: bone marrow; CR: complete marrow remission (blast decrease by ≥ 50% from baseline and reduction of blasts in the bone marrow to ≤5%); DRM: disease-related mortality; F: female; HLH: hemophagocytic lymphohistiocytosis; HSCT: hematopoietic stem cell transplantation; M: male; N: neutropenia; NA: not applicable; PD: progressive disease (blast increase by ≥ 50% from baseline or development of acute leukemia); PR: partial marrow response (blast decrease by ≥ 50% from baseline but >5% blasts); SD: stable disease; T: thrombocytopenia; TRM: treatment-related mortality; SAA: severe aplastic anemia.

1510

haematologica | 2016; 101(12)

Azacitidine in advanced pediatric MDS

additional patient was found to have multi-organ dysfunction following previous chemotherapies for a malignant solid tumor. The patient was deemed not to be in a condition to undergo HSCT at the time of initial evaluation and was excluded from our analysis. Ultimately, 22 patients were included in our analysis.

Azacitidine treatment Since October 2010, off-label treatment with AZA was offered to all patients diagnosed with RCEB who did not have a matched related donor. Eight patients agreed to treatment with AZA and were compared to patients not having received this treatment (controls). Controls included children with RCEB before AZA was offered at the treating institutions (n=10), and after its introduction (n=4). The latter 4 patients were not treated with AZA due to family preference, drug coverage matters, or provision of pre-HSCT care at outside centers where AZA was not offered upfront.

AZA was administered at a dose of 75mg/m2 subcutaneously or intravenously for seven consecutive days every 28 days analogous to the dosing scheme in major studies for adult patients.23,25,33 In 7/8 patients, pre-medication with ondansetron was given intravenously or orally prior to starting AZA, and concomitant IV hydration was administered. AZA treatment was started after a median of 41.5 days (range 19-99) from diagnosis and a median of 3 cycles (1-13) were administered.

Patient characteristics Age at presentation, sex, and bone marrow blast percentages at presentation did not differ significantly between the two groups (Table 1). Time to HSCT was significantly longer in the group of patients who underwent treatment with AZA (Online Supplementary Table S1). One patient with treatment-related MDS was included in each group: one patient after autologous HSCT for a solid tumor and one after chemotherapy and radiotherapy for a lymphoma, respectively. One patient in the AZA treatment group, who had a previous short episode of hemophagocytic lymphohistiocytosis (HLH), had been treated successfully with corticosteroids alone and was in remission from HLH without ongoing treatment at the time of diagnosis of RCEB. A control group patient had Behรงet disease and was treated with colchicine and corticosteroids at the time of diagnosis of RCEB. Constitutional RUNX1 mutations were identified in 1/5 tested AZA-treated and 1/5 tested control patients. One patient in the AZA treatment group was found to have a syndromic disease (not further classified) with short stature, developmental delay, and anhidrosis. Most patients were screened for Fanconi anemia by chromosome fragility testing in the blood or on skin fibroblasts. Of those tested, all were within the normal range (n=7 tested in the AZA treatment group and n=9 in the control group).

showed stable myeloid blast counts in repeated bone marrow aspirations within 21 days, and thus the decision was taken not to proceed to AML-type treatment but offer AZA treatment instead, as suggested previously by other authors.35,36 Four patients in the AZA treatment group and 12 patients in the untreated group had clonal marrow cytogenetic abnormalities. There were no significant group differences.

Evolution of additional clones from the time of RCEB diagnosis to HSCT was seen in two patients in the AZA treatment group before the initiation of treatment and in one control after 16, 35, and 28 days, respectively.

Adverse events on azacitidine treatment In total, 31 cycles of AZA were administered. Treatment was delayed or dose reduced in four cycles (Table 2). Severe bilineage cytopenia (CTCAE grade 4) led to a prolonged delay of treatment in one patient, and after one cycle the patient proceeded directly to HSCT with improved blood counts and markedly reduced blast percentage in the bone marrow (reduction from 28% to 5%). Nausea and vomiting (CTCAE grade 3) led to treatment delay after three cycles in one patient with subsequent HSCT. One patient with severe infection (appendicitis with abdominal abscess, CTCAE grade 4) already had evidence of severe neutropenia prior to the initiation of AZA treatment, and another patient had a transient increase in creatinine (CTCAE grade 2). Treatment with the full dose of AZA was resumed in both, the patient with severe infection and the patient with transient rise in creatinine, subsequently. Both patients did not show further major toxicity. All other adverse events were managed with standard supportive care and without modifications or a significant delay in AZA treatment.

P=0.008 P=0.5 P=0.02

Hematological and bone marrow findings Peripheral blood counts at diagnosis did not differ significantly between the two groups. Hypocellular bone marrow specimens (cellularity <50% of age-adjusted reference, n=1 in each group) and prominent dysplasia in at least two cell lineages were detected in a similar subgroup of patients (n=5 in the AZA-treated and n=8 in control patients). Two patients in the AZA treatment group had bone marrow findings consistent with acute erythroleukemia of mixed erythroid/myeloid subtype (AML M6a).34 Both patients haematologica | 2016; 101(12)

Figure 1. Bone marrow blast percentages at diagnosis, before start of azacitidine (AZA) treatment, and before hematopoietic stem cell transplantation (HSCT) in azacitidine-treated patients (n=8).

1511

N. Waespe et al.

Table 2. Adverse events identified in patients treated with azacitidine (n=8).

Grade 4 Hematological toxicity Fever and infection Nausea and vomiting Urticaria and rash Acute kidney injury Diarrhea Fatigue Insomnia Fluid overload with arterial hypertension

Highest CTCAE grade toxicity in a single patient Grade 3 Grade 2

†

2 1†

Grade 1

1 1†

1 1

1 2 1† 1 1 1 1

with treatment modification/delay (n=4, one patient with hematological toxicity and infection); CTCAE: common terminology criteria for adverse events.

†

Morphological response in the bone marrow Of the AZA-treated patients, four (50%) achieved bone marrow remission (reduction in bone marrow blast counts by ≥50% from baseline and absolute bone marrow blasts ≤5% of nucleated cells) after a median of 2.5 cycles (range 1-6). Three patients (37.5%) had partial bone marrow response (reduction in bone marrow blast counts by ≥50% from baseline, but bone marrow blasts >5% of nucleated cells). One patient (12.5%) had stable disease, and none had progressive disease.

Among the controls, one patient did not have repeat bone marrow assessment of sufficient quality. Of the remaining 13, one patient (7.7%) with a history of hepatitis-associated severe aplastic anemia, immunosuppressive treatment, and intermittent G-CSF administration had a decline in bone marrow blasts from 9% to 3% after discontinuation of GCSF without cytotoxic treatment; interestingly, this patient continued to have circulating blasts in the peripheral blood and the cytogenetic clones persisted in the bone marrow.

Nine patients (69.2%) had stable disease, and three (23.1%) had progressive disease with an increase in marrow blast percentages (increase by ≥50% in bone marrow blast counts from baseline; n=2) or progression to MDR-AML (blast percentage in the bone marrow of >30%; n=1). One patient developed marked marrow erythroid hyperplasia with >20% myeloblasts in the non-erythroid precursors, and was diagnosed with AML M6a two weeks after presentation. This patient and the patient with progression to MDR-AML were subsequently treated with two cycles of AML-type induction chemotherapy, including intravenous cytarabine, daunorubicin, etoposide, and intrathecal cytarabine. Both patients achieved bone marrow morphological remission prior to HSCT. One control patient was treated with low-dose cytarabine (30mg/m2/d SC) for five out of seven days per week. On repeat bone marrow testing before HSCT, blast counts were unchanged by morphological assessment but increased by flow cytometry.

Changes in bone marrow blast counts among patients treated with AZA were assessed. The comparison of blast percentages before starting AZA treatment and before HSCT showed a statistically significant reduction in blast counts for the AZA treatment group from a median of 15% (range 9-31%) to 5.5% (range 0-12%; Figure 1).

Patients without AZA treatment had similar medians of bone marrow blast percentages at presentation (9%; range 5-22%) and before HSCT or at progression (10%; range 3-93%; Online Supplementary Figure S1).

1512

Peripheral blood counts and bone marrow cytogenetic response One patient was treated with AZA for 13 cycles in total.

The patient became transfusion independent after six treatment cycles, and had morphological and complete cytogenetic remission (no clonal abnormalities detected with metaphase cytogenetics). In this case, the cytogenetic anomalies reappeared after cycle eleven; no increase in bone marrow blasts was seen. Hematological response was observed in two additional patients undergoing AZA treatment after two and three cycles respectively; one had an improvement in hemoglobin and platelet counts, and one showed normalization of absolute neutrophil counts from severe neutropenia at diagnosis. Despite a decrease in bone marrow blast percentages in four of the remaining five patients, no improvement in peripheral blood counts or transfusion requirements was seen after a median of three cycles of AZA (range 1-4). Clonal marrow cytogenetic abnormalities either persisted in these patients or information about repeat testing was not available.

Hematopoietic Stem Cell Transplantation All patients included in this study proceeded to allogeneic HSCT. There was no statistical difference in the degree of HLA-matching and hematopoietic stem cell graft source between the groups (Online Supplementary Table S1). A majority of patients in both groups received busulfan-based myeloablative conditioning regimens with or without anti-thymocyte globulin, depending on the stem cell graft source. One patient in the AZA treatment group had primary engraftment failure after 5/8 mismatched unrelated cord blood transplantation. This patient proceeded to a second HSCT from a haploidentical parent, engrafted, and is alive and disease-free 15.5 months post diagnosis. No primary graft failures occurred in the control group. There were no significant differences in time to engraftment between the AZA treatment and control groups. One patient in the AZA treatment group was undergoing HSCT at the time of data analysis and did not have evaluable engraftment data available.

Survival The estimated 4-year EFS was significantly higher in the AZA treatment group, with 100% of patients surviving in hematological remission at a median follow-up time of 32.5 months from diagnosis (range 7-54.4). EFS was 45.4% haematologica | 2016; 101(12)

Azacitidine in advanced pediatric MDS

A

Log rank: P=0.04

B

Log rank: P=0.1 Figure 2. (A) Event-free survival of patients with azacitidine treatment compared to those without azacitidine treatment. (B) Overall survival of patients with azacitidine treatment compared to those without azacitidine treatment.

in control patients (P=0.04) and median time to event or last follow-up was 31.7 months (range 4.9-98; Figure 2A).

Events in controls included progression to MDR-AML prior to HSCT (n=1), disease relapse, or progression to MDR-AML post HSCT (n=5), and treatment-related mortality (n=1). Two controls, who relapsed post-HSCT, were salvaged with AML-type induction chemotherapy followed by a second HSCT and are disease-free 108 and 111 months after diagnosis, respectively.

The estimated 4-year OS was 100% in the treatment group compared to 69.3% in the control group; however, the difference did not reach statistical significance (P=0.1, Figure 2B). We performed a subgroup analysis excluding all patients with matched related donors, which were only present in controls (n=9), and assessed OS and EFS compared to AZA-treated patients. The results were similar with significantly better EFS in AZA-treated patients (100% with AZA treatment vs. 40% without AZA treatment, P=0.03) with a trend towards better OS in patients with AZA treatment (100% with AZA treatment vs.

64.8% without AZA treatment, P=0.09). haematologica | 2016; 101(12)

Discussion This study is the first to suggest that treatment with AZA prior to HSCT in pediatric patients with advanced MDS could decrease the percentage of bone marrow blasts in an important subset of patients. In our patient group this was associated with significantly improved 4-year EFS. The improved EFS may suggest that pre-HSCT treatment with AZA can reduce the incidence of relapse after HSCT without adding major toxicity.

AZA treatment was administered primarily in an outpatient setting and was well tolerated. Four patients needed treatment modifications and two of them resumed initial treatment dosage in subsequent courses without further delays or modifications. Thus, only two patients (25%) showed adverse events leading to prolonged treatment delay. All other adverse events were managed with standard measures (Table 2). This favorable side effect profile is in contrast to adverse events in AML induction chemotherapy, with typically high toxicity leading to grade 3 and 4 toxicities in about 80% of patients and treatment-related 1513

N. Waespe et al. mortality of around 10-20%.37 Our findings are consistent with a recent report of good tolerability of AZA in 24 children and young adults with different types of MDS and MDR-AML.29 There was a trend towards higher bone marrow blast percentages at presentation in patients treated with AZA. A bias towards treating physicians favoring therapy with AZA in patients with higher blast counts at presentation cannot be completely excluded. The median blast percentage of patients included during the time period where AZA was available was not statistically different between the AZA-treated and untreated patients.

There was a trend towards more patients with chromosome 7 aberrations in the bone marrow in controls and a trend towards more normal cytogenetics in AZA-treated patients. Chromosome 7 or 7q abnormalities were not associated with poorer survival in pediatric MDS compared to other karyotypes in previous studies, which is in contrast to adult MDS.38,39 Indeed, when we analyzed the differences in OS and EFS in controls with or without 7 or 7q abnormalities, we did not find significant differences. Normal cytogenetics were associated with better survival in one report from Japan7 but not in a report from a European group.6 In our series, the outcome of controls without cytogenetic changes (1/2 died of relapse) was not different from controls with cytogenetic abnormalities in the bone marrow.

Complex karyotype is the main cytogenetic aberration associated with poor prognosis in children. Distribution in this regard was not different in the AZA treatment group compared to controls.

A significant reduction in median bone marrow blast percentages from diagnosis to HSCT was seen after AZA treatment. Half of the included patients achieved a morphological bone marrow remission prior to HSCT and another 37.5% of the children showed a partial response to AZA treatment, adding up to 87.5% responders. These are surprising results given the fact that a median of only three cycles were administered, and maximum response is expected in adults after four to six cycles.21 Furthermore, two patients achieved complete marrow remission after only one cycle and another one after two cycles. Unless these differences are due to chance in a small cohort, these observations suggest that pediatric MDS may be more responsive to first-line azacitidine than adult MDS; possibly due to an absence of mutations in epigenetic genes. It is still debated if response to azacitidine treatment stems only from epigenetic changes or an additional cytotoxic effect.

Recent data on chronic myelomonocytic leukemia in adult patients pointed towards epigenetic changes.40 The only patient on long-term treatment with >4 cycles of azacitidine administered showed complete marrow and cytogenetic remission and became transiently transfusion-independent, but reappearance of the clone was detected after cycle eleven. In adults, approximately 60-70% of patients experienced partial response or complete remission after a median of four cycles of AZA.21,26 These results suggest that AZA can induce an excellent bone marrow response in children, comparable to, or possibly better than in adults with advanced MDS. EFS was significantly higher in the AZA treatment group.

This was seen despite a trend towards a higher median percentage of blast cells in the bone marrow at presentation in the AZA treatment group and the longer time interval from diagnosis to HSCT, the latter having been previously associated with poorer survival.13 OS showed similar trends 1514

towards better survival after AZA treatment, although not reaching statistical significance. In adults, survival was improved with AZA treatment compared to conventional care,23 low-dose cytarabine prior to HSCT,23 and was similar or possibly better than induction chemotherapy.25,26 The increased EFS in our series in AZA-treated patients compared to no AZA treatment before HSCT was unexpected, since induction chemotherapy was not previously shown to positively influence the outcome in children unless they had progressed to MDR-AML.8 The survival benefit might stem from the marked reduction in bone marrow blast loads without causing excess toxicity. It is also possible that AZA inhibits subclones that otherwise may contribute to relapse.

This study has several limitations. First, the study is retrospective and non-randomized. Therefore, we cannot exclude the non-random selection of patients. Nevertheless, there seems to be no major selection bias that would favor patients in the AZA treatment group. Particularly, patient characteristics were either similar (age at presentation, sex, and baseline hematological findings) or more favorable in control patients (shorter time to HSCT, trend towards lower blast counts at presentation, and less mismatched stem cell grafts). Second, all patients in the AZA group were treated after 2010, and only four patients that were not treated with AZA were cared for in the same time period. Thus, the possibility of improved outcome due to advances in HSCT donor selection and supportive care cannot be excluded. In this regard, it should be emphasized that HSCT protocols did not differ significantly between the AZA and the control group. Third, the differences in bone marrow cytogenetics might play a role in patient outcome, even though the differences were not statistically different. Subgroup analyses of patients affected with chromosome 7 or 7q abnormalities and normal cytogenetics did not show differences compared to other controls. Lastly, definite conclusions about drug efficacy should be drawn from prospective, randomized, multicenter trials with large numbers of patients; however, due to the rarity of pediatric advanced MDS, it is unlikely that an efficacy study of AZA would be feasible in a large number of children with RCEB. In summary, this study suggests that introducing AZA as a bridging treatment while awaiting HSCT in advanced pediatric MDS is associated with a favorable side effect profile and reduction or stabilization of bone marrow blasts with remission induction in an important subset of patients.

Our report shows for the first time that treatment with AZA could improve EFS in children with advanced MDS, and thus provides a basis to implement the use of AZA in pediatric patients with advanced MDS who are at risk of progression to MDR-AML and do not have a readily available bone marrow stem cell donor.

Acknowledgments The authors would like to thank the participating patients and their families who contributed to this study.

Funding This study was supported by the C17 Research Network Canada with a research grant in partnership with the Aplastic Anemia and Myelodysplasia Association of Canada (AAMAC) and Alexion Pharmaceuticals; the Garron Family Cancer Centre and Sears Canada provided a fellowship and travel grant to NW. Publication funding was provided partly through a special contribution by the Aplastic Anemia and Myelodysplasia Association of Canada (AAMAC). haematologica | 2016; 101(12)

Azacitidine in advanced pediatric MDS

References 1. Locatelli F, Zecca M, Pession A, Maserati E, Stefano PD, Severi F. Myelodysplastic syndromes: the pediatric point of view.

Haematologica. 1995;80(3):268–279.

2. Mandel K, Dror Y, Poon A, Freedman MH. A practical, comprehensive classification for pediatric myelodysplastic syndromes: the CCC system. J Pediatr Hematol Oncol.

2002;24(5):343–352.

3. Cada M, Segbefia CI, Klaassen R, et al. The impact of category, cytopathology and cytogenetics on development and progression of clonal and malignant myeloid transformation in inherited bone marrow failure syndromes. Haematologica. 2015;100(5): 633– 642.

4. 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.

5. Niemeyer CM, Kratz CP. Paediatric myelodysplastic syndromes and juvenile myelomonocytic leukaemia: molecular classification and treatment options. Br J Haematol. 2008;140(6):610–624.

6. Hasle H, Niemeyer CM. Advances in the prognostication and management of advanced MDS in children. Br J Haematol.

2011;154(2):185–195.

7. Moriwaki K, Manabe A, Taketani T, Kikuchi A, Nakahata T, Hayashi Y. Cytogenetics and clinical features of pediatric myelodysplastic syndrome in Japan. Int J Hematol.

2014;100(5):478–484.

8. Strahm B, Nöllke P, Zecca M, et al.

Hematopoietic stem cell transplantation for advanced myelodysplastic syndrome in children: results of the EWOG-MDS 98 study.

Leukemia. 2011;25(3):455–462.

9. Woodard P, Carpenter PA, Davies SM, et al.

Unrelated Donor Bone Marrow Transplantation for Myelodysplastic Syndrome in Children. Biol Blood Marrow Transplant. 2011;17(5):723–728.

10. Strahm B, Locatelli F, Bader P, et al. Reduced intensity conditioning in unrelated donor transplantation for refractory cytopenia in childhood. Bone Marrow Transplant.

2007;40(4):329–333.

11. Hasegawa D, Chen X, Hirabayashi S, et al.

Clinical characteristics and treatment outcome in 65 cases with refractory cytopenia of childhood defined according to the WHO 2008 classification. Br J Haematol. 2014;166 (5):758–766.

12. Creutzig U, Heuvel-Eibrink MM van den, Gibson B, et al. Diagnosis and management of acute myeloid leukemia in children and adolescents: recommendations from an international expert panel. Blood. 2012;120 (16):3187–3205.

13. Smith AR, Christiansen EC, Wagner JE, et al.

Early hematopoietic stem cell transplant is associated with favorable outcomes in children with MDS. Pediatr Blood Cancer.

2013;60(4):705–710.

14. Vidal DO, Paixão VA, Brait M, et al.

Aberrant methylation in pediatric myelodysplastic syndrome. Leuk Res. 2007;31(2):175–181.

15. Hasegawa D, Manabe A, Kubota T, et al.

haematologica | 2016; 101(12)

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Methylation status of the p15 and p16 genes in paediatric myelodysplastic syndrome and juvenile myelomonocytic leukaemia. Br J Haematol. 2005;128(6):805–812.

Aggerholm A, Holm MS, Guldberg P, Olesen LH, Hokland P. Promoter hypermethylation of p15INK4B, HIC1, CDH1, and ER is frequent in myelodysplastic syndrome and predicts poor prognosis in early-stage patients.

Eur J Haematol. 2006;76(1):23–32.

Zhang T, Zhou J, Yang D, et al.

Hypermethylation of DLX4 predicts poor clinical outcome in patients with myelodysplastic syndrome. Clin Chem Lab Med. 2016;54(5):865–871.

Zhao Y, Guo J, Zhang X, et al.

Downregulation of p21 in Myelodysplastic Syndrome Is Associated With p73 Promoter Hypermethylation and Indicates Poor Prognosis. Am J Clin Pathol. 2013;140(6): 819–827.

Helbo AS, Treppendahl M, Aslan D, et al.

Hypermethylation of the VTRNA1-3 Promoter is Associated with Poor Outcome in Lower Risk Myelodysplastic Syndrome Patients. Genes. 2015;6(4):977–990.

Saunthararajah Y. Key clinical observations after 5-azacytidine and decitabine treatment of myelodysplastic syndromes suggest practical solutions for better outcomes.

ASH Educ Program Book. 2013;2013(1): 511–521.

Voso MT, Niscola P, Piciocchi A, et al.

Standard dose and prolonged administration of azacitidine are associated with improved efficacy in a real-world group of patients with myelodysplastic syndrome or low blast count acute myeloid leukemia. Eur J Haematol. 2016;96(4):344–351.

Silverman LR, Demakos EP, Peterson BL, et al. Randomized Controlled Trial of Azacitidine in Patients With the Myelodysplastic Syndrome: A Study of the Cancer and Leukemia Group B. J Clin Oncol. 2002;20(10):2429–2440.

Fenaux P, Mufti GJ, Hellstrom-Lindberg E, et al. Efficacy of azacitidine compared with that of conventional care regimens in the treatment of higher-risk myelodysplastic syndromes: a randomised, open-label, phase III study. Lancet Oncol. 2009;10(3):223–232.

Gurion R, Vidal L, Gafter-Gvili A, et al. 5azacitidine prolongs overall survival in patients with myelodysplastic syndrome - a systematic review and meta-analysis.

Haematologica. 2010;95(2):303–310.

Gerds AT, Gooley TA, Estey EH, Appelbaum FR, Deeg HJ, Scott BL. Pretransplantation Therapy with Azacitidine vs Induction Chemotherapy and Posttransplantation Outcome in Patients with MDS. Biol Blood Marrow Transpl.

2012;18(8):1211–1218.

Damaj G, Duhamel A, Robin M, et al.

Impact of Azacitidine Before Allogeneic Stem-Cell Transplantation for Myelodysplastic Syndromes: A Study by the Société Française de Greffe de Moelle et de Thérapie-Cellulaire and the GroupeFrancophone des Myélodysplasies. J Clin Oncol. 2012;30(36):4533–4540.

Götze K, Platzbecker U, Giagounidis A, et al. Azacitidine for treatment of patients with myelodysplastic syndromes (MDS): practical recommendations of the German MDS

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

Study Group. Ann Hematol. 2010;89(9): 841–850.

Inoue A, Kawakami C, Takitani K, Tamai H. Azacitidine in the Treatment of Pediatric Therapy-related Myelodysplastic Syndrome After Allogeneic Hematopoietic Stem Cell Transplantation. J Pediatr Hematol. 2014;36 (5):e322–324.

Cseh AM, Niemeyer CM, Yoshimi A, et al.

Therapy with low-dose azacitidine for MDS in children and young adults: a retrospective analysis of the EWOG-MDS study group. Br J Haematol. 2016;172(6):930–936.

Lange BJ, Kobrinsky N, Barnard DR, et al.

Distinctive Demography, Biology, and Outcome of Acute Myeloid Leukemia and Myelodysplastic Syndrome in Children With Down Syndrome: Children’s Cancer Group Studies 2861 and 2891. Blood.

1998;91(2):608–615.

Cheson BD, Greenberg PL, Bennett JM, et al.

Clinical application and proposal for modification of the International Working Group (IWG) response criteria in myelodysplasia.

Blood. 2006;108(2):419–425.

Cancer therapy evaluation program.

Common terminology criteria for adverse events v4.0. National Institutes of Health.

2010 Jun. Available from: http://ctep.cancer.gov/protocolDevelopment/electronic_ap plications/ctc.htm.

Fenaux P, Gattermann N, Seymour JF, et al.

Prolonged survival with improved tolerability in higher-risk myelodysplastic syndromes: azacitidine compared with low dose ara-C. Br J Haematol. 2010; 149(2): 244–249.

Swerdlow SH, Campo E, Harris NL, et al., editors. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues.

Lyon, France; International Agency for Research on Cancer; 2008.

Santos FPS, Bueso-Ramos CE, Ravandi F. Acute erythroleukemia: diagnosis and management. Expert Rev Hematol. 2010;3(6): 705–718.

Peng J, Hasserjian RP, Tang G, et al.

Myelodysplastic Syndromes Following Therapy with Hypomethylating Agents (HMAs): Development of Acute Erythroleukemia May Not Influence Assessment of Treatment Response. Leuk Lymphoma. 2016;57(4):812–819.

Lange BJ, Smith FO, Feusner J, et al.

Outcomes in CCG-2961, a Children’s Oncology Group Phase 3 Trial for untreated pediatric acute myeloid leukemia: a report from the Children’s Oncology Group.

Blood. 2008;111(3):1044–1053.

Hasle H, Baumann I, Bergsträsser E, et al.

The International Prognostic Scoring System (IPSS) for childhood myelodysplastic syndrome (MDS) and juvenile myelomonocytic leukemia (JMML). Leukemia. 2004;18(12): 2008–2014.

Göhring G, Michalova K, Beverloo HB, et al.

Complex karyotype newly defined: the strongest prognostic factor in advanced childhood myelodysplastic syndrome.

Blood. 2010;116(19):3766–3769.

Merlevede J, Droin N, Qin T, et al. Mutation allele burden remains unchanged in chronic myelomonocytic leukaemia responding to hypomethylating agents. Nat Commun.

2016;7:10767.

1515

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Myeloid Leukemia

Ferrata Storti Foundation

Haematologica 2016 Volume 101(12):1516-1523

Chromosome abnormalities at onset of complete remission are associated with worse outcome in patients with acute myeloid leukemia and an abnormal karyotype at diagnosis: CALGB 8461 (Alliance) Christian Niederwieser,1 Deedra Nicolet,1,2 Andrew J. Carroll,3 Jonathan E. Kolitz,4 Bayard L. Powell,5 Jessica Kohlschmidt,1,2 Richard M. Stone,6 John C. Byrd,7 Krzysztof MrĂłzek1* and Clara D. Bloomfield1*

The Ohio State University Comprehensive Cancer Center, Columbus, OH; 2Alliance for Clinical Trials in Oncology Statistics and Data Center, Mayo Clinic, Rochester, MN; 3 University of Alabama at Birmingham, Birmingham, AL; 4Monter Cancer Center, Hofstra North Shore-Long Island Jewish School of Medicine, Lake Success, NY; 5 Comprehensive Cancer Center, Wake Forest University, Winston-Salem, NC; 6 Dana-Farber Cancer Institute, Boston, MA and 7Division of Hematology, Department of Internal Medicine, The Ohio State University Comprehensive Cancer Center, Columbus, OH, USA

1

*KM and CDB contributed equally to this work as senior authors.

ABSTRACT

Correspondence: krzysztof.mrozek@osumc.edu or clara.bloomfield@osumc.edu

Received: May 18, 2016.

Accepted: July 26, 2016.

Pre-published: July 28, 2016. doi:10.3324/haematol.2016.149542

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/12/1516

Š2016 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.

1516

A

chievement of complete remission is essential for long-term survival of acute myeloid leukemia patients. We evaluated the prognostic significance of cytogenetics at complete remission in 258 adults with de novo acute myeloid leukemia and abnormal pre-treatment karyotypes, treated on Cancer and Leukemia Group B front-line studies, with cytogenetic data at onset of morphological complete remission.

Thirty-two patients had abnormal karyotypes at time of initial complete remission. Of these, 28 had at least 1 abnormality identified pre-treatment, and 4 acute myeloid leukemia-related abnormalities not detected pre-treatment. Two hundred and twenty-six patients had normal remission karyotypes. Patients with abnormal remission karyotypes were older (P<0.001), had lower pre-treatment white blood counts (P=0.002) and blood blast percentages (P=0.004), were less often classified as Favorable and more often as Adverse among European LeukemiaNet Genetic Groups (P<0.001), and had shorter disease-free survival (median 0.6 vs. 0.9 years; P<0.001) and overall survival (median 1.2 vs. 2.2 years; P<0.001) than patients with normal remission karyotypes.

Sixteen patients with normal remission karyotypes also harbored nonclonal abnormalities unrelated to pre-treatment karyotypes. They had shorter overall survival than 210 patients with only normal metaphases (P=0.04). Forty-eight patients with any clonal or non-clonal chromosome abnormality at complete remission had worse disease-free survival (median 0.6 vs. 1.0 years; P<0.001) and overall survival (median 1.2 vs. 2.5 years; P<0.001) than 210 patients with exclusively normal metaphases. In multivariable analyses, after adjustment for age, the presence of any remission abnormality was associated with shorter disease-free survival (P=0.03) and overall survival (P=0.01). We conclude that detection of any abnormality at complete remission is an adverse prognostic factor. (clinicaltrials.gov identifier: 00048958)

haematologica | 2016; 101(12)

Chromosome abnormalities at CR confer poor outcome

Introduction Acute myeloid leukemia (AML) is a result of acquisition of somatic genetic alterations, both submicroscopic and those detectable microscopically as numerical or structural chromosome abnormalities, in the leukemic blasts.1,2 One or more chromosome abnormalities are detected in 55%60% of adults with AML at diagnosis,3 and pre-treatment cytogenetic findings are among the most important prognostic factors in AML.4-9 Occasionally, chromosome abnormalities are detected in patients who are considered to have achieved a complete remission (CR) based on morphological assessment of their bone marrow. A few studies have assessed the prognostic significance of persistence of an abnormal karyotype at the time of morphological CR,10-14 and AML patients with chromosome abnormalities in CR samples were found to have worse disease-free survival (DFS),11,12 relapse-free survival,14 cumulative incidence of relapse,12 an increased rate of relapse,10 and worse overall survival (OS)12-14 compared with patients who had a normal karyotype at CR. However, some of the previous studies were relatively small,10,13 and included patients with acute promyelocytic leukemia (APL),11-13 high-risk myelodysplastic syndrome (MDS),13 secondary AML evolving from antecedent MDS,13,14 therapy-related AML,12,14 or patients who underwent allogeneic hematopoietic stem-cell transplantation (HSCT) in first CR in addition to those treated with chemotherapy post remisson.11,13,14 Importantly, cytogenetic remission (CRc), defined as “reversion to a normal karyotype at CR”, has been proposed to constitute a separate category of CR.15 However, because of insufficient data from prospective trials, it has been suggested that this should primarily be used in clinical research studies.15 The prognostic significance of chromosome abnormalities at CR that differ from the ones found at diagnosis is not clear.

Therefore, we analyzed the clinical outcomes of a relatively large cohort of AML patients with a long follow up who were enrolled onto Cancer and Leukemia Group B (CALGB)/Alliance for Clinical Trials in Oncology (Alliance) front-line treatment studies. Bone marrow (BM) and/or blood samples of all patients successfully underwent cytogenetic analysis both at diagnosis and at onset of CR. To avoid the confounding effects of AML type (de novo vs. secondary) and the kind of post-remission therapy (chemotherapy vs. allogeneic HSCT), we only included patients with de novo AML who had not received allogeneic HSCT in first CR. This means that, to our knowledge, this study is the first to be performed on a patient population of this kind.

Methods Patients and cytogenetic analysis We reviewed the cytogenetics database containing AML patients enrolled onto the prospective companion study 8461 carried out by CALGB (now part of the Alliance for Clinical Trials in Oncology group, National Clinical Trials Network) since 1984.

Among 2837 newly diagnosed patients with de novo AML (excluding APL) enrolled between 1987 and 2013, 396 patients achieved a morphological CR and had successful cytogenetic analyses performed both pre-treatment and no later than 30 days after achieving morphological CR. Two hundred and fifty-eight patients carhaematologica | 2016; 101(12)

ried at least one clonal chromosome abnormality at diagnosis and thus their pre-treatment karyotype was abnormal, whereas pretreatment karyotypes of 138 patients were normal (i.e. did not contain any clonal chromosome abnormality) (Figure 1). Only patients with de novo AML as defined by World Health Organization (WHO) criteria,16 who did not undergo allogeneic HSCT in first CR, are included in this study. Details of CALGB treatment protocols are provided in the Online Supplementary Appendix. All protocols were approved by the institutional review boards of participating institutions, and all patients provided written informed consent before enrollment in accordance with the Declaration of Helsinki.

Cytogenetic analyses were performed on BM and/or blood samples using unstimulated short-term [24- to 48-hours (h)] cultures in the institutional, CALGB-designated cytogenetics laboratories, and the results were reviewed centrally.17 Determination of a normal karyotype required analysis of at least 20 metaphases from cultured BM specimens. The clonality criteria and the karyotype interpretation procedure followed the recommendations of the International System for Human Cytogenetic Nomenclature.18 Since in some cases abnormalities detected at CR occurred in a single cell, descriptions of the patients’ karyotypes reported in Online Supplementary Tables S1 and S2 contain both the clonal and nonclonal aberrations.

All patients were classified according to their cytogenetic findings at the time of CR into either a cytogenetically abnormal or cytogenetically normal CR group. However, since only 2 of 138 patients with a normal pre-treatment karyotype acquired chromosome abnormalities at CR, all analyses were performed only on 258 patients who harbored a clonal chromosome abnormality or abnormalities at diagnosis.

Patients with a single cell at CR with the same abnormality(ies) as those detected at diagnosis (thereafter referred to as “pre-treatment-related abnormalities”) were considered cytogenetically abnormal at CR, as were patients harboring clonal abnormalities at CR. These abnormalities at CR included both pre-treatmentrelated abnormalities and clonal abnormalities that differed from the pre-treatment abnormalities but are known to be recurrent in AML. In contrast, patients with a non-clonal abnormality(ies) at CR that were not found in the pre-treatment sample (thereafter referred to as “non-clonal pre-treatment-unrelated abnormalities”) were considered to have a normal CR karyotype for the initial analyses. Subsequently, outcomes of these patients with a normal CR karyotype who nevertheless harbored non-clonal pre-treatment-unrelated abnormalities were compared with outcomes of patients with the entirely normal CR karyotype that consisted of 100% of normal metaphase cells.

Statistical analysis Baseline characteristics were compared between CR cytogenetic patient groups using the Wilcoxon rank-sum and Fisher’s exact tests for continuous and categorical variables, respectively.19 For time-to-event analyses, survival estimates were calculated using the Kaplan-Meier method,20 and the CR cytogenetic patient groups were compared using the log-rank test. The Cox proportional hazards model was used to calculate hazard ratios (HR) for DFS and OS.19 Multivariable proportional hazards models were constructed for DFS and OS using a forward selection procedure.19 Variables significant at α=0.20 from the univariable analyses were considered for multivariable analyses. For time-to-event end points, the proportional hazards assumption was checked for each variable individually. All statistical analyses were performed by the Alliance Statistics and Data Center on a database locked 1517

C. Niederwieser et al.

on September 21, 2015, using SAS 9.4 and TIBCO Spotfire S+ 8.2 software.

Results Pre-treatment cytogenetic and clinical characteristics, and clinical outcome of patients based on initial CR karyotypes Among 258 AML patients with an abnormal karyotype at diagnosis, 32 (12%) patients had an abnormal karyotype at CR. They included 28 patients with at least one pre-treatment-related abnormality identical to those observed at diagnosis, 18 of whom had an abnormal clone and 10 a single abnormal cell at CR, and 4 patients with AML-related clonal abnormalities that differed from those present at diagnosis (Online Supplementary Table S1). The CR karyotype was considered to be normal in 226 patients. Two-hundred and ten of these patients had only normal metaphase cells, whereas 16 patients with a normal CR karyotype had 1 (n=15) or 2 (n=1) metaphase cells with non-clonal pre-treatment-unrelated abnormalities in addition to the remaining metaphase cells that were entirely normal (Online Supplementary Table S2).

The distribution of specific chromosome abnormalities found at diagnosis differed between 32 patients who had an abnormal and 226 patients who had a normal karyotype at CR (P<0.001) (Table 1). Both inv(16)(p13.1q22) or t(16;16)(p13.1;q22) and t(8;21)(q22;q22) [abnormalities present in patients with core-binding factor AML who constitute a Favorable Genetic Group in the European LeukemiaNet (ELN) classification9,21] were almost four times less common in patients with an abnormal CR karyotype compared with those with a normal CR karyotype.

Conversely, abnormalities defining the ELN Adverse Genetic Group, predominantly a complex karyotype, were twice as frequent among patients with an abnormal

CR karyotype compared with those with a normal CR karyotype. Abnormalities denoting the ELN IntermediateII Genetic Group were more evenly represented among patients with an abnormal and those with a normal CR karyotype (44% vs. 31%), although no patient with t(9;11)(p22;q23) at diagnosis had chromosome abnormalities at CR (Table 1).

Compared with 226 patients with a normal CR karyotype, 32 patients who had an abnormal CR karyotype were older (median 63 vs. 48 years; P<0.001), had lower white blood cell (WBC) counts (median 4.3 vs. 13.5; P=0.002) and a lower percentage of blood blasts (median 25% vs. 44%; P=0.004) (Table 2). There were no significant differences in the remaining pre-treatment characteristics between the CR cytogenetic groups.

Median follow up for 89 patients alive was 8.3 years (range 3.1-16.0 years). DFS of 32 patients with an abnormal CR karyotype was shorter than that of 226 patients with a normal CR karyotype (median 0.6 vs. 0.9 years; P<0.001), with the 3-year DFS rates of 6% versus 33% (Table 3 and Figure 2A). Similarly, patients with an abnormal CR karyotype had a shorter OS (median 1.2 vs. 2.2 years; P<0.001), with 3-year OS rates of 19% versus 46% (Table 3 and Figure 2B).

The presence of non-clonal pre-treatment-unrelated chromosome abnormalities at CR influences the patientsâ&#x20AC;&#x2122; prognosis Among 226 patients considered to have a normal karyotype at CR, 16 had either a single metaphase cell (n=15) or 2 metaphase cells (n=1) with non-clonal pre-treatmentunrelated abnormality(ies) that were completely different from those detected at diagnosis (Online Supplementary Table S2). These non-clonal pre-treatment-unrelated abnormalities were detected with a similar frequency among patients under 60 years of age (7%, 12 of 179

Figure 1. Overview of the study design. AML: acute myeloid leukemia; CALGB: Cancer and Leukemia Group B; CR: complete remission.

1518

haematologica | 2016; 101(12)

Chromosome abnormalities at CR confer poor outcome

patients) and those aged 60 years or over (9%, 4 of 47 patients; P=0.75), and occurred with a similar frequency among the ELN Genetic Groups (P=0.39) (Online Supplementary Table S3).

Non-clonal pre-treatment-unrelated abnormalities included: reciprocal translocations (n=7), deletions (n=5), exclusively numerical abnormalities (n=3), and a supernumerary marker chromosome (n=1). A search of the Mitelman Database of Chromosome Aberrations and Gene Fusions in Cancer22 revealed that none of the 7 nonclonal pre-treatment-unrelated reciprocal translocations detected at CR in our patients has been reported as a clonal abnormality in AML patients in the literature.

Moreover, the incidental nature of these non-clonal pre-treatment-unrelated abnormalities is emphasized by the fact that none of them was present at the time of relapse in 10 patients for whom relapse cytogenetic data are available (Online Supplementary Table S2). This also includes non-clonal deletions of 6q, 11q and 13q, which are known to be recurrent chromosome abnormalities in AML.3-8,22 We then tested whether the presence of these nonclonal pre-treatment-unrelated abnormalities found in CR samples may have influenced clinical outcome.

Whereas no significant difference in DFS was observed

A

between patients with and without non-clonal pre-treatment-unrelated abnormalities at CR (P=0.18), OS of patients with non-clonal pre-treatment-unrelated abnormalities at CR was shorter than that of patients with an entirely normal CR karyotype (median 1.4 vs. 2.5 years; 3-year rates 25% vs. 48%; P=0.04) (Online Supplementary Table S4 and Figure S1).

Prognostic significance of any chromosome abnormality at CR Since we found that the presence of non-clonal pretreatment-unrelated chromosome abnormalities at CR is associated with adverse outcome, we combined 16 patients with these non-clonal pre-treatment-unrelated abnormalities with 32 patients who had an abnormal CR karyotype, and compared their outcome with that of 210 patients who at CR had only normal metaphase cells.

Disease-free survival of 48 patients with any abnormality at CR, either pre-treatment-related or pre-treatmentunrelated, was shorter than the DFS of 210 patients with an entirely normal CR karyotype (median 0.6 vs. 1.0 years; P<0.001), with 3-year DFS rates of 10% versus 35%.

Similarly, OS of patients harboring any chromosome aberration at the time of morphological CR was shorter than that of patients with only normal metaphase cells at CR

Table 1. Frequencies of specific clonal pre-treatment cytogenetic abnormalities in acute myeloid leukemia patients whose karyotype was abnormal or normal at complete remission.

Cytogenetic abnormalities at diagnosis

B

Figure 2. Disease-free survival (A) and overall survival (B) of patients with de novo acute myeloid leukemia and abnormal pre-treatment karyotypes according to the presence or absence of chromosome abnormalities at the time of complete remission (CR). In these analyses, patients with non-clonal abnormalities unrelated to abnormalities detected at diagnosis are considered to have a normal karyotype at complete remission.

haematologica | 2016; 101(12)

Abnormal Normal CR karyotype CR karyotype (n=32) (n=226) n. (%) n. (%)

ELN Favorable* 4 (12) t(8;21)(q22;q22) 2 (6) inv(16)(p13q22)/t(16;16)(p13;q22) 2 (6) ELN Intermediate-II* 14 (44) t(9;11)(p22;q23) 0 Other abnormalities, including 14 (44) Sole +8 4 (12) Other sole trisomy 0 Sole chromosome loss other than –5 or –7 0 Sole del(7q) or add(7q) 2 (6) Sole del(9q) 0 Other sole unbalanced abnormalities 2 (6) Sole reciprocal translocation or inversion 2 (6) Two abnormalities 4 (12) ELN Adverse* 14 (44) inv(3)(q21q26)/t(3;3)(q21;q26) 0 t(6;9)(p23;q34) 1 (3) t(v;11)(v;q23) 1 (3) –5 or del(5q) 0 –7 2 (6) Complex karyotype 10 (31)

106 (47) 42 (19) 64 (28) 69 (31) 8 (4) 61 (27) 13 (6) 10 (4) 5 (2) 6 (3) 5 (2) 5 (2) 7 (3) 10 (4) 51 (23) 2 (1) 1 (<1) 8 (4) 2 (1) 2 (1) 36 (16)

CR: complete remission; ELN: European LeukemiaNet Genetic Groups. *The ELN Favorable Genetic Group comprises cytogenetically abnormal-AML patients with t(8;21)(q22;q22)/RUNX1-RUNX1T1 or inv(16)(p13.1q22) or t(16;16) (p13.1;q22)/CBFB-MYH11). Intermediate–II and ELN Adverse Genetic Groups contain the remaining cytogenetically abnormal patients. The ELN Intermediate-II Genetic Group consists of patients with t(9;11)(p22;q23)/MLLT3-KMT2A or those with chromosome abnormalities not classified in the Favorable or Adverse Genetic Group. The ELN Adverse Genetic Group is defined by patients with inv(3)(q21q26.2) or t(3;3)(q21;q26.2)/GATA2–MECOM(EVI1); t(6;9)(p23;q34)/DEK-NUP214; t(v;11)(v;q23)/KMT2A rearranged; –5 or del(5q); –7; abnormalities of 17p; and a complex karyotype containing ≥3 cytogenetic abnormalities in the absence of one of the World Health Organization–designated recurring translocations or inversions: t(8;21), inv(16) or t(16;16), t(9;11), t(v;11)(v;q23), t(6;9), and inv(3)/t(3;3).21

1519

C. Niederwieser et al.

(median 1.2 vs. 2.5 years; P<0.001), with 3-year OS rates of 21% versus 48% (Table 4 and Figure 3).

To determine whether the presence of any chromosome abnormality at CR remained associated with outcome when controlling for other clinical prognostic factors, we constructed multivariable models for DFS and OS. For DFS, patients with any cytogenetic abnormality at CR, pre-treatment-related or pre-treatment-unrelated, had an approximately 50% higher risk of relapse or death (P=0.03), after adjustment for age (P<0.001). Similarly, the risk of death was 59% higher for patients who harbored any chromosome abnormality at CR than for those with

Table 2. Pre-treatment clinical characteristics of cytogenetically abnormal acute myeloid leukemia patients whose karyotype was abnormal or normal at complete remission.

Characteristic

Abnormal CR Normal CR karyotype karyotype (n=32) (n=226)

Age, years Median Range Age group, n. (%) <60 years ≥60 years Sex, n. (%) Male Female Race, n. (%) White Non-white Hemoglobin (g/dL) Median Range Platelet count (x109/L) Median Range WBC count (x109/L) Median Range Blood blasts, % Median Range Bone marrow blasts, % Median Range Extramedullary involvement, n. (%) ELN Genetic Group,* n. (%) Favorable Intermediate-II Adverse

P

<0.001 63 25-84

48 17-84

14 (44) 18 (56)

179 (79) 47 (21)

24 (75) 8 (25)

130 (58) 96 (42)

<0.001

Discussion Our cytogenetic analyses performed at diagnosis and at the time of first morphological CR on a relatively large series of de novo AML patients receiving induction treatment with cytarabine and an anthracycline (7+3) or similar regimens with long follow up demonstrated that the presence of an abnormal karyotype at CR is associated with adverse prognosis. Both DFS and OS of patients who had at least one cytogenetic abnormality at CR that was identical to abnormalities found in pre-treatment samples or had clonal abnormalities different from those found at diagnosis but known to be recurrent in AML were significantly shorter than DFS and OS of patients with a normal CR karyotype. These results are in line with those of earlier studies, both smaller10-13 and similar in size14 to our current series, demonstrating the adverse prognostic significance of persistence of an abnormal karyotype following achievement of morphological CR after completion of induction chemotherapy. However, to our knowledge, only our study included exclusively de novo AML patients who did not undergo allogeneic HSCT in first CR.

0.08

A 0.28

30 (94) 2 (6)

185 (84) 34 (16)

9.3 6.6-13.9

9.1 2.3-14.7

62 9-177

49 5-387

4.3 0.7-68.9

13.5 0.6-276.8

25 0-83

44 0-98

50 5-90 5 (18)

60 1-96 44 (22)

4 (13) 14 (44) 14 (44)

106 (47) 69 (31) 51 (23)

0.50 0.31

0.002

0.004

B 0.28

0.81 <0.001

AML: acute myeloid leukemia; ELN: European LeukemiaNet; WBC: white blood cell. *The ELN Favorable Genetic Group comprises cytogenetically abnormal-AML patients with t(8;21)(q22;q22)/RUNX1-RUNX1T1 or inv(16)(p13.1q22) or t(16;16)(p13.1;q22)/CBFB-MYH11). Intermediate–II and ELN Adverse Genetic Groups contain the remaining cytogenetically abnormal patients. The ELN Intermediate-II Genetic Group consists of patients with t(9;11)(p22;q23)/MLLT3-KMT2A or those with chromosome abnormalities not classified in the Favorable or Adverse Genetic Group.

The ELN Adverse Genetic Group is defined by patients with inv(3)(q21q26.2) or t(3;3)(q21;q26.2)/GATA2–MECOM(EVI1); t(6;9)(p23;q34)/DEK-NUP214; t(v;11)(v;q23)/KMT2A rearranged; –5 or del(5q); –7; abnormalities of 17p; and a complex karyotype containing ≥3 cytogenetic abnormalities in the absence of one of the World Health Organization–designated recurring translocations or inversions: t(8;21), inv(16) or t(16;16), t(9;11), t(v;11)(v;q23), t(6;9), and inv(3)/t(3;3).21

1520

an entirely normal CR karyotype (P=0.01), (after adjustment for age P<0.001) (Table 5).

Figure 3. Disease-free survival (A) and overall survival (B) of patients with de novo acute myeloid leukemia and abnormal pre-treatment karyotypes according to the presence or absence of any chromosome abnormality or abnormalities, both clonal and non-clonal, pre-treatment-related and pre-treatment-unrelated, at the time of complete remission (CR).

haematologica | 2016; 101(12)

Chromosome abnormalities at CR confer poor outcome

In contrast to all10,12-14 but one11 of the previous studies, we have also analyzed outcome of patients who at the time of CR acquired non-clonal chromosome abnormalities that were unrelated to the clonal chromosome abnormalities detected at diagnosis. Such non-clonal pre-treatment-unrelated abnormalities are usually not considered to be important and the karyotype in such cases is determined to be normal. However, in our study, the presence of non-clonal pre-treatment-unrelated abnormalities was found to portend a significantly shorter OS of patients who harbored them compared with OS of patients whose CR specimens contained 100% of normal metaphase cells.

This finding differs from the results of Grimwade et al.11 who identified 11 patients with cytogenetic abnormalities in their CR specimens that were not detected prior to induction treatment. DFS of these patients was virtually

identical to that of patients with an abnormal karyotype at diagnosis who had an entirely normal karyotype at CR, although OS of these patients was not assessed. However, our study is not directly comparable to that of Grimwade et al.11 because their cytogenetically analyzed CR samples were obtained not at the onset of CR but in CR at the time of bone marrow harvest for autologous HSCT, which took place between the third and fourth cycle of consolidation therapy. Furthermore, 7 of 11 patients analyzed in their study had a normal karyotype at diagnosis and 3 of the 11 patients had clonal, not non-clonal, aberrations at CR.11 The biological significance of cells with non-clonal pre-treatment-unrelated abnormalities at CR is unclear.

One possibility is that they might represent small cell populations undetected at diagnosis that were more resistant to chemotherapy, survived induction treatment and subse-

Table 3. Treatment outcomes of cytogenetically abnormal acute myeloid leukemia patients whose karyotype was abnormal or normal at complete remission.

Outcome end point

Disease-free survival Median, years Disease-free at 3 years, % (95% CI) Disease-free at 5 years, % (95% CI) Overall survival Median, years Alive at 3 years, % (95% CI) Alive at 5 years, % (95% CI)

Abnormal CR karyotype (n=32)

Normal CR karyotype (n=226)

0.6 6 (1-18) 6 (1-18)

0.9 33 (27-40) 30 (24-36)

1.2 19 (8-34) 12 (3-26)

2.2 46 (39-52)

P

HR (95% CI)

<0.001

2.17 (1.46-3.21)

<0.001

2.18 (1.46-3.26)

41 (35-48)

CI: confidence interval; CR: complete remission; HR: hazard ratio.

Table 4. Treatment outcomes of cytogenetically abnormal acute myeloid leukemia patients according to the presence or absence of any clonal or non-clonal, pre-treatment-related or pre-treatment-unrelated chromosome abnormality or abnormalities at complete remission.

Outcome end point Disease-free survival Median, years Disease-free at 3 years, % (95% CI) Disease-free at 5 years, % (95% CI) Overall survival Median, years Alive at 3 years, % (95% CI) Alive at 5 years, % (95% CI)

Any abnormality present at CR (n=48)

Entirely normal CR karyotype (n=210)

0.6 10 (4-21) 10 (4-21)

1.0 35 (28-41) 31 (25-37)

1.2 21 (11-33) 16 (7-28)

2.5 48 (41-54) 42 (36-49)

P

HR (95% CI)

<0.001

1.92 (1.36-2.71)

<0.001

2.12 (1.49-3.01)

CI: confidence interval; CR: complete remission; HR: hazard ratio.

Table 5. Multivariable analyses of outcome in cytogenetically abnormal acute myeloid leukemia patients according to the presence or absence of any clonal or non-clonal, pre-treatment-related or pre-treatment-unrelated chromosome abnormality or abnormalities at complete remission.

End point/variables in final models Disease-free survivalâ&#x20AC; Any abnormality at CR versus an entirely normal karyotype at CR Age, continuous, 10-year increase Overall survivalâ&#x20AC; Any abnormality at CR versus an entirely normal karyotype at CR Age, continuous, 10-year increase

Hazard ratio*

95% CI

P

1.49 2.19

1.04-2.15 1.58-3.04

0.03 <0.001

1.59 2.62

1.10-2.30 1.86-3.68

0.01 <0.001

CI: confidence interval; CR: complete remission. *A hazard ratio greater than 1 corresponds to a higher risk for the first category listed of dichotomous variables and higher values of continuous variables. â&#x20AC; Variables considered for model inclusion and evaluated in univariable models were: CR cytogenetic group (any abnormality at CR versus an entirely normal karyotype at CR), age (as a continuous variable, in 10-year increments), sex (male vs. female), ethnic group (white vs. non-white), white blood cell (WBC) count (as a continuous variable, in 50-unit increments), hemoglobin (as a continuous variable, in 1-unit increments), platelet count (as a continuous variable, in 50-unit increments) and extramedullary involvement (present vs. absent).

haematologica | 2016; 101(12)

1521

C. Niederwieser et al.

quently gave rise to an abnormal clone or clones responsible for the disease relapse. However, this scenario is highly unlikely since we did not detect any of these non-clonal pre-treatment-unrelated abnormalities again in samples of 10 patients who relapsed and had a successful cytogenetic analysis at that time. Another potential mechanism is that the occurrence of non-clonal pre-treatment-unrelated abnormalities could indicate increased predisposition to genomic instability in leukemic blasts,23 which might increase the likelihood of generation of therapy-resistant clones that lead to relapse and poor outcome. Regardless of the exact mechanism, our findings, if corroborated, suggest that detection of any chromosome abnormality at CR, whether related to the pre-treatment karyotype or not, should be regarded as a sign of increased risk of relapse or death, and that the patients with such cytogenetically abnormal CR samples should be considered as candidates for more intensive therapy, including allogeneic HSCT14,24 and/or alternative treatment regimens.

Whereas both our study and previous ones10-14 show that cytogenetic analysis of CR samples can identify patients who have an increased risk of relapse or death, the resolution of cytogenetic analysis is relatively low.

Other, more sensitive methodologies, such as mutation detection using DNA- or RNA-based real-time quantitative polymerase chain reaction (RQ-PCR) and/or next-generation sequencing,25,26 or multiparameter flow cytometry become available. immunophenotyping,27-29 have However, these techniques also have limitations. In the case of molecular analyses by RQ-PCR and/or next-generation sequencing, optimal sensitivity thresholds for mutation clearance at CR still need to be established,26 especially because these sensitive techniques are capable of detecting low levels of fusion transcripts, such as RUNX1-

References 1. Döhner H, Weisdorf DJ, Bloomfield CD. Acute myeloid leukemia. N Engl J Med. 2015;373(12):1136-1152.

2. Mrózek K, Bloomfield CD. Chromosome abnormalities in acute myeloid leukaemia and their clinical importance. In: Rowley JD, Le Beau MM, Rabbitts TH, eds.

Chromosomal Translocations and Genome Rearrangements in Cancer. Cham, Heidelberg, New York, NY, Dordrecht, London: Springer International Publishing; 2015:275-317.

3. Mrózek K, Heerema NA, Bloomfield CD. Cytogenetics in acute leukemia. Blood Rev. 2004;18(2):115-136.

4. Grimwade D, Walker H, Oliver F, et al. The importance of diagnostic cytogenetics on outcome in AML: analysis of 1,612 patients entered into the MRC AML 10 trial. Blood.

1998;92(7):2322-2333.

5. Slovak ML, Kopecky KJ, Cassileth PA, et al.

Karyotypic analysis predicts outcome of preremission and postremission therapy in adult acute myeloid leukemia: a Southwest Oncology Group/Eastern Cooperative Oncology Group study. Blood. 2000;96(13): 4075-4083.

6. Byrd JC, Mrózek K, Dodge RK, et al.

Pretreatment cytogenetic abnormalities are predictive of induction success, cumulative incidence of relapse, and overall survival in adult patients with de novo acute myeloid

1522

7.

8.

9.

10.

11.

12.

RUNX1T1 or CBFB-MYH11, or mutations in the DNMT3A gene, which are known to persist in patients remaining in durable CR.26,30 Immunophenotyping using multiparameter flow cytometry can be technically challenging and is relatively expensive, with varying threshold levels proposed for risk stratification.28 One of the advantages of cytogenetics, despite its limited sensitivity, is that this assay is widely available, also in less developed countries.

In summary, our cytogenetic study performed on a relatively large cohort of patients with de novo AML, none of whom had received allogeneic HSCT in first CR, with a long follow up has shown that persistence of at least one cell with an abnormality or abnormalities seen in the pretreatment sample at the time of morphological CR is associated with poor outcome independently of other clinical prognosticators. Moreover, we found that acquisition of non-clonal abnormalities, seemingly unrelated to the diagnostic karyotype, may also have adverse prognostic consequences. Obviously, this new finding should be prospectively validated in future, large studies. If confirmed, detection of any chromosome abnormality at CR should be considered as a factor in clinical decision making.

Acknowledgments The authors would like to thank Lisa J. Sterling and Christine Finks for data management. This work was supported in part by the National Cancer Institute of the National Institutes of Health [grants: U10CA180821 and U10CA180882 (to the Alliance for Clinical Trials in Oncology), U10CA101140, U10CA077658, U10CA180861, P50CA140158, P30CA016058], and the Leukemia Clinical Research Foundation. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

leukemia: results from Cancer and Leukemia Group B (CALGB 8461). Blood.

2002;100(13):4325-4336.

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 younger adult patients treated in the United Kingdom Medical Research Council trials. Blood. 2010;116 (3):354-365.

Grimwade D, Mrózek K. Diagnostic and prognostic value of cytogenetics in acute myeloid leukemia. Hematol Oncol Clin North Am. 2011;25(6):1135-1161.

Mrózek K, Marcucci G, Nicolet D, et al.

Prognostic significance of the European LeukemiaNet standardized system for reporting cytogenetic and molecular alterations in adults with acute myeloid leukemia. J Clin Oncol. 2012;30(36):45154523.

Freireich EJ, Cork A, Stass SA, et al.

Cytogenetics for detection of minimal residual disease in acute myeloblastic leukemia. Leukemia. 1992;6(6):500-506.

Grimwade D, Walker H, Oliver F, et al.

What happens subsequently in AML when cytogenetic abnormalities persist at bone marrow harvest? Results of the 10th UK MRC AML trial. Bone Marrow Transplant.

1997;19(11):1117-1123.

Marcucci G, Mrózek K, Ruppert AS, et al.

Abnormal cytogenetics at date of morpho-

13.

14.

15.

16.

17.

logic complete remission predicts short overall and disease-free survival, and higher relapse rate in adult acute myeloid leukemia: results from Cancer and Leukemia Group B study 8461. J Clin Oncol. 2004;22(12):2410-2418.

Balleisen S, Kuendgen A, Hildebrandt B, Haas R, Germing U. Prognostic relevance of achieving cytogenetic remission in patients with acute myelogenous leukemia or high-risk myelodysplastic syndrome following induction chemotherapy. Leuk Res. 2009;33(9):1189-1193.

Chen Y, Cortes J, Estrov Z, et al. Persistence of cytogenetic abnormalities at complete remission after induction in patients with acute myeloid leukemia: prognostic significance and the potential role of allogeneic stem-cell transplantation. J Clin Oncol.

2011;29(18):2507-2513.

Cheson BD, Bennett JM, Kopecky KJ, et al.

Revised recommendations of the International Working Group for Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia. J Clin Oncol.

2003;21(24):4642-4649.

Vardiman JW, Thiele J, Arber DA, et al. The 2008 revision of the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia: rationale and important changes. Blood.

2009;114(5):937-951.

Mrózek K, Carroll AJ, Maharry K, et al.

haematologica | 2016; 101(12)

Chromosome abnormalities at CR confer poor outcome

18.

19.

20.

21.

22.

Central review of cytogenetics is necessary for cooperative group correlative and clinical studies of adult acute leukemia: the Cancer and Leukemia Group B experience.

Int J Oncol. 2008;33(2):239-244.

Mitelman F, ed. ISCN (1995): An International System for Human Cytogenetic Nomenclature. Basel, Karger, 1995.

Vittinghoff E, Glidden DV, Shiboski SC, McCulloch CE. Regression Methods in Biostatistics: Linear, Logistic, Survival and Repeated Measures Models. New York, NY, Springer, 2005.

Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53(282):457-481.

Döhner 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.

Mitelman F, Johansson B, Mertens F, eds.

haematologica | 2016; 101(12)

23.

24.

25.

26.

Mitelman database of chromosome aberrations and gene fusions in cancer, 2016.

Available from: http://cgap.nci.nih.gov/ Chromosomes/Mitelman. Last accessed February 22, 2016.

Heng HHQ, Regan SM, Liu G, Ye CJ. Why it is crucial to analyze non clonal chromosome aberrations or NCCAs? Mol Cytogenet. 2016;9:15.

Dvorak P, Lysak D, Vokurka S, Karas M, Subrt I. Allogeneic stem cell transplantation can improve outcome of AML patients without complete cytogenetic response after induction and consolidation treatment. Neoplasma. 2015;62(1):140145.

Krönke J, Schlenk RF, Jensen KO, et al.

Monitoring of minimal residual disease in NPM1-mutated acute myeloid leukemia: a study from the German-Austrian Acute Myeloid Leukemia Study Group. J Clin Oncol. 2011;29(19):2709-2716.

Klco JM, Miller CA, Griffith M, et al.

Association between mutation clearance

27.

28.

29.

30.

after induction therapy and outcomes in acute myeloid leukemia. JAMA. 2015; 314(8):811-822.

Inaba H, Coustan-Smith E, Cao X, et al.

Comparative analysis of different approaches to measure treatment response in acute myeloid leukemia. J Clin Oncol.

2012;30(29):3625-3632.

Terwijn M, van Putten WLJ, Kelder A, et al.

High prognostic impact of flow cytometric minimal residual disease detection in acute myeloid leukemia: data from the HOVON/SAKK AML 42A study. J Clin Oncol. 2013;31(31):3889-3897.

Vidriales MB, Pérez-López E, Pegenaute C, et al. Minimal residual disease evaluation by flow cytometry is a complementary tool to cytogenetics for treatment decisions in acute myeloid leukaemia. Leuk Res. 2016;40:1-9.

Pløen GG, Nederby L, Guldberg P, et al.

Persistence of DNMT3A mutations at longterm remission in adult patients with AML. Br J Haematol. 2014;167(4):478-486.

1523

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Lymphoblastic Leukemia

Ferrata Storti Foundation

Haematologica 2016 Volume 101(12):1524-1533

International reference analysis of outcomes in adults with B-precursor Ph-negative relapsed/refractory acute lymphoblastic leukemia Nicola Gökbuget,1 Hervè Dombret,2 Jose-Maria Ribera,3 Adele K. Fielding,4 Anjali Advani,5 Renato Bassan,6 Victoria Chia,7 Michael Doubek,8 Sebastian Giebel,9 Dieter Hoelzer,1 Norbert Ifrah,10 Aaron Katz,7 Michael Kelsh,7 Giovanni Martinelli,11 Mireia Morgades,3 Susan O’Brien,12 Jacob M. Rowe,13 Julia Stieglmaier,14 Martha Wadleigh15 and Hagop Kantarjian12

University Hospital, Goethe University, Frankfurt, Germany; 2Hôpital Saint-Louis, Paris, France; 3ICO-Hospital Germans Trias I Pujol, Jose Carreras Research Institute, Barcelona, Spain; 4UCL Cancer Institute, London, UK; 5Cleveland Clinic, Ohio, USA; 6 UOC Ematologia, Ospedale dell'Angelo, Mestre-Venezia, Italy; 7Center for Observational Research, Amgen, USA; 8University Hospital, Brno, Czech Republic; 9Maria Sklodowska Curie Memorial Cancer Center and Institute of Oncology, Gliwice, Poland; 10Center Hospitalier Universitaire, Angers, France; 11Policlinico Sant’Orsola, Istituto Seragnoli, Bologna, Italy; 12University of Texas, MD Anderson Cancer Center, Houston, USA; 13 Rambam Medical Center, Haifa, Israel; 14Clinical Development, Amgen, Germany and 15Dana Farber Cancer Institute, Boston, Massachusetts, USA

1

ABSTRACT

A

Correspondence: goekbuget@em.uni-frankfurt.de

Received: February 16, 2016.

Accepted: August 23, 2016.

Pre-published: September 23, 2016. doi:10.3324/haematol.2016.144311

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/100/12/1524

©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.

1524

dults with relapsed/refractory acute lymphoblastic leukemia have an unfavourable prognosis, which is influenced by disease and patient characteristics. To further evaluate these characteristics, a retrospective analysis of 1,706 adult patients with Ph-negative relapsed/refractory B-precursor acute lymphoblastic leukemia diagnosed between 1990-2013 was conducted using data reflecting the standard of care from 11 study groups and large centers in Europe and the United States. Outcomes included complete remission, overall survival, and realization of stem cell transplantation after salvage treatment. The overall complete remission rate after first salvage was 40%, ranging from 35%-41% across disease status categories (primary refractory, relapsed with or without prior transplant), and was lower after second (21%) and third or greater (11%) salvage. The overall complete remission rate was higher for patients diagnosed from 2005 onward (45%, 95% CI: 39%-50%). One- and three-year survival rates after first, second, and third or greater salvage were 26% and 11%, 18% and 6%, and 15% and 4%, respectively, and rates were 2%-5% higher among patients diagnosed from 2005. Prognostic factors included younger age, longer duration of first remission, and lower white blood cell counts at primary diagnosis. This large dataset can provide detailed reference outcomes for patients with relapsed/refractory Ph-negative B-precursor acute lymphoblastic leukemia. clinicaltrials.gov identifier: 02003612

Introduction Overall prognosis among adult acute lymphoblastic leukemia (ALL) patients has improved by optimisation of front-line therapy,1 but outcomes remain poor for patients who relapse or are refractory to initial treatment. Reported rates of complete remission (CR) after salvage treatment range from 18%-45%, and median survival times range from 2-8 months, with less than 10% survival after 5 years in most studies.2-7 Achievement of CR and subsequent HSCT is the only curative option in relapsed adult ALL;2,3,6 however, this can only be achieved in a subgroup haematologica | 2016; 101(12)

Outcome of relapsed/refractory acute lymphoblastic leukemia

of patients. Longer duration of first CR is associated with improved survival; however, there is no established cutoff point for CR duration that predicts long-term survival.2,4-6 Options are even more limited for patients who do not respond to first-line of salvage, and probabilities of survival decline substantially with successive lines of treatment.3,4,7 For these reasons, new treatment options are needed for adult patients with relapsed/refractory (r/r) ALL, and a number of promising compounds are currently under clinical evaluation.8,9 The rarity of adult r/r ALL,10 combined with the very poor outcomes and non-standardised approaches of salvage therapies, make it difficult to conduct randomised trials of new compounds. Approvals of new treatments are therefore often based on evidence from phase 2 single-arm trials.11-14 To attain a more precise estimate of clinical practice outcomes in adult r/r ALL, and to evaluate important patient subgroups, we pooled data from major national study groups and large individual sites treating adult ALL from Europe and the United States to create the most extensive clinical dataset available in this population.

The analysis aimed to describe patient characteristics and outcome parameters (achievement of CR, overall survival [OS] and realization of allogeneic HSCT) among adult patients with Ph-negative B-precursor r/r ALL treated according to standard of care in Europe and the United States before the introduction of new targeted therapies, such as blinatumomab. Additional objectives were to evaluate prognostic factors, define subgroups, and to provide a reference population for the assessment of new compounds in this setting.

Methods Study conduct This observational study was designed jointly by the sponsor, Amgen and the authors. All authors were involved in the preparation, revision, and formal approval of the manuscript. All authors met the criteria of the International Committee of Medical Journal Editors (http://www.icmje.org) for authorship, and take responsibility for the content and accuracy of data presented. Co-author roles are described in the Appendix.

Patients in relapse or in the refractory setting were included, independent of treatment strategy. Patients treated with blinatumomab were excluded, but all other regimens were permitted, including palliative care; type of salvage therapy was not recorded in all databases. Primary refractory disease was defined by study group criteria as persistent disease after standard induction and consolidation therapy. Relapse was defined by standard criteria as reappearance of disease after previous achievement of CR.

Outcome measures The primary outcome measure was the rate of CR after salvage therapy. CR was defined by the study groups/sites based on general clinical practice guidelines, with some variation across the individual study groups.

Secondary outcomes included OS and realization of HSCT after salvage treatment (allogeneic HSCT only). OS was defined as the time from the start of last salvage therapy to death from any cause.

Subgroups and covariates For analysis of outcomes, three clinically-relevant patient “disease status” subgroups were defined: “Primary Refractory”, “Relapsed, with prior HSCT”, and “Relapsed, without prior HSCT”. Four study groups did not have eligible patients with primary refractory disease in their database. Patients were also grouped according to their line of salvage treatment, classified as first, second, or third or greater salvage.

Statistical analysis Generalized estimating equations were used to compare CR proportions across lines of salvage. For OS, the Kaplan Meier (KM) median and KM proportions at 6, 12, and 36 months were estimated with corresponding 95% CI.15,16 Univariate and multivariate logistic regression models were used to evaluate the potential relationships between patient and disease characteristics. CR and Cox models were used to evaluate relationships for OS. Mantel-Byar17 and Simon-Makuch methods were used for statistical evaluation of time-dependent variables (see Appendix). Sensitivity analyses were conducted to account for potential changes in outcomes across the study time period; these analyses were limited to patients who were diagnosed from 2005 onward to provide a cohort of patients receiving current standard of care therapy.

Results Study design Details of the study design, outcomes, and analyses procedures are described in the Appendix. Briefly, individual databases were prepared by the participating study groups or centers to include all consecutive adult patients with r/r ALL. These databases were transferred and evaluated centrally for inclusion in the pooled analysis. After individual database review for data coding, consistency, and characterization of missing data, the clinical data were then harmonized.

All patients provided informed consent to each study site investigator to use their data for clinical research purposes.

Eligibility criteria Eligibility criteria were defined across the databases as follows: adult patients with r/r B-precursor ALL; age ≥15 years at time of first diagnosis; Ph/BCR-ABL negative; initial diagnosis of ALL in the year 1990-2013; no isolated extramedullary relapse or central nervous system (CNS) involvement at relapse; and available endpoint data (CR, HSCT, or OS). Other than Ph/BCR-ABL chromosome status, no other cytogenetics data were collected. haematologica | 2016; 101(12)

Patient population: demographic and clinical characteristics Adult Ph-negative, B-precursor r/r ALL patient data were pooled across 11 groups or sites (6 European study groups and 2 centers, and 3 US centers). The number of patients provided by groups or sites that met the inclusion criteria ranged from 33 to 427 (Online Supplementary Table S1), with 1706 patients providing data for at least one outcome (CR, OS, or HSCT). Survival data were available for 1695 patients. Information was available on results from first salvage in 1628 patients, second salvage in 374 patients, and from third or greater salvage in 160 patients.

A single salvage record was available from 1336 patients (which could be from first or later salvage) and multiple records were available from 370 patients. Overall, 2210 salvage records were included (Figure 1, Online Supplementary Table S1). Information on treatment intensity was not available for most patients (68% with OS data and 47% with CR data in first salvage), and only 8 patients 1525

N. Gökbuget et al.

Table 1. Adult Ph-negative relapsed/refractory B-precursor ALL patient data by selected demographic and clinical factors, summarized by outcomes and HSCT status.

All patients In 1st salvage In 2ndsalvage (N=1 618) (N=372) n (%) Age at salvage treatment, years 15-17 65 (4.0) 18-34 758 (46.8) 35-54 578 (35.7) 55-64 170 (10.5) ≥65 47 (2.9) Sex Male 950 (58.7) Female 668 (41.3) WBC at diagnosis <30,000/ mL 918 (69.3) ≥30,000/ mL 406 (30.7) Period of salvage treatment 1990-1994 107 (6.6) 1995-1999 342 (21.1) 2000-2004 538 (33.3) 2005 or later 631 (39.0) Disease status Primary refractory 115 (7.1) Relapsed, without prior HSCT 1291 (79.9) Relapsed, with prior HSCT 210 (13.0) Time from CR1 to first relapse (in months)* <12 months 725 (56.9) ≥12 months 550 (43.1) <6 months 421 (33.0) 6 to <12 months 304 (23.8) 12 to <18 months 162 (12.7) 18 to <24 months 125 (9.8) 24 or more months 263 (20.6) Time from prior alloHSCT to first salvage*** <12 months 151 (71.9) ≥12 months 59 (28.1)

Patients with CR data available

Patients with information on HSCT status after salvage available In 1st salvage (N=1 337)

In 1st salvage (N=901)

In 2nd salvage (N=332)

n (%)

In 3rd or greater salvage (N=160)** n (%)

n (%)

n (%)

In 3rd or greater salvage (N=159) n (%)

12 (3.2) 182 (48.9) 125 (33.6) 33 (8.9) 20 (5.4)

4 (2.5) 88 (55.0) 43 (26.9) 15 (9.4) 10 (6.3)

39 (4.3) 399 (44.3) 320 (35.5) 97 (10.8) 46 (5.1)

11 (3.3) 163 (49.1) 107 (32.2) 31 (9.3) 20 (6.0)

4 (2.5) 87 (54.7) 43 (27.0) 15 (9.4) 10 (6.30)

59 (4.4) 618 (46.2) 490 (36.6) 126 (9.4) 44 (3.3)

225 (60.5) 147 (39.5)

99 (61.9) 61 (38.1)

536 (59.5) 365 (40.5)

201 (60.5) 131 (39.5)

98 (61.6) 61 (38.4)

790 (59.1) 547 (40.9)

125 (69.1) 56 (30.9)

19 (65.5) 10 (34.5)

428 (69.5) 188 (30.5)

98 (68.5) 45 (31.5)

18 (64.3) 10 (35.7)

744 (70.4) 313 (29.6)

44 (11.8) 71 (19.1) 102 (27.4) 155 (41.7)

26 (16.3) 48 (30.0) 32 (20.0) 34 (33.8)

85 (9.4) 178 (19.8) 314 (34.9) 324 (36.0)

44 (13.3) 71 (21.5) 80 (24.2) 137 (41.1)

26 (16.4) 48 (30.2) 32 (20.1) 53 (33.3)

100 (7.5) 323 (24.2) 491 (36.7) 423 (31.6)

69 (18.6) 240 (64.9) 61 (16.5)

35 (21.9) 108 (67.5) 17 (10.6)

111 (12.4) 707 (78.6) 81 (9.0)

57 (17.3) 223 (67.6) 50 (15.1)

34 (21.4) 108 (67.9) 17 (10.7)

105 (7.9) 1074 (80.4) 157 (11.7)

118 (66.7) 59 (33.3) 60 (33.9) 58 (32.8) 23 (13.0) 11 (6.2) 25 (14.1)

58 (55.2) 47 (44.8) 32 (30.5) 26 (24.8) 17 (16.2) 8 (7.6) 22 (20.9)

415 (59.6) 281 (40.4) 232 (33.3) 183 (26.3) 87 (12.5) 61 (8.8) 133 (19.1)

115 (65.7) 60 (34.3) 57 (32.6) 58 (33.1) 23 (13.1) 11 (6.3) 26 (14.9)

58 (55.2) 47 (44.8) 32 (30.5) 26 (24.8) 17 (16.2) 8 (7.6) 22 (20.9)

588 (55.4) 473 (44.6) 334 (31.5) 254 (23.9) 137 (12.9) 100 (9.4) 236 (22.2)

47 (77.1) 14 (22.9)

13 (65.0) 7 (35.0)

57 (70.4) 24 (29.6)

38 (76.0) 12 (24.0)

13 (65.0) 7(35.0)

108 (68.8) 49 (31.2)

n (%)

No significant differences in patient characteristics were observed between the different analysis groups defined by availability of data on response to salvage or subsequent HSCT, or for the analysis groups defined by different lines of salvage. *excludes refractory patients and those with prior HSCT. ** Earliest of 3rd or greater salvage treatment.*** Among the 217 patients who had a prior HSCT.

were documented as receiving palliative care.

Patient demographic data are presented in Table 1. The median age of the overall population was 34 years (range, 15-83), and approximately one-third were aged 35-54 years. Most patients (59%) were male, and over 70% received salvage therapy after the year 2000. A large majority of patients (80%) had relapsed ALL without prior HSCT; 13% had relapse after HSCT, and 7% had primary refractory disease. In patients who had relapsed without prior HSCT, 30% relapsed within six months, 24% within 6-12 months and 21% after more than two years from diagnosis. In patients who had relapsed after HSCT, 69% relapsed within one year.

No differences in patient characteristics were observed between the analysis groups defined by availability of data on response to salvage or subsequent HSCT, or for the analysis groups defined by different lines of salvage (Table 1).

1526

Response to salvage therapy Response to first salvage. Data on CR following first salvage therapy were available for 901 patients. In total, 361 patients achieved a CR after first salvage (40%; 95% CI: 37%-43%). CR rates in first salvage ranged from 35%41% depending on disease status (primary refractory, relapsed with or without prior HSCT; Table 2), but these differences were not statistically significant.

Prognostic factors for response to first salvage. Significantly higher CR rates were observed among younger patients (P=0.008) and patients with longer times to relapse from either prior HSCT (P<0.001) or remission (P<0.001) (Table 2 and Online Supplementary Table S2). Interestingly, patients with a WBC count of >30,000/ml at first diagnosis had a significantly lower CR rate after first salvage (34% vs. 45%; P=0.011). CR rates increased during each 5-year period from 1990 onward, from 29% in patients who received salvage therapy from 1990-1994 to 45% in haematologica | 2016; 101(12)

Outcome of relapsed/refractory acute lymphoblastic leukemia

Table 2. Complete remission and overall survival among adult Ph-negative relapsed/refractory B-precursor ALL patients in first salvage treatment by selected demographic and clinical factors.

n/N

Overall 361/901 Disease status Primary refractory 44/111 Relapsed, without prior alloHSCT 289/707 Relapsed, with prior alloHSCT 28/81 P-value from univariate model P-value from multivariate model Time from CR1 to first relapse (in months)* <6 months 78/232 6 to <12 months 56/183 12 to <18 months 35/87 18 to <24 months 30/61 24 or more months 87/134 P-value from univariate model P-value from multivariate model Months from prior HSCT to first salvage <12 12/57 12 or more months 16/24 P-value from univariate model P-value from multivariate model** Period of salvage treatment 1990-1994 25/85 1995-1999 65/178 2000-2004 126/314 2005 or later 145/324 P-value from univariate model*** P-value from multivariate model*** Age at salvage treatment (in years) 15-17 22/39 18-34 176/399 35-54 118/320 55-64 33/97 â&#x2030;Ľ65 12/46 P-value from univariate model P-value from multivariate model Sex Male 202/536 Female 159/365 P-value from univariate model P-value from multivariate model WBC at diagnosis <30,000/ mL 193/428 â&#x2030;Ľ30,000/ mL 63/188 P-value from univariate model P-value from multivariate model

Complete remission CRsg % N (95% CI)

Overall survival Median OS, 6-month in months OS KM% (95% CI) (95% CI)

12-month OS KM% (95% CI)

36-month OS KM% (95% CI)

40 (37, 43)

1618

5.8 (5.5, 6.2) 49 (46, 51)

26 (24, 28)

11 (10, 13)

40 (30, 49) 41 (37, 45) 35 (24, 46) 0.55 0.37

115 1291 210

8.2 (6.6, 11.0) 64 (55, 72) 5.8 (5.4, 6.2) 49 (46, 51) 4.4 (3.7, 5.8) 42 (36, 49) 0.03 <0.001

38 (29, 47) 25 (23, 28) 23 (18, 29)

11 (6, 18) 11 (10, 13) 10 (7, 15)

34 (28, 40) 31 (24, 38) 40 (30, 51) 49 (36, 62) 65 (56, 73) <0.001 <0.001

421 304 162 125 264

4.4 (4.0, 4.9) 5.0 (4.1, 5.6) 6.1 (5.2, 7.3) 7.4 (5.6, 8.3) 9.3 (8.1, 11.2) <0.001 <0.001

39 (34, 43) 40 (35, 46) 52 (44, 59) 57 (47, 65) 68 (62, 73)

18 (15, 22) 19 (15, 24) 26 (19, 33) 29 (21, 37) 42 (36, 48)

8 (5, 11) 6 (4, 10) 10 (6, 16) 10 (6, 16) 24 (19, 30)

21 (10, 32) 67 (48, 86) <0.001 NA

151 59

4.2 (3.2, 4.6) 36 (28, 44) 7.5 (4.3, 11.2) 58 (44, 70) 0.014 NA

19 (13, 26) 34 (22, 47)

5 (2, 10) 15 (7, 27)

29 (20, 40) 37 (29, 44) 40 (35, 46) 45 (39, 50) 0.003 0.025

107 342 538 631

3.5 (2.7, 6.6) 4.6 (4.1, 5.3) 6.3 (5.6, 6.7) 6.5 (5.8, 7.3) <0.001 <0.001

42 (33, 51) 38 (33, 43) 53 (48, 57) 53 (49, 57)

21 (14, 29) 17 (13, 21) 27 (24, 31) 31 (27, 35)

7 (3, 13) 5 (3, 8) 13 (11, 16) 13 (11, 17)

56 (40, 72) 44 (39, 49) 37 (32, 42) 34 (25, 44) 26 (14, 41) <0.001 0.008

65 758 578 170 47

7.6 (6.0, 13.0) 6.7 (6.3, 7.4) 4.6 (4.2, 5.2) 4.8 (3.7, 5.9) 2.9 (2.0, 4.7) <0.001 <0.001

62 (49, 72) 56 (52, 60) 42 (38, 46) 41 (34, 49) 34 (21, 47)

43 (31, 55) 30 (27, 34) 21 (18, 25) 18 (13, 25) 17 (8, 29)

16 (8, 26) 15 (12, 18) 8 (6, 11) 6 (3, 11) 3 (0, 12)

38 (34, 42) 44 (38, 49) 0.078 0.56

950 668

6.1 (5.6, 6.6) 51 (47, 54) 5.5 (4.8, 6.1) 47 (43, 50) 0.034 0.045

28 (25, 31) 24 (20, 27)

12 (10, 15) 9 (7, 12)

45 (40, 50) 34 (27, 41) 0.013 0.011

918 406

6.3 (5.8, 6.7) 52 (49, 56) 4.6 (4.2, 5.3) 41 (36, 46) 0.006 0.001

29 (26, 32) 20 (17, 25)

13 (11, 16) 9 (6, 12)

P-values were generated from univariate and multivariate logistic regression models for complete remission and Cox models for survival . *Excludes refractory patients and those with prior HSCT. **Not included in the multivariable models due to small patient numbers. ***Calendar period categorized as 1990-1999 and 2000 or later in univariate and multivariate models.

patients who received therapy from 2005-2013 (P=0.025) (Table 2). Three study groups/sites provided data across all four calendar periods, and for these the increase in CR over time was not significant (P=0.077). In sensitivity analyses (Online Supplementary Table S3) limited to patients treated for r/r ALL from 2005-2013, the overall CR was slightly higher at 45% (95% CI: 39%-50%) than for the entire cohort of patients treated from 1990 onward. Time from CR to first relapse and from HSCT to first salvage remained significant prognostic factors for CR in this subhaematologica | 2016; 101(12)

group. Most importantly, there was little improvement for unfavourable subtypes of relapse â&#x20AC;&#x201C; e.g., relapses within 6 months from diagnosis with 34% CR rate in the whole cohort, and 32% CR rate in the more recent cohort.

Response to second or later salvage. Data on response to second salvage were available in 332 patients, and on third or later salvage in 159 patients. The overall CR rates (excluding patients refractory to primary treatment) decreased with line of salvage, from 40% (95% CI: 37%44%) in first salvage to 21% (95% CI: 16%-26%) in sec1527

N. Gรถkbuget et al.

ond salvage and 11% (95% CI: 6%-18%) in third or greater salvage (Table 3). Similarly, CR rates decreased with line of salvage in patients without prior HSCT: 41% (95% CI: 37%-44%) for first salvage, 19% (95% CI: 14%24%) for second salvage, 11% (95% CI: 6%-19%) for third or later salvage (P<0.001), and in patients with prior HSCT: 35% (95% CI: 24%-46%) for first salvage, 32% (95% CI: 20%-47%) for second salvage, 12% (95% CI: 1%-36%) for third or later salvage (P=0.367). In the cohort of patients treated from 2005 onward, CR was slightly higher in the second (27%; 95% CI: 19%-37%) and third or greater salvages (14%; 95% CI: 5%-29%) than for the overall patient cohort (Online Supplementary Table S4).

Overall survival after salvage therapy Survival after first salvage. The median duration of OS among the 1618 patients in first salvage with available survival data was 5.8 months (95% CI: 5.5-6.2 months; Figure 2A). The proportion of patients alive after 6 months was 49% (95% CI: 46%-51%), after 1 year 26% (95% CI: 24%-28%), and after 3 years 11% (95% CI: 10%-13%) (Table 2), while the median length of follow up in the 193 patients alive at last observation was 43.6 months (range, < 1-247 months). The median OS was 8.2 months (95% CI: 6.6-11.0 months) in refractory patients, 5.8 months (95% CI: 5.4-6.2 months) in patients who had relapsed without prior HSCT, and 4.4 months (95% CI: 3.7-5.8 months) in patients who had relapsed with prior HSCT

Figure 1. Flow chart of patients and patient records analysed.

Patients with more than one line of salvage in the dataset could contribute multiple records; the numbers in the bottom row correspond to the number of records available for each line of salvage. There were 11 patients without survival information that were included in the analysis: 10 had CR information only and 1 had HSCT information only. The total number of patients (who met eligibility criteria and had outcome data available) was 1706. Overall survival data was missing from 11 patients. Two study group databases did not collect information on CR (n = 590 patients); CR data was missing from a further 140 patients. HSCT data were missing from 290 patients.

Table 3. Complete remission and overall survival among adult Ph-negative relapsed* B-precursor ALL patients by line of salvage treatment by selected clinical factors.

In 1st salvage Without prior HSCT With prior HSCT Combined P-value In 2nd salvage Without prior HSCT With prior HSCT Combined P-value In 3rd+ salvage Without prior HSCT With prior HSCT Combined P-value

Complete remission N Median OS, in months (95% CI)

6-month OS KM% (95% CI)

Overall survival 12-month OS KM% (95% CI)

36-month OS KM% (95% CI)

5.8 (5.4, 6.2) 4.4 (3.7, 5.8) 5.7 (5.2, 6.0) 0.17

49 (46, 51) 42 (36, 49) 48 (45, 50)

25 (23, 28) 23 (18, 29) 25 (23, 27)

11 (10, 13) 10 (7, 15) 11 (10, 13)

242 61 303

3.3 (2.5, 4.0) 4.5 (2.4, 5.3) 3.4 (2.8, 4.2) 0.50

30 (25, 36) 35 (23, 47) 31 (26, 36)

14 (10, 19) 17 (9, 28) 14 (11, 19)

6 (3, 10) 4 (1, 12) 5 (3, 8)

108 17 125

2.8 (2.2, 3.1) 4.0 (1.2, 7.0) 2.9 (2.6, 3.4) 0.41

22 (14, 30) 39 (17, 61) 24 (17, 32)

11 (6, 17) 20 (5, 42) 12 (7, 18)

5 (2, 11) 0 4 (1, 9)

n/N

CRsg %

289/709 28/81 317/790

41 (37, 44) 35 (24, 46) 40 (37, 44) 0.28

1293 210 1503

42/225 16/50 58/275

19 (14, 24) 32 (20, 47) 21 (16, 26) 0.04

12/108 2/17 14/125

11 (6, 19) 12 (1, 36) 11 (6, 18) 0.94

*excludes primary refractory patients. P-values comparing patients without prior HSCT to those with prior HSCT were generated from logistic regression models for complete remission and Cox models for survival.

1528

haematologica | 2016; 101(12)

Outcome of relapsed/refractory acute lymphoblastic leukemia

(P<0.001 for the difference between the groups; Figure 2B and Table 2). In patients without prior HSCT, median survival was longer in patients with a longer time to relapse after first CR (Figure 2C), increasing from 4.4 months (95% CI: 4.0-4.9 months) in patients with a first CR lasting <6 months to 9.3 months (95% CI: 8.1-11.2 months) in patients with a first CR lasting â&#x2030;Ľ24 months.

Simon-Makuch curves depict the impact of CR after first salvage on survival (Figure 3). There was a significant association between achieving CR and improved overall survival (Mantel-Byar, P<0.001).

Prognostic factors for survival after first salvage Factors associated with OS in a multivariate analysis of patients in first salvage followed similar patterns to those observed for CR. Time to relapse after first CR was significantly associated with survival, with median OS increasing with each additional 6 months of remission (Table 2; P<0.001). Different cut-offs for time to first relapse all showed a significantly improved OS with longer duration of first remission (Online Supplementary Table S2).

Similarly, in patients who had previously received HSCT, a remission duration of >12 months was significantly

A

B

C

D

haematologica | 2016; 101(12)

Figure 2. (A) Overall survival from time of first salvage among adult Ph-negative relapsed/refractory B-precursor ALL patients. (B) Overall survival from time of first salvage among adult Ph-negative relapsed/refractory B-precursor ALL patients, by disease status at the time of salvage. (C) Overall survival from time of first salvage among adult Ph-negative relapsed B-precursor ALL patients without prior HSCT, by time to relapse (refractory patients excluded). (D) Overall survival from time of salvage among adult Ph-negative relapsed/refractory B-precursor ALL patients, by line of salvage (all disease statuses included).

1529

N. Gรถkbuget et al.

Mantel-Byar, P<0.0001 Mantel-Byar, P<0.0001

Figure 3. Overall survival by CR status among adult Ph-negative relapsed/refractory B-precursor ALL patients in first salvage. The CR survival curve includes only those patients who achieved CR with the first salvage therapy. Patients who achieved CR but for whom no date of CR was available were excluded from the analysis. Survival between groups is assessed beginning at 36 days after the start of first salvage (the median time to CR) and therefore patients who died or whose data were censored before 36 days are not included in the comparison. At 36 days, 137 patients had achieved CR with first salvage and 563 patients had not. Thirty patients who achieved CR with first salvage and 17 patients who did not remained alive and uncensored at 4 years.

Figure 4. Overall survival among adult R/R ALL patients by receipt of HSCT during first salvage. The HSCT survival curve includes only those patients who received HSCT following first salvage and before any subsequent salvage treatment. Patients who received HSCT but for whom no date of HSCT was available were excluded from the analysis. Survival between groups is assessed beginning at 110 days after the start of first salvage (the median time to HSCT); therefore, patients who died or whose data were censored before 110 days are not included in the comparison. At 110 days, 67 patients had received HSCT following first salvage and 581 patients had not. Twenty-two patients who received HSCT following first salvage and 40 patients who did not remained alive and uncensored at 4 years.

associated with longer median OS (7.5 months vs. 4.2 months; P=0.014). Median OS increased over time, and was greater in patients who received first salvage therapy from the year 2000 onward than those who received it from 1990-1999 (6.4 vs. 4.5 months; P<0.001). As with CR, the increase in median OS among the three study groups/sites that provided data across all four calendar periods was not significant (P=0.14). Survival was better in younger patients (P<0.001) and the difference in median OS between males and females was significant (6.1 vs. 5.5 months; P=0.045). Finally, the better outcome of patients with lower WBC counts (<30,000/mL at first diagnosis) that was observed for CR rates was also apparent for OS (6.3 vs. 4.6 months; P=0.001). In sensitivity analyses (Online Supplementary Table S3) limited to patients treated for r/r ALL from 2005-2013, the overall median survival was higher at 6.5 months (95% CI: 5.8-7.3 months) than for the entire cohort of patients treated from 1990 onward.

Disease status, time from CR to first relapse, and age remained significant prognostic factors for survival in this subgroup.

13%) for first salvage to 14% (95% CI: 11%-19%) and 5% (95% CI: 3%-8%) for second salvage and 12% (95% CI: 7%-18%) and 4% (95% CI: 1%-9%) for third or greater salvage. No statistically significant difference was detected for OS comparing patients with or without prior HSCT (Table 3). In the cohort of patients treated from 2005 onward, median survival was slightly higher in the second (4.4 months; 95% CI: 3.1-4.9) and third or greater salvages (3.5 months; 95% CI: 2.3-5.4) than for the overall cohort of patients (Online Supplementary Table S4).

Survival after second or later salvage Median duration of OS after salvage therapy decreased with each subsequent line of therapy (Figure 2D), from 5.8 months (95% CI: 5.5-6.2 months) after first salvage, to 3.4 months (95% CI: 2.8-4.2 months) after second salvage, and 2.9 months (95% CI: 2.6-3.4 months) after third or greater salvage. Among patients who were not refractory to primary treatment, one- and three-year survival rates also decreased with each subsequent salvage therapy, from 25% (95% CI: 23%-27%) and 11% (95% CI: 10%1530

HSCT after relapse Realization of HSCT after relapse. Information on HSCT realization after relapse was available for 1337 patients in first salvage. Twenty-eight percent of patients received HSCT after their first salvage treatment, with a range of HSCT in different countries between 15% and 54% (Table 4). The median time from first salvage treatment to HSCT was 3.6 months (range, <1-15 months). HSCT realization was more common in younger patients, and the proportion of patients who received HSCT increased over time (Table 4). Approximately half the patients who received HSCT were known to be in CR at the time of transplant (183/375 patients, 49%). A similar proportion of patients who achieved a CR received HSCT after their first salvage (183/337 patients, 54%; Table 4). Patients with a prior HSCT were less likely to receive a HSCT after salvage treatment; however, if they achieved CR, they received a second HSCT as often as all other patients.

Overall survival according to receipt of HSCT after relapse.

Simon-Makuch curves were used to graphically depict the effect on survival of HSCT during first salvage, using haematologica | 2016; 101(12)

Outcome of relapsed/refractory acute lymphoblastic leukemia

Table 4. HSCT after salvage therapy in Ph-negative relapsed/refractory B-precursor ALL patients in first salvage treatment only.

n/N

In 1st salvage All countries Ranges across all countries Period of salvage treatment 1990-1994 1995-1999 2000-2004 2005 or later P-value Disease status Relapsed without prior HSCT Relapsed with prior HSCT P-value Age at salvage treatment (in years) 15-17 18-34 35-54 55-64 ≥65 P-value

Among all patients receiving salvage treatment Proportion of any patients n/N with HSCT received after first salvage treatment (%, 95% CI)

Among patients achieving CR Proportion of patients receiving HSCT in CR after first salvage treatment* (%, 95% CI)

375/1 337 --

28 (26, 31) 15-54

183/337 --

54 (49, 60) 40-83

15/100 59/323 151/491 150/423

15 (9, 24) 18 (14, 23) 31 (27, 35) 35 (31, 40) <0.001

5/24 20/63 70/118 88/132

21 (7, 42) 32 (21, 45) 59 (50, 68) 67 (58, 75) <0.001

318/1 074 27/157

30 (27, 33) 17 (12, 24) <0.001

150/271 13/26

55 (49, 61) 50 (30, 70) 0.63

19/59 204/618 128/490 23/126 1/44

32 (21, 46) 33 (29, 37) 26 (22, 30) 18 (12, 26) 2 (0, 12) <0.001

10/20 96/162 61/112 15/31 1/12

50 (27, 73) 59 (51, 67) 54 (45, 64) 48 (30, 67) 8 (0, 38) 0.01

P-values were generated from univariate Cox models. *Number of patients who achieved CR and who had data on HSCT after salvage.

HSCT as a time-dependent covariate (Figure 4). Duration of survival from 110 days (the median time to transplant) was significantly longer if HSCT was received after first salvage (Mantel-Byar, P<0.001). In this analysis, 40 patients who did not receive HSCT during first salvage were still alive after 3 years; of these, 18 patients were transplanted after a subsequent salvage therapy. This method does not account for differences in patient characteristics that are associated with receiving HSCT.

Discussion This pooled historical data analysis summarizes the largest dataset on patient outcomes for adults with Phnegative, B-precursor r/r ALL assembled so far, thereby providing a comprehensive benchmark for standard of care in Europe and the US against which emerging therapies can be judged. By harmonising and pooling the data we were able to obtain more precise estimates, and the relatively large size of the dataset for this rare disease provides greater opportunity for comparing outcomes between important patient subgroups. Rates of remission and survival were found to vary significantly across lines of salvage and between other clinically-relevant subgroups.

Among patients in first salvage, 40% achieved a CR, and the median duration of overall survival was 5.8 months.

The one- and three-year survival rates were 26% and 11%, respectively, indicating that one-year of follow-up may be informative to detect a potential survival difference with novel agents, although longer follow up is required to provide full information on prognosis in r/r ALL. Patients receiving therapy 2005 or later had modesthaematologica | 2016; 101(12)

ly better CR and OS outcomes. After failure of first salvage, outcomes declined significantly. This reflects the likelihood that with each line of salvage, more chemotherapy-resistant subpopulations of ALL blasts will be selected. For clinical trials with new compounds, patients refractory to one or more lines of salvage are usually chosen; our analysis provides the appropriate comparator data for each subset of patients. It furthermore confirms that achievement of CR with the first salvage approach is prognostic: OS was significantly longer in patients who responded to their first salvage treatment than those who did not (P<0.001). It remains to be demonstrated whether the poor prognostic impact of refractory relapses may change when more potent, targeted therapies become available.

Overall, the data underline that relapses in adult ALL typically occur early – more than half occurring within 1 year, during ongoing intensive treatment. Duration of first CR is a known prognostic factor for achievement of subsequent responses to salvage therapy and for survival outcomes2-5,7,18 and was confirmed here, independently of the cutpoints selected for analysis. This was demonstrated for patients with relapse after chemotherapy, and in patients with relapse after HSCT. In both settings, patients with very late relapse—more than 24 months after first diagnosis—represent a favourable subgroup with better chances to achieve a CR and to obtain long-term survival. The reason for the poor prognosis of early relapses is probably different disease biology.19 Whereas late relapses may arise from slowly-cycling still chemotherapy-sensitive subclones of ALL, early relapses probably arise from selected, chemotherapy-resistant subclones that proliferate despite ongoing standard chemotherapy.19,20 Interestingly, the rate of CR and OS among patients 1531

N. GĂśkbuget et al.

receiving salvage therapy after first relapse increased over the calendar period (1990-2013), although no new compounds with extraordinary performance entered the market during this period. The observed improvements over time may be the result of various factors including the more frequent application of more intensive treatments, better supportive care, more aggressive use of HSCT, and more attempts to treat. Sensitivity analyses, limited to patients treated from 2005 onward, showed improvements in CR and survival compared with the entire cohort. However, an analysis limited to those study groups or sites that provided data over the entire calendar period (i.e., from 1990 onward) showed no significant trends over time. CR and OS also differed between study groups or centers, which may be due to differences in the distribution of patient or prognostic factors, such as US sites having patients with more lines of salvage and more patients treated from 1990â&#x20AC;&#x201C;2000 than the European study groups or centers.

An association with lower WBC counts at first diagnosis and increased chance of CR has been found in some studies6 but no significant association was found with either CR or OS in others.2-5 Higher WBC counts reflect a more aggressive disease biology, which is associated with higher relapse rates; patients with higher WBC are therefore considered as high-risk in many study groups. The inferior rates of CR and OS associated with this biologic subtype of B-precursor ALL has now also been confirmed for the r/r setting. Gender has been investigated as a prognostic factor with conflicting results: no significant relationship was observed with OS in some studies,3,6,18 but our analysis was consistent with previous findings that although females have no significant difference in probability of achieving CR,3,5 they may have a shorter duration of survival than males.2,5 The reasons for this are unclear, but an increased mortality rate during or after re-induction may be a contributing factor.5 Limited or missing data did not allow us to examine CR or OS results by specific type of induction or salvage therapy.

The overall rate of HSCT after relapse was 28% and ranged from 15% to 54% across different study groups/centers, likely reflecting different practices between countries, donor availabilities and options provided by healthcare systems. However, it reached 50% in patients achieving a CR after first salvage. It is also noteworthy that the rate of HSCT increased over time. This is clinically important because our analysis, although potentially limited by selection factors of who receives and who does not receive HSCT, confirms the long-term benefits of HSCT after relapse from initial therapy. We considered HSCT as a time-dependent covariate to account for the positive selection bias that applies to patients who survive long enough to receive HSCT. Based on these analyses, we show that receipt of HSCT during first salvage significantly increased OS and confirmed that long-term survival in r/r ALL is strongly influenced by whether or not a patient receives HSCT. However, the observed difference in survival may also be influenced by differences in patient characteristics, which may have influenced receipt of HSCT, and which were not collected in this study. These potential factors, such as presence of comorbidities or poor performance status, may also be strong predictors of survival.

A significant association of older age with decreased CR 1532

and poorer survival was confirmed in our pooled analysis, consistent with published findings.2,3,7,18 This may reflect the reduced use of intensive chemotherapy, which has expected toxicities and is not needed for older patients who are likely to be ineligible for transplant. These patients have a specific need for less toxic, targeted therapies with the option to obtain prolonged survival without subsequent HSCT. In the conduct of pooled analysis across different countries and clinical centers, heterogeneity between measured and unmeasured variables may have influenced our findings. The definition of CR, as described in the supplementary material, was based on general clinical practice guidelines and could have varied across the study groups or centers. Our pooled analysis was also limited by data availability on HSCT realization and date of HSCT for a number of patients. If patients with missing HSCT data were different from patients with available data this could introduce bias, but comparisons of the distribution of available demographic and clinical factors between groups with and without HSCT data did not reveal meaningful differences. Additional limitations of patient or prognostic factors included lack of information on minimal residual disease, high-risk biomarkers, or other indications of upfront HSCT which may have decreased the incidence of relapse.

The vast majority of ALL relapses occur in the bone marrow. Patients with CNS involvement or other isolated extramedullary relapses were not included in the analysis because these patients typically receive different treatments and represent a subgroup often reported separately in prospective or retrospective studies, and are usually excluded from trials of new drugs in r/r ALL. This pooled retrospective analysis is the largest to date that characterizes outcomes among patients with Ph-negative, B-precursor r/r ALL. The rates of CR and OS were comparable with published studies, which is not unexpected since several of the datasets included have already been published separately.2-5,18 Newly-approved and emerging therapies offer the promise of improved response rates and opportunities for long-term survival, with the potential for less toxicity than chemotherapy.21-23 The data underline the necessity to closely describe subgroups of r/r ALL in clinical trials with new drugs in order to make published results comparable. Our results will therefore be important to provide context when evaluating outcomes with these new therapies, particularly with respect to the important role of composition of patient characteristics in different published trials for r/r ALL. However, investigators should carefully consider selection bias and other limitations of historical data when making such comparisons. These historical real-world clinical data are not equivalent to clinical trial data; nonetheless, by pooling datasets we were able to provide a more accurate estimate of patient outcomes among different subgroups of patients receiving salvage chemotherapy, which can be valuable across all stages of drug development to assess outcomes with current standard of care.

Acknowledgments The authors would like to thank James Oâ&#x20AC;&#x2122;Kelly, an employee of Amgen, for providing medical writing assistance; and Sharon Dunn and Richard Bo, consultants paid by Amgen, who provided statistical programming support. haematologica | 2016; 101(12)

Outcome of relapsed/refractory acute lymphoblastic leukemia

References 9.

1. Bassan R, Hoelzer D. Modern therapy of acute lymphoblastic leukemia. J Clin Oncol.

2011;29(5):532-543.

2. Fielding AK, Richards SM, Chopra R, et al.

Outcome of 609 adults after relapse of acute lymphoblastic leukemia (ALL); an MRC UKALL12/ECOG 2993 study. Blood.

2007;109(3):944-950.

3. Gokbuget N, Stanze D, Beck J, et al.

Outcome of relapsed adult lymphoblastic leukemia depends on response to salvage chemotherapy, prognostic factors, and performance of stem cell transplantation.

Blood. 2012;120(10):2032-2041.

4. O'Brien S, Thomas D, Ravandi F, et al.

Outcome of adults with acute lymphocytic leukemia after second salvage therapy.

Cancer. 2008;113(11):3186-3191.

5. Oriol A, Vives S, Hernandez-Rivas JM, et al.

Outcome after relapse of acute lymphoblastic leukemia in adult patients included in four consecutive risk-adapted trials by the PETHEMA Study Group. Haematologica.

2010;95(4):589-596.

6. Tavernier E, Boiron JM, Huguet F, et al.

Outcome of treatment after first relapse in adults with acute lymphoblastic leukemia initially treated by the LALA-94 trial.

Leukemia. 2007;21(9):1907-1914.

7. Kantarjian HM, Thomas D, Ravandi F, et al.

Defining the course and prognosis of adults with acute lymphocytic leukemia in first salvage after induction failure or short first remission duration.

Cancer.

2010;116(24):5568-5574.

8. Bhojwani D, Pui CH. Relapsed childhood

haematologica | 2016; 101(12)

10.

11.

12.

13.

14.

15.

acute lymphoblastic leukaemia. Lancet Oncol. 2013; 14(6):e205-217.

Hoelzer D. Targeted therapy with monoclonal antibodies in acute lymphoblastic leukemia. Curr Opin Oncol. 2013;25(6):701706.

Howlader N, Noone AM, Krapcho M,et al. (eds). SEER Cancer Statistics Review, 19752011, National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975_2011/, based on November 2013 SEER data submission, posted to the SEER web site, April 2014.

DeAngelo DJ, Yu D, Johnson JL, et al.

Nelarabine induces complete remissions in adults with relapsed or refractory T-lineage acute lymphoblastic leukemia or lymphoblastic lymphoma: Cancer and Leukemia Group B study 19801. Blood.

2007;109(12):5136-5142.

Jeha S, Gaynon PS, Razzouk BI, et al. Phase II study of clofarabine in pediatric patients with refractory or relapsed acute lymphoblastic leukemia. J Clin Oncol. 2006; 24(12):1917-1923.

Ottmann OG, Druker BJ, Sawyers CL, et al.

A phase 2 study of imatinib in patients with relapsed or refractory Philadelphia chromosome-positive acute lymphoid leukemias.

Blood. 2002;100(6):1965-1971.

Topp MS, Gokbuget N, Zugmaier G, et al.

Phase II trial of the anti-CD19 bispecific T cell-engager blinatumomab shows hematologic and molecular remissions in patients with relapsed or refractory B-precursor acute lymphoblastic leukemia. J Clin Oncol.

2014;32(36):4134-4140.

Brookmeyer R, Crowley J. A. K-sample median test for censored data. J Am

Statistical Assoc. 1982;77:433-440.

16. Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. John Wiley and Sons, New York, 1980.

17. Simon R, Makuch RW. A non-parametric graphical representation of the relationship between survival and the occurrence of an event: application to responder versus nonresponder bias. Stat Med. 1984;3(1):35-44.

18. Thomas DA, Kantarjian H, Smith TL, et al.

Primary refractory and relapsed adult acute lymphoblastic leukemia: characteristics, treatment results, and prognosis with salvage therapy. Cancer. 1999;86(7):1216-1230.

19. Mullighan CG. Molecular genetics of B-precursor acute lymphoblastic leukemia. J Clin Invest. 2012;122(10):3407-3415.

20. Choi S, Henderson MJ, Kwan E, et al.

Relapse in children with acute lymphoblastic leukemia involving selection of a preexisting drug-resistant subclone. Blood.

2007;110(2):632-639.

21. Kantarjian H, Thomas D, Jorgensen J, et al.

Inotuzumab ozogamicin, an anti-CD22calecheamicin conjugate, for refractory and relapsed acute lymphocytic leukaemia: a phase 2 study. Lancet Oncol. 2012; 13(4):403-411.

22. Maude SL, Frey N, Shaw PA, et al. Chimeric antigen receptor T cells for sustained remissions in leukemia. N Engl J Med. 2014;371 (16):1507-1517.

23. Topp MS, Gokbuget N, Stein AS, et al.

Safety and activity of blinatumomab for adult patients with relapsed or refractory Bprecursor acute lymphoblastic leukaemia: a multicentre, single-arm, phase 2 study.

Lancet Oncol. 2015;16(1): 57-66.

1533

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Lymphoblastic Leukemia

Ferrata Storti Foundation

CLIC5: a novel ETV6 target gene in childhood acute lymphoblastic leukemia Benjamin Neveu,1,2 Jean-François Spinella,1,3 Chantal Richer,1 Karine Lagacé,1,2 Pauline Cassart,1 Mathieu Lajoie,1 Silvana Jananji,1 Simon Drouin,1 Jasmine Healy,1 Gilles R.X. Hickson1,4 and Daniel Sinnett1,2,5 CHU Sainte-Justine Research Center, Montreal; 2Department of Biochemistry and Molecular Medicine, Faculty of Medicine, University of Montreal; 3Molecular biology program, Faculty of Medicine, University of Montreal; 4Department of Pathology and Cellular Biology, Faculty of Medicine, University of Montreal and 5Department of Pediatrics, Faculty of Medicine, University of Montreal, Montreal, Canada

1

Haematologica 2016 Volume 101(12):1534-1543

ABSTRACT

T

Correspondence: daniel.sinnett@umontreal.ca

Received: May 19, 2016.

Accepted: August 11, 2016.

Pre-published: August 18, 2016. doi:10.3324/haematol.2016.149740

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/12/1534

©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.

1534

he most common rearrangement in childhood precursor B-cell acute lymphoblastic leukemia is the t(12;21)(p13;q22) translocation resul-ting in the ETV6-AML1 fusion gene. A frequent concomitant event is the loss of the residual ETV6 allele suggesting a critical role for the ETV6 transcriptional repressor in the etiology of this cancer.

However, the precise mechanism through which loss of functional ETV6 contributes to disease pathogenesis is still unclear. To investigate the impact of ETV6 loss on the transcriptional network and to identify new transcriptional targets of ETV6, we used whole transcriptome analysis of both pre-B leukemic cell lines and patients combined with chromatin immunoprecipitation. Using this integrative approach, we identified 4 novel direct ETV6 target genes: CLIC5, BIRC7, ANGPTL2 and WBP1L. To further evaluate the role of chloride intracellular channel protein CLIC5 in leukemogenesis, we generated cell lines overexpressing CLIC5 and demonstrated an increased resistance to hydrogen peroxide-induced apoptosis. We further described the implications of CLIC5’s ion channel activity in lysosomal-mediated cell death, possibly by modulating the function of the transferrin receptor with which it colocal-izes intracellularly. For the first time, we showed that loss of ETV6 leads to significant overexpression of CLIC5, which in turn leads to decreased lysosome-mediated apoptosis. Our data suggest that heightened CLIC5 activity could promote a permissive environment for oxidative stressinduced DNA damage accumulation, and thereby contribute to leukemogenesis.

Introduction ETV6 is a known transcriptional repressor1 involved in hematopoiesis.2 ETV6 rearrangements are frequently observed in multiple hematological diseases, including precursor B-cell acute lymphoblastic leukemia (pre-B ALL), the most common pediatric cancer.3 In fact, the t(12;21)(p13;q22) translocation, which generates an inframe ETV6-AML1 fusion product,4 is the most frequent chromosomal abnormality in childhood pre-B ALL, pres-ent in 20% of cases.5 The expression of ETV6-AML1 is systematically observed in t(12;21)-positive pre-B ALL,6,7 indicating a possible function for this chimeric protein in pre-B ALL etiology. However, it was shown that the frequency of the t(12;21) translocation is 100 times greater than that of preB ALL,8 suggesting that its presence alone is insufficient to induce leukemia. It has been demonstrated that the second non-rearranged ETV6 allele is frequently deleted or inactivated in t(12;21)-positive pre-B ALL,7,9-11 and recent studies have reported germline ETV6 loss-of-function mutations that were shown to be associated with familial hematological disorders, including ALL.12 Although these data suggest that ETV6 plays a key tumor suppressor role and that its complete inactivation may be haematologica | 2016; 101(12)

CLIC5 in childhood leukemia

required for leukemogenesis,13 little is known about the function of ETV6 in normal hematopoiesis and leukemic transformation.

Given its role in transcriptional repression, we postulated that loss of ETV6 could result in deregulated expression of downstream target genes and perturb key cellular processes and pathways leading to oncogenesis. Only two transcriptional targets of ETV6 have been identified to date: the MMP3 matrix metallopeptidase14 and the antiapoptotic protein BcL-xL.15 To comprehensively identify novel ETV6 target genes, we combined whole transcriptome analysis and chromatin immunoprecipitation (ChIP) assays. Using an in vitro cell-based system combined with data from childhood pre-B ALL patient tumors, we identified 4 genes (CLIC5, BIRC7, ANGPTL2 and WBP1L) whose expression was directly regulated by ETV6.

Functional interrogation of CLIC5 revealed its implication in lysosome-mediated cell death, possibly by regulating iron homeostasis through the transferrin receptor with which it colocalizes intracellularly. In this study, we provide the first evidence of a role for CLIC5-mediated resistance to oxidative stress that may contribute to leukemogenesis.

mapped to the hg19 reference genome using STAR with default settings,16 and read counts per genes were determined using HTSeq-count.17 To identify differentially expressed genes (DEGs), we used the R bioconductor package edgeR18 with BenjaminiHochberg P-value adjustment. The two clones were considered as biological replicates.

The patient cohort used for RNA sequencing was composed of 9 hyperdiploid and 9 t(12;21) patients. Total RNA was extracted from leukemic bone marrow samples of all patients and from control pre-B cells (CD19+CD10+) isolated from healthy cord blood samples. cDNA libraries were prepared using the SOLiD Total RNA-seq kit and sequenced on the SOLiD 4/5500 System (Life Technologies). Reads were aligned to the hg19 reference genome and read counts per gene obtained using LifeScope Genomic Analysis Software with default parameters.

Quantitative real-time PCR 350ng of total RNA were retro-transcribed with M-MLV reverse transcriptase (Life Technologies). cDNA was then subjected to quantitative real-time PCR using the primer sets listed in the Online Supplementary Table S1. Relative expression was determined by the 2-(DDCt) comparative method19 using GAPDH as the reference gene.

Chromatin immunoprecipitation Methods Complete methods can be found in the Online Supplementary Methods section.

Expression profiling by RNA-sequencing Total RNA from two different Reh clones (generated in methylcellulose media) each stably expressing ETV6-His and ETV6DETS_NLS-His (and pLENTI control) were processed through the TruSeq Stranded Total RNA protocol and sequenced on the HiSeq 2500 system (Illumina). Reads for each sample were

Chromatin immunoprecipitation (ChIP) was performed on 10x106 transduced Reh cells cross-linked directly in cell medium for 10min with 1% methanol-free formaldehyde (Polysciences, Inc.). Immunoprecipitation of sheared chromatin was carried out using anti-HA magnetic beads (Thermo Fisher Scientific). Beads were eluted twice with HA peptides (Thermo Fisher Scientific) before reverse cross-linking. DNA was purified twice by standard phenol/chloroform/isoamyl alcohol (Sigma-Aldrich) extraction prior to qRT-PCR analysis (primers are listed in the Online Supplementary Table S2).

Figure 1. Schematic representation of the transcriptome-based design to detect putative direct ETV6 target genes. To identify direct targets of ETV6, we first designed an in vitro RNA-seq experiment using ETV6–/– Reh-derived clones. Cells were transduced with lentiviral constructs to express ETV6His and ETV6DETS_NLS-His. Total RNA was extracted from stable cell populations and RNA-seq libraries were sequenced.

Expression profiles were analyzed using EdgeR. Gene expression profiles in ETV6-His cells were first compared with ETV6DETS_NLS-His and pLENTI cells to identify repressed genes (FDR ≤ 0.1). We then included data from the ETV6ΔETS_NLS-His vs. pLENTI comparison and further considered genes whose expression remains constant (P-value≥0.05 or logFC≥-0.5) which are more likely to be direct ETV6 targets. Finally, only genes that showed a specific overexpression in t(12;21)-positive childhood pre-B ALL (pre-B acute lymphoblastic leukemia) patients were considered. ALL: acute lymphoblastic leukemia; FDR; false discovery rate; logCPM: log counts per million reads. haematologica | 2016; 101(12)

1535

B. Neveu et al.

Table 1. Expression status of the 6 putative direct ETV6 target genes.

ETV6-His vs.

ETV6DETS_NLS-His ETV6-His vs. pLENTI

In vitro

ETV6DETS_NLS-His vs. pLENTI ALL patients t(12;21) vs. B-cells

logFC logCPM PValue FDR logFC logCPM PValue FDR logFC logCPM PValue FDR logFC logCPM PValue FDR

CLIC5

BIRC7

Gene Symbol ANGPTL2

WBP1L

LRRC4

SLC51A

-3.23 4.66 7.00E-52 9.59E-48 -3.20 3.75 6.40E-39 8.09E-35 0.03 4.62 0.88 1 6.36 10.55 1.27E-30 6.91E-29

-1.93 4.02 3.44E-11 1.45E-07 -1.39 3.02 2.30E-05 4.76E-03 0.53 3.85 0.03 1 6.79 9.74 1.57E-13 1.55E-12

-1.31 5.00 4.24E-11 1.45E-07 -1.68 4.37 3.34E-14 6.41E-11 -0.38 4.81 0.02 1 5.82 9.42 6.39E-08 3.11E-07

-0.79 5.78 8.04E-06 9.18E-03 -1.10 5.18 1.06E-09 6.71E-07 -0.32 5.49 0.06 1 1.92 10.68 4.87E-07 2.17E-06

-1.02 3.90 9.18E-05 6.29E-02 -1.16 3.23 3.99E-06 1.20E-03 -0.15 3.63 0.55 1 4.25 9.30 7.67E-12 6.25E-11

-1.31 3.11 1.14E-04 6.79E-02 -1.50 2.43 1.45E-05 3.28E-03 -0.20 2.90 0.50 1 4.27 8.26 1.91E-06 7.59E-06

ALL: acute lymhoblastic leukemia; logFC: log fold change; logCPM: log counts per million reads; FDR: false discovery rate.

Apoptosis assays Apoptosis was induced by treating cells for 20h with hydrogen peroxide (PRXD; Sigma-Aldrich), camptothecin (CPT; Tocris Bioscience) or doxorubicin (DOXO; Sigma-Aldrich) and assayed

by Alexa Fluor 488-coupled Annexin V and propidium iodide (PI) double staining. 1x104 stained cells were analyzed by flow cytometry. Total apoptosis includes Annexin V+/ PI - (early apoptotic), Annexin V+/PI + (late apoptotic) and Annexin V-/PI+ (necrotic) cells.

Immunofluorescence microscopy Reh cells were seeded at 2x106 cells/mL in a 96 well glass plate (Whatman) and fixed in a 3.7% formaldehyde solution.

Immunostaining was performed overnight with CLIC5A antibody (ab191102 dil. 1:1000; Abcam) and transferrin receptor antibody (ab84036 dil. 1:200; Abcam). Goat anti-Mouse Alexa Fluor 488 (dil.

1:500; Thermo Fisher Scientific) was used to detect anti-CLIC5A, and Goat anti-Rabbit Alexa Fluor 546 (dil. 1:500; Thermo Fisher Scientific) was used to detect anti-transferrin receptor. Hoechst 33258 DNA stain (dil. 1:500; Thermo Fisher Scientific) was included to stain nuclei.

Statistical tests The significance of observations was assessed using one or twotailed Fisher's exact test or Mann-Whitney U test when appropriate.

Ethics statement The CHU Sainte-Justine Research Ethics Board approved the protocol. Informed consent was obtained from the parents of the patients to participate in this study and for publication of this report and any accompanying images.

patient tumor samples (Figure 1). Based on the expression profiles of transduced t(12;21)-positive pre-B ALL Reh clones (Online Supplementary Figure S1), we found 331 genes repressed in His-tagged ETV6 (ETV6-His) cells compared to control cells (pLENTI empty vector; P-value≤0.05), of which 88 remained significant after multiple testing corrections (FDR≤0.1; Online Supplementary Table S3). 18 genes were significantly repressed by ETV6-His compared to its DNA-binding deficient mutant ETV6DETS_NLS-His (FDR≤0.1; Online Supplementary Table S4), of which 11 were both confidently expressed (logCPM≥1) and also present in the above-mentioned list of 88 genes. These genes are thus more likely to be direct targets of ETV6 since their repression depends on ETV6's DNA-binding domain. However, only 7 of these genes absolutely required the DNA-binding domain for repression (ETV6DETS_NLS-His vs. pLENTI; P-value≥0.05 or logFC≥-0.5), further confirming their direct regulation by ETV6: CLIC5, BIRC7, DDIT4L, ANGPTL2, WBP1L, LRRC4, and SLC51A. We then assessed whether the ETV6-dependent transcriptional repression observed in vitro translated to childhood pre-B ALL patient tumor samples. Expression of the 331 genes repressed by ETV6 in vitro was evaluated in transcriptome data from 9 t(12;21)-positive samples (ETV6 negative) and compared to 9 hyperdiploid cases as well as to 3 normal pre-B cell (CD19+/CD10+) samples (ETV6 positive).

We identified 45 genes that were downregulated in vitro (13.6%) and that were also specifically overexpressed in t(12;21)-positive patients (Figure 2).

Interestingly, these include 6 of the 7 genes identified as putative direct ETV6 targets in vitro (CLIC5, BIRC7, ANGPTL2, WBP1L, LRRC4 and SLC51A, but not DDIT4L), further supporting a role for ETV6 in their regulation.

Results ETV6 represses the expression of 6 genes in pre-B ALL cell lines and patient samples To identify novel direct ETV6-regulated genes, we carried out a transcriptome analysis in both cell lines and 1536

CLIC5, BIRC7, ANGPTL2 and WBP1L are direct targets of ETV6 To validate ETV6-dependent expression of these 6 genes (Table 1), we used quantitative real-time PCR (qRT-PCR) in both Reh clones and the original Reh pophaematologica | 2016; 101(12)

CLIC5 in childhood leukemia

Figure 2. Expression of putative ETV6 targets in pediatric pre-B ALL (pre-B acute lymphoblastic leukemia) patients. An unsupervised clustering heatmap was generated with the 45 genes specifically overexpressed (yellow) in t(12;21)-positive ALL (acute lymhoblastic leukemia) patients (t12.21, n=9) compared to hyperdiploid ALL patients (HD, n=9) and normal B-cells controls (B-cell, n=3). Among the 7 genes identified as putative direct ETV6 targets in our cell line model, CLIC5, BIRC7, ANGPTL2, WBP1L, LRRC4 and SLC51A were specifically overexpressed in t(12;21)-positive pediatric pre-B ALL patient tumor samples (Arrows).

ulation overexpressing ETV6 WT, ETV6-HA or GFP as a control (Online Supplementary Figure S2). Both ETV6 WT and ETV6-HA efficiently repressed expression of these genes except for LRRC4 (Figure 3A). To assess the physical interaction between ETV6 and the proximal promoters of these 5 putative ETV6 targets we performed ChIP experiments in Reh cells overexpressing ETV6-HA or ETV6 WT as a negative control. Importantly, ETV6-HA behaves similarly to ETV6 WT in our qRT-PCR experiments (Figure 3A), indicating that the epitope tag does not negatively interfere with normal ETV6 repressor function. As shown in Figure 3B, we successfully enriched the proximal promoters of CLIC5, BIRC7, ANGPTL2 and WBP1L, but not SLC51A, further confirming that these 4 genes are indeed direct targets of ETV6.

CLIC5A reduces hydrogen peroxide-induced apoptosis We pursued functional interrogation of our strongest candidate, the chloride intracellular channel CLIC5, to investigate its cellular function and potential contribution to childhood pre-B ALL. The CLIC5 locus encodes two major isoforms, CLIC5A and CLIC5B20 (Online Supplementary Figure S3A), transcribed by two alternative promoters and differing only from their first exon. Using ChIP experiments (as above), we showed that the CLIC5A promoter was specifically enriched, whereas the CLIC5B promoter showed no significant enrichment compared to the negative control region (Online Supplementary Figure S3B). ETV6 overexpression was also shown to lead to a marked decrease of the CLIC5A protein, whereas CLIC5B levels remained constant (Online Supplementary Figure haematologica | 2016; 101(12)

S3C). Together, these results confirm specific ETV6-mediated repression of CLIC5A. In light of these results, we overexpressed the CLIC5A isoform in Reh cells (Figure 4A) in order to investigate its role on B-lymphoblast function. Importantly, the overexpression of CLIC5A in our Reh cells is highly similar to that observed in a validation cohort of t(12;21) ALL patients (Online Supplementary Figure S4). Given that changes in migration were observed upon silencing of CLIC5A,21 we first evaluated this phenotype. We found no particular difference in migration toward stromal cellderived factor 1 (SDF-1, also known as CXCL12) in a classic transwell experiment between control and CLIC5A overexpressing cells (Online Supplementary Figure S5).

Although CLIC5A had never been shown to be associated with apoptosis, suppression of its closely related family member CLIC4 had previously been shown to enhance hydrogen peroxide-induced apoptosis.22 To test this hypothesis in our cell model, we treated CLIC5A overexpressing cells with hydrogen peroxide, camptothecin, or doxorubicin, and evaluated apoptosis. Of note, peroxide induces an apoptotic cell death in these conditions (Figure 4B), rather than necrosis. We observed a modest yet consistent reduction in hydrogen peroxide-induced apoptosis compared to control cells (Figure 4C), suggesting a potential role in the intracellular response to free radicals. A similar reduction in apoptosis following peroxide treatment was observed with CLIC5A overexpression in the IM9 B-lymphoblastoid cell line (Figure 4D-F) endogenously expressing wild-type ETV6, further confirming that CLIC5A overexpression specifically reduces hydrogen peroxide-induced apoptosis across cellular backgrounds.

1537

B. Neveu et al.

A

B

Figure 3. Quantitative real-time PCR and chromatin immunoprecipitation validation of putative ETV6 target genes. Reh cells together with the two Reh derivated clones were infected with pCCL GFP, ETV6 WT or ETV6-HA. (A) Total RNA was extracted from these cells and complementary DNA was generated. This cDNA was submitted to qRT-PCR (quantitative real time PCR) analysis to quantify relative expression of putative ETV6 targets. Expression of all genes but LRRC4 is repressed by ETV6 WT (wild-type) and ETV6HA. (B) ChIP (chromatin immunoprecipitation) experiments were performed in ETV6-WT and ETV6-HA cells. Putative ETV6 target gene proximal promoter enrichment was determined by qRT-PCR using promoter-specific primers. Results are presented as the ratio of the input percentage obtained in ETV6-HA cells compared to ETV6 WT cells, corrected by the background enrichment obtained with an unbound region (Neg). CLIC5, BIRC7, ANGPTL2 and WBP1L promoters are enriched. For both qRT-PCR and ChIP experiments, 4 values were calculated for each of the 3 cell lines and were merged for a total of 12 values (n=12). Error bars represent the standard deviation. Statistical significance is calculated by two-tailed and one-tailed Mann-Whitney U test for qRT-PCR and ChIP analysis, respectively.

CLIC5A is an endosomal ionic channel involved in lysosome-mediated apoptosis Given that hydrogen peroxide is known to trigger lysosomal membrane permeabilization (LMP) and initiate the lysosomal-mediated apoptosis pathway (Figure 5A),23,24 we hypothesized that CLIC5A’s role in protecting cells against apoptosis may function through the modulation of LMP. Since LMP has a direct impact on mitochondrial outer membrane permeabilization (MOMP) that can be assessed through mitochondrial membrane potential (MMP), we evaluated MMP in our Reh cellular model treated with hydrogen peroxide. We observed a significant reduction of MMP loss correlated with CLIC5A overexpression compared to control, indicating that substantially more mitochondria remained intact following peroxide exposure when CLIC5A was overexpressed (Online Supplementary Figure S6). This result supports a role for CLIC5A in protecting cells against peroxide-induced apoptosis and suggests that it functions upstream of MOMP, which indeed corroborates a possible implication of CLIC5A in LMP regulation.

Deleterious effects of hydrogen peroxide on lysosome membranes are strongly dependent on lysosomal Fe2+ availability since it dictates its conversion to the highly reactive hydroxyl radicals.25 We thus investigated CLIC5A’s impact on lysosome-mediated cell death by modulating lysosomal Fe2+ availability prior to hydrogen peroxide exposure by pre-treating cells with the iron chelator deferoxamine mesylate salt (DFO).26,27 As shown in Figure 5B, cells treated with DFO prior to peroxide showed a drastic reduction in apoptosis. The residual apoptotic activity observed in DFO-treated cells could be driven by DNA damage,28 which appears to be CLIC5A-independent given the results obtained with the two DNA damaging agents camptothecin and doxoru1538

bicin (Figure 4C).

Inversely, we used ferric ammonium citrate (FAC) to positively modulate Fe2+ concentration in lysosomes.29 Increased lysosomal Fe2+ concentration favors hydroxyl radical production from hydrogen peroxide and hence is expected to increase LMP and apoptosis. Accordingly, cells pre-treated with FAC showed increased peroxide-induced apoptosis and the protective effect of CLIC5A overexpression was completely lost (Figure 5C), indicating that CLIC5A plays a role upstream of LMP. This specific role further corroborates CLIC5A’s inability to protect cells against DNA damage.

To evaluate CLIC5A’s ion channel activity30 in this context we pre-treated CLIC5A overexpressing Reh cells with indanyloxyacetic acid 94 (IAA-94), a known CLIC-specific ion channel inhibitor.30,31 IAA-94 treatment increased peroxide-induced apoptosis and completely prevented CLIC5A-mediated protection (Figure 5D), similar to that which was observed in the presence of increased Fe2+ availability following FAC treatment (Figure 5C).

Endogenous CLIC5A inhibition contributes to this phenotype and explains the increased apoptosis level of control cells. Together, these data confirm that CLIC5A’s ion channel activity is required to protect cells against peroxide-induced apoptosis, perhaps by limiting Fe2+ availability in the lysosomal pathway.

To further investigate the functional implications of CLIC5A in lysosomal apoptosis, we examined the intracellular localization of CLIC5A using immunofluorescence. Although we did not observe colocalization of CLIC5A with lysosomes (Online Supplementary Figure S7), we did show positive colocalization with transferrin receptor (Figure 6), which is in line with CLIC5A's postulated role in modulating lysosomal Fe2+ availability.

Transferrin receptors (TFRC gene) are responsible for celhaematologica | 2016; 101(12)

CLIC5 in childhood leukemia

A

B

C

D

E

F

Figure 4. CLIC5A protects cells from hydrogen peroxide-induced apoptosis. (A) pLENTI control and CLIC5A stably infected Reh cells were challenged for 20h with 35mM hydrogen peroxide (PRXD), 250nM camptothecin (CPT), or 150nM doxorubicin (DOXO) and analyzed by flow cytometry with Alexa Fluor 488-coupled Annexin V and propidium iodide (PI) staining. (B) A representative example of staining following PRXD treatment is presented with PI on the x axis and Alexa Fluor 488 on the y axis. (C) Total apoptosis is calculated for each sample. CLIC5A overexpressing cells displayed reduced apoptosis compared to control only when treated with PRXD. Each experiment was performed 3 times in triplicates (n=9). (D) IM9 cells transduced with CLIC5A or pLENTI empty vector control were challenged with 50mM PRXD and processed similarly to assess apoptosis. (E) A representative example of staining and (F) total apoptosis is shown for the IM9 cell line. Again, CLIC5A overexpression leads to reduced apoptosis. Two independent experiments carried out in triplicate and an additional single test were performed (n=7). Error bars in (C) and (F) represent the standard deviation. Statistical significance is calculated by two-tailed Mann-Whitney U test. For (A) and (D) adjustments of brightness and contrast were applied to the whole image. DMSO: dimethyl sulfoxide.

lular iron intake through the binding and internalization of its iron-bound ligand transferrin (TF gene).32,33 CLIC5A could perturb this process and thereby impact cellular iron concentrations and modulate lysosome sensitivity to hydrogen peroxide. Interestingly, double positive staining appears to be particularly present in distinct vesicle-like structures that are likely transferrin receptor-containing recycling endosomes, further suggesting that CLIC5A may interfere with normal iron homeostasis, thus reducing lysosome sensitivity to oxidative stress.

Taken together, our data strongly support a role for the newly identified ETV6 transcriptional target CLIC5A in modulating lysosome-mediated apoptosis. We propose that CLIC5A overexpression in t(12;21)-positive, ETV6 depleted pre-B cells could contribute to increased resistance to oxidative stress and therefore promote cell survival.

Discussion Approximately 20% of childhood pre-B ALL patients harbor the t(12;21) translocation, yielding an ETV6-AML1 fusion protein that is, however, insufficient to initiate leukemia.8,34,35 This process often requires further loss of the remaining wild-type ETV6 allele,7,9-11,13 suggesting that deregulation of the ETV6 transcriptional machinery could play an important role in leukemogenesis. Unfortunately, haematologica | 2016; 101(12)

very few ETV6-regulated transcriptional targets are known and the mechanisms through which they are involved in leukemogenesis remain elusive. Herein, we combined both in vitro (cell lines) and ex vivo (pre-B ALL patients) transcriptome data to identify candidate ETV6 target genes, and through additional cellular assays identified 4 novel direct ETV6 target genes: CLIC5, BIRC7, ANGPTL2 and WBP1L. The CLIC5 gene was previously associated with the t(12;21)-positive ALL molecular signature,36 and our pediatric pre-B ALL expression data corroborated this result with strong CLIC5 overexpression shown to be specific to the t(12;21)-positive subgroup within our cohort. Thus CLIC5 overexpression in these patients is likely due to ETV6 loss. Unfortunately, very little is known about CLIC5's potential contribution to leukemogenesis. Using engineered cell lines, we demonstrated an increased resistance to hydrogen peroxide-induced apoptosis following overexpression of CLIC5A. Notably, Reh cells already express the endogenous CLIC5A isoform, which could explain the modest effect of the overexpression of CLIC5A (mean=7.46%, P-value=8.32x10-11, merged n=36).

Given that normal B-cells show no expression of CLIC5 (mean FPKM<0.1 over our 3 normal CD19+/CD10+ pre-B cell samples isolated from human cord blood), an eventual stronger impact of CLIC5A re-expression in a pre-B cell upon ETV6 depletion can be expected.

We further linked this phenotype to CLIC5Aâ&#x20AC;&#x2122;s ion chan1539

B. Neveu et al.

A

B

C

D

Figure 5. CLIC5A is implicated in lysosome-mediated cell death. (A) Schematic illustration of the lysosomal apoptosis pathway and DNA damage pathway. DFO: deferoxamine mesylate salt; FAC: ferric ammonium citrate; LMP: Lysosomal membrane permeabilization; MOMP: Mitochondrial outer membrane permeabilization.

(B) pLENTI control and CLIC5A overexpressing Reh cells were pre-treated with 10mM DFO to chelate lysosomal ferrous iron and apoptosis was induced using 35mM PRXD for 20h followed by flow cytometry quantification. PRXD-induced apoptosis was greatly reduced with DFO treatment. (C) Similarly, 1mM FAC was used to increase ferrous iron concentration which led to higher PRXD-induced apoptosis with no protective effect of CLIC5A overexpression. (D) Cells were pre-treated with 50mM IAA-94 CLIC-specific ion channel inhibitor and apoptosis was assayed after a subsequent PRXD treatment. In these conditions, CLIC5A overexpression did not reduce apoptosis. Each experiment was carried out 3 times in triplicate (n=9). Error bars represent the standard deviation. Statistical significance is calculated by two-tailed Mann-Whitney U test.

nel activity in lysosomal-mediated cell death. The presence of CLIC5A on transferrin receptor-containing endosomes potentially negatively impacts lysosomal iron availability, thus reducing lysosome sensitivity to oxidative stress. However, the exact mechanism by which CLIC5A modulates lysosomal iron to prevent peroxide-induced apoptosis remains unclear. It has been demonstrated that changes in the concentration of several ions can modulate endosomal pH through ion dependent H+ pumps.37 With CLIC5A being a poorly selective ion channel,38 we can hypothesize a somewhat similar function: slight differences in endosome acidification could modulate transferrin iron release or trafficking and therefore impact lysosomal iron concentration leading to differential peroxide sensitivity.

Additional experiments should be performed to further dissect CLIC5Aâ&#x20AC;&#x2122;s role in lysosome-mediated apoptosis.

Nonetheless, the observed effects of CLIC5A in peroxide resistance could play a key role in ALL initiation.

A recent study highlighted the high oxidative stress levels of leukemic blasts in the bone marrow niche induced by bone marrow stromal cell signaling.39 Furthermore, it has been shown that increased levels of ROS in t(12;21)-positive cells was associated with DNA damage 1540

accumulation.40 High oxidative stress levels are thought to trigger apoptotic cell death through the lysosomal pathway, thus preventing DNA damage accumulation.

However, in t(12;21)-positive ALL, we have shown that loss of ETV6 expression leads to significant overexpression of CLIC5A. This overexpression leads to decreased lysosome-media-ted apoptosis, which in turn can promote a more permissive environment to heighten ROS levels (Figure 7). Cells evading apoptosis can thus accumulate ROS-induced mutations at a greater rate. Moreover, this mutational process of t(12;21)-positive pre-B cells could be facilitated not only by CLIC5A overexpression, but together with several ETV6 deregulated targets such as the inhibitor of apoptosis BIRC7. Although the overall mutational burden of ALL is low compared to other cancers,41 the difference in ROS-induced DNA damage accumulation can contribute, over time, to promote leukemic transformation when impacting key cancer genes.

It remains challenging, however, to evaluate the contribution of this pathway to the total amount of DNA damage found in pre-B ALL patients. Although ROS-mediated alterations of nucleotides are well characterized, their signature in sequencing data remains unclear.42,43 Interrogation of our patient sequencing data did not reveal haematologica | 2016; 101(12)

CLIC5 in childhood leukemia

Figure 6. Colocalization of CLIC5A and transferrin receptors. pLENTI control and CLIC5A overexpressing Reh cells were used for co-localization studies.

Immunostaining of both CLIC5A and transferrin receptor were performed in the same cells simultaneously with Hoechst DNA staining. Results were obtained at 100X magnification (upper panels; scale bar = 10mm). A strong colocalization was observed between CLIC5A and transferrin receptors. Additional enlargement for the marked region of the initial image is presented in lower panels (scale bar = 2mm). A merged image was generated (right panels).

any particular mutational profile that could undoubtedly be attributed to ROS. However, increased ROS-mediated DNA double strand breaks have been observed in t(12;21)positive B-cells when assessed by comet assays.40 These alterations do often lead to deletions or rearrangements due to aberrant repair, that are frequent events in leukemia, thus supporting a role for ROS-induced DNA damage in ALL. With a similar CLIC5A-mediated protective effect on peroxide-induced apoptosis obtained in both Reh and IM9 cell lines, we demonstrated a phenotype that is independent from a particular genetic background.

Interestingly, CLIC5 expression has been associated with haematologica | 2016; 101(12)

poor prognosis in breast cancer,44 and expression arrays of a broad range of normal and tumoral tissues obtained through the GENT database45 showed an overexpression of CLIC5 in ovarian cancers (Online Supplementary Figure S8).

Based on our results, and given the importance of oxidative stress resistance in these solid tumors that require angiogenesis to promote growth, it suggests that CLIC5A protection against oxidative stress may not be limited to pre-B cell leukemia. The unfavorable prognosis associated with CLIC5 deregulation may be due to increased oxidative stress resistance, thus reducing the necessity of angiogenesis and fostering an environment prone to ROS-induced DNA damage. Although ETV6 alterations have not been 1541

B. Neveu et al.

Figure 7. Proposed mechanism of CLIC5A involvement in ETV6-associated childhood pre-B ALL. The t(12;21) translocation occurs early and contributes to an increased level of reactive oxygen species (ROS). Lysosome-mediated apoptosis is sensitive to this excess of ROS and therefore prevents the accumulation of ROS-mediated DNA damage. With the subsequent deletion of the residual ETV6 allele (LOH), CLIC5A expression is drastically upregulated. This overexpression of CLIC5A has a negative impact on the lysosomal apoptosis pathway and thus creates a permissive environment for the accumulation of mutations driven by high oxidative stress. Over time, some of these mutations may impact key cellular biological processes and pathways, which will ultimately lead to full leukemic transformation and development of childhood pre-B acute lymphoblastic leukemia (pre-B ALL).

reported in these solid tumors, other mechanisms could drive CLIC5 overexpression. Interestingly, ETS factor binding sites (EBS) were found to be highly enriched for small non-coding mutations across a wide variety of cancers.46 These mutations disrupt the consensus binding site of ETS factors, such as ETV6, and thus may prevent it's binding and the repression of its targets. Despite the presence of wild-type ETV6 in these cases, some of its key targets (CLIC5, BIRC7 or ANGPTL2) may be overexpressed, thus leading to a selective advantage.

While we focused on CLIC5, the other identified targets were also of interest. Notably, the caspases inhibitor BIRC7 has previously been shown to be part of a t(12;21)-positive pre-B ALL molecular signature.36 Overexpression of BIRC7 was also observed in a variety of tumors, and contributes to oncogenesis through the inhibition of apoptosis,47 suggesting a similar involvement for BIRC7 in t(12;21)-positive childhood pre-B ALL. The ANGPTL2 gene encodes a secreted protein with, among others, pro-angiogenic and anti-apoptotic properties.

Although its expression has been linked to a broad range of diseases, including cancers, it has been shown to increase survival and expansion of hematopoietic stem cells.48 ANGPTL2’s contribution to leukemogenesis remains unknown. Lastly, WBP1L, also known as OPAL1 (Outcome Predictor in Acute Leukemia 1), is associated with the t(12;21) favorable outcome in ALL.49 It is still unclear whether WBP1L plays a role in leukemia, since its molecular function is yet to be characterized.

In conclusion, for the first time, we describe a role for CLIC5-mediated resistance to oxidative stress that could

References 1. Lopez RG, Carron C, Oury C, Gardellin P, Bernard O, Ghysdael J. TEL is a sequencespecific transcriptional repressor. J Biol Chem. 1999;274(42):30132-30138.

1542

promote cell survival and contribute to leukemogenesis.

We propose a mechanism in which complete loss of wildtype ETV6 expression in a pre-leukemic blast leads to CLIC5A overexpression, thus creating a permissive environment for the accumulation of mutations driven by high oxidative stress, eventually giving rise to full leukemic transformation.

Acknowledgments The authors would like to thank the patients and their parents for participating in this study. We are grateful to Dr. Christian Beauséjour who kindly provided lentiviral-related material.

Funding This study was supported by research funds provided by the Terry Fox Research Institute and the Canadian Institutes of Health Research. BN is the recipient of a Cole Foundation scholarship. JFS is the recipient of a Réseau de médecine génétique appliquée (RMGA) scholarship. DS holds the FrançoisKarl Viau Research Chair in Pediatric Oncogenomics. Whole transcriptome sequencing was performed at the Integrated Clinical Genomic Center in Pediatrics at the CHU Sainte-Justine Research Center. Computations were performed on the Briarée supercomputer at the Université de Montréal, managed by Calcul Québec and Compute Canada. The operation of this supercomputer is funded by the Canada Foundation for Innovation (CFI), NanoQuébec, RMGA and the Fonds de recherche du Québec - Nature et technologies (FRQNT). GH thanks the CFI, the Cole Foundation and the Fonds de recherche du Québec-Santé for infrastructure, transition award and salary support, respectively.

2. Wang LC, Swat W, Fujiwara Y, et al. The TEL/ETV6 gene is required specifically for hematopoiesis in the bone marrow. Genes Dev. 1998;12(15):2392-2402.

3. Bohlander SK. ETV6: a versatile player in leukemogenesis. Semin Cancer Biol.

2005;15(3):162-174.

4. Golub TR, Barker GF, Bohlander SK, et al.

Fusion of the TEL gene on 12p13 to the AML1 gene on 21q22 in acute lymphoblastic leukemia. Proc Natl Acad Sci USA. 1995; 92(11):4917-4921.

5. Tasian SK, Loh ML, Hunger SP. Childhood acute lymphoblastic leukemia: Integrating

haematologica | 2016; 101(12)

CLIC5 in childhood leukemia

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

genomics into therapy. Cancer. 2015; 121(20):3577-3590.

Agape P, Gerard B, Cave H, et al. Analysis of ETV6 and ETV6-AML1 proteins in acute lymphoblastic leukaemia. Br J Haematol.

1997;98(1):234-239.

Poirel H, Lacronique V, Mauchauffe M, et al. Analysis of TEL proteins in human leukemias. Oncogene. 1998;16(22):28952903.

Mori H, Colman SM, Xiao Z, et al.

Chromosome translocations and covert leukemic clones are generated during normal fetal development. Proc Natl Acad Sci USA. 2002;99(12):8242-8247.

Patel N, Goff LK, Clark T, et al. Expression profile of wild-type ETV6 in childhood acute leukaemia. Br J Haematol. 2003; 122(1):94-98.

Lilljebjorn H, Soneson C, Andersson A, et al. The correlation pattern of acquired copy number changes in 164 ETV6/RUNX1-positive childhood acute lymphoblastic leukemias. Hum Mol Genet. 2010; 19(16):3150-3158.

Montpetit A, Larose J, Boily G, Langlois S, Trudel N, Sinnett D. Mutational and expression analysis of the chromosome 12p candidate tumor suppressor genes in pre-B acute lymphoblastic leukemia. Leukemia.

2004;18(9):1499-1504.

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.

Anderson K, Lutz C, van Delft FW, et al.

Genetic variegation of clonal architecture and propagating cells in leukaemia. Nature.

2011;469(7330):356-361.

Fenrick R, Wang L, Nip J, et al. TEL, a putative tumor suppressor, modulates cell growth and cell morphology of ras-transformed cells while repressing the transcription of stromelysin-1. Mol Cell Biol.

2000;20(16):5828-5839.

Irvin BJ, Wood LD, Wang L, et al. TEL, a putative tumor suppressor, induces apoptosis and represses transcription of Bcl-XL. J Biol Chem. 2003;278(47):46378-46386.

Dobin A, Davis CA, Schlesinger F, et al.

STAR: ultrafast universal RNA-seq aligner.

Bioinformatics. 2013;29(1):15-21.

Anders S, Pyl PT, Huber W. HTSeq--a Python framework to work with highthroughput sequencing data.

Bioinformatics. 2015;31(2):166-169.

Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26 (1):139-140.

Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001;25(4):402408.

Shanks RA, Larocca MC, Berryman M, et

haematologica | 2016; 101(12)

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

al. AKAP350 at the Golgi apparatus. II. Association of AKAP350 with a novel chloride intracellular channel (CLIC) family member. J Biol Chem. 2002;277(43):4097340980.

Flores-Tellez TN, Lopez TV, Vasquez Garzon VR, Villa-Trevino S. Co-Expression of Ezrin-CLIC5-Podocalyxin Is Associated with Migration and Invasiveness in Hepatocellular Carcinoma. PLoS One. 2015;10(7):e0131605.

Xu Y, Kang J, Yuan Z, et al. Suppression of CLIC4/mtCLIC enhances hydrogen peroxide-induced apoptosis in C6 glioma cells.

Oncol Rep. 2013;29(4):1483-1491.

Turk B, Turk V. Lysosomes as "suicide bags" in cell death: myth or reality? J Biol Chem.

2009;284(33):21783-21787.

Aits S, Jaattela M. Lysosomal cell death at a glance. J Cell Sci. 2013;126(Pt 9):1905-1912.

Kruszewski M. Labile iron pool: the main determinant of cellular response to oxidative stress. Mutat Res. 2003;531(1-2):81-92.

Kurz T, Leake A, Von Zglinicki T, Brunk UT. Relocalized redox-active lysosomal iron is an important mediator of oxidativestress-induced DNA damage. Biochem J. 2004;378(Pt 3):1039-1045.

Boya P, Kroemer G. Lysosomal membrane permeabilization in cell death. Oncogene.

2008;27(50):6434-6451.

Oller AR, Thilly WG. Mutational spectra in human B-cells. Spontaneous, oxygen and hydrogen peroxide-induced mutations at the hprt gene. J Mol Biol. 1992;228(3):813826.

Repnik U, Cesen MH, Turk B. The endolysosomal system in cell death and survival. Cold Spring Harb Perspect Biol.

2013;5(1):a008755.

Berryman M, Bruno J, Price J, Edwards JC. CLIC-5A functions as a chloride channel in vitro and associates with the cortical actin cytoskeleton in vitro and in vivo. J Biol Chem. 2004;279(33):34794-34801.

Landry DW, Akabas MH, Redhead C, Edelman A, Cragoe EJ, Jr., Al-Awqati Q. Purification and reconstitution of chloride channels from kidney and trachea. Science.

1989;244(4911):1469-1472.

Bresgen N, Eckl PM. Oxidative stress and the homeodynamics of iron metabolism.

Biomolecules. 2015;5(2):808-847.

Maxfield FR, McGraw TE. Endocytic recycling. Nat Rev Mol Cell Biol. 2004;5(2):121132.

Andreasson P, Schwaller J, Anastasiadou E, Aster J, Gilliland DG. The expression of ETV6/CBFA2 (TEL/AML1) is not sufficient for the transformation of hematopoietic cell lines in vitro or the induction of hematologic disease in vivo. Cancer Genet Cytogenet. 2001;130(2):93-104. van der Weyden L, Giotopoulos G, Rust AG, et al. Modeling the evolution of ETV6RUNX1-induced B-cell precursor acute lymphoblastic leukemia in mice. Blood.

2011;118(4):1041-1051.

36. Ross ME, Zhou X, Song G, et al.

Classification of pediatric acute lymphoblastic leukemia by gene expression profiling. Blood. 2003;102(8):2951-2959.

37. Scott CC, Gruenberg J. Ion flux and the function of endosomes and lysosomes: pH is just the start: the flux of ions across endosomal membranes influences endosome function not only through regulation of the luminal pH. Bioessays. 2011;33(2): 103-110.

38. Singh H, Cousin MA, Ashley RH. Functional reconstitution of mammalian 'chloride intracellular channels' CLIC1, CLIC4 and CLIC5 reveals differential regulation by cytoskeletal actin. FEBS J. 2007;274(24):6306-6316.

39. Liu J, Masurekar A, Johnson S, et al.

Stromal cell-mediated mitochondrial redox adaptation regulates drug resistance in childhood acute lymphoblastic leukemia.

Oncotarget. 2015.

40. Kantner HP, Warsch W, Delogu A, et al.

ETV6/RUNX1 induces reactive oxygen species and drives the accumulation of DNA damage in B cells. Neoplasia.

2013;15(11):1292-1300.

41. Alexandrov LB, Nik-Zainal S, Wedge DC, et al. Signatures of mutational processes in human cancer. Nature. 2013;500(7463):415421.

42. Cooke MS, Evans MD, Dizdaroglu M, Lunec J. Oxidative DNA damage: mechanisms, mutation, and disease. FASEB J. 2003;17(10):1195-1214.

43. Helleday T, Eshtad S, Nik-Zainal S. Mechanisms underlying mutational signatures in human cancers. Nat Rev Genet.

2014;15(9):585-598.

44. Yau C, Sninsky J, Kwok S, et al. An optimized five-gene multi-platform predictor of hormone receptor negative and triple negative breast cancer metastatic risk.

Breast Cancer Res. 2013;15(5):R103.

45. Shin G, Kang TW, Yang S, Baek SJ, Jeong YS, Kim SY. GENT: gene expression database of normal and tumor tissues. Cancer Inform. 2011;10:149-157.

46. Araya CL, Cenik C, Reuter JA, et al.

Identification of significantly mutated regions across cancer types highlights a rich landscape of functional molecular alterations. Nat Genet. 2015.

47. Wang L, Zhang Q, Liu B, Han M, Shan B. Challenge and promise: roles for Livin in progression and therapy of cancer. Mol Cancer Ther. 2008;7(12):3661-3669.

48. Thorin-Trescases N, Thorin E. Angiopoietin-like-2: a multifaceted protein with physiological and pathophysiological properties. Expert Rev Mol Med. 2014; 16:e17.

49. Holleman A, den Boer ML, Cheok MH, et al. Expression of the outcome predictor in acute leukemia 1 (OPAL1) gene is not an independent prognostic factor in patients treated according to COALL or St Jude protocols. Blood. 2006;108(6):1984-1990.

1543

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Lymphoblastic Leukemia

Ferrata Storti Foundation

A sequential approach with imatinib, chemotherapy and transplant for adult Ph+ acute lymphoblastic leukemia: final results of the GIMEMA LAL 0904 study

Sabina Chiaretti,1* Antonella Vitale,1* Marco Vignetti,2 Alfonso Piciocchi,2 Paola Fazi,2 Loredana Elia,1 Brunangelo Falini,3 Francesca Ronco,4 Felicetto Ferrara,5 Paolo De Fabritiis,6 Mario Luppi,7 Giorgio La Nasa,8 Alessandra Tedeschi,9 Catello Califano,10 Renato Fanin,11 Fausto Dore,12 Franco Mandelli,2 Giovanna Meloni1 and Robin Foà1

Hematology, Department of Cellular Biotechnologies and Hematology, Policlinico Umberto 1, “Sapienza” University of Rome; 2GIMEMA Data Center, Rome; 3Institute of Hematology, University of Perugia; 4Hematology Unit, Azienda Ospedaliera Bianchi Melacrino Morelli, Reggio Calabria; 5Division of Hematology and Stem Cell Transplantation Unit, Cardarelli Hospital, Naples; 6Hematology Unit and Department of Pharmacy Services, Sant'Eugenio Hospital, Rome; 7Hematology Division, Department of Medical and Surgical Sciences, University of Modena and Reggio Emilia, Modena; 8 Hematology Unit, Department of Medical Sciences, University of Cagliari; 9Department of Oncology/Hematology, Niguarda Cancer Center, Niguarda Ca' Granda Hospital, Milano; 10Oncohematology Unit A. Tortora Hospital, Pagani, Salerno; 11Division of Hematology and Bone Marrow Transplantation, University Hospital, Udine and 12 Department of Biomedical Sciences, University of Sassari, Italy

1

Haematologica 2016 Volume 101(12):1544-1552

*SC and AV contributed equally to this work. Presented in part at the 18th EHA Congress (2013).

ABSTRACT

Correspondence: rfoa@bce.uniroma1.it or chiaretti@bce.uniroma1.it

Received: February 16, 2016.

Accepted: August 9, 2016.

Pre-published: August 11, 2016. doi:10.3324/haematol.2016.144535

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/12/1544

©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.

1544

I

n the GIMEMA LAL 0904 protocol, adult Philadelphia positive acute lymphoblastic leukemia patients were treated with chemotherapy for induction and consolidation, followed by maintenance with imatinib. The protocol was subsequently amended and imatinib was incorporated in the induction and post-remission phase together with chemotherapy. Due to the toxicity of this combined approach, the protocol was further amended to a sequential scheme based on imatinib plus steroids as induction, followed by consolidation with chemotherapy plus imatinib and, when applicable, by a hematopoietic stem cell transplant. Fifty-one patients (median age 45.9 years) were enrolled in the final sequential protocol. At the end of induction (day +50), 96% of evaluable patients (n=49) achieved a complete hematologic remission; after consolidation, all were in complete hematologic remission. No deaths in induction were recorded. Overall survival and disease-free survival at 60 months are 48.8% and 45.8%, respectively. At day +50 (end of imatinib induction), a more than 1.3 log-reduction of BCR-ABL1 levels was associated with a significantly longer disease-free survival (55.6%, 95%CI: 39.0-79.3 vs. 20%, 95%CI: 5.8-69.1; P=0.03), overall survival (59.1%, 95%CI: 42.3-82.6 vs. 20%, 95%CI: 5.8-69.1; P=0.02) and lower incidence of relapse (20.5%, 95%CI: 7.2-38.6 vs. 60.0%, 95%CI: 21.6-84.3; P=0.01). Mean BCR-ABL1 levels remained significantly higher in patients who subsequently relapsed. Finally, BCRABL1p190 patients showed a significantly faster molecular response than BCR-ABL1p210 patients (P=0.023). Though the study was not powered to evaluate the role of allogeneic stem cell transplant, allografting positively impacted on both overall and disease-free survival. In conclusion, a sequential approach with imatinib alone in induction, consolidated by chemotherapy plus imatinib followed by a stem cell transplant is a feasible, well-tolerated and effective strategy for adult Philadelphia positive acute lymphoblastic leukemia, leading to the best long-term survival rates so far reported. (clinicaltrials.gov identifier: 00458848). haematologica | 2016; 101(12)

Final results of the GIMEMA LAL 0904 for Ph+ ALL

Introduction

Methods Study design and therapy

The Philadelphia (Ph) chromosome represents the most frequent cytogenetic alteration in adult acute lymphoblastic leukemia (ALL). The incidence of the Philadelphia positive (Ph+) ALL increases with age, from approximately 2%-5% in children/adolescents, to 22% among adult patients aged 21-50 years, and to over 50% in patients over 50 years of age.1-4 The presence of the Ph chromosome has historically defined a subgroup of ALL with a particularly unfavorable prognosis. The advent of tyrosine kinase inhibitors (TKI) profoundly changed the management and prognosis of this high-risk group of patients, and the management of Ph+ ALL is described as pre- or post the TKI era.5-21 Prior to the introduction of TKI, prognosis was very poor, with virtually no adult patients (<5%) cured with standard chemotherapy; median survival was 8-10 months unless an allogeneic hematopoietic stem cell transplant (allo-SCT), the only potentially curative strategy, could be performed.22-25 Today, treatment with TKI, with5-10,12-15,18-21 or without11,16,17 systemic chemotherapy, represents the most appropriate first-line management of patients with Ph+ ALL in terms of rates of complete hematologic remission (CHR) and disease-free survival (DFS).

Imatinib has been incorporated into different schedules either in induction5-7,10,12-14,18-21 or following induction8,10 in cohorts including also elderly patients,8 with CHR rates varying from 72% to 96%. Moreover, the use of imatinib can also act as a “bridge” to an allo-SCT for those patients eligible.26-30 In the first TKI-based GIMEMA protocol (LAL 0201), imatinib alone (plus steroids) was used as induction treatment for elderly (>60 years) Ph+ ALL patients.11 The results showed for the first time that a treatment strategy based on a TKI alone [plus steroids and central nervous system (CNS) prophylaxis] and no systemic chemotherapy was associated with a CHR in virtually all elderly patients with no deaths in induction;11 some patients are alive ten years later (S Chiaretti, personal data, 2016). These data provided the starting point for the design of the subsequent GIMEMA LAL 1205 protocol, based on the use of the second-generation TKI dasatinib alone for 12 weeks as firstline induction treatment for all Ph+ ALL over 18 years of age, with no upper age limit; all evaluable patients obtained a CHR with an overall good compliance and no deaths and relapses during induction.16 In the GIMEMA LAL 1205 protocol, post-induction therapy was left to the investigator’s choice. The issue that still remained after this study was how best to consolidate patients who were in CHR following TKI induction, and this is being addressed in the current dasatinib-based GIMEMA 1509 total therapy protocol.17 The GIMEMA LAL 0904 trial was initially designed for both Philadelphia negative (Ph–) and Ph+ ALL cases.

When available, imatinib was added for Ph+ ALL patients. The protocol underwent different amendments (see Methods). In the last amended protocol (3rd amendment), imatinib plus steroids was used as induction without systemic chemotherapy, followed by a uniform intensive consolidation chemotherapy plus imatinib and a subsequent hematopoietic SCT [allo-SCT or autologous (auto)-SCT if a donor was not available]. We report the final results of the 3rd amendment for the management of adult Ph+ ALL. haematologica | 2016; 101(12)

Between October 2004 and April 2010, 100 patients with de novo Ph+ ALL, aged 15-60 years, were enrolled in the GIMEMA 0904 protocol. In the first version of the protocol, Ph+ patients received chemotherapy as induction and consolidation. When imatinib became available, it was administered as maintenance.

The protocol underwent a first (1st) amendment based on the use of a novel asparaginase formulation. The protocol was subsequently amended a second time (2nd) and imatinib was incorporated into the induction and post-remission phase. This scheme was associated with an unacceptably high rate of induction deaths: 2 deaths were recorded (due to fungal infection and liver toxicity, respectively) among the first 9 patients enrolled. The protocol underwent a final amendment (3rd) and imatinib (600 mg/day for 50 days) plus steroids (60 mg/m2/day) were used as induction, followed by a HAM cycle plus imatinib (Figure 1 and Online Supplementary Appendix) and a hematopoietic SCT, either allogeneic or autologous if no donor was available. A donor search was carried out first among family members, and if a related donor was not available an alternative donor search was made according to the policy of each center.

The enrollment periods, centers, overall survival (OS) and DFS of the 3 amendments are detailed in the Online Supplementary Appendix, while the results presented here refer to the patients enrolled in the 3rd amendment only. The objective of this study was to evaluate the role of imatinib as induction treatment followed by a uniform consolidation treatment based on the HAM regimen and a transplant procedure, when possible.

The study was approved by the Ethics Committees of all participating centers and all patients gave their written informed consent in accordance with the Declaration of Helsinki.

Response assessment Bone marrow (BM) evaluations and molecular monitoring were performed at baseline, at day +35, +50 (end of induction) and post consolidation. Patients were considered in CHR in the presence of 5% or less BM blasts, absence of blasts in the peripheral blood (PB), no extramedullary involvement and a full PB count recovery (i.e. polymorphonucleates >1.5x109/L and platelets >100x109/L).

Steroid response was based on the PB blast reduction (threshold ≥75%) after the steroid pre-phase. Hematologic relapse was defined as the presence of blasts in the PB or any non-hematologic site, or 5% or more BM blasts.

Molecular diagnosis and minimal residual disease monitoring Molecular analyses were performed at “La Sapienza” University of Rome, Italy. Total RNA was extracted from BM samples using the TRizol reagent or the Qiagen extraction kit. A reverse transcriptase multiplex polymerase chain reaction (RT-PCR-multiplex)31 was performed to detect the p190 or p210 forms of the BCR-ABL1 fusion product within the 7-day steroid pre-phase.

Minimal residual disease (MRD) monitoring was performed by quantitative real-time PCR (Q-RT-PCR).32,33 BCR-ABL1 transcript levels were normalized to the number of the ABL1 control gene and expressed as BCR-ABL1/ABL1 x100; the level of BCR-ABL1 expression was then converted into a base 10 logarithmic scale. A complete molecular response was defined as a BCR-ABL1/ABL1 ratio equal to zero.

Statistical analysis Overall survival and DFS were estimated using the KaplanMeier method. Cumulative incidence of relapse (CIR) was calcu1545

S. Chiaretti et al.

Imatinib maintenance

lated using the cumulative incidence method (Online Supplementary Appendix).

A 1.3 log-reduction cut off was chosen on the martingale residual analysis on the univariate Cox model; an increasing smoothed martingale residual plot indicated a prognostic effect of log BCRABL1 reduction levels on DFS. The statistical significance for reduction of BCR-ABL1 levels was assessed using the MannWhitney test.

The log-rank test was used to compare risk-factor categories for the Kaplan-Meier curves and the Gray test for the incidence curves. Multivariate analysis was performed by the Cox model; results were expressed as Hazard Ratios (HR) Âą 95% confidence intervals (95%CI). The role of transplant was evaluated in a Cox model with a time-dependent covariate. All tests were two-sided; P<0.05 was considered significant. Analyses were performed using SAS v.9.4 software (SAS Institute, Cary, NC, USA).

Results Patients From July 2007 to April 2010, 51 adult Ph+ ALL were enrolled in the 3rd amended protocol; 28 were females and 23 males. Median age was 45.9 years (range 16.9-59.7) and median white blood cell (WBC) count was 28.0x109/L (range 1.4-597.0). Thirty-nine patients had the p190 form of BCR-ABL1, 7 the p210 form, and 5 both p190 and p210 (for all analyses, these latter two groups were considered together).

Induction treatment response and toxicity Of the 51 patients enrolled, 49 were evaluable for response, while 2 discontinued treatment for medical decision. After the steroid pre-phase, 79% of patients showed a PB blast reduction of 75% or over and 21% less than 1546

Figure 1. Schematic representation of the GIMEMA 0904 3rd amendment. Steroid prephase: oral prednisone at increasing doses (1060 mg/m2/day) for seven days. Induction therapy: oral imatinib at a dose of 600 mg daily for 50 days; prednisone (60 mg/m2/day) until day +24, then tapered and stopped at day +32; intrathecal methotrexate (15 mg) on days +21 and +35.

Consolidation treatment: HAM regimen plus oral imatinib at a dose of 600 mg daily. Post consolidation: a hematopoietic SCT, either allogeneic or autologous (if no donor was available) was offered; otherwise, patients continued treatment with imatinib. Patient flow-chart is also provided. pts: patients; CR: complete remission; Progr: progression; allo-SCT: allogeneic stem cell transplantation; Auto-SCT: autologous stem cell transplantation; AraC: cytarabine.

75%. At the end of the induction treatment with imatinib plus steroids, 47 of 49 evaluable patients (96%) achieved a CHR; of the remaining 2 patients, 1 had a partial response and the other was refractory to induction and both were rescued with the HAM chemotherapy scheme and achieved a CHR. Overall, treatment was well tolerated; adverse events (AE) were recorded in 24 patients and were grade 3 or over only in 7 (Table 1). The most frequent AE were represented by grade 1-2 gastrointestinal disorders (n=6) and increase in alanine/aspartate aminotransferase levels (n=4). No deaths were recorded during the induction phase.

Post-remission treatment Of the 47 patients in CHR after the induction phase with only imatinib plus steroids, 43 performed the planned consolidation therapy with the HAM regimen. Of the remaining 4 patients, 1 refused further treatment and 3 did not undergo the treatment scheduled by medical decision: 2 of 3 these patients directly underwent an alloSCT without additional chemotherapy and the other continued treatment with imatinib.

Both patients who achieved a CHR after HAM underwent an allo-SCT; one after receiving a consolidation cycle with high-dose cytarabine (HD-ARA-C) and idarubicin, as per protocol guidelines, while the other proceeded directly to transplant.

After the planned consolidation chemotherapy, 20 patients underwent an allo-SCT with a myeloablative conditioning regimen [8 siblings, 10 matched unrelated donor (MUD), 2 haploidentical]. Twenty-four patients did not receive a transplant for the following reasons: medical decision in 8 patients, relapse in 7, toxicity in 5, refusal in 1, and no donor in 3. An auto-SCT was performed in the 3 patients for whom no donor was available. haematologica | 2016; 101(12)

Final results of the GIMEMA LAL 0904 for Ph+ ALL

The induction and post-remission results are shown in Figure 1.

Considering the entire cohort, including also the patients (n=5) who did not perform the planned therapy, 9 patients have died in CHR, either due to complications following chemotherapy (6 of 26, 23.07%) or allo-SCT (3 of 23, 13.04%). Among patients who received only chemotherapy, 3 experienced a fatal infection: 1 had a hemorrhage and 1 a fatal neurotoxicity, while cause of death is unknown for the other. Among allografted patients, 1 experienced multiorgan failure, 1 died of infection and 1 of widespread graft-versus-host disease.

Minimal residual disease monitoring BCR-ABL1 transcript levels decreased during the imatinib plus steroids induction therapy. The highest reduction was observed between the onset and day +35 of therapy (P<0.0001), while reduction did not reach statistical significance (P=0.393) between days +35 and +50, thus indicating that the greatest activity of treatment in terms of MRD reduction is observed during the first four weeks of treatment. Consolidation chemotherapy (HAM regimen) induced a further reduction of the BCR-ABL1 levels compared to those obtained at the end of induction, even if the log-reduction was not statistically significant (P=0.103) (Figure 2). In particular, a disease reduction greater than 1.3 log was observed in 72% of patients at day +35, in 72% of patients at day +50, and in 90% of patients after the post-consolidation therapy; 3% of patients achieved a complete molecular remission at day +50 of the current protocol.

Mean BCR-ABL1 levels remained significantly higher in patients who subsequently relapsed compared with those who remained free from relapse at the last follow up (P=0.038). BCR-ABLp190 patients showed a more rapid molecular response compared to BCR-ABLp210 patients and this difference was statistically significant at day +50 (end of induction) (P=0.023).

Relapses

Q-PCR levels (log scale)

At the last follow up (median 51.8 months, range 2.775.3), 17 relapses had occurred [11 hematologic, 5 central nervous system (CNS) and 1 not defined], all after completing the induction phase. In those patients who received the planned therapy, 5 relapses occurred in the

haematologica | 2016; 101(12)

20 allo-SCT patients (25%), 11 in the 21 non-transplanted group of patients (52.4%), and an additional relapse was recorded among the 4 patients (25%) who did not undergo consolidation therapy. Overall, the median time to relapse from achievement of 1st CHR was 7.6 months (range 2.1-32.9): this was 3.5 months (range 2.1-32.9) in non-transplanted patients, while in transplanted cases it was 8.6 months (range 5.7-12.2). In both subgroups, no relapse was observed later than 33 months.

Among relapsed patients, there was a statistically significant difference in median WBC count at diagnosis between relapsed versus non-relapsed cases (59.2x109/L vs.

19.5x109/L; P=0.01). Furthermore, relapse was associated with female sex (5 men and 12 women in relapsed cases vs. 18 men and 14 women among non-relapsed cases), although this was not significant (P=0.07). Finally, fewer relapses [not significant (n.s.)] occurred in patients with the BCR-ABL1p190 form (12 of 39, 30.7%) than in patients with the BCR-ABL1p210 form (5 of 12, 41.7%). We observed no differences in age among the two groups (median age 43.0 vs. 46.3 years, respectively).

We also analyzed the impact of MRD in relation to

Table 1. Adverse events in induction.

System organ class

Grade Grade Patients 1-2 â&#x2030;Ľ3 with grade â&#x2030;Ľ3 AE

Gastrointestinal disorders Vascular disorders, thrombosis Nervous system disorders Cardiac disorders Respiratory, thoracic and mediastinal disorders General disorders Skin and subcutaneous tissue disorders Infections and infestations Laboratory investigations Alanine and aspartate aminotransferase Fibrinogen increased Fibrinogen decreased

6 1 2 1 4 5 4 1 -

1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1

AE: adverse events.

Figure 2. Minimum residual disease (MRD) monitoring during treatment. MRD monitoring was performed by quantitative real-time PCR (Q-RT-PCR) and BCR-ABL1 expression levels converted into a logarithmic (base 10) scale. A highly significant (P<0.0001) disease reduction was observed between the onset and day (d) +35, and an additional decrease [P=not significant (ns)] between days +35 and +50. Consolidation chemotherapy induced a further reduction (P=ns) of the BCR-ABL1 levels compared to those obtained at the end of induction.

1547

relapse and observed a significant difference in terms of CIR between patients who had reached MRD levels below 1.3 log at day +50 and those who did not (60.0%, 95%CI: 21.6-84.3 vs. 20.5%, 95%CI: 7.2-38.6, respectively; P=0.01) (Figure 3).

Disease-free survival and overall survival Median follow up of the study is 51.8 months (range 2.7-75.3). DFS at 60 months is 45.8% (95%CI: 33.6-62.5), with a median DFS of 40.1 months (Figure 4A). OS at 60 months is 48.8% (95%CI: 36.4-65.3), with a median OS of 48.8 months (Figure 4B).

Overall survival and DFS were analyzed taking into account MRD levels and post-consolidation treatment (i.e. allo-SCT vs. no allo-SCT). With regard to MRD, DFS was analyzed according to the molecular response at day +50, i.e. the induction end point: estimations at 60 months were 20% (95%CI: 5.8-69.1) for patients with BCR-ABL1 log reduction levels less than 1.3 log and 55.6% (95%CI: 39.0-79.3) for patients with log reduction levels of 1.3 log or over (P=0.03) (Figure 5A). Accordingly, OS was significantly worse (P=0.02) for patients with BCR-ABL1 log reduction levels less than 1.3 log (20%, 95%CI: 5.8-69.1) than for patients with log reduction levels 1.3 log or over (59.1%, 95%CI: 42.3-82.6) (Figure 5B).

Disease-free survival and OS were also analyzed on the basis of the allo-SCT procedure (allo-SCT, n=23; no alloSCT, n=26). Patients undergoing an auto-SCT (n=3) were not considered given the small number of patients. AlloSCT, considered as a time-dependent covariate in patients undergoing HAM therapy as consolidation treatment, impacted on all survival end points (DFS, P=0.06; OS, P=0.03; CIR, P=0.06).

Univariate and multivariate DFS analysis Taking into account in univariate analysis WBC count at diagnosis (as continuous variable), age (as continuous variable), sex, response to the steroid pre-phase, type of BCR-ABL1 transcript, BCR-ABL1 log-reduction at day +35, day +50 and post consolidation, and allo-SCT, a significant correlation was found with DFS for the WBC count (P=0.03), with DFS and OS for response to the steroid prephase (P=0.004 and 0.0035, respectively) and BCR-ABL1 1.3 log-reduction at day +50 (P=0.04 and 0.028); concerning OS, a trend towards significance was only observed according to age (P=0.06).

In multivariate analysis, only response to the steroid pre-phase and a BCR-ABL1 1.3 log-reduction at day +50 correlated with DFS (P=0.002 and P=0.05, respectively) and OS (P=0.002 and P=0.008, respectively).

Discussion We report the final results of the GIMEMA LAL 0904 protocol for adult Ph+ ALL patients. This protocol was initially based on an induction and post-remission phase based on chemotherapy, followed by imatinib administration as maintenance. Subsequently, it was modified to a combined imatinib-chemotherapy induction treatment which, due to toxicity, was amended to a sequential program schedule based on imatinib plus steroids in induction, followed by the HAM chemotherapy regimen and, when applicable, a transplant procedure. The results obtained indicate that not only in the elderly,11 but also in 1548

Probability of relapse

S. Chiaretti et al.

Figure 3. Cumulative incidence of relapse (CIR) on the basis of reduction in minimum residual disease (MRD). Patients were stratified according to the 1.3 log reduction at day +50 (cut-point 1.3 log). CIR was significantly lower in patients with an MRD reduction of 1.3 log or over (continuous line) at day +50 vs. those with an MRD reduction less than 1.3 log (dashed line) (60.0%, 95%CI: 21.6-84.3 vs. 20.5%, 95%CI: 7.2-38.6, respectively; P=0.01). CR: complete remission.

young adult ALL patients (age 15-60 years), imatinib plus steroids alone as induction treatment result in a marked debulking of the neoplastic clone. In fact, at the end of the induction phase, 47 of 49 of the evaluable patients (96%) achieved a CHR and the 2 remaining patients obtained a CHR with the HAM regimen. In terms of CHR, these results compare favorably with those reported by other studies in which imatinib was administered concomitantly to chemotherapy or in various schedules during induction or consolidation, with CHR rates ranging from 72% to 96%.5,6,10,12,13,15,18-21,34,35 Furthermore, our schedule has the advantage that there are fewer deaths during induction treatment, in contrast to the majority of the combination studies in which, with few exceptions,5 toxic deaths were recorded in 2%-7% of cases.6,8,12-14,18-20 Indeed, toxicity was recorded in the initial combination protocol (imatinib+chemotherapy) that led to the final amendment to a sequential strategy. These results indicate that the induction treatment for adult Ph+ ALL can be effectively based on the administration of imatinib plus steroids, without systemic chemotherapy, which enables a CHR to be obtained in virtually all patients with no deaths in induction. The toxicity of the combination of a tyrosine kinase inhibitor (TKI) with conventional chemotherapy has also been reported by the PETHEMA group,19 and more recently by Chalandon et al.36 who have shown that imatinib can be effectively and more safely combined with reduced intensity chemotherapy.

As expected, at day +50 (end of the induction), the molecular disease was still present at low levels in the majority of patients. Mean BCR-ABL1 levels appeared to be more rapidly reduced in p190+ cases than in p210+ patients, as observed in our previous study based on dasatinib,16 confirming a significantly greater susceptibility of BCR-ABLp190-expressing cells to TKI. In our analysis, patients who reached lower levels of disease at day +50 had a significantly better outcome than those with higher levels of residual disease at the end of induction, both in terms of DFS (20.0% vs. 55.6%; P=0.03), reduced CIR haematologica | 2016; 101(12)

Final results of the GIMEMA LAL 0904 for Ph+ ALL

B

Survival probability

Survival probability

A

Figure 4. Survival of the whole study population. Median follow up is 51.8 months.

(A) Disease-free survival (DFS) at 60 months is 45.8% (95%CI: 33.6-62.5), with a median DFS of 40.1 months. (B) Overall survival (OS) at 60 months is 48.8% (95%CI: 36.4-65.3), with a median OS of 48.8 months. CR: complete remission.

(20.5% vs. 60.0%; P=0.01), and OS (20.0% vs. 64.0%; P=0.02) at 60 months. In particular, patients who maintained their remission state over time had achieved significantly lower levels of MRD at day +50 than patients who subsequently relapsed. Therefore, MRD levels post induction or in consolidation is a significant risk factor for treatment failure. These results confirm the observations made in our previous study based on the use of dasatinib in induction,16 and are in line with the current general knowledge on the prognostic impact of MRD.12,19,37-43 Furthermore, they strengthen the notion that MRD negativity should be regarded as a major goal in Ph+ ALL treatment. Indeed, with few exceptions,20,38 it is now well established that BCR-ABL1 transcript levels correlate with response. Lee et al.37 were able to demonstrate that a 3-log reduction in BCR-ABL1 transcripts after one month of imatinib treatment strongly predicted a reduced risk of relapse and confirmed these results in a subsequent study.40 A correlation between MRD levels and outcome has also been reported by Ravandi et al. in patients treated either with imatinib or dasatinib and chemotherapy.41 Two studies44,45 analyzing the outcome of patients with Ph+ ALL who underwent a transplant showed that the persistent expression of BCR-ABL1 during the first 100 days post transplant was associated with a higher incidence of relapse and a lower DFS. Both studies argue in favor of a maintenance therapy with imatinib after transplant in patients with a positive MRD evaluation, and this has also been suggested in more recent studies.46,47 In contrast, Yanada et al.38 observed no association between rapid achievement of BCR-ABL1 negativity and long-term outcome after an initial imatinib/chemotherapy induction regimen, and the GRAAL group also reported that early MRD evaluation did not significantly influence patient outcome, either in terms of OS or DFS.20 One issue that remains open to discussion concerning the role of MRD is represented by the substantial differences in the way PCR for BCR-ABL1 detection is performed, how results are reported in different laboratories worldwide, and the timing of MRD evaluation. The indications suggested by White et al.48 will hopefully lead to an international standardization of the assessment of the levhaematologica | 2016; 101(12)

els of BCR-ABL1. With these premises in mind, the MRD clearance obtained in the current study is inferior to that reported by other groups in which the reported levels of MRD, evaluated at different time points and with different cut-off points, ranged from 26% to 86%.5,7,10,13,18,19 This could be explained by the fact that, in this trial, chemotherapy was not part of the induction treatment.

However, in the 2nd amended version of this trial, where both chemotherapy and TKI were simultaneously administered, 2 deaths in induction were recorded in the first 9 treated patients (1 fungal infection and 1 liver toxicity).

The protocol was, therefore, discontinued and modified into a sequential strategy. In addition, it should be remembered that imatinib has per se a less profound molecular debulking effect compared to dasatinib, which in vitro has a 325-fold greater potency in inhibiting BCR-ABL1.49 In line with this, in our previous GIMEMA LAL 1205 trial16 based on dasatinib administration and steroids as induction (without a uniform post-induction treatment), we documented a more pronounced MRD clearance at the end of induction; in fact, 16% of patients achieved a complete molecular response at day +57 and 20% at day +85 (end of induction) compared with 3% at day +50 in the current protocol.

One of the aims of the present study was to see whether a uniform post-induction therapy was capable of improving outcome. The comparison between this regimen and the dasatinib study leads to two main conclusions. 1) Dasatinib exerts a more potent anti-leukemic activity compared to imatinib, mostly in terms of MRD clearance.

2) It confirms the importance of a uniform consolidation chemotherapy regimen (HAM) and transplant when possible; in fact, this translated into significantly better DFS and OS rates compared to the LAL 1205 protocol in which the post-consolidation phase was open (Online Supplementary Figure S2).

A formal comparison with other studies, with the limitation that the follow-up period is extremely heterogeneous, shows that our data compare favorably in terms of both short- and, in particular, long-term DFS and OS. In fact, in our study, OS and DFS at 24 months were 66.4% and 54.6%, respectively; these results are in line with 1549

S. Chiaretti et al.

1550

tinib). OS and leukemia-free survival (LFS) significantly increased among the 3 categories, being 16% and 11%, 48% and 39%, and 57% and 52%, respectively. In our study, only 3 patients underwent an auto-SCT and it is, therefore, impossible to draw any conclusion on this topic; nevertheless, 2 of 3 patients are in continuous complete remission at 42.8 and 57.2 months, respectively, confirming that this procedure may indeed be effective in this subset of patients. In conclusion, this study shows that in adult Ph+ ALL a sequential strategy based on an induction with ima-

Survival probability

A

B

Survival probability

studies from the MDACC,5 GRAAL8 and GRAAPH-200310 trials, and with a more recent report from the MDACC14 based on dasatinib administration. With regard to the long-term follow up, our data show that the OS and DFS at 60 months are 48.8% and 45.8%, respectively. In the Italian NILG 09/00,12 Spanish CSTIBES02,13 French GRAAL,20 English UKALLXII/ECOG299321 and MDACC35 studies, the 4-5 year OS and DFS rates were 38% and 39%,12 30% for both,13 52% and 44%,20 38% of OS and 50% at four years of relapse-free survival,21 and 43% for both.35 Thus, our long-term outcome results appear superior; this could be due to the fact that we have had no deaths in induction nor major toxicities associated to chemotherapy, which was delivered to patients (96% of cases) already in hematologic remission. All relapses occurred after completing the imatinib plus steroids induction phase and took place during/after the post-consolidation therapy; in fact, the median time to relapse from achievement of the first CHR was 7.6 months. Five of the 17 relapses were at the CNS level. Notably, no CNS relapses were observed in our previous dasatinib-based study, thus confirming the scarce penetration capability of imatinib;50-52 indeed, Pfeifer et al. reported that treatment with imatinib without CNS prophylaxis is associated with meningeal leukemia in 12% of cases.50 In line with this, Takayama et al.51 showed that the concentration of imatinib in the cerebrospinal fluid is roughly 92-fold lower than that in the blood.51 Finally, Porkka et al. reported that dasatinib is capable of increasing survival in a K562 intracranial chronic myeloid leukemia (CML) mouse model, whereas imatinib is not.52 Overall, these results indicate that a more active CNS prophylaxis must be administered when using imatinib as front-line therapy.

This study was not powered to define the efficacy of allo-SCT but rather the feasibility of a sequential therapeutic strategy that contemplated also a transplant. (alloSCT was administered to patients with an available donor, mostly siblings, and to those deemed fit to proceed to transplant procedures.) Nevertheless, allo-SCT, that is still the standard curative approach in this subset of patients, was associated with a better DFS, OS or CIR in this imatinib-based protocol. It must be noted that the non-transplanted group included patients who experienced an early relapse, the main reason for not undergoing an allo-SCT, and older patients. It is also worth underlining that in both transplanted and non-transplanted patients relapses were not observed later than 33 months, suggesting that a subgroup of patients might be spared transplant procedures and the related morbidities.

Finally, it has recently been proposed12,36,53,54 that auto-SCT might have a role in Ph+ ALL since the introduction of TKI and the precise quantification of MRD. The NILG group12 reported a 5-year cumulative survival of 67% for 9 autografted patients. Chalandon et al.36 showed that, among the 29 autografted patients in major molecular response, there was no statistically significant difference in outcome from that of allografted cases. Similarly, the CALGB Study 10001 (Alliance)53 proved that an auto-SCT, performed in 19 individuals, gave results similar to those obtained with an alloSCT (n=15). Finally, Giebel et al.54 extensively evaluated the role of auto-SCT in a cohort of 177 adult Ph+ ALL. Patients were subdivided into 3 categories, according to the period in which the procedure was performed: 1996-2001, 20022006 (during which sporadic cases received TKI during therapy), and 2007 onwards (when all patients received ima-

Figure 5. Survival at 60 months on the basis of the reduction in minimal residual disease (MRD). (A) Disease-free survival (DFS) was significantly better for patients with BCR-ABL1 log reduction levels of 1.3 log or over (55.6%, 95%CI: 39.0-79.3, continuous line) than for patients with log reduction levels less than 1.3 log (20.0%, 95%CI: 5.8-69.1, dashed line) (P=0.03). (B) Overall survival (OS) at 60 months is 59.1% (95%CI: 42.3-82.6) for patients with BCR-ABL1 log reduction levels of 1.3 log or over, and 20% for patients with log reduction levels less than 1.3 log (95%CI: 5.8-69.1). CR: complete remission.

haematologica | 2016; 101(12)

Final results of the GIMEMA LAL 0904 for Ph+ ALL

tinib and steroids followed by chemotherapy and, when possible, by a SCT is feasible, well tolerated and provides superior results in terms of CHR achievement, long-term DFS and OS compared to other studies in which imatinib/dasatinib are administered together with chemotherapy during the induction phase or following chemotherapy. Based on a case series of almost 200 adult patients with Ph+ ALL, with no upper age limit, the GIMEMA group has observed that, with the use of a TKI alone (plus steroid and intrathecal treatment), more than 95% of cases can obtain a CHR with no deaths in induction. We continue, therefore, to support a strategy that allows a CHR to be obtained in virtually all patients, including the elderly, based on this chemotherapy-free approach.

Furthermore, our study confirms the importance of MRD, the negativity of which should be a major objective in this disease: a higher rate of MRD negativity is likely to be achieved with the inclusion of 2nd- and, more likely, 3rd-generation TKI such as ponatinib, which have been proven to have a greater debulking effect,9,14,16,55 and might possibly be more active in patients with the BCR-ABLp210 form, given the lower MRD clearance observed in this subgroup of patients. As far as ponatinib is concerned, a recent study from the MDACC55 on 37 patients has indeed shown that its use as upfront therapy, combined with chemotherapy, is able to induce extremely promising results with 2-year event-free survival (EFS) and OS of 81% and 80%, respectively, although 6 toxic deaths were recorded.55

References 1 Mancini M, Scappaticci D, Cimino G, et al.

A comprehensive genetic classification of adult acute lymphoblastic leukemia (ALL): analysis of the GIMEMA 0496 protocol.

Blood. 2005;105(9):3434-3441.

2 Moorman AV, Harrison CJ, Buck GA, et al.

Karyotype is an independent prognostic factor in adult acute lymphoblastic leukemia (ALL): analysis of cytogenetic data from patients treated on the Medical Research Council (MRC) UKALLXII/ Eastern Cooperative Oncology Group (ECOG) 2993 trial. Blood. 2007; 109(8):3189-3197.

3 Burmeister T, Schwartz S, Bartram CR, et al.

Patients’ age and BCR–ABL frequency in adult B-precursor ALL: a retrospective analysis from the GMALL study group. Blood.

2008;112(3):918-919.

4 Chiaretti S, Vitale A, Cazzaniga G, et al.

Clinico-biological features of 5202 patients with acute lymphoblastic leukemia enrolled in the Italian AIEOP and GIMEMA protocols and stratified in age cohorts.

Haematologica. 2013;98(11):1702-1710.

5 Thomas DA, Faderl S, Cortes J, et al.

Treatment of Philadelphia chromosomepositive acute lymphocytic leukemia with hyper-CVAD and imatinib mesylate. Blood.

2004;103(12):4396-4407.

6 Yanada M, Takeuchi J, Sugiura I, et al. High complete remission rate and promising outcome by combination of imatinib and chemotherapy for newly diagnosed BCRABL-positive acute lymphoblastic leukemia: a phase II study by the Japan Adult Leukemia Study Group. J Clin Oncol.

2006;24(3):460-466.

7 Wassmann B, Pfeifer H, Goekbuget N, et al.

haematologica | 2016; 101(12)

8

9

10

11

12

Other strategies, based on the use of monoclonal antibodies or other immunologic approaches, need to be tested with the goal of offering an overall chemotherapy-free approach that might eradicate/control residual leukemic clones. This is particularly challenging for the elderly/less fit patients, particularly as the prevalence of Ph+ ALL cases increases with age.4 Finally, an allo-SCT, that is still the only potentially curative strategy, might be offered to patients who have been spared prior chemotherapy. This approach should help to reduce side effects, thus sparing further toxicity.

Acknowledgments The authors wish to thank Sandra De Simone for administrative support and data management.

Funding This work was supported by Associazione Italiana per la Ricerca sul Cancro (AIRC) Special Program Molecular Clinical Oncology, 5x1000 (MCO1007), Milan, Italy; Ministero dell’Università e Ricerca (MIUR), Fondo per gli Investimenti della Ricerca di Base (FIRB), Rome, Italy; Progetti di Ateneo Sapienza Università di Roma, 2015 (#C26A15F9WW); Progetto Giovani Ricercatori 2010, Policlinico di Modena (#GR2010-2313609). The GIMEMA Foundation, sponsor of the study, received unrestricted research grants from Novartis.

Finally, we also would like to acknowledge AIL (Associazione Italiana contro le Leucemie-Linfomi e Mielomi).

Alternating versus concurrent schedules of imatinib and chemotherapy as front-line therapy for Philadelphia-positive acute lymphoblastic leukemia (Ph+ALL). Blood.

2006;108(5):1469-1477.

Delannoy A, Delabesse E, Lhéritier V, et al.

Imatinib and methylprednisolone alternated with chemotherapy improve the outcome of elderly patients with Philadelphia-positive acute lymphoblastic leukemia: results of the GRAALL AFR09 study. Leukemia.

2006;20(9):1526-1532.

Talpaz M, Shah NP, Kantarjian H, et al.

Dasatinib in imatinib-resistant Philadelphia chromosome-positive leukemias. N Engl J Med. 2006;354(24):2531-2541. de Labarthe A, Rousselot P, Huguet-Rigal F, et al. Group for Research on Adult Acute Lymphoblastic Leukemia (GRAALL).

Imatinib combined with induction or consolidation chemotherapy in patients with de novo Philadelphia chromosome-positive acute lymphoblastic leukemia: results of the GRAAPH-2003 study. Blood. 2007;109 (4):1408-1413.

Vignetti M, Fazi P, Cimino G, et al. Imatinib plus steroids induces complete remissions and prolonged survival in elderly Philadelphia chromosome-positive patients with acute lymphoblastic leukemia without additional chemotherapy: results of the Gruppo Italiano Malattie Ematologiche dell'Adulto (GIMEMA) LAL0201-B protocol.

Blood. 2007;109(9):3676-3678.

Bassan R, Rossi G, Pogliani EM, et al.

Chemotherapy-phased imatinib pulses improve long-term outcome of adult patients with Philadelphia chromosomepositive acute lymphoblastic leukemia: Northern Italy Leukemia Group protocol 09/00. J Clin Oncol. 2010;28(22):3644-3652.

13 Ribera JM, Oriol A, Gonzalez M, et al.

Concurrent intensive chemotherapy and imatinib before and after stem cell transplantation in newly diagnosed Philadelphia chromosome-positive acute lymphoblastic leukemia. Final results of the CSTIBES02 trial. Haematologica. 2010;95(1):87-95.

14 Ravandi F, O'Brien S, Thomas DA, et al. First report of phase 2 study of dasatinib with hyper-CVAD for the frontline treatment of patients with Philadelphia chromosome– positive (Ph+) acute lymphoblastic leukemia. Blood. 2010;116(12):2070-2077.

15 Mizuta S, Matsuo K, Yagasaki F, et al. Pretransplant imatinib-based therapy improves the outcome of allogeneic hematopoietic stem cell transplantation for BCR-ABL-positive acute lymphoblastic leukemia.

Leukemia. 2011;25:41-47.

16 Foà R, Vitale A, Vignetti M, et al. Dasatinib as first-line treatment for adult patients with Philadelphia chromosome-positive acute lymphoblastic leukemia. Blood. 2011;118 (25):6521-6528.

17 Chiaretti S, Vitale A, Elia L, et al. Multicenter total therapy Gimema LAL 1509 protocol for de novo adult Ph+ acute lymphoblastic leukemia (ALL) patients. Updated results and refined genetic-based prognostic stratification. Abstract 81, 57th ASH Annual Meeting & Exposition December 5-8, 2015, Orlando, Florida; USA. 18 Thyagu S, Minden MD, Gupta V, et al.

Treatment of Philadelphia chromosomepositive acute lymphoblastic leukaemia with imatinib combined with a paediatricbased protocol. Br J Haematol. 2012; 158(4):506-514.

19 Ribera JM, García O, Montesinos P, et al.

Treatment of young patients with Philadelphia chromosome-positive acute

1551

S. Chiaretti et al.

20

21

22

23

24

25

26

27

28

29

30

31

1552

lymphoblastic leukaemia using increased dose of imatinib and deintensified chemotherapy before allogeneic stem cell transplantation. Br J Haematol. 2012; 159(1):78-81.

Tanguy-Schmidt A, Rousselot P, Chalandon Y, et al. Long-term follow-up of the imatinib GRAAPH-2003 study in newly diagnosed patients with de novo Philadelphia chromosome-positive acute lymphoblastic leukemia: a GRAALL study. Biol Blood Marrow Transplant. 2013;19(1):150-155.

Fielding AK, Rowe JM, Buck G, et al.

UKALLXII/ECOG2993: addition of imatinib to a standard treatment regimen enhances long-term outcomes in Philadelphia positive acute lymphoblastic leukemia. Blood.

2014;123(6):843-850.

Dombret H, Gabert J, Boiron JM, et al.

Outcome of treatment in adults with Philadelphia chromosome-positive acute lymphoblastic leukemia--results of the prospective multicenter LALA-94 trial.

Blood. 2002;100(7):2357-2366.

Laport GG, Alvarnas JC, Palmer JM, et al.

Long-term remission of Philadelphia chromosome-positive acute lymphoblastic leukemia after allogeneic hematopoietic cell transplantation from matched sibling donors: a 20-year experience with the fractionated total body irradiation-etoposide regimen. Blood. 2008;112(3):903-909.

Fielding AK, Goldstone AH. Allogeneic haematopoietic stem cell transplant in Philadelphia-positive acute lymphoblastic leukemia. Bone Marrow Transplant.

2008;41(5):447-453.

Fielding AK, Rowe JM, Richards SM, et al.

Prospective outcome data on 267 unselected adult patients with Philadelphia chromosome–positive acute lymphoblastic leukemia confirms superiority of allogeneic transplantation over chemotherapy in the pre-imatinib era: results from the International ALL Trial MRC UKALLXII/ECOG2993. Blood. 2009;113 (19):4489-4496.

Gruber F, Mustjoki S, Porkka K. Impact of tyrosine kinase inhibitors on patient outcomes in Philadelphia chromosome-positive acute lymphoblastic leukaemia. Br J Haematol. 2009;145(5):581-597.

Ottmann OG, Pfeifer H. First-line treatment of Philadelphia chromosome-positive acute lymphoblastic leukaemia in adults. Cur Opin Oncol. 2009;21:S43-S46.

Mathisen MS, O'Brien S, Thomas DA, et al.

Role of tyrosine kinase inhibitors in the management of Philadelphia chromosomepositive acute lymphoblastic leukemia. Curr Hematol Malig Rep. 2011;6(3):187-194.

Liu-Dumlao T, Kantarjian H, Thomas DA, et al. Philadelphia-positive acute lymphoblastic leukemia: Current treatment options. Curr Oncol Rep. 2012;14(5):387-394.

Thomas X. Philadelphia chromosome-positive leukemia stem cells in acute lymphoblastic leukemia and tyrosine kinase inhibitor therapy. World J Stem Cells. 2012; 26;4(6):44-52.

Elia L, Mancini M, Moleti ML, et al. A multiplex reverse transcriptase-polymerase chain reaction strategy for the diagnostic molecular screening of chimeric genes: a clinical evaluation on 170 patients with acute lymphoblastic leukemia.

Haematologica. 2003;88(3):275-279.

32 Beillard E, Pallisgaard N, van der Velden VH, et al. Evaluation of candidate control genes for diagnosis and residual disease detection in leukemic patients using 'real-time' quantitative reverse-transcriptase polymerase chain reaction (RQ-PCR) - a Europe Against Cancer program. Leukemia. 2003;17(12): 2474-2486.

33 Gabert J, Beillard E, van der Velden VH, et al.

Standardization and quality control studies of 'real-time' quantitative reverse transcriptase polymerase chain reaction of fusion gene transcripts for residual disease detection in leukemia - a Europe Against Cancer program.

Leukemia. 2003; 17(12):2318-2357.

34 Ravandi F, Jorgensen JL, Thomas DA, et al.

Detection of MRD may predict the outcome of patients with Philadelphia chromosome– positive ALL treated with tyrosine kinase inhibitors plus chemotherapy. Blood.

2013;122(7):1214-1221.

35 Daver N, Thomas D, Ravandi F, et al. Final report of a phase II study of imatinib mesylate with hyper-CVAD for the front-line treatment of adult patients with Philadelphia chromosome-positive acute lymphoblastic leukemia. Haematologica.

2015;100(5):653-661.

36 Chalandon Y, Thomas X, Hayette S, et al.

Randomized study of reduced-intensity chemotherapy combined with imatinib in adults with Ph-positive acute lymphoblastic leukemia. Blood. 2015;125(24):3711-3719.

37 Lee S, Kim DW, Kim YJ, et al. Minimal residual disease-based role of imatinib as a firstline interim therapy prior to allogeneic stem cell transplantation in Philadelphia chromosome-positive acute lymphoblastic leukemia. Blood. 2003;102(8):3068-3070.

38 Yanada M, Sugiura I, Takeuchi J, et al.

Prospective monitoring of BCR-ABL1 transcript levels in patients with Philadelphia chromosome-positive acute lymphoblastic leukaemia undergoing imatinib-combined chemotherapy. Br J Haematol. 2008; 143(4):503-510.

39 Mizuta S, Matsuo K, Maeda T, et al.

Prognostic factors influencing clinical outcome of allogeneic hematopoietic stem cell transplantation following imatinib-based therapy in BCR–ABL-positive ALL. Blood Cancer Journal. 2012;2(5),e72.

40 Lee S, Kim D-W, Cho B-S, et al. Impact of minimal residual disease kinetics during imatinib-based treatment on transplantation outcome in Philadelphia chromosome-positive acute lymphoblastic leukemia.

Leukemia. 2012;26(11):2367-2374.

41 Ravandi F, Jorgensen JL, Thomas DA, et al.

Detection of MRD may predict the outcome of patients with Philadelphia chromosome– positive ALL treated with tyrosine kinase inhibitors plus chemotherapy. Blood.

2013;122(7):1214-1221.

42 Preudhomme C, Henic N, Cazin B, et al.

Good correlation between RT-PCR analysis and relapse in Philadelphia (Ph1)-positive acute lymphoblastic leukemia (ALL).

Leukemia. 1997;11(2):294-298.

43 Pane F, Cimino G, Izzo B, et al. Significant reduction of the hybrid BCR/ABL transcripts after induction and consolidation therapy is a powerful predictor of treatment response in adult Philadelphia-positive acute lymphoblastic leukemia. Leukemia. 2005;19(4): 628-635.

44 Radich J, Gehly G, Lee A, et al. Detection of

45

46

47

48

49

50

51

52

53

54

55

bcr-abl transcripts in Philadelphia chromosome-positive acute lymphoblastic leukemia after marrow transplantation. Blood.

1997;89(7):2602-2609.

Stirewalt DL, Guthrie KA, Beppu L, et al.

Predictors of relapse and overall survival in Philadelphia chromosome-positive acute lymphoblastic leukemia after transplantation. Biol Blood Marrow Transplant.

2003;9(3):206-212.

Chen H, Liu KY, Xu LP, et al. Administration of imatinib after allogeneic hematopoietic stem cell transplantation may improve disease-free survival for patients with Philadelphia chromosome-positive acute lymphoblastic leukemia. J Hematol Oncol.

2012;5:29.

Pfeifer H, Wassmann B, Bethge W, et al.

Randomized comparison of prophylactic and minimal residual disease-triggered imatinib after allogeneic stem cell transplantation for BCR-ABL1 positive acute lymphoblastic leukemia. Leukemia. 2013;27(6): 1254-1262.

White HE, Matejtschuk P, Rigsby P, et al.

Establishment of the first World Health Organization International Genetic Reference Panel for quantitation of BCRABL mRNA. Blood. 2010;116(22):e111-e117.

O'Hare T, Walters DK, Stoffregen EP, et al. In vitro activity of Bcr-Abl inhibitors AMN107 and BMS-354825 against clinically relevant imatinib-resistant Abl kinase domain mutants. Cancer Res. 2005; 65(11):45004505.

Pfeifer H, Wassmann B, Hofmann WK, et al.

Risk and prognosis of central nervous system leukemia in patients with Philadelphia chromosome-positive acute leukemias treated with imatinib mesylate. Clin Cancer Res. 2003;9(13):4674-4681.

Takayama N, Sato N, O'Brien SG, et al.

Imatinib mesylate has limited activity against the central nervous system involvement of Philadelphia chromosome-positive acute lymphoblastic leukaemia due to poor penetration into cerebrospinal fluid. Br J Haematol. 2002;119(1):106-108.

Porkka K, Koskenvesa P, Lundán T, et al.

Dasatinib crosses the blood-brain barrier and is an efficient therapy for central nervous system Philadelphia chromosome-positive leukemia. Blood. 2008;112(4):10051012.

Wetzler M, Watson D, Stock W, et al.

Autologous transplantation for Philadelphia chromosome-positive acute lymphoblastic leukemia achieves outcomes similar to allogeneic transplantation: results of CALGB Study 10001 (Alliance). Haematologica.

2014;99(1):111-115.

Giebel S, Labopin M, Gorin NC, et al.

Improving results of autologous stem cell transplantation for Philadelphia-positive acute lymphoblastic leukaemia in the era of tyrosine kinase inhibitors: a report from the Acute Leukaemia Working Party of the European Group for Blood and Marrow Transplantation. Eur J Cancer. 2014; 50(2):411-417.

Jabbour E, Kantarjian H, Ravandi F, et al.

Combination of hyper-CVAD with ponatinib as first-line therapy for patients with Philadelphia chromosome-positive acute lymphoblastic leukaemia: a single-centre, phase 2 study. Lancet Oncol. 2015; 16(15):1547-1555.

haematologica | 2016; 101(12)

ARTICLE

Chronic Lymphocytic Leukemia

SLP76 integrates into the B-cell receptor signaling cascade in chronic lymphocytic leukemia cells and is associated with an aggressive disease course

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Nili Dezorella,1,2,* Ben-Zion Katz,1,2,* Mika Shapiro,1 Aaron Polliack,3 Chava Perry1 and Yair Herishanu1,2

1 Department of Hematology, Tel-Aviv Sourasky Medical Center; 2Sackler Faculty of Medicine, Tel-Aviv University and 3Department of Hematology, Hadassah University Hospital and Hebrew University Medical School, Jerusalem, Israel

*ND and B-Z Katz contributed equally to this work

Haematologica 2016 Volume 101(12):1553-1562

ABSTRACT

I

In the last decade, the B-cell receptor has emerged as a pivotal stimulus in the pathogenesis of chronic lymphocytic leukemia, and a very feasible therapeutic target in this disease. B-cell receptor responsiveness in chronic lymphocytic leukemia cells is heterogeneous among patients and correlates with aggressiveness of the disease. Here we show, for the first time, that SLP76, a key scaffold protein in T-cell receptor signaling, is ectopically expressed in chronic lymphocytic leukemia cells, with variable levels among patients, and correlates positively with unmutated immunoglobulin heavy chain variable gene status and ZAP70 expression. We found that SLP76 was functionally active in chronic lymphocytic leukemia cells. A SYK-dependent basal level of phosphorylated SLP76 exists in the cells, and upon B-cell receptor engagement, SLP76 tyrosine phosphorylation is significantly enhanced concomitantly with increased physical association with BTK. B-cell receptor-induced SLP76 phosphorylation is mediated by upstream signaling events involving LCK and SYK. Knockdown of SLP76 in the cells resulted in decreased induction of BTK, PLCγ2 and IκB phosphorylation, as well as cell viability after B-cell receptor activation with anti-IgM. Consistent with our biochemical findings, high total SLP76 expression in chronic lymphocytic leukemia cells correlated with a more aggressive disease course. In conclusion: SLP76 is ectopically expressed in chronic lymphocytic leukemia cells where it plays a role in B-cell receptor signaling.

Correspondence: yairh@tlvmc.gov.il

Received: November 9, 2015.

Accepted: July 12, 2016.

Pre-published: July 21, 2016. doi:10.3324/haematol.2015.139154

Introduction Chronic lymphocytic leukemia (CLL) is characterized by the progressive accumulation of monoclonal, CD5+ B cells in the peripheral blood, bone marrow and secondary lymphoid organs.1 Despite the fact that CLL is currently incurable by standard chemo-immunotherapy, impressive clinical responses can be obtained which prolong overall survival.2 B-cell receptor (BCR) signaling is a crucial component of normal B-cell development, and plays an important role in the differentiation, survival, proliferation and antibody secretion of these cells.3 In mature B cells, antigen engagement of the BCR induces coordinated downstream signaling cascades. These initial events include the recruitment and activation of Lyn to phosphorylate the immunoreceptor tyrosine-based activation motifs of the Igα/Igβ components of the BCR. These events are followed by further recruitment and activation of additional kinases and adaptor molecules such as SYK, Bruton tyrosine kinase (BTK), phosphatidylinositol 3kinases (PI3K), B-cell linker (BLNK or SLP65) and PLCγ2 which form a micro-signalosome that enables the amplification and propagation of the signal through a haematologica | 2016; 101(12)

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/12/1553

©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.

1553

N. Dezorella et al.

number of downstream cascades.3 BCR signaling also plays a critical role in the pathogenesis of CLL, and antigen engagement is presumed to be a key regulator of CLL cell survival and proliferation in vivo.4,5 There are two main subgroups of CLL based on immunoglobulin heavy chain variable (IGHV) gene mutational status.6-8 CLL with mutated IGHV is characterized by stable or slowly progressive disease, while the unmutated IGHV CLL subtype has a more aggressive clinical course.6,7 In vitro studies have shown that activation of the BCR protects CLL cells from apoptosis9,10 and promotes entry into the cell cycle.11,12 However, responsiveness of CLL cells to BCR activation is heterogeneous.13 CLL cells with unmutated IGHV are usually BCR-signaling competent, while those with mutated IGHV generally respond weakly to BCR activation.8 The zeta chain-associated protein kinase of 70 kD (ZAP-70), which is normally expressed in T cells, is involved in T-cell receptor (TCR) signaling. ZAP-70 is ectopically expressed in most cases of CLL with unmutated IGHV CLL and less often with mutated IGHV.14-16 Expression of ZAP-70 in CLL cells is associated with an augmented response to BCR activation17 and correlates with a more aggressive clinical course.16,18-20 Given the essential role of BCR signaling in the pathogenesis of CLL, this pathway has now become a target for anti-CLL therapy and small-molecules directed against kinases such as SYK, BTK, or PI3K have impressive clinical activity.21,22 The SH2 domain-containing leukocyte protein of 76 kDa (SLP76, also known as LCP2) is a hematopoietic adaptor protein known to be important in multiple biochemical signaling pathways and is expressed in all hematopoietic lineages except for mature B cells.23,24 In T cells, SLP76 functions as a critical signal transducer downstream of the TCR.25 Engagement of the TCR leads to consecutive activation of kinases including LCK, which phosphorylates the immunoreceptor tyrosine-based activation motifs of this receptor.26 These phosphorylated motifs recruit and activate ZAP-70 which then phosphorylates SLP76 and the transmembrane adaptor LAT.27 Activated LAT recruits SLP76 from the cytosol to the cell membrane to form a multimolecular complex comprising a number of signaling molecules,28 including PLCγ1, VAV and NCK (which bind to SLP76 tyrosine residues Y112 and Y128), ITK (which binds SLP76 Y145),29 ADAP, LCK and HPK1.30 Formation of the SLP76-LAT multimolecular complex allows proximity between the signaling molecules and facilitates an efficient propagation of various TCR signaling processes.30 Since CLL cells express ZAP-70 and LCK,15,16,31,32 we examined whether they also express additional TCR-associated molecules. Here we report, for the first time, the aberrant expression of SLP76 in CLL cells, which plays a role in the BCR signaling pathway and correlates with an aggressive clinical course.

Antibodies and reagents The antibodies and reagents used in this study are detailed in Online Supplement S2.

CD19 enrichment Peripheral blood mononuclear cells were magnetically labeled using CD19 microbeads (Miltenyi Biotec, Inc., Auburn, CA, USA), and separated (more than 95% purity) on a magnetic cell separation LS column (Miltenyi Biotec, Inc.) according to the manufacturer's instructions.

Western blotting and co-immunoprecipitation Western blot and co-immunoprecipitation protocols are described in Online Supplement S2.

mRNA extraction and cDNA synthesis RNA was extracted using an RNeasy kit (Qiagen, CA, USA), and reverse transcription was performed using a Verso cDNA kit (Thermo Fisher Scientific/ABgene, Epsom, UK), according to the manufacturers’ instructions.

Quantitative reverse transcriptase polymerase chain reaction Gene transcripts were quantified by quantitative polymerase chain reaction using the Absolute Blue QPCR SYBR Green ROX mix (Thermo Fisher Scientific/ABgene) and a Rotor-Gene RG6000 apparatus (Corbett Research, Mortlake, Australia). The primers used are presented in Online Supplement S2.

Immunoglobulin heavy chain variant gene and ZAP-70 analysis The IGHV gene was amplified as described elseswhere.7 The protocol is available in Online Supplement S2. ZAP-70 expression was assessed by western blotting of CD19+ purified CLL cells.

Flow cytometry Cells were stained using Cytofix fixation buffer and Phosflow Perm/Wash buffer I according to the manufacturer’s instructions.

Samples were acquired by a FACSCalibur and analyzed using CellQuest software (Becton Dickinson, San Jose, CA, USA).

In vitro B-cell receptor stimulation CLL cells (1x107/mL) were stimulated with goat F(ab′)2 anti– human IgM (10 mg/mL) at 37°C for the indicated times. For inhibition assays, cells were incubated prior to IgM stimulation in the absence or presence of the following: 10 mM PP2 for 15 min, 10 mM SYK inhibitor II for 15 min, 0.5 mM ibrutinib for 1 h, 20 mM cytochalasin B for 30 min, 10 mM MβCD for 30 min, 40, 200, and 1000 nM LCK inhibitor for 2 h, and 0.2, 1, and 5 mM R406 for 30 min. These concentrations were chosen on the basis of previous publications,32,34-37 and in this study were titrated to obtain a maximal effect without killing the cells. Inhibitors were dissolved in dimethylsulfoxide, while controls were treated accordingly with dimethylsulfoxide.

Short interfering RNA transfection Methods Patients and samples After signing an informed consent form approved by the Institutional Review Board according to the Helsinki declarations, blood samples were collected from patients fulfilling the standard criteria for CLL33 and also from healthy controls. The patients’ characteristics are shown in Online Supplement S1. The sample handling protocol is available in Online Supplement S2.

1554

Cells were transfected with siRNA using the 4DNucleofectordevice (Lonza Group Ltd, Basel, Switzerland) according to the manufacturer's instructions. The protocol is described in Online Supplement S2.

Cell apoptosis assay Cell apoptosis was detected using a MEBCYTO Apoptosis Kit (MBL Co., Ltd. Nagoya, Japan) according to the manufacturer's instructions. The protocol is available in Online Supplement S2. haematologica | 2016; 101(12)

Aberrant expression of SLP76 in CLL

Statistical analysis

that the scaffold protein SLP76 is aberrantly expressed in CLL cells, and not in the control, mature B-cell lines, Raji (Figure 1A) and Daudi (data not shown). The levels of SLP76 expression in CLL cells varied among patients (Figure 1A), and correlated with SLP76 mRNA levels (Online Supplement S3). To exclude any possibility that SLP76 expression originated from contaminating T cells in the CD19+ selected B-cell samples, the blots were also analyzed for CD3ε, a T-cell specific molecule. As shown in Figure 1A, CD3ε was undetectable in the purified CLL samples, but readily found in the Jurkat T-cell line. BLNK, the counterpart adaptor molecule of SLP76 which is normally expressed in B cells, was detected in all CLL cell samples at comparable levels (Figure 1A). To further verify these findings, SLP76 expression was also measured using intracellular flow cytometry. As shown in Figure 1B,C,

We compared continuous variables between groups using the Student t-test. A P value <0.05 was considered statistically significant. Survival curves were created using the method of Kaplan and Meier, and the log-rank test was used to assess differences between the subgroups. A P value <0.05 was considered statistically significant. All statistical analyses were performed using Graphpad Prism 5.0 software (GraphPad Software, San Diego, CA, USA).

Results SLP76 is aberrantly expressed in chronic lymphocytic leukemia cells Western blotting of CD19+ purified CLL cells revealed

A

D

E

B C

Figure 1. SLP76 is ectopically expressed in CLL cells and is associated with unfavorable prognostic markers. (A) Detection of total SLP76 as well as ZAP-70, LCK, LAT, ITK, PLCγ1, PLCγ2, BLNK (SLP65) and CD3ε levels by western blot in primary CD19+ purified CLL cells and in cell lines including Jurkat (positive control for Tcell-related proteins) and Raji (negative control for T-cell-related proteins). Actin was used to verify equal loading (n=22). A representative image is shown. (B-C) Flow cytometry analysis of SLP76 levels in T cells and B-cells in peripheral blood samples of healthy individuals and CLL patients. (B) Representative cases showing SLP76 expression in normal and CLL blood samples. (C) A summary of SLP76 mean fluorescence intensity (average ± SD) normalized to isotype control in gated B cells of healthy subjects (n=6) and CLL patients (n=6). (D) Detection of ZAP70 levels in four samples of primary CD19+ purified CLL cells with low SLP76 levels and four with high SLP76 levels by western blot. The Jurkat cell line was used as a positive control. CD3ε was used to verify that there was no T-cell contamination. Actin was used to verify equal loading. (E) Correlation between SLP76 protein levels, in CD19+ purified CLL cells (detected by western blot then quantified and normalized to actin), and IGHV mutational status and ZAP-70 expression (detected by western blot, n=22).

haematologica | 2016; 101(12)

1555

N. Dezorella et al.

SLP76 was detected in both CD19+ and CD3+ gated populations in CLL samples, but levels were lower in CLL cells than in T cells present in the same sample. In contrast to CLL cells, in peripheral blood lymphocytes of healthy individuals SLP76 was only detected in T cells, and was not present in B cells (Figure 1B).

Since SLP76 is a downstream signaling molecule of the TCR pathway, we examined whether other components of this pathway are also expressed in CLL cells. As previously reported, the proximal signaling kinases LCK and ZAP-70 were detected in CLL cells at varying levels (Figure 1A). However the more distal downstream TCR molecules (LAT and ITK) were not expressed in CLL samples, while both PLCγ2 and PLCγ1 were detected at near comparable levels (Figure 1A). In most cases examined, patients’ samples containing high levels of SLP76 proteins also had high levels of ZAP-70 expression (Figure 1D).

Quantitative analysis of SLP76 expression at the protein level showed that CLL cells with unmutated IGHV or which were positive for ZAP-70 expressed higher levels of SLP76 than CLL cells which had mutated IGHV or were negative for ZAP-70 (Figure 1E)

B-cell receptor engagement phosphorylates SLP76 in chronic lymphocytic leukemia cells independently of ZAP-70 In order to examine whether aberrant SLP76 expression in CLL cells plays a role in BCR signaling, we activated CLL cells by BCR cross-linking. Upon TCR engagement, tyrosine residues Y113, Y128, and Y145 on the SH2 domain binding motifs in the N-terminal of SLP76 are phosphorylated.23 In contrast, SLP76 phosphorylation on serine 376 generates a delayed negative signal that regulates T-cell activation.38 As shown in Figure 2A,B, after surface IgM engagement, SLP76 undergoes tyrosine phosphorylation (Y128) in a time-dependent manner, with this process reaching maximal intensity at 15 min, decreasing after 45 min, and returning to near basal levels by 120 min (Figure 2A). In contrast, serine 376 (S376) phosphorylation had increased by 15 min, peaked and remained elevated at 45 and 120 min, respectively (Figure 2A), concomitant with the decrease in Y128 phosphorylation. Activation of ERK and AKT followed SLP76 Y128 residue phosphorylation, and diminished concomitantly with the phosphorylation of SLP76 S376 residue (Figure 2A). Overall, Y128 phosphorylation following IgM engagement was variable among patients, with an average increase of over 30% (Figure 2C). Despite the correlation between total SLP76 and ZAP-70 expression, there was no correlation between ZAP-70 expression and the degree of SLP76 phosphorylation after BCR activation (Figure 2D,E). In fact, significant SLP76 phosphorylation also occurred in some ZAP-70negative CLL cells (Figure 2E), indicating that ZAP-70 is not essential for SLP76 activation in CLL cells. In a similar manner phosphorylation of the S376 residue did not correlate with ZAP-70 expression (Online Supplement S4).

B-cell receptor-triggered phosphorylation of SLP76 in chronic lymphocytic leukemia cells is mediated through LCK and SYK Since SRC family kinases, in particular, LYN, play a pivotal role in the proximal signaling cascade triggered by BCR engagement,36 we examined whether SLP76 phosphorylation was dependent on SRC family kinases activity. As shown in Figure 3A, pre-incubation of CLL cells 1556

with PP2, an inhibitor of SRC family kinases, blocked SLP76 phosphorylation almost completely. In a similar manner, the SYK inhibitor-SYKII abrogated the phosphorylation of SLP76, and even depleted the basal level. PP2 and SYKII also partially inhibited the phosphorylation of SLP76 S376 residue (Figure 3B). Cytochalasin B (a cytoskeleton inhibitor) and methyl-β-cyclodextrin (MβCD, extracts membrane-associated cholesterol) had no effect on SLP76 tyrosine phosphorylation, indicating that the integrity of the cytoskeleton or the cholesterolrich membrane domains is not necessary to initiate activation of SLP76 in response to BCR engagement (Figure 3A).

As expected, PP2 and SYKII abrogated AKT phosphorylation, which was used to confirm the activity of each one of these inhibitors (Figure 3). Due to the critical role of LCK in TCR upstream signaling, and its ectopic expression in CLL cells, we also determined its role in the activation of SLP76 after BCR triggering. LCKi, a highly selective inhibitor of LCK that does not affect LYN activity,32 reduced SLP76 phosphorylation in response to BCR engagement to baseline level (Figure 3C). The clinicallyused SYK inhibitor R406 completely abolished SLP76 phosphorylation (Figure 3D), in a manner similar to the activity of SYKII (Figure 3A).

SLP76 binds BTK, an association enhanced

by B-cell receptor engagement It is established that in CLL cells, BTK associates with the adaptor protein BLNK.39 As shown in Figure 4A,B, immunoprecipitation of SLP76 in unstimulated CLL cells resulted in co-precipitation of BTK. Moreover, engagement of the BCR led to a significant increase in the association of SLP76/BTK. Although, SLP76 and BTK are physically associated in CLL cells, BCR-mediated induction of SLP76 phosphorylation was only slightly affected by the potent BTK inhibitor ibrutinib (Figure 3A).

SLP76 downregulation partially abrogates B-cell receptor signaling In an attempt to examine the direct role of SLP76 in BCR signaling, SLP76 expression was down-regulated using siRNA. Transfection of CLL cells with siRNA against SLP76 resulted in a decrease of this gene after 24 h of culture, as measured by reverse transcriptase polymerase chain reaction (data not shown), while a maximum decline in protein level was evident after 36 h (approximately 50%, Figure 4C). Modulations in BCR signaling were also examined using the knockdowns and control CLL cells after anti-IgM cross-linking. As shown in Figure 4C,E, down-regulation of SLP76 significantly decreased BTK phosphorylation in BCR-activated CLL cells. In addition, the BCR-mediated activation of PLCγ2, a downstream kinase in the BCR cascade, was also significantly inhibited by the down regulation of SLP76 (Figure 4C,F). We further investigated the role of SLP76 in the regulation of distal signaling events following BCR cross-linking. Knockdown of SLP76 significantly reduced IκB activation in BCR-stimulated CLL cells, while phosphorylation of ERK and AKT was left intact (Figure 5A,B). In addition, knockdown of BLNK, the counterpart adaptor molecule of SLP76 in B cells, resulted in a similar effect of reduction in BTK (P=0.04) and IκB activation in anti-IgM-stimulated CLL cells (Figure 4D and Figure 5A,B), without a significant impact on ERK and AKT activation. Thus, our data indicate that SLP76 as well as BLNK are specifically required haematologica | 2016; 101(12)

Aberrant expression of SLP76 in CLL

for BCR-induced activation of NF-κB. In light of the above, the viability of control CLL cells cultured for 48 h in the presence of anti-IgM antibody was improved compared to that of unstimluated cells, but this protective effect was partially abrogated by SLP76 knockdown (Figure 5C).

subgroups with low and high expression. As shown in Figure 5D, patients with CLL who had higher SLP76 levels had a shorter time to disease progression or first treatment compared to patients with lower SLP76 levels.

SLP76 expression correlates with progression of chronic lymphocytic leukemia

Discussion

Based on the observation that SLP76 expression is heterogeneous among patients with CLL, and considering that SLP76 is phosphorylated after BCR activation, we also evaluated the possible clinical significance of SLP76 expression in CLL. In this analysis, the median value of SLP76 protein levels was used to separate patients into

In this study we demonstrate that the scaffold protein SLP76, which plays a critical role in the TCR signaling pathway,23 is aberrantly expressed in CLL cells. The level of SLP76 varies among patients and correlates with unmutated IGHV gene status, positive ZAP-70 expression, and a shorter time to disease progression. Similar to SLP76, the

A

B

C

D

E

Figure 2. SLP76 is phosphorylated upon BCR activation in CLL cells. Peripheral blood CLL cells were incubated with goat F(ab’)2 anti-human IgM (10 mg/mL) for different times. (A) A representative western blot analysis showing SLP76 tyrosine (Y128) and serine (S376) phosphorylation, as well as AKT (S473), ERK (T202/Y204) activation and IκB expression (n=4). Actin was used to verify equal loading. (B) A representative analysis of SLP76 tyrosine phosphorylation (Y128) determined by flow cytometry in CD19+ gated CLL cells after BCR stimulation with goat F(ab’)2 anti-human IgM (10 mg/mL) for 5, 15, and 45 min compared to unstimulated cells.

(C) A summary of the phosphorylation levels of SLP76 (Y128) after IgM stimulation with goat F(ab’)2 anti-human IgM (10 mg/mL) for 15 min, determined by flow cytometry in CD19+ gated CLL cells compared to unstimulated cells (n=20). (D) SLP76 tyrosine phosphorylation (Y128) determined by flow cytometry, in ZAP-70 negative (n=7) and ZAP-70 positive (n=11) patients. (E) Western blot analysis of SLP76 tyrosine phosphorylation (Y128) following 15 min of IgM stimulation, in ZAP-70 negative (n=4) and ZAP-70 positive (n=4) patients.

haematologica | 2016; 101(12)

1557

N. Dezorella et al.

proximal TCR signaling molecules LCK and ZAP-70 are ectopically expressed in CLL cells, with varying levels among patients, while being mostly undetectable in peripheral blood B cells from healthy individuals.31,40 Like SLP76, ZAP-70 expression is also associated with more aggressive disease and correlates with unmutated IGHV status. LCK, a SRC family protein kinase that targets ZAP70 as one of its principal substrates, as well as ZAP-70 itself, both appear to mediate BCR signaling in CLL cells.17,32,41,42 ZAP-70 has also been shown conclusively to enhance BCR response in CLL cells.17,41,42

In normal T cells, phosphorylation of SLP76 at multiple tyrosine residues (Y112, Y128, and Y145) is required for a functional TCR,23 and here we demonstrate that SLP76 is not only ectopically expressed in CLL cells, but is also functional. Engagement of the BCR in CLL cells induces immediate transient tyrosine phosphorylation of SLP76 followed by phosphorylation of an inhibitory serine residue, in a similar manner to that occurring on TCR activation in T cells. In agreement with previously published data regarding BCR signaling, BCR-mediated activation of SLP76 was independent of cholesterol-rich lipid rafts.43 We

A

B

C

D

Figure 3. BCR-induced SLP76 phosphorylation is mediated by LCK and SYK. (A-B) Peripheral blood CLL cells were pre-incubated with the specified inhibitors (PP2: 10 mM for 15 minutes, SYKII: 10 μM SYK inhibitor II for 15 minutes, Ibr: 0.5 mM Ibrutinib for 1 hour CyB: 20 μM Cytochalasin B for 30 minutes, mβCD, 10mM methyl β cyclodextrin for 30 minutes), followed by activation with goat F(ab’)2 anti-human IgM (10 mg/mL). Protein was extracted and analyzed by western blot.

Representative analyses showing in A-Left: induction of SLP76 tyrosine phophorylation (Y128), 15 min after BCR activation and in B-left: SLP76 serine phophorylation (S376), 45 min after anti-IgM engagement, in the presence or absence of each one of the inhibitors. AKT phosphorylation was used to confirm the inhibitory activity of each inhibitor. A-right and B-right: A summary of three experiments showing quantification of SLP76 tyrosine and serine phophorylation in BCR-activated CLL cells pre-incubated with the indicated inhibitors and normalized to control (BCR-activated CLL cells without inhibitors). (C-D) Peripheral blood CLL cells were incubated with: (C) 0.04, 0.2, or 1 mM of LCKi inhibitor for 2 h. DMSO treated cells served as controls. (D) 0.2, 1, or 5 mM of the SYK inhibitor R406 for 30 min. DMSO treated cells served as controls. Both control and inhibitor-treated cells were then activated with goat F(ab’)2 anti-human IgM (10 mg/mL) for 15 min. SLP76 tyrosine phosphorylation (Y128), was then determined by flow cytometry in CD19+ gated cells. C-left and D-left: One representative analysis. C-right and D-right: A summary of SLP76 phosphorylation levels (C) n=4, (D) n=3.

1558

haematologica | 2016; 101(12)

Aberrant expression of SLP76 in CLL

also showed that phosphorylation of SLP76 after BCR engagement occurs regardless of ZAP-70 expression, as it is also induced in ZAP-70-negative CLL cells. However, both LCK and SYK participate in the upstream signaling of SLP76 in CLL cells, as evident from the inhibition of SLP76 tyrosine phosphorylation using inhibitors directed to LCK or Syk.30,32 In addition, SYK but not LCK, appears to maintain a basal level of SLP76 phosphorylation independently of BCR engagement. Our findings are compatible with previous reports which show that SLP76 is phosphorylated in a SYK-dependent manner in several non-T-cell subtypes.29,44 Furthermore, LCK has been reported to mediate phosphorylation of SYK but not ZAP-70 in BCR-stimulated CLL cells,32 whereas tyrosine residues required for ZAP70 kinase activation are not phosphorylated in CLL cells following BCR stimulation.41,42 Another key signaling adaptor protein of the TCR pathway is LAT, which is constitutively localized to the membrane glycolipid-enriched microdomains.23 Upon TCR stimulation, phosphorylation of LAT recruits SLP76 through GADS to the cell membrane, and the two adaptor proteins serve as a scaffold for the formation of a multimolecular signaling complex.23 In contrast to T cells, we show that LAT

A

C

E

haematologica | 2016; 101(12)

is not expressed in CLL. In T cells, LAT is essential for activation of PLCγ1 and the Ras pathway upon TCR ligation,45 while in other cell types signaling involving SLP76 may propagate independently of LAT. In this respect, Mizuno et al. showed that murine B-cell lines express SLP76 and that knockdown of SLP76 attenuated BCR signaling, despite the fact that these cells do not express LAT.46 Additionally, there seems to be a differential requirement of LAT in platelets, and LAT-deficient mice show no bleeding diathesis unlike those animals that are SYK-/- or SLP76-/-.47,48 Here we also show that ITK is not expressed in CLL cells, while BTK is constitutively physically linked to SLP76, an association which increases significantly in response to BCR engagement. In a similar manner, BTK has also been shown to be constitutively associated with SLP76 in mast cells, which increased association after FcεRI ligation.49 It seems that SLP76 operates as an adaptor protein that mediates SYK-dependent BTK activation in CLL cells. Accordingly, the BTK inhibitor ibrutinib had minimal effect on BCR-mediated SLP76 phosphorylation, while knockdown of SLP76 resulted in down-regulation of BTK phosphorylation after BCR stimulation. SLP76 knockdown also significantly inhibited the BCR-induced

B

D

F

Figure 4. SLP76 associates with BTK and modulates BCR signaling. (A, B) IgM activated (for 15 min) and control CLL cell extracts were immunoprecipitated with anti-SLP76 antibody and probed with anti-SLP76 and anti-BTK antibodies. (A) Representative western blot analysis of two patients’ samples.

(B) Summary of immunoprecipitation results. BTK levels are indicated as percentage of control (n=17). (C-F) Peripheral blood CLL cells were transfected with either control or SLP76-specific siRNA or with BLNK-specific siRNA. After 36 h, cells were stimulated with 10 mg/mL goat F(ab’)2 anti-human IgM for 15 min. Following stimulation, cells were lysed and SLP76 levels, BLNK levels, BTK phosphorylation (Y223) levels and PLCγ2 phosphorylation (Y759) levels were detected by western blot. (C, D) Representative western blot analysis showing control and SLP76-knockdown cells (C) or control and BLNK-knockdown cells (D) either stimulated or unstimulated with goat F(ab’)2 anti-human IgM (10 mg/mL). (E-F) The average + SE of the phosphorylation responses of BTK (E) n=12 and PLCγ2 (F) n=10 in control and SLP76-knockdown cells either stimulated or unstimulated with goat F(ab’)2 anti-human IgM (10 mg/mL).

1559

N. Dezorella et al.

phosphorylation of PLCγ2, a more distal signaling molecule of the BCR cascade. Similarly, SLP76 has also been shown to mediate PLCγ2 activation in different cell types.50 Examination of the various distal BCR-activated signaling pathways in SLP76 knockdown CLL cells revealed intact activation of ERK and AKT activation, but impaired activation of the canonical NF-κB pathway, which plays a critical role in CLL pathogenesis.5,51 Our finding, that BLNK knockdown resulted in a similar inhibitory pattern on BTK and IκB, further suggests that SLP76 and BLNK have overlapping or complementary activity, namely orchestrating the formation of the BTKPLCγ2 axis that regulates NF-κB activation in BCR-stimulated CLL cells. Similarly, Tan et al. previously reported that in BCR-stimulated mouse BLNK-/- B cells, activation of NF-kB is impaired, while induction of AKT and mitogenactivated protein kinases are intact.52 Given that the NF-κB pathway has been implicated in regulation of genes essential for B-cell survival, it is not surprising that knockdown of SLP76 partially blocked the pro-survival protection

A

C

induced by BCR activation in CLL cells.

In this study we also demonstrate that higher SLP76 protein levels correlate with a shorter time from diagnosis to disease progression or the initiation of first treatment, which provides support for the biological role of SLP76 in the pathogenesis of CLL. The strong association between SLP76 and other BCR pathway components, such as ZAP70 and IGHV mutational status, suggests that the clinical correlation between SLP76 levels and disease progression is not solely related to SLP76, but is part of a variety of factors that interact to enhance BCR signaling. Some of these components and in particular those indicative of "T-cell origin", may share a common transcriptional regulation involving ZAP-70, LCK, and SLP76.

Experimental evidence accumulated over recent years indicates that BCR-mediated responses in CLL cells are probably very different from those occurring in normal mature B cells. Although key components of the TCR signaling complex are apparently expressed by CLL cells, which are B cells, their biochemical function in CLL cells

B

D

Figure 5. SLP76 expression correlates with disease progression and cell survival. (A, B) Peripheral blood CLL cells were transfected with either control or SLP76specific siRNA or with BLNK-specific siRNA. After 36 h, cells were stimulated with 10 mg/mL goat F(ab’)2 anti-human IgM for 15 min. Following stimulation, cells were lysed and SLP76, BLNK, ERK (T202/Y204) phosphorylation, AKT (S473) phosphorylation and IκB (S32) phosphorylation levels were detected by western blot.

Actin was used to verify equal loading (A) Representative western blot analysis showing control and SLP76-knockdown cells (left) and control and BLNK-knockdown cells (right) either stimulated or unstimulated with goat F(ab’)2 anti-human IgM (10 mg/mL). (B) The average + SE of the phosphorylation responses following IgM stimulation of IkB in SLP76-knockdown (n=6) and BLNK-knockdown (n=4) cells. (C) Peripheral blood CLL cells were transfected with control or SLP76-specific siRNA. After 36 h, cells were stimulated with 10 mg/mL goat F(ab’)2 anti-human IgM for 48 h. Following stimulation, cells were analyzed for apoptosis by flow cytometry.

To determine the percentage of apoptotic cells, annexin V-positive cells were gated out of CD23+ CD5+ cells (n=8). (D) Kaplan-Meier analysis of time to first disease progression (n=22) (left) and time to first treatment (right) dichotomized according to whether patients had low or high SLP76 protein expression. SLP76 expression was determined in purified CLL cells using western blot, quantified and normalized to actin. Separation into low and high subgroups was based on the median SLP76 level. Details of the patients’ characteristics are available in Online Supplement S1; the low SLP76 subgroup includes patients CLL_01 through CLL_11 and the high SLP76 subgroups consists of patients CLL_12 through CLL_22.

1560

haematologica | 2016; 101(12)

Aberrant expression of SLP76 in CLL

is fundamentally different from that evident in the context of the TCR cascade. Our study shows that, unlike in T cells, the scaffold protein SLP76 in CLL cells operates independently of LAT, ZAP-70 or ITK, and also promotes BTK, PLCÎł2 and IÎşB activation upon BCR engagement. Thus, BCR activation, which is essential in CLL pathophysiology, as well as the subsequent signaling cascade it induces are enhanced in CLL cells via this novel modality.

References 1. Caligaris-Cappio F, Hamblin TJ. B-cell chronic lymphocytic leukemia: a bird of a different feather. J Clin Oncol. 1999;17(1): 399-408.

2. Hallek M, Fischer K, Fingerle-Rowson G, et al. Addition of rituximab to fludarabine and cyclophosphamide in patients with chronic lymphocytic leukaemia: a randomised, open-label, phase 3 trial. Lancet.

2010;376(9747):1164-1174.

3. Kurosaki T, Shinohara H, Baba Y. B cell signaling and fate decision. Annu Rev Immunol. 2010;28:21-55.

4. Stevenson FK, Krysov S, Davies AJ, Steele AJ, Packham G. B-cell receptor signaling in chronic lymphocytic leukemia. Blood.

2011;118(16):4313-4320.

5. Herishanu Y, Perez-Galan P, Liu D, et al.

The lymph node microenvironment promotes B-cell receptor signaling, NF-kappaB activation, and tumor proliferation in chronic lymphocytic leukemia. Blood.

2011;117(2):563-574.

6. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38 expression as novel prognostic indicators in chronic lymphocytic leukemia. Blood. 1999;94(6):18401847.

7. Hamblin TJ, Davis Z, Gardiner A, Oscier DG, Stevenson FK. Unmutated Ig V(H) genes are associated with a more aggressive form of chronic lymphocytic leukemia.

Blood. 1999;94(6):1848-1854.

8. Lanham S, Hamblin T, Oscier D, Ibbotson R, Stevenson F, Packham G. Differential signaling via surface IgM is associated with VH gene mutational status and CD38 expression in chronic lymphocytic leukemia. Blood. 2003;101(3):1087-1093.

9. Longo PG, Laurenti L, Gobessi S, Sica S, Leone G, Efremov DG. The Akt/Mcl-1 pathway plays a prominent role in mediating antiapoptotic signals downstream of the B-cell receptor in chronic lymphocytic leukemia B cells. Blood. 2008;111(2):846-855.

10. Petlickovski A, Laurenti L, Li X, et al.

Sustained signaling through the B-cell receptor induces Mcl-1 and promotes survival of chronic lymphocytic leukemia B cells. Blood. 2005;105(12):4820-4827.

11. Deglesne PA, Chevallier N, Letestu R, et al.

Survival response to B-cell receptor ligation is restricted to progressive chronic lymphocytic leukemia cells irrespective of Zap70 expression. Cancer Res. 2006;66(14):7158-7166.

12. Krysov S, Potter KN, Mockridge CI, et al.

Surface IgM of CLL cells displays unusual glycans indicative of engagement of antigen in vivo. Blood. 2010;115(21):41984205.

13. Zupo S, Isnardi L, Megna M, et al. CD38 expression distinguishes two groups of Bcell chronic lymphocytic leukemias with

haematologica | 2016; 101(12)

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

Acknowledgments This research was performed in partial fulfillment of the requirements for a Ph. D. degree of Nili Dezorella, at the Sackler Faculty of Medicine, Tel-Aviv University, Israel.

Funding This work was supported by the Varda and Boaz Dotan Foundation (YH, BK) and the Israeli Science Foundation (YH).

different responses to anti-IgM antibodies and propensity to apoptosis. Blood.

1996;88(4):1365-1374.

Rosenwald A, Alizadeh AA, Widhopf G, et al. Relation of gene expression phenotype to immunoglobulin mutation genotype in B cell chronic lymphocytic leukemia. J Exp Med. 2001;194(11):1639-1647.

Wiestner A, Rosenwald A, Barry TS, et al.

ZAP-70 expression identifies a chronic lymphocytic leukemia subtype with unmutated immunoglobulin genes, inferior clinical outcome, and distinct gene expression profile. Blood. 2003;101(12):4944-4951.

Crespo M, Bosch F, Villamor N, et al. ZAP70 expression as a surrogate for immunoglobulin-variable-region mutations in chronic lymphocytic leukemia. N Engl J Med. 2003;348(18):1764-1775.

Chen L, Widhopf G, Huynh L, et al.

Expression of ZAP-70 is associated with increased B-cell receptor signaling in chronic lymphocytic leukemia.

Blood.

2002;100(13):4609-4614.

Orchard JA, Ibbotson RE, Davis Z, et al.

ZAP-70 expression and prognosis in chronic lymphocytic leukaemia. Lancet.

2004;363(9403):105-111.

Rassenti LZ, Huynh L, Toy TL, et al. ZAP70 compared with immunoglobulin heavychain gene mutation status as a predictor of disease progression in chronic lymphocytic leukemia. N Engl J Med. 2004;351(9):893901.

Rassenti LZ, Jain S, Keating MJ, et al.

Relative value of ZAP-70, CD38, and immunoglobulin mutation status in predicting aggressive disease in chronic lymphocytic leukemia. Blood. 2008;112(5): 1923-1930.

Wiestner A. Emerging role of kinase-targeted strategies in chronic lymphocytic leukemia. Blood. 2012;120(24):4684-4691.

Woyach JA, Johnson AJ, Byrd JC. The B-cell receptor signaling pathway as a therapeutic target in CLL. Blood. 2012;120(6):11751184.

Koretzky GA, Abtahian F, Silverman MA. SLP76 and SLP65: complex regulation of signalling in lymphocytes and beyond. Nat Rev Immunol. 2006;6(1):67-78.

Clements JL, Ross-Barta SE, Tygrett LT, Waldschmidt TJ, Koretzky GA. SLP-76 expression is restricted to hemopoietic cells of monocyte, granulocyte, and T lymphocyte lineage and is regulated during T cell maturation and activation. J Immunol.

1998;161(8):3880-3889.

Motto DG, Ross SE, Wu J, HendricksTaylor LR, Koretzky GA. Implication of the GRB2-associated phosphoprotein SLP-76 in T cell receptor-mediated interleukin 2 production. J Exp Med. 1996;183(4):1937-1943.

Zamoyska R, Basson A, Filby A, Legname G, Lovatt M, Seddon B. The influence of the src-family kinases, Lck and Fyn, on T

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

cell differentiation, survival and activation.

Immunol Rev. 2003;191:107-118.

Zhang W, Sloan-Lancaster J, Kitchen J, Trible RP, Samelson LE. LAT: the ZAP-70 tyrosine kinase substrate that links T cell receptor to cellular activation. Cell.

1998;92(1):83-92.

Boerth NJ, Sadler JJ, Bauer DE, Clements JL, Gheith SM, Koretzky GA. Recruitment of SLP-76 to the membrane and glycolipidenriched membrane microdomains replaces the requirement for linker for activation of T cells in T cell receptor signaling.

J Exp Med. 2000;192(7):1047-1058.

Hendricks-Taylor LR, Motto DG, Zhang J, Siraganian RP, Koretzky GA. SLP-76 is a substrate of the high affinity IgE receptorstimulated protein tyrosine kinases in rat basophilic leukemia cells. J Biol Chem.

1997;272(2):1363-1367.

Quiroga MP, Balakrishnan K, Kurtova AV, et al. B cell antigen receptor signaling enhances chronic lymphocytic leukemia cell migration and survival: specific targeting with a novel Syk inhibitor, R406. Blood.

2009;114(5):1029-1037.

Majolini MB, D'Elios MM, Galieni P, et al.

Expression of the T-cell-specific tyrosine kinase Lck in normal B-1 cells and in chronic lymphocytic leukemia B cells. Blood.

1998;91(9):3390-3396.

Talab F, Allen JC, Thompson V, Lin K, Slupsky JR. LCK is an important mediator of B-cell receptor signaling in chronic lymphocytic leukemia cells. Mol Cancer Res. 2013;11(5):541-554.

Hallek M, Cheson BD, Catovsky D, et al.

Guidelines for the diagnosis and treatment of chronic lymphocytic leukemia: a report from the International Workshop on Chronic Lymphocytic Leukemia updating the National Cancer Institute-Working Group 1996 guidelines. Blood. 2008;111 (12):5446-5456.

Buchner M, Fuchs S, Prinz G, et al. Spleen tyrosine kinase is overexpressed and represents a potential therapeutic target in chronic lymphocytic leukemia. Cancer Res. 2009;69(13):5424-5432.

Cheng S, Ma J, Guo A, Lu P, et al. BTK inhibition targets in vivo CLL proliferation through its effects on B-cell receptor signaling activity. Leukemia. 2014;28(3):649657.

Contri A, Brunati AM, Trentin L, et al.

Chronic lymphocytic leukemia B cells contain anomalous Lyn tyrosine kinase, a putative contribution to defective apoptosis. J Clin Invest. 2005;115(2):369-378.

Tibaldi E, Brunati AM, Zonta F, et al. Lynmediated SHP-1 recruitment to CD5 contributes to resistance to apoptosis of B-cell chronic lymphocytic leukemia cells.

Leukemia. 2011;25(11):1768-1781.

Wang X, Li JP, Chiu LL, et al. Attenuation of T cell receptor signaling by serine phospho-

1561

N. Dezorella et al.

39.

40.

41.

42.

1562

rylation-mediated lysine 30 ubiquitination of SLP-76 protein. J Biol Chem. 2012;287 (41):34091-34100.

Hashimoto S, Iwamatsu A, Ishiai M, et al.

Identification of the SH2 domain binding protein of Bruton's tyrosine kinase as BLNK--functional significance of Btk-SH2 domain in B-cell antigen receptor-coupled calcium signaling. Blood. 1999;94(7):23572364.

Nolz JC, Tschumper RC, Pittner BT, Darce JR, Kay NE, Jelinek DF. ZAP-70 is expressed by a subset of normal human B-lymphocytes displaying an activated phenotype.

Leukemia. 2005;19(6):1018-1024.

Gobessi S, Laurenti L, Longo PG, Sica S, Leone G, Efremov DG. ZAP-70 enhances Bcell-receptor signaling despite absent or inefficient tyrosine kinase activation in chronic lymphocytic leukemia and lymphoma B cells. Blood. 2007;109(5):20322039.

Chen L, Huynh L, Apgar J, et al. ZAP-70 enhances IgM signaling independent of its kinase activity in chronic lymphocytic leukemia. Blood. 2008;111(5):2685-2692.

43. Blery M, Tze L, Miosge LA, Jun JE, Goodnow CC. Essential role of membrane cholesterol in accelerated BCR internalization and uncoupling from NF-kappa B in B cell clonal anergy. J Exp Med. 2006;203(7):1773-1783.

44. Reeve JL, Zou W, Liu Y, Maltzman JS, Ross FP, Teitelbaum SL. SLP-76 couples Syk to the osteoclast cytoskeleton. J Immunol.

2009;183(3):1804-1812.

45. Finco TS, Kadlecek T, Zhang W, Samelson LE, Weiss A. LAT is required for TCRmediated activation of PLCgamma1 and the Ras pathway. Immunity. 1998;9(5):617626.

46. Mizuno K, Tagawa Y, Watanabe N, Ogimoto M, Yakura H. SLP-76 is recruited to CD22 and dephosphorylated by SHP-1, thereby regulating B cell receptor-induced c-Jun N-terminal kinase activation. Eur J Immunol. 2005;35(2):644-654.

47. Clements JL, Lee JR, Gross B, et al. Fetal hemorrhage and platelet dysfunction in SLP-76-deficient mice. J Clin Invest.

1999;103(1):19-25.

48. Poole A, Gibbins JM, Turner M, et al. The

49.

50.

51.

52.

Fc receptor gamma-chain and the tyrosine kinase Syk are essential for activation of mouse platelets by collagen. EMBO J. 1997;16(9):2333-2341.

Kettner A, Pivniouk V, Kumar L, et al.

Structural requirements of SLP-76 in signaling via the high-affinity immunoglobulin E receptor (Fc epsilon RI) in mast cells. Mol Cell Biol. 2003;23(7):2395-2406.

Hidano S, Kitamura D, Kumar L, Geha RS, Goitsuka R. SLP-76 is required for highaffinity IgE receptor- and IL-3 receptormediated activation of basophils. Int Immunol. 2012;24(11):719-727.

Hewamana S, Alghazal S, Lin TT, et al. The NF-kappaB subunit Rel A is associated with in vitro survival and clinical disease progression in chronic lymphocytic leukemia and represents a promising therapeutic target. Blood. 2008;111(9):4681-4689.

Tan JE, Wong SC, Gan SK, Xu S, Lam KP. The adaptor protein BLNK is required for b cell antigen receptor-induced activation of nuclear factor-kappa B and cell cycle entry and survival of B lymphocytes. J Biol Chem. 2001;276(23):20055-20063.

haematologica | 2016; 101(12)

ARTICLE

Chronic Lymphocytic Leukemia

Ibrutinib for relapsed/refractory chronic lymphocytic leukemia: a UK and Ireland analysis of outcomes in 315 patients

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

UK CLL Forum

ABSTRACT

Haematologica 2016 Volume 101(12):1563-1572

I

n 2014, ibrutinib was made available for relapsed/refractory chronic lymphocytic leukemia patients. The UK Chronic Lymphocytic Leukaemia Forum collected data from UK/Ireland patients with a minimum of 1 year follow-up with pre-planned primary endpoints; the number of patients still on therapy at 1 year “discontinuation-free survival” and 1 year overall survival. With a median of 16 months follow up, data on 315 patients demonstrated a 1 year discontinuation-free survival of 73.7% and a 1 year overall survival of 83.8%. Patients with better pre-treatment performance status (0/1 vs. 2+) had superior discontinuation-free survival (77.5% vs. 61.3%; P<0.0001) and overall survival (86.3% vs. 76.0%; P=0.0001). In univariable analysis, overall survival and discontinuation-free survival were not associated with the number of prior lines of therapy or 17p deletion. However, multivariable analysis identified an interaction between prior lines of therapy, age and 17p deletion, suggesting that older patients with 17p deletion did worse when treated with ibrutinib beyond the second line. Overall, 55.6% of patients had no first year dose reductions or treatment breaks of >14 days and had an overall survival rate of 89.7%, while 26% of patients had dose reductions and 13% had temporary treatment breaks of >14 days. We could not demonstrate a detrimental effect of dose reductions alone (1 year overall survival: 91.7%), but patients who had first year treatment breaks of >14 days, particularly permanent cessation of ibrutinib had both reduced 1 year overall survival (68.5%), and also a statistically significant excess mortality rate beyond one year. Although outcomes appear inferior to the RESONATE trial (1 year overall survival; 90%: progression-free survival; 84%), this may partly reflect the inclusion of performance status 2+ patients, and that 17.5% of patients permanently discontinued ibrutinib due to an event other than disease progression.

Introduction The RESONATE trial established the efficacy and tolerability of ibrutinib in relapsed/refractory chronic lymphocytic leukemia (CLL) and led to the licensing of ibrutinib for this indication in the USA and Europe.1,2 In 2014, a named patient scheme (NPS) made ibrutinib available for relapsed/refractory CLL patients in the UK and Ireland who broadly matched RESONATE trial entry criteria. Following closure of the scheme, the UK CLL Forum initiated a service evaluation of data from patients who commenced treatment on the scheme in 2014 with a minimum follow-up of 1 year. Accepting the limitations of retrospective data analysis, the UK CLL Forum executive committee pre-planned the two most objective primary endpoints for the evaluation: 1. Percentage of patients alive and still taking ibrutinib at 1 year (discontinuation-free survival; DFS) and 2. Percentage of 1 year overall survival (OS). As data collection was >12 months after all patients commenced ibrutinib, the 1 year DFS and OS are therefore absolute values that cannot change with further follow-up. The broad proposal with this service evaluation was to assess how the primary endpoints were influenced by basic patient demographics and perhaematologica | 2016; 101(12)

Correspondence: george.follows@addenbrookes.nhs.uk

Received: April 15, 2016.

Accepted: October 10, 2016.

Pre-published: October 18, 2016. doi:10.3324/haematol.2016.147900

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/12/1563

©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.

1563

UK CLL Forum

formance status, aspects of CLL biology and treatmentrelated variables.

Methods All clinicians entering patients into the CLL ibrutinib NPS were asked whether they wished to contribute anonymized data to the UK CLL Forum ibrutinib service evaluation. To meet entry criteria for the evaluation, patients had to have relapsed/refractory CLL having received prior immunochemotherapy, and had at least 1 day of ibrutinib treatment in the NPS, commencing treatment in 2014. Twelve months after the closure of the scheme, the participating clinicians were sent a questionnaire requesting 25 data points per patient, roughly grouped into 9 categories, as set out in the Online Supplementary Table S1.

Clinicians were given a further opportunity to update their data in March 2016. Clinicians were asked to report any clinically significant adverse event (AE) which was possibly related to ibrutinib, and to provide a best response to therapy. Inevitably, there are limitations to the accuracy of AE reporting and response assessments in retrospective analysis, particularly as there is very variable use of CT scanning and bone marrow assessments in nontrial practice. Defining accurate complete and partial remission rates was therefore not possible. Patients were grouped as ‘responder’ if clinicians graded the response to therapy as partial remission (PR) (including PR + lymphocytosis), or 'better', or ‘nonresponder’ for stable disease or worse. Kaplan-Meier survival

analyses, Cox regression and log-rank tests were used for time-toevent analyses, and the assumption of proportional hazards was checked using Schoenfeld residuals. Where the assumption of proportional hazards did not hold, 16 month rates are presented. Data were analyzed using Stata version 14.1.

Results Demographics, disease and patient characteristics Patient data were returned on 315 patients who met entry criteria from 62 hospitals from across the UK (England, Scotland, Wales, Northern Ireland) and the Republic of Ireland. Contributing hospitals, patient numbers contributed and responsible clinicians are detailed in the Online Supplementary Table S2. The median age of patients on the first day of treatment was 69 (range: 4293), with 69% male. The median prior lines of therapy was 2 (range: 1-14), with 48% of patients having received 3 or more prior lines. Specific data on types of prior therapy were not collected. Fluorescence in situ hybridization (FISH) data were provided for 263/315 patients (83.5%).

All 263 patients had FISH for 17p deletion, but testing for other loci was variable between centers. Testing for mutation of TP53 was limited to a small number of academic centers, and the number of patients tested for this mutation is not known, although 3 patients were reported with a TP53 mutation. In total, 90 patients were identified with

A

B

C

D

Figure 1. Kaplan-Meier plots of (A) discontinuation-free survival (DFS) and (B) overall survival (OS) for the whole cohort of 315 patients. Patient outcomes as per preibrutinib performance status showing (C) DFS and (D) OS. PS: performance status.

1564

haematologica | 2016; 101(12)

Ibrutinib for relapsed CLL

a 17p deletion (90/263; 34.2%); Clinician assessed ECOG performance status (PS) was 0/1 in 240 (76.2%) patients (0=78, 1=162) and 2/3 in 74 (23.5%) patients (2=62, 3=12).

One patient was PS 4.

Discontinuation-free and overall survival data From the entire cohort, 73.7% (232/315) of patients were still on therapy at 1 year with an absolute one year survival rate of 83.8% (264/315) (Figure 1A,1B). At the median follow-up of 16 months, OS was 77.4% (95% CI: 71.9–81.9). The primary endpoints of 1 year DFS and OS were then analyzed by demographic, disease-specific and treatment-related criteria. The hazard ratios for this data with 95% confidence limits are presented in Table 1.

Patients with a better performance status pre-treatment had better outcomes, with poorer performance status patients having more than double the risk of discontinuation and/or death: 1 year DFS for PS 0/1 was 77.5% and for PS 2+ was 61.3%; P<0.0001, and OS rates were 86.3% and 76.0% respectively; P=0.0001 (Figure 1C,1D).

Younger patients (median age of 69 or below) fared better in terms of both DFS (1 year rates: 80.7% and 68.2%; P=0.024) and OS (86.7% and 81.2%; P=0.10, Figure 2A) although this did not reach significance for OS. When age was analyzed as a continuous variable, the detrimental consequences for each additional 10 years was statistically significant for both DFS and OS (DFS HR=1.43 (1.14 – 1.80), P=0.01; OS HR=1.51 (1.15 – 1.98), P=0.0025). Male and female patients had no difference in DFS and OS. Although 1 year DFS and OS appeared inferior for 17p- patients compared with wild-type 17p, this was not statistically significant (DFS: 71.1% vs. 77.5%; P=0.74, OS: 84.4% vs. 86.7%; log-rank P=0.86, Figure 2B). It is noteworthy that patients with no FISH data available had a worse DFS and OS. There is no clear explanation for this observation. When the effect of prior therapies was analyzed, no differences could be demonstrated for either DFS or OS for patients treated with 1 prior, 2 prior or 3+ prior lines of therapy (OS: 83.5% 1 line, 82.9% 2 lines and 84.3% 3+ lines; P=0.997, Figure 2C). Furthermore, there

was no suggestion of any separation of the DFS or OS Kaplan-Meier survival curves beyond one year. No data were available on types of prior therapy. Response assessments were available for 311 patients, with 266/311 (85.5%) classified as ‘responder’ by their clinician and 45/311 (14.5%) classified as ‘non-responder’. Responding patients had a markedly superior 1 year DFS and OS compared with non-responding patients (OS: 90.2% vs.

46.7%, P<0.0001, Figure 2D).

All five pre-treatment variables from Table 1 were included in a mutivariable model. When fitted, it became apparent that there were significant interactions for DFS between age and number of prior lines and 17p and number of prior lines (Online Supplementary Table S3). If patients had received 1 line of prior therapy, then the older group patients had similar DFS and OS outcomes to younger patients. However, for patients with 2 prior lines of therapy, age was significantly associated with inferior DFS (a more than 4-fold increase in risk) and showed the same trend with OS (a 2-fold increase, P=0.17). For patients receiving 3 or more prior lines the same trend was seen, but the effect size was much smaller and did not reach statistical significance (a 71% increase in risk of discontinuation or death (P=0.26) and a 76% in the risk of death, P=0.13). The Kaplan-Meier OS plots for younger and older patients separated by prior lines of therapy are shown in Figure 3A and Figure 3B, respectively, while the corresponding DFS curves are shown in the Online Supplementary Figures S1A and S1B. 17p deletion showed a very similar pattern; if patients had received only 1 prior line of therapy, there was no evidence that 17p deletion had a detrimental effect, but with 2 prior lines the risk of discontinuation or death was 4 times higher (P=0.006) and the risk of death was more than double (P=0.13). For 3+ prior lines there was a non-significant increase of 71% in the risk of discontinuation or death (P=0.12) and 82% in the risk of death (P=0.13). It is not clear why the effect was less marked in 3+ prior lines compared with 2 prior lines, and there remains a possibility that there are unknown confounding factors. The Kaplan-Meier OS plots for wild-

Table 1. Univariable analysis of pre-treatment parameters for DFS and OS. *Compares patients with 17p results only **One patient was performance status (PS) 4.

Variable

DFS Events/N

HR(95% CI)

p-value

OS Events/N

HR(95% CI)

P

39/150 55/148

1.00 1.60 (1.06 – 2.42)

0.024

27/150 40/148

1.00 1.50 (0.92 – 2.43)

0.10

50/173 30/90 22/52

1.00 1.08 (0.68 – 1.71) 1.56 (0.94 – 2.57)

0.74*

34/173 19/90 18/52

1.00 1.05 (0.60 – 1.84) 1.95 (1.10 – 3.45)

0.86*

25/85 26/76 47/146

1.00 1.25 (0.72 – 2.18) 1.09 (0.67 – 1.77)

0.71

19/85 17/76 34/146

1.00 0.97 (0.51 – 1.87) 0.98 (0.56 – 1.73)

0.997

64/240 38/74

1.00 2.30 (1.54 – 3.44)

<0.0001

42/240 29/75

1.00 2.47 (1.54 – 3.96)

0.0001

26/93 66/203

1.00 1.12 (0.71 – 1.76)

0.63

19/93 49/203

1.00 1.13 (0.67 – 1.92)

0.65

Age ≤median (69 years) >median TP53 No 17p deletion 17p deletion Missing Prior therapies 1 2 3+ Performance status 0-1 2+** Sex Female Male

HR: hazard ratio; OS: overall survival; CI: confidence interval; DFS: discontinuation-free survival.

haematologica | 2016; 101(12)

1565

UK CLL Forum

type 17p and 17p deleted patients separated by prior lines of therapy are shown in Figure 3C and Figure 3D, respectively, while the corresponding DFS curves are shown in the Online Supplementary Figures S1C and S1D. The association of prior lines with DFS and OS is complicated by the two interactions of age and 17p deletion status. Given the small numbers of events in the subsets of patients it is hard to draw firm conclusions, though it appears clear that for patients who are older and have 17p deletion, the risk of death or discontinuation increases dramatically with more lines of prior therapy (at least a 4-fold increase, HRs range from 4.34 to 17.04). The same more than 2-fold increase in the risk for both DFS and OS for PS 2+ patients was seen in both the multivariable and univariable analysis. There was no evidence of an association (or any interactions) with sex in the multivariable model. As there was missing data for a small group of patients, this variable has not been included in the model presented in the Online Supplementary Table S3. A number of clinicians included individual case histories describing marked quality of life (QoL) improvements in their patients, and 85.2% of patients (248/291) were reported to have an improved QoL with ibrutinib therapy. Clinical suspicion of Richter’s transformation was reported in 9.2% of the whole patient cohort (29/315). Of these 29 patients, the transformation was biopsy-confirmed in 18 patients, i.e., 5.7% of all patients, with 13 biopsy-proven in the first year. Of the 29 patients clinically suspected of

A

Richter’s transformation, 22 (76%) had died by the time of data collection.

Table 2. Dominant reason given for permanently stopping ibrutinib in 83 patients who stopped the drug within the first year of treatment.

Dominant reasons given for stopping ibrutinib before 1 year

Number of patients

Infection 15 Progressive or refractory disease 14 Richter’s transformation 14 (biopsy proven in 12) Hemorrhage/bleeding-related/anticoagulation-related 9 General debility 6 2nd cancer 6 Lower / upper GI toxicity 2/1 Cytopenias 2 Cardiac issues 2 Dermatological 1 Neuropathy 1 Reason for stopping ibrutinib 10 (including 3 not provided patients who died on therapy) GI: gastro-intestinal.

B

C

D

Figure 2. Kaplan-Meier plots of (A) overall survival (OS) of patients older than median age and median age or younger, (B) OS with or without 17p deletion (*P-value for the comparion of patients with and without 17p deletion) (C) OS of patients by number of prior lines of therapy and (D) OS of patients classified by local clinician as ‘responder’ or ‘non-responder’ to ibrutinib therapy. FISH: fluorescence in situ hybridization.

1566

haematologica | 2016; 101(12)

Ibrutinib for relapsed CLL

Treatment discontinuation and dose reduction

rhage/bruising (9 cases), cytopenias (4 cases), lower gastro-intestinal (GI) toxicity (3 cases), and skin rash/dermatological conditions (3 cases). Dose reductions were relatively common, with 26% of patients (82/315) being reduced to 280mg (42 patients) or 140mg (40 patients) lasting from 1 week to permanent dose reduction (median = 6 months). 32 of these 82 patients also had additional treatment breaks ranging from 15 days to permanent discontinuation. The primary reasons given for dose reductions are given in Table 3.

Overall, clinicians reported clinically significant AEs in 56.5% of patients, although a number of these events did not require either dose reduction or treatment breaks. The overall profile of AEs was similar to those in published studies, and included atrial fibrillation (AF) in 5.1% of patients.

We wanted to analyze whether any alterations in therapy potentially compromised outcomes. To assess whether dose reductions/treatment breaks could impact on outcome, we defined a reference group of patients (group A) who had minimal alterations to therapy, defined as having received standard dose ibrutinib with no dose reductions and total treatment breaks no greater than 14 days in the first year. Group B were patients with any dose reductions but no treatment breaks greater than 14 days. Group C included any patient where ibrutinib was withheld for longer than 14 days, either temporarily or permanently, whether or not the patient had any ibrutinib dose reduc-

Survival was poor for the 83 patients who stopped ibrutinib permanently within the first year. Of these patients, 11 died on therapy (median 153 days from the first dose (range: 46-363)), and of the remaining 72 patients, 40 died before 1 year and 8 died within the period of data collection. The median survival for these 72 patients was 95 days after stopping ibrutinib and 319 days from the first dose. Of the 83 patients who permanently stopped ibrutinib in the first year, 28 were broadly due to disease (refractory disease, progressive disease or Richterâ&#x20AC;&#x2122;s transformation) and 55 due to other causes (summarized in Table 2).

Clinicians were asked if the drug was stopped permanently due to an ibrutinib-related AE. For 56/83 patients, the local clinician classified the main reason for stopping was due to an ibrutinib-related AE, while for 27/83 patients the local clinician did not classify the reason for stopping as AE-related. There was a striking difference in the 1 year OS between these 2 groups. Of the patients who stopped for a â&#x20AC;&#x2DC;clinician-definedâ&#x20AC;&#x2122; AE, 29/56 (51.8%) died before 1 year, but for the patients who stopped the drug for reasons other than an AE, mortality was much higher, with 22/27 (81.5%) dying before 1 year.

Thirty four patients had treatment breaks of 14 days or less. Temporary treatment breaks between 15 days and 6 months (median = 28 days) were reported in 41 patients.

The five commonest primary reasons given for these longer treatment breaks were: infection (12 cases), hemor-

A

B

C

D

Figure 3. Kaplan-Meier plots of overall survival (OS) for (A) patients equal to or younger than the median age and (B) patients older than the median age and (C) patients without 17p deletion and (D) patients with 17p deletion stratified by the number of prior lines of therapy.

haematologica | 2016; 101(12)

1567

UK CLL Forum

tions. Kaplan-Meier DFS and OS curves for the 3 groups are presented in Figure 4.

The total number of group A patients was 175, which included 136 patients who continued on ibrutinib unchanged for the whole year, and 21 patients who had up to 14 days temporarily off therapy. PS 0/1 patients were over-represented in group A with 141 PS 0/1 patients (80.6%), compared with 38 (79.2%) in group B and 61 (66.3%) in group C (Chi square P=0.03). There were 18 deaths in group A before 1 year, with 8 patients dying on therapy and 10 patients dying within 14 days of stopping therapy. Of the 18 deaths in group A, major AEs associated with the final illness were: infection (6), progressive CLL (4), Richter’s transformation (2), cardiac problems (1), hemorrhage (1), general debility (1), and not given (3). Group A DFS and OS were both 89.7%. The total number of group B patients was 48, with 18 patients having dose reductions for less than or equal to 6 months (lowest dose 140mg in 11 and 280mg in 7) and 30 patients with dose reductions of >6 months (lowest dose 140mg in 16 and 280mg in 15). There were 4 deaths before 1 year in group B, 2 patients dying on therapy and 2 within 14 days of stopping. Major AEs associated with the final illness were infection (1), upper GI toxicity (1), and not given (2). All 4 deaths occurred in patients who were dose reduced to 140mg. Group B DFS and OS were 89.6% and 91.7%, respectively. There were 92 patients in group C, which included 58 patients with treatment breaks but no dose reductions, and 34 patients who had breaks in therapy and dose reductions. From group C, 32/92 patients were still on ibrutinib at 1 year with 29 patients having died before 1 year. Of the 92 group C patients, 42 were identified by their clinician as having temporary treatment breaks of >14 days in the first year. Of these 42 patients, 8 died before 1 year. Group C DFS and OS were 34.8% and 68.5%, respectively.

Assessing the consequences of dose modifications in a retrospective analysis is inevitably challenging owing to multiple confounding factors, primarily due to the fact that the most ill patients are inevitably the most likely to ‘self-select’ themselves to be dose reduced/stopped early.

In an attempt to control for this, we carried out a post 1

year analysis of patients from group A, B and C, only analyzing patients who were alive in their specific group at 1 year. To be included in this post 1 year analysis, group A patients had to be alive on ibrutinib at the 1 year point with no modifications or breaks of >14 days in the first year, group B had to be alive on ibrutinib at 1 year having had (or having ongoing) dose reductions but no breaks of >14 days. With this prospective analysis from 1 year, we could also split group C into group C1, who were patients who had had temporary breaks of >14 days in the first year, but were alive and taking ibrutinib at 1 year, and group C2, who were patients who had stopped ibrutinib permanently before 1 year, but were alive at 1 year. The split of these patient groups are shown in a flow chart (Online Supplementary Figure S2). Patient numbers were: A=157, B=44, C1=32 and C2=31. The hazard ratios for DFS and OS beyond 1 year are shown in Table 4 and the Kaplan-Meier plots for DFS and OS beyond 1 year are shown in Figure 5.

Patients who have had dose reductions in the first year (group B), rather than treatment breaks (groups C1 and C2) appear to have very similar outcomes to patients who have been treated with no dose reductions (group A), within the constraints of the limited follow-up of this study. However, patients who have had temporary treatment breaks (>14 days) within the first year (group C1) appear to have an almost 4-fold increase in the risk of stopping ibrutinib beyond one year. The same trend is seen for the risk of death post 1 year (P<0.0001), though the assumption of proportional hazards does not hold for this comparison (P=0.015) so the hazard ratios are not valid; at the median follow-up of 16 months the OS rates in groups A and B are very similar (96.5% and 100%, respectively), but these drop to 85.4% in group C1 and just 68.1% in group C2. These combined results suggest that post 1 year survival does not appear to be compromised by dose reductions in ibrutinib, but does appear to be compromised by both temporary and permanent breaks in ibrutinib therapy.

By analyzing the patients alive at one year, we were also able to see whether the number of prior lines of therapy or pre-treatment performance status had any correla-

Table 3. Dominant reason given for ibrutinib dose reductions in 82 patients who dose reduced within the first year of treatment.

Dominant reasons given for dose reducing ibrutinib Lower / upper GI toxicity Cytopenias Infection Physician decision due to general debility Abnormal liver function tests Atrial fibrillation / coagulation issues Hemorrhage / bruising Arthralgias / musculo-skeletal Mouth ulcers Dermatological Cardiac failure Deterioration of Parkinson’s disease Not specified

Number of patients 15 / 2 14 14 10 6 6 5 4 2 1 1 1 1

GI: gastro-intestinal.

1568

haematologica | 2016; 101(12)

Ibrutinib for relapsed CLL

tion with the first year dose reductions and treatment breaks. We could not demonstrate any statistically significant association between the number of prior lines of therapy and either dose reductions or treatment breaks.

However, there did appear to be a correlation between poorer performance status and higher frequency of treatment breaks. Of the 207 PS 0/1 patients alive at one year, they were split between groups A to C2 as follows: 62.8% (A); 17.4% (B); 10.1% (C1); 9.7% (C2). The 58 PS2+ patients alive at one year were split as follows: 48.3% (A); 13.8% (B); 19% (C1); 19% (C2), indicating

that less than half of the poor performance status patients had no treatment modifications by one year, and twice as many had temporary and permanent treatment breaks in the first year compared to PS 0-1 (P=0.033).

Discussion The RESONATE trial established ibrutinib as an effective therapy for relapsed/refractory CLL,1 and ibrutinib is now a recommended therapy in this setting in European

A

B

Figure 4. Kaplan-Meier plots of discontinuation-free survival (DFS) (A) and overall survival (OS) (B) of patients divided into group A, B or C as per definition in the text.

Table 4. Hazard ratios with 95% confidence intervals for OS and DFS for the 4 separate treatment compliance groups.

Treatment group A B C1 C2

Events/N

DFS HR(95% CI)

11/157 4/44 5/32 -

1.00 1.58 (0.49 – 5.06) 3.76 (1.24 – 11.39) -

P

Events/N

OS* HR(95% CI)

0.045

5/157 2/44 5/32 8/31

1.00 1.38 (0.27 – 7.12) 5.76 (1.65 – 20.08) 9.30 (3.04 – 28.45)

P <0.0001

For this analysis, the origin time for DFS and OS was taken as the 1 year time point. *Fails the proportional hazards assumption – HR can only be interpreted as an average over time. HR: hazard ratio; OS: overall survival; CI: confidence interval; DFS: discontinuation-free survival.

haematologica | 2016; 101(12)

1569

UK CLL Forum and US clinical guidelines.3,4,5 There is considerable interest in real-world experience with this drug outside clinical trials, and this UK/Ireland evaluation represents the largest multi-center dataset of ibrutinib patients treated off-trial with a median follow-up of 16 months for surviving patients. Patients in this analysis were treated in 62 centers, ranging from small district general hospitals to large university teaching centers. The 1 year overall survival for the cohort was strikingly better than patients treated in historical relapsed/refractory CLL trials,6,7 however, the patients in this evaluation appeared to fare less well than patients treated in the RESONATE trial. The patients in this study were similar in terms of age, number of prior therapies and 17p deletion status to the patients recruited into the RESONATE trial, which has now been presented with 16 months median follow-up with a 12 month progression-free survival (PFS) of 84% and overall survival of 90%.2 Although our UK/Ireland dataset does not have a PFS, the 1 year absolute survival was inferior to the RESONATE 1 year PFS. Although DFS and PFS are only approximate equivalents, it does appear that the real-world rate of ibrutinib discontinuation rate and death rate appear higher than patients treated within the RESONATE trial. This real-world observation is not limited to the UK/Ireland data. The single-center Mayo clinic data included 124 relapsed/refractory CLL patients with a median follow-up of 6.4 months and has been presented as an abstract.8 The estimated proportion of patients continuing ibrutinib at 6 months was 84% (95% CI: 77-92%) and at 12 months was 70% (95% CI: 5983%), both figures being similar to the UK/Ireland data.

Furthermore, the multi-center Swedish experience presented data on 95 CLL patients treated for a median of 10.2 months, with a 10 month PFS of 77% and OS of 83%.9 There are a number of potential reasons why the rates of ibrutinib discontinuation and survival are likely to be worse in a real-world setting than in a clinical trial.

Patients treated outside of a clinical trial are more likely to have poorer performance status and more comorbidities.

Nearly a quarter of the UK/Ireland patients had a pretreatment performance status that would have excluded them from the RESONATE trial, and 45% of the Swedish patients had pre-treatment criteria that would have excluded them from RESONATE. If only PS 0/1 patients from the UK data are considered, then DFS of 77.5% and OS of 86.3% are closer to the figures from the RESONATE trial. Our data are the first to confirm that a poorer pre-treatment PS (2+) is significantly associated with reduced discontinuation-free and overall survival (16.2% and 9.3% lower at 1 year, respectively). We have also shown that of the patients who were alive at 1 year, the PS 2+ group were significantly more likely to have had treatment breaks during the first year of therapy.

Interestingly, there appears to be an ongoing divergence of survival curves beyond 1 year for good and poor PS patients, suggesting ongoing consequences for patients who are less well when therapy commences.

Patients treated within a clinical trial have more stringent rules for dose modifications/dose interruptions that are likely to translate into higher levels of drug compliance. Dose reductions/breaks were reported in 4% of patients in the RESONATE trial and 10% in the Ohio State series,10 whereas 26% of the UK/Ireland cohort had a dose reduction of ibrutinib (with or without treatment 1570

A

B

Figure 5. Kaplan-meir plots of discontinuation-free survival (DFS) (A) and overall survival (OS) (B) for groups A, B, C1 and C2 showing survival beyond one year.

breaks), and 19% of patients had treatment breaks (temporary and permanent) with no dose reductions. It is difficult to compare the relative frequency of dose modifications with the clinical trial data exactly, as treatment breaks in particular can be classified in different ways.

However, it seems clear that the extent of dose modification was much higher in this UK/Ireland series than in the published trials. The reasons given for dose reductions and treatment breaks predominantly fit within the expected AE profile of relapsed/refractory CLL patients treated with ibrutinib, with infection, cytopenias, bleeding issues, and gastro-intestinal toxicity being the recurring reasons cited for both temporary and permanent dose reductions and therapy breaks. It is not clear why the rates of modifications were so high in our study, although the inclusion of poorer PS patients was a likely contributing factor, and there was variation in practice between centers. We could not, however, demonstrate clear differences in outcomes between centers grouped by size/number treated/university hospital status etc. (data not shown). It seems unlikely that these dose modifications would have been permitted within the context of a clinical trial, and although a direct causal link between dose modifications and inferior outcomes cannot be made from our data, it does appear from our data that treatment breaks in particular are associated with inferior outcomes both at 1 year, and beyond 1 year for patients who were alive and had re-started ibrutinib by the 1 year haematologica | 2016; 101(12)

Ibrutinib for relapsed CLL

time point. In contrast, we could not identify any statistically significant inferiority for DFS and OS up to and beyond 1 year for patients who were only dose reduced but had minimal treatment breaks. Our data therefore suggest that continuous therapy with ibrutinib (excepting minimal breaks) throughout the first year is required for optimal outcomes, but raises the question as to whether 420mg is required to gain maximal benefit from the drug.

Therefore, if clinicians feel the need to dose modify therapy due to an AE, dose reduction may potentially be preferable to treatment cessation. Of course, there are major limitations to our retrospective dataset, particularly the limited follow-up, thus, whether or not dose reductions compromise longer term outcomes will only be answered by prospective clinical trials which are currently recruiting.

With regards to the permanent discontinuation of ibrutinib, it is clear that the drug was stopped in far fewer patients due to an AE in the RESONATE and Ohio State trials (4% and 12% with 9.4 months and 3 years followup, respectively) than in real-world datasets. Despite shorter follow-up in the Swedish and Mayo Clinic datasets, 10.5% and 12.1%, respectively, of patients in these real-world series stopped ibrutinib for an AE other than progressive disease, although both these figures are smaller than the 17.5% observed in the UK/Ireland series.

Together these results suggest that higher rates of ibrutinib discontinuation are to be expected when patients are treated off-trial. When the reasons for permanent discontinuation of ibrutinib are compared between real-world datasets, there are some similarities. In the UK/Ireland, Swedish and Mayo clinic series, infection is the commonest single reason other than Richterâ&#x20AC;&#x2122;s transformation/progressive CLL for permanent discontinuation of ibrutinib, and infection is also the dominant cause of death, other than Richterâ&#x20AC;&#x2122;s transformation/progressive CLL in the UK/Ireland and Swedish series. After stopping ibrutinib within the first year of treatment, a notable feature of our dataset is the short OS. If patients who died while still taking the drug are excluded from the analysis, the median survival was 95 days, which appears shorter than reported in other series.8 The reasons for this are not clear, but the lack of access to alternative non-chemotherapy treatments in the UK/Ireland after ibrutinib discontinuation could be a contributing factor.

Although our data suggested a slightly inferior 1 year DFS for 17p deleted patients (71.1% vs. 77.7%), this was not statistically significant, and OS at 1 year was similar (84.4% vs. 86.7%). This contrasts with published data, where, with longer follow-up, patients with TP53 disruption have worse PFS and OS.10 Potentially, this separation could be seen with our data with a longer followup period. Our data contrasts markedly with the Swedish data, where Kaplan-Meier plots of PFS and OS show very early divergence for patients with 17p deletion. The reasons for these differences are not clear. We also looked at the effect of prior lines of therapy on 1 year outcomes. With the updated abstract presentation of RESONATE at 16 months median follow-up, there is a suggestion that patients treated with 1 prior line of therapy compared with 2+ prior lines of therapy had a statistically meaningful PFS advantage at 12 months (94% vs. 82%). Although it would be reasonable to expect a more heavily pre-treated group of patients to be haematologica | 2016; 101(12)

enriched for poorer prognostic features such as poorer PS and higher levels of 17p deletion, with univariable analysis we could not see any outcome differences for more or less heavily pre-treated patients. With our data, DFS and OS were highly similar for patients treated with 1, 2 or 3+ prior lines of therapy with no suggestion of divergence of survival curves beyond 1 year, although these curves could potentially separate with longer follow-up.

However, when pre-treatment variables of age, sex, PS, 17p status and prior lines of therapy were subject to multivariable analysis, significant interactions were uncovered. PS remains statistically significant, but it also appears that older patients and those with 17p deletion have inferior DFS and OS when treated beyond first relapse. These results are biologically plausible. It is highly likely that a 17p- patient treated with ibrutinib beyond the second line of therapy would have had a subclone of 17p- CLL cells when treated with earlier lines of chemotherapy. Potentially, these earlier lines of treatment could contribute to more genomic complexity and worse outcomes when treated with ibrutinib including and beyond the third line of therapy, although this remains speculation at this stage.

As response assessments in routine practice do not include bone marrow biopsy and CT scan assessments, it was not possible to accurately verify remission status in this evaluation. We therefore grouped all patients who achieved at least a partial remission (or PR + lymphocytosis) together as responding patients. Overall, the response rate of 85% in this study was identical to the investigatorassessed response rate in the RESONATE trial. As expected, patients who were classified by their clinician as responding to therapy demonstrated a markedly superior DFS and OS compared with non-responding patients.

Although we could not demonstrate any clear differences in the incidence of dose reductions/temporary treatment breaks between patients classified as responder or nonresponder (data not shown), the DFS and OS rates for responding patients who had no dose reductions and no treatment break of >14 days were excellent, with 95% (152/160) of patients in this group alive and continuing on ibrutinib treatment at 1 year.

In conclusion, with this presentation of the largest nontrial multi-center dataset of ibrutinib-treated relapsed/refractory CLL patients, we confirm that ibrutinib is a highly effective, generally well tolerated drug in this population, although our data and other real-world datasets suggest overall outcomes in routine clinical practice are inferior to those observed in the pivotal clinical trials. While it seems likely that some of the inadequecy reflects the treatment of poorer PS patients in the nontrial setting, it also remains possible that the unexpectedly high incidence of treatment breaks in the UK/Ireland practice could have been contributory. The lack of access to other CLL therapies in the UK/Ireland could also have contributed to the short OS observed following ibrutinib cessation.

Acknowledgments The UK CLL Forum would like to thank and specifically acknowledge Amy Kirkwood, Statistician, University College London, who has coordinated and run the statistical analysis for this project, and the CLL Chapter of the Lymphoma Forum of Ireland for their collaboration with this project.

1571

UK CLL Forum

References 1. Byrd JC, Brown JR, O'Brien S, et al.

Ibrutinib versus ofatumumab in previously treated chronic lymphoid leukemia. N Engl J Med. 2014;371(3):213-223.

2. Brown JR, Hillmen P, O'Brien S,et al.

Updated efficacy including genetic and clinical subgroup analysis and overall safety in the phase 3 RESONATE trial of ibrutinib versus ofatumumab in previously treated chronic lymphocytic leukemia/small lymphocytic lymphoma. Blood. 2014; 124(21): 3331-3331.

3. Dutch/Belgium HOVON CLL working group. Dutch guidelines for the diagnosis and treatment of chronic lymphocytic leukaemia. Neth J Med. 2016; 74(2):68-74.

4. Interim statement from the BCSH CLL Guidelines Panel. Follows G, Bloor A,

1572

Dearden C, et al. Available from: http://www.bcshguidelines.com/documents/Interim_statement_CLL_guidelines_version6.pdf. Last accessed: 17th October 2016.

5. Zelenetz AD, Gordon LI, Wierda WG, Abramson, et al. Chronic lymphocytic leukemia/small lymphocytic lymphoma, version 1.2015. J Natl Compr Canc Netw.

2015;13(3):326-362.

6. Keating MJ, Flinn I, Jain V, et al.

Therapeutic role of alemtuzumab (Campath-1H) in patients who have failed fludarabine: results of a large international study. Blood. 2002;99(10):3554-3561.

7. Fischer K, Cramer P, Busch R, et al.

Bendamustine combined with rituximab in patients with relapsed and/or refractory chronic lymphocytic leukemia: a multicenter phase II trial of the German Chronic

Lymphocytic Leukemia Study Group. J Clin Oncol. 2011;29(26):3559-3566.

8. Parikh SA, Chaffee KR, Call TG, et al.

Ibrutinib Therapy for Chronic Lymphocytic Leukemia (CLL): An Analysis of a Large Cohort of Patients Treated in Routine Clinical Practice. Blood. 2015; 126(23):2935-2935.

9. Winqvist M, Asklid A, Andersson P, et al.

Real-world results of ibrutinib in patients with relapsed or refractory Chronic Lymphocytic Leukemia: Data from 95 consecutive patients treated in a compassionate use program. Haematologica. 2016 May 19. [Epub ahead of print].

10. Byrd JC, Furman RR, Coutre SE, et al.

Three-year follow-up of treatment-naive and previously treated patients with CLL and SLL receiving single-agent ibrutinib.

Blood. 2015;125(16):2497-2506.

haematologica | 2016; 101(12)

ARTICLE

Chronic Lymphocytic Leukemia

Real-world results of ibrutinib in patients with relapsed or refractory chronic lymphocytic leukemia: data from 95 consecutive patients treated in a compassionate use program. A study from the Swedish Chronic Lymphocytic Leukemia Group

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Maria Winqvist,1,2 Anna Asklid,2,3 PO Andersson,4 Karin Karlsson,5 Claes Karlsson,1,2 Birgitta Lauri,6 Jeanette Lundin,1,2 Mattias Mattsson,7 Stefan Norin,1 Anna Sandstedt,8 Lotta Hansson*1,2 and Anders Österborg*1,2 *LH and AÖ contributed equally to this work

Department of Hematology, Karolinska University Hospital, Stockholm; 2Department of Oncology-Pathology, Karolinska Institutet, Stockholm 3Department of Oncology, Karolinska University Hospital, Stockholm; 4Department of Hematology, South Älvsborg Hospital, Borås; 5Department of Hematology and Vascular Disorders, Skåne University Hospital, Lund; 6Department of Hematology, Sunderby Hospital, Sunderbyn Luleå; 7 Department of Hematology, Uppsala University Hospital, Uppsala and 8Department of Hematology, Linköping University Hospital, Linköping, Sweden 1

Haematologica 2016 Volume 101(12):1573-1580

ABSTRACT

I

brutinib, a Bruton’s tyrosine kinase inhibitor is approved for relapsed/refractory and del(17p)/TP53 mutated chronic lymphocytic leukemia. Discrepancies between clinical trials and routine healthcare are commonly observed in oncology. Herein we report real-world results for 95 poor prognosis Swedish patients treated with ibrutinib in a compassionate use program. Ninety-five consecutive patients (93 chronic lymphocytic leukemia, 2 small lymphocytic leukemia) were included in the study between May 2014 and May 2015. The median age was 69 years. 63% had del(17p)/TP53 mutation, 65% had Rai stage III/IV, 28% had lymphadenopathy ≥10cm. Patients received ibrutinib 420 mg once daily until progression. At a median follow-up of 10.2 months, the overall response rate was 84% (consistent among subgroups) and 77% remained progression-free. Progression-free survival and overall survival were significantly shorter in patients with del(17p)/TP53 mutation (P=0.017 and P=0.027, log-rank test); no other factor was significant in Cox proportional regression hazards model.

Ibrutinib was well tolerated. Hematomas occurred in 46% of patients without any major bleeding. Seven patients had Richter's transformation. This real-world analysis on consecutive chronic lymphocytic leukemia patients from a well-defined geographical region shows the efficacy and safety of ibrutinib to be similar to that of pivotal trials. Yet, del(17p)/TP53 mutation remains a therapeutic challenge. Since not more than half of our patients would have qualified for the pivotal ibrutinib trial (RESONATE), our study emphasizes that real-world results should be carefully considered in future with regards to new agents and new indications in chronic lymphocytic leukemia.

Introduction Chronic lymphocytic leukemia (CLL), the most common leukemia in adults, is characterized by a clonal expansion of CD5+ and CD23+ B-lymphocytes which accumulate in blood, bone marrow and lymphoid tissues. Chemoimmunotherapy is the standard first-line treatment, but patients with del(17p) or TP53 gene mutahaematologica | 2016; 101(12)

Correspondence: lotta.hansson@karolinska.se

Received: February 17, 2016.

Accepted: May 18, 2016.

Pre-published: May 19, 2016. doi:10.3324/haematol.2016.144576

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/12/1573

©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.

1573

M. Winqvist et al.

tion have a poor prognosis and inferior clinical outcomes with such regimens.1,2 Patients who are refractory to multi-agents have a very poor prognosis.3,4 The CLL cell receives survival and proliferation signals from the microenvironment, and the B-cell receptor (BCR) is a key factor in this interaction. Bruton´s tyrosine kinase (BTK) is a non-receptor tyrosine kinase and plays a crucial role in BCR signalling. Ibrutinib is an oral, selective and irreversible inhibitor of BTK. It binds to the cysteine-481 amino acid of the BTK enzyme.5 Next-generation BTK inhibitors are under clinical development with promising early results.6 In the first phase 2 study of ibrutinib, 71% of patients with relapsed or refractory CLL achieved a partial response (PR) according to the International Workshop on Chronic Lymphocytic Leukemia (IWCLL) 2008 criteria,7 and an additional 18% of patients had PR with lymphocytosis (PR-L).8 The response was independent of clinical and genomic risk factors including del(17p). The 3-year followup of this study reported no late occurring toxicity, and progression was uncommon even though a shorter progression-free survival (PFS) for patients with del(17p) was found.9 In a pivotal phase 3 study (RESONATE), patients with previously treated CLL were randomized between treatment with ibrutinib and ofatumumab with a significantly longer PFS and overall survival (OS) in favor of ibrutinib.10 As a consequence of these results, ibrutinib received US Food and Drug Administration approval in February 2014 for patients with CLL who received at least one prior therapy, and approval in July 2014 for all patients with del(17p). Approval by The European Medicines Agency for patients who had received at least one prior therapy, or for all patients with del(17p) or TP53 mutation was granted in October 2014. It was recently shown in a phase 3 study that ibrutinib was superior to chlorambucil in previously untreated patients with CLL, regarding PFS, OS, overall response rate (ORR), and improvement in hematologic variables.11 This might result in a broader indication for ibrutinib therapy.

However, there is often a discrepancy between data obtained from patients strictly included in clinical trials and those obtained outside trials in routine health care, or in trials conducted in a community-based setting, as shown in CLL patients with the fludarabine, cyclophosphamide and rituximab (FCR) regimen.1,12 Trials in CLL generally enrolled younger patients with fewer co-morbidities than in actual clinical practice.13 Thus, knowledge about real-world results in hematology14,15 becomes increasingly important for the optimal usage of new agents in various disorders, including CLL, especially if such agents are to be used continuously until progression.

We report herein real-world results for 95 consecutive Swedish patients with poor prognosis CLL who received ibrutinib treatment in a compassionate use program (CUP), outside the setting of clinical trials.

Methods Study design and participants This retrospective analysis was conducted at 27 Swedish hospitals that included at least one CLL patient in the CUP, which was open for inclusion between May 15, 2014 and May 31, 2015. The program offered free drug access for patients with CLL until ibrutinib was generally available on the market, after which patients 1574

were prescribed ibrutinib capsules according to normal Swedish health care regulations. All patients included in the CUP could be identified. Data was extracted from each patient’s individual medical file and entered into case record forms (CRF) by each treating physician for statistical analysis. The monitoring of data from individual patient files was performed by the academic study team and cross-checked vs. CRFs. The procedure was approved by the regional ethics committee and conducted in accordance with the Declaration of Helsinki.

Patients with CLL were eligible for the CUP if they had: a confirmed diagnosis of CLL or small lymphocytic lymphoma (SLL)

Table 1. Patient characteristics (%) prior to start of ibrutinib (n=95).

Median age, years (range) 69 (42-86) ≥ 70 years 48 ≥ 75 years 23 ≥ 80 years 14 Sex Male 72 Female 28 ECOG 0 17 1 54 2 20 3 9 CIRS 0-3 63 4-5 18 ≥6 19 Bulky disease ≥ 5 cm 50 ≥10 cm 28 Rai stage 0-II 35 III 19 IV 46 Cytogenetics* del(17p)/TP53mut 63 del(11q) 18 trisomy 12 5 del(13q) 8 normal 6 Median ANC, x109/L (range) 2.6 (0-14) Mean hemoglobin, g/L (range) 113 (76-179) Mean platelets x109/L (range) 104 (11-299) Median ALC, x109/L (range) 52 (0-595) Median number of prior therapies (range) 3 (0-9) 0 1 1-2 35 3-4 40 ≥5 24 Median time since last therapy, months (range) 7 (0-116) Prior therapies FCR/FC/F 98 BR/bendamustine 56 Chlorambucil 31 Alemtuzumab 43 Ofatumumab 24 Allogeneic SCT 8 Other agents 53 ANC: absolute neutrophil count; ALC: absolute lymphocyte count; ECOG: Eastern Cooperative Oncology Group; CIRS: Cumulative Illness Rating Scale; FCR: fludarabine + cyclophosphamide + rituximab; FC: fludarabine + cyclophosphamide; F: fludarabine; BR: bendamustine + rituximab. *78 patients tested with complete FISH and 2 patients tested for del(17p)TP53mut only.

haematologica | 2016; 101(12)

Real-world results of ibrutinib therapy in CLL

according to the IWCLL criteria,7 a need for treatment, and a highrisk disease that did not respond to a chemoimmunotherapy regimen or that progressed within 24 months after completion of the regimen. Also included in the CUP were patients with del(17p) or TP53 mutation, who could be included at any time point, patients with an Eastern Cooperative Oncology Group (ECOG) performance status of ≤ 2, neutrophil count ≥ 0.5x109/L, platelet count ≥ 30x109/L, serum creatinine ≤ 2 times, liver enzymes ≤ 3 times and total bilirubin ≤ 1.5 times the upper limit of normal. Key exclusion criteria were: treatment with a strong CYP3A inhibitor or warfarin, allogeneic stem cell transplantation (SCT) within the past 6 months, ongoing active infection, uncontrolled autoimmune hemolytic anemia or immune-mediated thrombocytopenia, Richter’s transformation (RT), and other active malignancies or uncontrolled cardiovascular disease.

Treatment Patients received 420 mg oral ibrutinib once daily until progression or occurrence of unacceptable side effects. Individual dose modifications were allowed as recommended by the manufacturer and as decided by the treating physician.

A

Assessments Endpoints were ORR, which included PR, PR-L and complete remission (CR) rate, as well as PFS, OS and the safety of ibrutinib.

Treatment response was classified according to the IWCLL response criteria of 2008,7 with the exception that lymphocytosis was not the sole criterion for disease progression.10 Response evaluation in patients with SLL was based on the 2007 International Working Group Criteria for non-Hodgkin lymphoma.16 Lymph node response was evaluated clinically and/or by CT scan, as decided by the physician. Evaluation of bone marrow response was done at the discretion of the physician. CR required evaluation by CT scan and bone marrow examination. Cumulative Illness Rating Scale (CIRS) was used to define comorbidities at baseline.17,18 Treatment toxicity was evaluated using the National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events v3.0 except for anemia, thrombocytopenia and neutropenia, which were graded according to the IWCLL grading scale for hematologic toxicity in studies.7 Adverse events (AE) of grade 3 or higher were generally recorded, whereas any grade of AE was reported for hemorrhage, diarrhea, arthralgia, atrial fibrillation and blood counts.

Time to response was defined as the time from start of ibrutinib until the date of fulfilling criteria for PR-L or PR. PFS was defined as the time from start of treatment to progression, the start of new anti-cancer treatment or death from any cause. Two patients proceeded to allogeneic SCT after having responded to ibrutinib and

Table 2. Response rates (%) according to subgroup.

B

Figure 1. Time kinetics of recovery of blood counts. Hemoglobin concentration (A) and platelet count (B) during ibrutinib therapy in patients with baseline cytopenias. Mean +/-SEM at each timepoint is shown. haematologica | 2016; 101(12)

Subgroup

n

ORR*

95% CI

All patients Age < 70 yr ≥ 70 yr < 75 yr ≥ 75 yr Sex Male Female Rai stage 0-II III IV No of previous regimens < 3 regimens ≥ 3 regimens Del(17p)/TP53mut Positive Negative ECOG 0-1 2-3 CIRS <6 ≥6 Bulky disease < 5 cm ≥ 5 cm < 10 cm ≥ 10 cm

95

84

(75.3, 90.9)

49 46 73 22

88 80 85 82

(75.2, 95.3) (66.1, 90.6) (74.6, 92.2) (59.7, 94.8)

68 27

84 85

(72.9, 91.6) (66.3, 95.8)

32 18 43

91 78 81

(75.0, 98.0) (52.4, 93.6) (66.6, 91,6)

34 61

85 84

(69.0, 95.0) (71.9, 91.8)

50 30

80 93

(66.3, 90.0) (77.9, 99.2)

66 26

85 88

(73.9, 92.5) (69.8, 97.6)

76 18

88 67

(78.7, 94.4) (41.0, 86.7)

47 47 68 26

83 85 85 81

(69.2, 92.4) (71.7, 93.8) (74.6, 92.7) (60.6, 93.4)

*Overall response rate (ORR): complete response (CR) + partial response (PR) + partial response with lymphocytosis (PR-L). ECOG; Eastern Cooperative Oncology Group; CIRS: Cumulative Illness Rating Scale.

1575

M. Winqvist et al.

were excluded from PFS analysis. OS was defined as the time from start of ibrutinib treatment to death or latest follow-up.

Statistical analysis Response rates are presented with 95% confidence intervals (CI). Survival (PFS, OS) was estimated by the Kaplan-Meier method. Univariate and multivariate analyses on time to failure were estimated using the Cox proportional regression hazards model. Results are presented as hazard ratios (HR) and 95% CI. All statistics were performed with IBM SPSS, version 23.0.

Results

ibrutinib and were excluded from analysis. Two additional CLL patients were excluded as they were diagnosed with acute myeloid leukemia (AML) at the time of inclusion.

Ninety-five patients (93 with CLL, 2 with SLL) were included in the analysis. Baseline characteristics are shown in Table 1. The median age was 69 years and 23% were 75 years of age or older. Del(17p) and/or TP53 mutation tests were performed in 80 patients and 63% were positive.

Del(11q) without del(17p) was present in 18% of the patients. Sixty-five per cent had Rai stage III or IV. Twenty-eight per cent had lymph nodes â&#x2030;Ľ10 cm. Twentynine per cent had ECOG performance status grade 2 or 3 and 19% had CIRS â&#x2030;Ľ6. A median of 3 prior therapies (range 0-9) was reported.

Patients One hundred and one patients were enrolled between May 15, 2014 and May 31, 2015. Four patients died early or were diagnosed upfront with RT; they never received

A

Efficacy and safety Response rates are shown in Table 2. ORR was 84% (61% PR, 20% PR-L, 3% CR). The ORR among the 42

B

P=0.017

C

D

P=0.027

Figure 2. Overall survival (OS) and progression-free survival (PFS). Kaplan-Meier curves of PFS in (A) all patients and (B) patients with or without del(17p)/TP53 mutation and OS in (C) all patients and (D) patients with or without del(17p)/TP53 mutation.

1576

haematologica | 2016; 101(12)

Real-world results of ibrutinib therapy in CLL

patients who performed a CT scan as part of their response evaluation was 83%. Responses were consistent across subgroups (Table 2). All 8 patients who had a previous allogeneic SCT responded to ibrutinib.

The median time to PR-L and PR was 1.5 (range 0.313.2) and 3.9 (range 0.9-10.9) months, respectively.

Hematologic recovery in patients with pre-existing cytopenias is shown in Figure 1. A gradual improvement of anemia (Figure 1A) and thrombocytopenia (Figure 1B) was observed during treatment. The mean neutrophil count showed a rapid increase early during therapy and thereafter stabilized (data not shown). The response in the bone marrow was analyzed in 26 patients who underwent a second bone marrow biopsy after a median treatment time of 9.5 months (range 2.1-16.4). A >50% reduction of CLL cells in the bone marrow was achieved in 62% of the patients (16/26, including 2 CRs). The ORR in the same 26 individuals was 88% (23/26).

Blood lymphocyte counts reached their peak value after a median time of 4 weeks (range: 2 days to 5.2 months) and thereafter declined slowly. Four patients developed a lymphocyte count of > 500x109/L; two of them developed symptoms of hyperviscosity (at a lymphocyte count of 899 and 917x109/L, respectively) which resolved following leukapheresis.

At a median follow-up time of 10.2 months, the median PFS has not yet been reached. The estimated PFS rate at 10 months was 77% (Figure 2A). PFS was significantly shorter in patients with del(17p)/TP53mut (71% at 10 months) than in those without this cytogenetic aberration (93% at 10 months) (P=0.017, log-rank test) (Figure 2B). CIRS ≥6 was also associated with a significantly shorter PFS (P=0.001, log-rank test). A Cox proportional regression hazards model was performed to check whether baseline characteristics (age, sex, Rai stage, ECOG, number of prior therapies, CIRS, bulky disease ≥5 cm or del(17p)/TP53mut) had an impact on PFS. In this analysis, only del(17p)/TP53mut reached statistical significance (HR 22.18, 95% CI 1.58-311.7, P=0.022) while all other baseline factors were statistically non-significant (Table 3).

Ten patients had progressive disease (PD) during ibrutinib therapy; 7 had RT and 3 had CLL progression. RT occurred after a median time of 9.5 months (range 3.813.1). Six out of 7 patients with RT had baseline del(17p)/TP53mut and 4 patients had bulky disease ≥10 cm.

Median OS was not reached, with a 10-month OS rate of 83% (Figure 2C). OS was significantly shorter in

patients with del(17p)/TP53mut (78% were alive at 10 months) than in non-del(17p) patients (97% were alive at 10 months) (P=0.027, log-rank test) (Figure 2D). Similar to the PFS analysis, CIRS ≥6 was associated with a significantly shorter OS (P=0.006, log-rank test), but del(17p)/TP53mut was the only significant factor in the Cox proportional regression hazards model (HR 14.74, 95% CI 1.05-207.3, P=0.046) (Table 3).

Twenty-three patients (24%) discontinued ibrutinib therapy due to PD (n=10), AE (n=10) or due to other reasons (n=3). Reasons for withdrawal other than PD were infection (n=4), secondary malignancy (n=2), generalized exanthema or blisters (n=2), sudden death of unknown origin (n=1), allogeneic SCT (n=1), diarrhea (n=1), need for strong antithrombotic therapy (n=1) and patient decision (n=1). Two patients were switched to treatment with idelalisib; one due to the need for strong antithrombotic therapy who developed cutaneous toxicity on idelalisib and was switched back to ibrutinib with a maintained response. The second patient discontinued ibrutinib after 3 weeks due to cutaneous toxicity, and thereafter reached a response on idelalisib + rituximab therapy.

Sixteen patients have died. The reasons for death were infection (n=6), RT (n=5), CLL progression (n=2), secondary malignancy (n=2) and sudden death of unknown origin (n=1).

All included patients (n=95) were assessed for safety.

AEs during ibrutinib therapy are listed in Table 4. Ibrutinib was generally well tolerated. The most common nonhematologic grade 3-4 AE was infection, which occurred in 39% of the patients. Diarrhea was reported in 25%; only 1 patient had grade 3. Arthralgia or muscle pain was reported in 15%, all but one were grade 1-2. Hemorrhage occurred in 46%; all were grade 1-2 and no major bleeding was observed. Nineteen patients received concomitant low molecular heparin and 5 patients received aspirin or clopidogrel. There was no increased incidence of bleeding in these patients. Atrial fibrillation occurred in 8 patients (8%) during ibrutinib treatment.

Six patients developed a secondary malignancy during therapy: 4 had squamous cell carcinoma of the skin, one had generalized adenocarcinoma, and one had a suspicion of myelodysplastic syndrome (MDS), which was not confirmed by cytogenetics.

Fifty-three patients (56%) were hospitalized at least once; the major reason was infection (32%). In the remaining patients the reasons for hospitalization varied.

Hematologic toxicity is shown in Table 4. Thirteen per-

Table 3. Cox proportional regression hazards model of factors on PFS and OS.

Baseline factor Age Sex Rai stage ECOG CIRS No. of prior therapies Bulky disease (≥ 5 cm) del (17p)/TP53mut

HR

PFS 95% CI

P

HR

OS 95% CI

P

0.97 0.78 0.53 1.03 2.80 1.20 1.80 22.18

0.89, 1.05 0.18, 3.38 0.13, 2.15 0.30, 3.53 0.79, 9.86 0.90, 1.60 0.44, 7.27 1.58, 311.7

0.45 0.74 0.37 0.97 0.11 0.21 0.41 0.02

0.98 0.72 0.54 0.68 2.58 1.25 1.50 14.74

0.90, 1.07 0.16, 3.33 0.11, 2.61 0.17, 2.69 0.63, 10.57 0.91, 1.72 0.32, 6.89 1.05, 207.3

0.68 0.67 0.44 0.58 0.19 0.16 0.60 0.046

ECOG; Eastern Cooperative Oncology Group; CIRS: Cumulative Illness Rating Scale.

haematologica | 2016; 101(12)

1577

M. Winqvist et al.

cent of the patients had grade 4 neutropenia and 18% had grade 3-4 thrombocytopenia. Eighteen patients (19%) received treatment with granulocyte-colony stimulating factor (G-CSF).

A dose reduction of ibrutinib was conducted in 21 patients; 11 were dose-reduced by one level and 10 by two levels to 140 mg. The reasons were: neutropenia (n=4), hematoma (n=4), infection (n=3), thrombocytopenia (n=2), diarrhea (n=2), the need of a strong CYP inhibitor (n=1), anemia (n=1), cutaneous AE (n=1), arthralgia (n=1), increasing serum creatinine (n=1), and aggravation of atrial fibrillation with fluid retention (n=1).

Cytopenias generally resolved or improved following dose reduction. To check whether dose reduction may impact on the outcome we compared PFS and OS in patients who had a dose reduction lasting longer than 3 months vs. patients with no or only a short, temporary dose reduction. No differences in PFS and OS were observed (data not shown).

Discussion This real-world report on ibrutinib was based on consecutively identified CLL patients from a geographically well-defined region. Ibrutinib was provided through the CUP (from May 2014 to May 2015) and thereafter prescribed as licensed indication. The size of the study group (n=95), in relation to the length of the recruitment period (12 months), is likely to be representative for a general population of advanced phase CLL patients in Sweden.

Nineteen of the 27 participating centers were non-university hospitals, and together our centers covered all health care and geographical regions in the country. Thirty-four of the 95 patients were treated at non-university hospitals.

Furthermore, practically all Swedish CLL patients are followed and treated at hospital-based clinics within the public health care sector, whereas private and office-based hematology is almost absent in our country. The active, nationwide network established within the Swedish CLL Study Group further reduced the risk of missing patients eligible for the CUP. Quality assurance of data was made by the on-site monitoring of individual patient files which were cross-checked. Taken together, we assume that our study represents a real-world population of CLL patients.

The importance of real-world results on patient cohorts, treated outside prospective clinical trials, has been increasingly recognized in hematologic malignancies,14,15 in which complex or toxic regimens are frequently used or new agents are introduced. It was reported that CLL patients included and treated in confirmatory community-based trials responded less well and experienced more side effects than reported in the preceding pivotal studies.1,12 Similar findings have been published on patients with advanced phase CLL when monoclonal antibodies were used as single-agent therapy in routine health care.19,20 Patients included in CLL trials are often younger and have less advanced disease than the general CLL population.13 This is also evident from the current real-world report, in which one fourth of the patients were 75 years or older, more than 60% had del(17p)/TP53mut, most had a reduced ECOG performance status and many had large lymph nodes. Patients included in the pivotal clinical trials on ibrutinib were often younger, had a better ECOG and 1578

Table 4. Adverse Events.1 Hematologic AEs Any Grade (%) Grade 3-4 (%) Anemia 52 0 Thrombocytopenia2 70 18 Neutropenia3 61 33 Non-hematologic AEs excluding infections Diarrhea 25 1 Hemorrhage 46 0 Atrial fibrillation 8 1 Arthralgia 15 1 Hyperviscosity syndrome4 2 Pulmonary embolism 1 Pleural fluid 1 Other5 11 Infections Pneumonia6 18 Febrile neutropenia or septicemia7 11 Other8 12 Expressed as percentage of patients. Grade 1-4 was recorded for hematologic AEs, diarrhea, bleeding, atrial fibrillation and arthralgia. Grade 3-4 only was recorded for other adverse events. 2If platelet count was <20x109/L before therapy the patient was not evaluable for toxicity, according to IWCLL criteria. 3If ANC was < 1x109/L before therapy the patient was not evaluable for toxicity, according to IWCLL criteria. 4Due to ALC of 899 and 917x109/L, respectively. Resolved following leukapheresis. 5Total 10 patients, including exanthema n=2, congestive heart failure n=1, cutaneous ulcers n=1, bone fracture/trauma n=3, subileus n=1, psychiatric disorder n=1 and choking n=1. 6Including 3 due to Pneumocystis jiroveci. 7Including 1 due to Cryptococcus. 8Including 1 cerebral nocardiosis.

1

less pronounced lymphadenopathy.8,10 The ORR of 84% reported in our study is similar to that of the multicenter pivotal trials.8,10 despite the fact that our patients had more poor prognosis characteristics. Many of our patients are still in the relatively early phase of ibrutinib treatment and may possibly demonstrate an improved response with time, as reported in the trials.8,9,21 However, one should be aware that response evaluation in our study was made retrospectively and thus should be viewed with some caution.

The PFS rate of 77% and OS rate of 83% at 10 months is also similar to that of prospective clinical trials,8,10 even though a longer follow-up is required to make a more reliable judgment upon the outcome (PFS and OS). Notably, the PFS and OS were significantly shorter in patients with del(17p)/TP53mut than in those without this aberration.

CIRS â&#x2030;Ľ6 showed a similar pattern but was not statistically significant in the Cox proportional regression hazards model. Similar findings in del(17p) patients were recently reported in the 3-year follow-up of a phase 1b-2 multicenter study9 as well in the RESONATE-17 study.22,23 Thus additional therapeutic actions appear to be required in this difficult-to-treat subgroup of patients. The outcome of del(17p) patients was similar among patients with vs. without bulky disease.

Only limited information is available on the bone marrow effects of ibrutinib. A gradual improvement of cytopenias was described by Byrd et al.,9 providing indirect evidence that ibrutinib induces tumor regression in the bone marrow compartment. This is further supported by our results on longitudinal follow-up of blood counts (Figure 1). In our study, the response rate in the bone marrow was 62% after a median treatment time of 9.5 months; more and deeper remissions are likely to occur over time during prolonged therapy, as recently reported in a follow-up of the NIH conducted clinical trial.24 Ibrutinib was well tolerated. AEs were consistent with haematologica | 2016; 101(12)

Real-world results of ibrutinib therapy in CLL

those reported in prospective clinical trials, with the exception of grade 3-4 infections, which occurred in almost half of our patients. This is likely to be attributable mainly to the advanced disease of our patients, as most infectious AEs appeared early during ibrutinib therapy.

The infection rate was similar in patients with or without del(17p) and was unrelated to CIRS. Well-known side effects such as diarrhea, arthralgia and atrial fibrillation were observed at the expected frequency and severity. A longer follow-up is required to obtain a complete picture of toxicity during ibrutinib therapy. It should be noted that patients with severe cardiovascular disease, warfarin treatment as well as severe neutropenia were excluded from participation in the CUP. RT occurred in 7 patients, but there is currently no evidence to suggest an increased risk of RT during ibrutinib therapy.9,25 Cutaneous hematomas were common, but no patient had a major bleed. When cross-checking individual patient files we often identified uncertainty among physicians on how to deal with anti-coagulative agents (other than warfarin) in various clinical situations and in relation to the platelet counts; guidelines for routine health care are warranted as these situations are not uncommon.

One may argue that the CUP inclusion criteria to some extent may parallel those of clinical trials. We therefore checked whether our patients fulfilled the inclusion criteria of the pivotal phase 3 study (RESONATE).10 Forty-five percent of our patients had one or more exclusion criteria for RESONATE, with poor performance status or cytopenias as the main reasons, further supporting the real-world representativity of our patient material.

In conclusion, our real-world results confirm and extend the data reported in prospective clinical trials on singleagent ibrutinib in high-risk or R/R CLL patients.8,10,22 Good tolerability, a high ORR and a promising PFS were found in our group of consecutive patients, despite the fact that many of them were old and had poor prognostic features.

The shorter PFS and OS in patients with del(17p)/TP53mut highlights the need for continued development of new therapeutic principles for this most difficult-to-treat subgroup of CLL patients. Our study emphasizes that real-world results should be carefully

References 1. Hallek M, Fischer K, Fingerle-Rowson G, et al. Addition of rituximab to fludarabine and cyclophosphamide in patients with chronic lymphocytic leukaemia: a randomised, open-label, phase 3 trial. Lancet. 2010;376 (9747):1164-1174.

2. Zenz T, Eichhorst B, Busch R, et al. TP53 mutation and survival in chronic lymphocytic leukemia. J Clin Oncol. 2010;28(29): 4473-4479.

3. Eketorp Sylvan S, Hansson L, Karlsson C, et al. Outcomes of patients with fludarabinerefractory chronic lymphocytic leukemia: a population-based study from a well-defined geographic region. Leuk Lymphoma.

2014;55(8):1774-1780.

4. Tam CS, O'Brien S, Lerner S, et al. The natural history of fludarabine-refractory chronic lymphocytic leukemia patients who fail alemtuzumab or have bulky lymphadenopathy. Leuk Lymphoma. 2007;48

haematologica | 2016; 101(12)

considered from now on regarding new agents6,26 and new indications11 in CLL. Appendix The following centers and physicians recruited and treated patients in the CUP: South Älvsborg Hospital, Borås (P-O Andersson); Central Hospital in Karlstad, Karlstad (B Hedlund, I Nilsson); Falun Hospital, Falun (M Flogegård); Gävle Hospital, Gävle (A Othzén); Hallands Hospital, Halmstad (C Karlsson, N Kuric); Hallands Hospital, Varberg (C Simonson); Kalmar County Hospital, Kalmar (J Häggström); Karolinska University Hospital, Stockholm (A Asklid, L Hansson, C Karlsson, J Lundin, M Machaczka, S Norin, M Palma, M Winqvist, A Österborg); Kristianstad Central Hospital, Kristianstad (E Horney); Ljungby Hospital, Ljungby (A Swinarska-Kluska); Linköping University Hospital, Linköping (A Bergendahl Sandstedt); Mälarsjukhuset, Eskilstuna (F Pasquariello); Norrland University Hospital, Umeå (M Erlanson, K Forsberg, C Isaksson, M Liljeholm,); Oskarshamns Hospital, Oskarshamn (P Hinnen); Sahlgrenska University Hospital, Gothenburg (M Sender); Skaraborgs Hospital, Lidköping (S Erdal); Skaraborg Hospital, Skövde (R Billström); Skåne University Hospital, Lund and Malmö (E Holm, M Jerkeman, G Juliusson, B Kapas, K Karlsson); Sunderby Hospital, Luleå (L Brandefors, M Johansson, C Kämpe Björkvall, B Lauri); Södersjukhuset, Stockholm (G Lärfars); Uddevalla Hospital, Uddevalla (P Johansson); University Hospital Örebro, Örebro (P Kozlowski, B Uggla); Uppsala University Hospital, Uppsala (M Höglund, M Mattsson); Visby Hospital, Visby (A Aldrin, K Ekman); Växjö Central Hospital, Växjö (J Bjereus, MS Carlsson Alle); Östersund Hospital, Östersund (A Asklund, U Sokolowska-Kolacz), Sweden.

Funding This study was supported by grants from The Swedish Cancer Society (Ref no: 15 0894), The Cancer Society in Stockholm (Ref no: 144142, 151313), King Gustav V Jubilee Fund (Ref no: 144193), The Cancer and Allergy Foundation (Ref no: 150 420, 150 431), StratCan Karolinska Institutet (Proj code: 2201), AFA Insurance (Ref no: 130054) and The Stockholm County Council (Ref no: 20150070), Sweden. We thank Ms Leila Relander for editorial assistance and Dr. Ben King for linguistic corrections.

(10):1931-1939.

5. Pan Z, Scheerens H, Li SJ, et al. Discovery of selective irreversible inhibitors for Bruton's tyrosine kinase. ChemMedChem. 2007;2 (1):58-61.

6. Byrd JC, Harrington B, O'Brien S, et al.

Acalabrutinib (ACP-196) in Relapsed Chronic Lymphocytic Leukemia. N Engl J Med. 2016;374(4):323-332.

7. Hallek M, Cheson BD, Catovsky D, et al.

Guidelines for the diagnosis and treatment of chronic lymphocytic leukemia: a report from the International Workshop on Chronic Lymphocytic Leukemia updating the National Cancer Institute-Working Group 1996 guidelines. Blood. 2008;111 (12):5446-5456.

8. Byrd JC, Furman RR, Coutre SE, et al.

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

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

10.

11.

12.

13.

14.

receiving single-agent ibrutinib. Blood.

2015;125(16):2497-2506.

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

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

Ibrutinib as Initial Therapy for Patients with Chronic Lymphocytic Leukemia. N Engl J Med. 2015;373(25):2425-2437.

Reynolds C, Di Bella N, Lyons RM, et al. A Phase III trial of fludarabine, cyclophosphamide, and rituximab vs. pentostatin, cyclophosphamide, and rituximab in B-cell chronic lymphocytic leukemia. Invest New Drugs. 2012;30(3):1232-1240.

Terasawa T, Trikalinos NA, Djulbegovic B, Trikalinos TA. Comparative efficacy of firstline therapies for advanced-stage chronic lymphocytic leukemia: a multiple-treatment meta-analysis. Cancer Treat Rev. 2013;39(4): 340-349.

Abrahamsson A, Albertsson-Lindblad A, Brown PN, et al. Real world data on primary

1579

M. Winqvist et al.

15.

16.

17.

18.

19.

1580

treatment for mantle cell lymphoma: a Nordic Lymphoma Group observational study. Blood. 2014;124(8):1288-1295.

Ellin F, Landstrom J, Jerkeman M, Relander T. Real-world data on prognostic factors and treatment in peripheral T-cell lymphomas: a study from the Swedish Lymphoma Registry. Blood. 2014;124(10):1570-1577.

Cheson BD, Fisher RI, Barrington SF, et al.

Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol.

2014;32(27):3059-3068.

Goede V, Bahlo J, Chataline V, et al.

Evaluation of geriatric assessment in patients with chronic lymphocytic leukemia: Results of the CLL9 trial of the German CLL study group. Leuk Lymphoma. 2015:1-8.

Parmelee PA, Thuras PD, Katz IR, Lawton MP. Validation of the Cumulative Illness Rating Scale in a geriatric residential population. J Am Geriatr Soc. 1995;43(2):130-137.

Fiegl M, Falkner A, Hopfinger G, et al.

Routine clinical use of alemtuzumab in

patients with heavily pretreated B-cell chronic lymphocytic leukemia: a nationwide retrospective study in Austria. Cancer.

2006;107(10):2408-2416.

20. Moreno C, Montillo M, Panayiotidis P, et al.

Ofatumumab in poor-prognosis chronic lymphocytic leukemia: a phase IV, noninterventional, observational study from the European Research Initiative on Chronic Lymphocytic Leukemia. Haematologica.

2015;100(4):511-516.

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

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

22. O'Brien S, Jones JA, Coutre S, et al. Efficacy and Safety of Ibrutinib in Patients with Relapsed or Refractory Chronic Lymphocytic Leukemia or Small Lymphocytic Leukemia with 17p Deletion: Results from the Phase II RESONATE (TM)-17 Trial. 57th ASH Annual Meeting and Exposition. Orlando, FL, USA. Blood 2014;124(21).

23. Stilgenbauer S, Jones JA, Coutre S, et al.

Outcome of Ibrutinib Treatment by Baseline Genetic Features in Patients with Relapsed or Refractory CLL/SLL with del17p in the Resonate-17 Study. 57th ASH Annual Meeting and Exposition. Orlando, FL, USA. Blood 2015:Abstract 833.

24. Farooqui M, Valdez J, Soto S, et al. Single Agent Ibrutinib in CLL/SLL Patients with and without Deletion 17p. 57th ASH Annual Meeting and Exposition. Orlando, FL, USA. Blood 2015:Abstract 2937.

25. Chanan-Khan A, Cramer P, Demirkan F, et al. Ibrutinib combined with bendamustine and rituximab (BR) in previously treated chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL): First results from a randomized, double-blind, placebocontrolled, phase III study. ASCO Annual Meeting, Chicago, Illinois, USA. 2015: Abstract LBA7005.

26. Roberts AW, Davids MS, Pagel JM, et al.

Targeting BCL2 with Venetoclax in Relapsed Chronic Lymphocytic Leukemia. N Engl J Med. 2016;374(4):311-322.

haematologica | 2016; 101(12)

ARTICLE

Non-Hodgkin Lymphoma

Non-Hodgkin lymphoma and pre-existing conditions: spectrum, clinical characteristics and outcome in 213 children and adolescents

Andishe Attarbaschi,1 Elisa Carraro,2 Oussama Abla,3 Shlomit Barzilai-Birenboim,4 Simon Bomken,5 Laurence Brugieres,6 Eva Bubanska,7 Birgit Burkhardt,8 Alan K.S. Chiang,9 Monika Csoka,10 Alina Fedorova,11 Janez Jazbec,12 Edita Kabickova,13 Zdenka Krenova,14 Jelena Lazic,15 Jan Loeffen,16 Georg Mann,1 Felix Niggli,17 Natalia Miakova,18 Tomoo Osumi,19 Leila Ronceray,1 Anne Uyttebroeck,20 Denise Williams,21 Wilhelm Woessmann,22 Grazyna Wrobel23 and Marta Pillon;2 on behalf of the European Intergroup for Childhood Non-Hodgkin Lymphoma (EICNHL) and the International Berlin-Frankfurt-Münster (i-BFM) Study Group

Pediatric Hematology and Oncology, St. Anna Children's Hospital, Medical University of Vienna, Austria; 2Pediatric Hematology and Oncology, University of Padova, Italy; 3 Department of Pediatrics, Division of Hematology and Oncology, Hospital for Sick Children, Toronto, Canada; 4Pediatric Hematology and Oncology, Schneider Children's Medical Center of Israel, Petah-Tivka, Israel and Sackler Faculty of Medicine, Tel Aviv University, Israel; 5Northern Institute for Cancer Research, Newcastle University, UK; 6 Department of Pediatric Oncology, Institute Gustave-Roussy, Villejuif, France; 7 Department of Pediatric Oncology and Hematology, University Children's Hospital, Banska Bystrica, Slovakia; 8Pediatric Hematology and Oncology, University of Münster, Germany; 9Department of Pediatrics and Adolescent Medicine, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong; 10 Pediatric Hematology and Oncology, Semmelweis University, Budapest, Hungary; 11 Belarusian Research Center for Pediatric Oncology, Hematology and Immunology, Minsk, Belarus; 12Division of Pediatrics, Hematology and Oncology, University Medical Center Ljubljana, Slovenia; 13Pediatric Hematology and Oncology, Charles University and University Hospital Motol, Prague, Czech Republic; 14Pediatric Hematology and Oncology, University Hospital, Brno, Czech Republic; 15Pediatric Hematology and Oncology, Medicine University of Belgrade, Serbia; University Children's Hospital, School of 16 Pediatric Hematology and Oncology, Erasmus MC - Sophia Children's Hospital, Rotterdam, the Netherlands; 17Pediatric Hematology and Oncology, University Hospital, Zurich, Switzerland; 18Pediatric Hematology and Oncology, Federal Center for Pediatric Hematology, Oncology and Immunology, Moscow, Russia; 19Children's Cancer Center, National Center for Child Health and Development, Tokyo, Japan; 20Pediatric Hematology and Oncology, University Hospitals Leuven, Belgium; 21Pediatric Hematology and Oncology, Cambridge University Hospitals Foundation Trust, Addenbrooke’s Hospital, Cambridge, UK; 22Pediatric Hematology and Oncology, Justus Liebig University, Giessen, Germany and 23Bone Marrow Transplantation and Pediatric Hematology and Oncology, Wroclaw Medical University, Poland

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2016 Volume 101(12):1581-1591

1

Correspondence: andishe.attarbaschi@stanna.at

Received: March 30, 2016.

Accepted: August 9, 2016.

Pre-published: August 11, 2016.

ABSTRACT

C

hildren and adolescents with pre-existing conditions such as DNA repair defects or other primary immunodeficiencies have an increased risk of non-Hodgkin lymphoma. However, largescale data on patients with non-Hodgkin lymphoma and their entire spectrum of pre-existing conditions are scarce. A retrospective multinational study was conducted by means of questionnaires sent out to the national study groups or centers, by the two largest consortia in childhood non-Hodgkin lymphoma, the European Intergroup for Childhood non-Hodgkin Lymphoma, and the international BerlinFrankfurt-Münster Study Group. The study identified 213 patients with non-Hodgkin lymphoma and a pre-existing condition. Four subcategories were established: a) cancer predisposition syndromes (n=124, 58%); b) primary immunodeficiencies not further specified (n=27, 13%); c) genetic diseases with no increased cancer risk (n=40, 19%); and d) non-classifiable conditions (n=22, 10%). Seventy-nine of 124 (64%) cancer predispositions were reported in groups with more than 20 patients: ataxia telangiectasia (n=32), Nijmegen breakage syndrome (n=26), constitutional mishaematologica | 2016; 101(12)

doi:10.3324/haematol.2016.147116

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/12/1581

©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.

1581

A. Attarbaschi et al.

match repair deficiency (n=21). For the 151 patients with a known cancer risk, 5-year event-free survival and overall survival rates were 40%±4% and 51%±4%, respectively. Five-year cumulative incidences of progression/relapse and treatment-related death as a first event were 22%±4% and 24%±4%, respectively. Ten-year incidence of second malignancy was 24%±5% and 7-year overall survival of the 21 patients with a second malignan-

cy was 41%±11%. Patients with non-Hodgkin lymphoma and pre-existing conditions have an inferior survival rate with a large proportion of therapy-related deaths compared to patients with non-Hodgkin lymphoma and no pre-existing conditions. They may require special vigilance when receiving standard or modified/reduced-intensity chemotherapy or when undergoing allogeneic stem cell transplantation.

Introduction

Methods

Although causes of childhood cancer remain largely unknown, recent studies have shown germline genetic variants in a variety of genes, and, in particular, mutations in cancer predisposition genes, to be associated with the occurrence of lymphoid malignant diseases such as acute lymphoblastic leukemia (ALL) and nonHodgkin lymphoma (NHL).1-4 Two studies found that 8.5%-10% of children and adolescents with cancer have pathogenetic genomic alterations in cancer predisposition genes, suggesting that family history alone is not a reliable marker for the likelihood of a pre-existing cancer predisposition in any patient with a newly diagnosed malignancy.2-4 A wide spectrum of clinically evident and/or molecularly defined pre-existing conditions carries an increased risk of cancer development, although the reasons for the specific predisposition for lymphoid malignancies such as leukemia and lymphoma have not been fully clarified.5-19 Primary immunodeficiencies (PID) such as DNA repair defects (Nijmegen breakage syndrome, Ataxia telangiectasia, Bloom syndrome or constitutional mismatch repair deficiency), severe combined immunodeficiencies (SCID), common variable immunodeficiencies (CVID), and immune-osseous dysplasias (Di George syndrome or cartilage hair hypoplasia) have an extraordinary risk of developing leukemia and lymphoma.5-17,20,21 Although these patients seem to have an inferior prognosis and an increased risk of treatment-related toxicity and death compared to patients with lymphoid malignancies without a PID, curative therapies including allogeneic stem cell transplantation (allo-SCT) have been repeatedly reported.5,6,10,11,15,16,22 Systematic data on the spectrum of common and rare pre-existing conditions associated with NHL in children and adolescents are scarce with respect to the type of the pre-existing conditions and the clinical characteristics and outcome of the associated NHL subtypes. Thus, the two largest childhood NHL consortia, the European Intergroup for Childhood NHL (EICNHL) and the international Berlin-Frankfurt-Münster (i-BFM) Study Group (SG), designed a retrospective multinational study to collect data on unselected types of pre-existing conditions among children and adolescents with NHL. The study was carried out in full recognition of the limitations imposed by the fact that such a large retrospective study may lack complete accuracy and could, therefore, make unequivocal interpretation difficult.

Between April 2014 and December 2015 we performed an international survey of children and adolescents aged 0-19 years presenting with pre-existing conditions and NHL. Pre-existing conditions were defined as any proven or suspected inherited medical condition. As a prerequisite, the analysis only included patients with nationally centrally-reviewed histopathology of the NHL subtype. Patients diagnosed between 1984 and 2015 (1984-2000: n=45; 2000-2015: n=168) were retrieved from 21 EICNHL and/or i-BFM Study Group members. The survey included questions about demographics and disease [pre-existing condition, NHL subtype, age, sex, stage of disease and pre-therapeutic lactate dehydrogenase (LDH) level], treatment (chemotherapy, radiotherapy and SCT), and outcome (date of progression/relapse, secondary malignancy and death). The letter of invitation and the complete questionnaire are provided at the beginning of the Online Supplementary Appendix and in Online Supplementary Table S1, respectively.

By 1st January 2016, 213 patients with NHL and a pre-existing condition had been identified. Diagnosis of the NHL subtype was based on the contemporary application of morphological and immunophenotypic criteria.23-25 Staging procedures are described in detail elsewhere.26 Patients were included in national studies or registries. In some of these, patients were treated according to treatment guidelines for children with PID/chromosomal breakage syndromes. Of the 213 patients, 174 (82%) were treated according to protocols of the EICNHL, NHL-BFM SG, AIEOP (Associazione Italiana Ematologia Oncologia Pediatrica), UKCCSG (United Kingdom Children's Cancer SG), SFOP (Société Française d'Oncologie Pédiatrique), EORTC (European Organisation for Research and Treatment of Cancer), COG (Children's Oncology Group), and JPLSG (Japanese Pediatric Leukemia/Lymphoma SG).27-42 The remaining 39 patients (18%) received CHOP±rituximab, COP±rituximab, rituximab only, or miscellaneous treatments. The survey asked for information about dose reductions, suspension of chemotherapy elements or single drugs, and individualized therapy approaches. Unfortunately, these were not reported in all patients making it impossible to assess whether modifications of therapy influenced therapy-related toxicity and mortality, response and outcome.

All patients were treated after informed consent from the patient, or the patient’s parents or legal guardian. Treatment studies were conducted according to the Declaration of Helsinki and approved by the respective ethics committees of the participating bodies.

1582

Statistical analysis Event-free survival (EFS) and overall survival (OS) were anahaematologica | 2016; 101(12)

Non-Hodgkin lymphoma and pre-existing conditions

lyzed by the Kaplan-Meier method and compared by the log-rank test. Event-free survival was defined as the time from diagnosis of NHL to the first adverse event (first relapse, progression, secondary malignancy, death from any cause) or date of last follow up.

Overall survival was defined as the time from diagnosis of NHL to death from any cause or date of last follow up. Cumulative incidence functions for competing events were constructed by the method of Kalbfleisch and Prentice and were compared with the Gray’s test. Cumulative incidences of relapse/progression (CIR) were estimated taking into account death without relapse/progression and secondary malignancies (SML) as competing events.

Cumulative incidence of treatment-related death (CID) as a first event was estimated taking into account relapse/progression and SMLs as competing events. Results are shown by percentages and include standard errors. Median follow up was calculated for surviving patients. Statistical analysis was performed using the SAS statistical program (SAS-PC, v.9.3; SAS Institute, Cary, NC, USA).

ResultsA. Attarbaschi et al.

Four subgroups were identified (Table 1): a) known cancer predisposition syndromes (n=124, 58%); b) PIDs not further specified (defined as having no definite genetic characterization; n=27, 13%); c) genetic diseases not known to be associated with an increased cancer risk (n=40, 19%); and d) non-classifiable pre-existing conditions (n=22, 10%), including pre-existing conditions of multifactorial etiology (n=7), organ malformations (n=5), and syndromes not further specified (n=10).

For the entire group of 213 patients, 5-year EFS and OS rates were 45%±4% and 54%±4%, respectively (Online Supplementary Figure S1A and B) after a median follow up of 4.95 years (range 0.33-18.80 years). Five-year CIR and CID were 26%±3% and 19%±3%, respectively (Online Supplementary Figure S1C and D). When restricting out-

Table 1. Four categories of pre-existing conditions in 213 patients with non-Hodgkin lymphoma.

Type of pre-existing condition

Condition (mode of transmission, incidence per live birth)

All patients Cancer predisposition syndrome (n>1 patient)

213 Ataxia telangiectasia (AR, 1:4000) Nijmegen breakage syndrome (AR, 1:100,000) Constitutional mismatch repair disease (AR and AD*, n.a.)

X-linked lymphoproliferative syndrome (recessive, 1:1,000,000) Wiskott-Aldrich syndrome (X-linked recessive, 1:100,000-250,000) Chromosomal breakage syndrome n. f. sp. (AR, n.a.)

Down syndrome (trisomy 21, 1:500-800) Neurofibromatosis type 1 (AD, 1:2500-3300) Cartilage hair syndrome (AR, 1:18.000-23,000) Hyper-IgM syndrome (AR, 0.2:100,000) Hyper-IgE syndrome (AR, 1:1,000,000) Others (each in 1 patient)

Immunodeficiencies n.f.sp.

Primary immunodeficiency n.f.sp.

Common variable immunodeficiency Genetic disease (n>1 patient)

N. patients

G6PD deficiency (X-linked recessive, incidence region-dependent) Cystic fibrosis (AR, 1:2000) Hemophilia A (X-linked recessive, 1:5000) α-1-antitrypsin deficiency (AR, 1:2000-5000) Williams-Beuren syndrome (del 7q11.23, 1:7500-20,000) Prader-Labhart-Willi syndrome (15p aberrations, 1:10,000-15,000) Others (each in 1 patient)

Non-classifiable conditions Multifactorial disease Familial myoclonic epilepsy Autism Gilles de la Tourette syndrome Congenital hearing loss Autoimmune enteropathy Organ malformation Congenital heart disease Malformation of the CNS Kidney agenesis Skeletal malformation Syndromes n.f.sp.

Complex developmental delay Others (each in 1 patient)

124 32 (26%) 26 (21%) 21 (17%) 11 (9%) 7 (6%) 3 (2.5%) 3 (2.5%) 3 (2.5%) 2 (1.5%) 2 (1.5%) 2 (1.5%) 12 (10%) 27 20 (74%) 7 (26%) 40 4 (10%) 2 (5%) 2 (5%) 2 (5%) 2 (5%) 2 (5%) 26 (65%) 22 7 2 (29%) 2 (29%) 1 (14%) 1 (14%) 1 (14%) 5 1 (20%) 2 (40%) 1 (20%) 1 (20%) 10 8 (80%) 2 (20%)

N.: number; n.a.: not available; AR: autosomal recessive inheritance; AD: autosomal dominant inheritance; n.f.sp.: not further specified; G6PD: glucose-6-phopsphate dehydrogenase; CNS: central nervous system.*Lynch syndrome.

haematologica | 2016; 101(12)

1583

A. Attarbaschi et al.

come analysis to the 151 patients with cancer predisposition syndromes and PIDs not further specified, 5-year EFS and OS rates were 40%±4% and 51%±4% (Figure 1A and B) and 5-year CIR and CID were 22%±4% and 24%±4%, respectively (Figure 1C and D) after a median follow up of 4.91 years (range 0.33-17.67 years).

Known cancer predisposition syndromes Among the 124 patients with a cancer predisposition syndrome, 112 (90%) diagnoses were reported in more than 1 patient: ataxia telangiectasia (AT, n=32), Nijmegen breakage syndrome (NBS, n=26), constitutional mismatch repair deficiency (CMMRD, n=21), X-linked lymphoproliferative disease (XLP, n=11), Wiskott-Aldrich syndrome (WAS, n=7), Down syndrome (DS, n=3), chromosomal breakage syndromes not further specified (n=3), neurofibromatosis type 1 (NF1, n=3), and 2 patients each with cartilage hair syndrome, hyper-IgM syndrome, and hyperIgE syndrome. Patients' characteristics and outcomes are shown in Table 2 (diagnoses >2 patients, n=106) and

Online Supplementary Table S2 (diagnoses <3 patients, n=18), respectively.

Male-to-female ratio was 2:1 and median age was 7.98 years (range 0.19-18.20 years). The distribution of histological subtypes was: mature B-cell NHL (n=75, 60%), lymphoblastic lymphoma [T-LBL, n=27, 22%, and B-cell precursor (BCP)-LBL, n=3, 2%], peripheral T-cell lymphoma (PTCL, n=9, 7%), anaplastic large cell lymphoma (ALCL, n=6, 5%), and other NHL subtypes (n=4, 3%).

Twenty-eight of 124 patients (23%) suffered from progression/relapse and 21 (17%) from an SML. In addition, one patient with XLP developed a secondary aplastic anemia and one patient with chromosomal breakage syndrome not further specified developed a hemophagocytic lymphohistiocytosis. Sixty-seven of 124 patients (54%) died, 18 (27%) from primary NHL, 14 (21%) from SML, 30 (45%) from therapy-related toxicity (including 3 SCTrelated deaths); information on the cause of death was not available for 3 patients (4%). Two patients (3%) died from AT-related complications.

A

Years from Diagnosis

B

Time from Diagnosis

1584

Figure 1. Five-year event-free and overall survival rates (A) and 5-year cumulative incidence rates of relapse and treatment-related death (B) of the 151 patients with cancer predisposition syndromes (n=124) and primary immunodeficiencies (PIDs) not further specified (n=27).

haematologica | 2016; 101(12)

Non-Hodgkin lymphoma and pre-existing conditions

Twenty-two of 124 patients (18%) underwent allo-SCT. Of the 16 patients transplanted in first complete remission (CR), 13 (81%) have not yet experienced any event while 3 (19%) died from toxicity. Three of the remaining 6 transplanted patients were transplanted for relapsed disease.

Among these 3 there were 2 deaths; one for aplastic anemia (alive) and one for hemophagocytic lymphohistiocytosis (dead). For the 1 surviving patient, the time point of SCT was not available. After a median follow up of 5.46 years (range 0.33-17.67 years), 5-year EFS and OS rates were 40%±5% and 53%±5%, respectively (Figure 2A and B). Five-year CIR and CID were 24%±4% and 21%±4%, respectively (Figure 3A and B).

Primary immunodeficiencies not further specified Among the 27 patients with a PID not further specified, there were 7 (26%) patients with a common variable immunodeficiency (CVID) and 20 (74%) with an immunodeficiency without a definite genetic characterization as reported by the treating physicians. Characteristics and outcomes of these patients are shown in Table 3.

Male-to-female ratio was 1.1:1 and median age was 5.83 years (range 0.70-17.84 years). The distribution of histological subtypes was: mature B-cell NHL (n=21, 78%), lymphoblastic lymphoma (T-LBL, n=1, 4%), PTCL (n=3, 11%), and ALCL (n=2, 7%). Four of 27 patients (15%) experienced progression/relapse. However, 14 (52%)

Table 2. Clinical and laboratory characteristics and outcome of non-Hodgkin lymphoma patients in 8 categories of cancer predisposition syndromes (n>2 patients).

N. of patients Sex Male Female Age (years) at dg.

Median Range Histology Mature B-NHL Burkitt lymphoma DLBCL PMLBCL B-cell NHL n.f.sp. others T-LBL BCP-LBL ALCL PTCL Others Stage of disease I II III IV Not available LDH ≥500 U/l <500 U/l Not available Relapse/progression Second malignancy Death Disease-related Therapy-related SML-related Other/not available In CR Follow-up (years) Median Range

AT

NBS

CMMRD

XLP

WAS

DS

NF1

CBS n.f.sp.

32

26

21

11

7

3

3

3

21 (66%) 11 (34%)

15 (58%) 11 (42%)

15 (71%) 6 (29%)

11 (100%) 0

7 0

2 1

2 1

2 1

9.88 3.30-17.85

9.95 2.65-18.20

7.1 2.87-17.97

6.04 2.38-16.42

7.52 0.53-17.24

7.64 3.38-14.45

4.47 3.38-10.33

15.57 12.01-15.89

27 (84%) 6 17 0 2 2 0 2 (6%) 2 (6%) 1 (3%) 0

12 (46%) 1 8 0 2 1 5 (19%) 0 1 (4%) 6 (23%) 2 (8%)

3 (14%) 1 2 0 0 0 17 (81%) 1 (5%) 0 0 0

11 (100%) 5 5 0 1 0 0 0 0 0 0

5 1 2 1 1 0 0 0 1 0 1

1 0 1 0 0 0 0 0 0 2 0

1 1 0 0 0 0 2 0 0 0 0

2 0 0 0 2 0 0 0 1 0 0

4 (13%) 10 (31%) 12 (38%) 4 (13%) 2 (6%)

1 (4%) 0 16 (62%) 8 (31%) 1 (4%)

0 0 20 (95%) 1 (5%) 0

0 2 (18%) 3 (27%) 2 (18%) 4 (36%)

1 1 4 0 1

0 0 2 0 1

0 0 3 0 0

0 1 2 0 0

6 (19%) 21 (66%) 5 (16%) 5 (16%) 3 (9%) 24 (75%) 3 16 2 3* 8 (25%)

12 (46%) 13 (50%) 1 (4%) 10 (38%) 4 (15%) 16 (62%) 7 6 2 1* 10 (38%)

5 (24%) 2 (10%) 14 (66%) 7 (33%) 11 (52%) 15 (71%) 6 0 9 0 6 (29%)

1 (9%) 6 (55%) 4 (36%) 0 1 (9%) 1 (9%) 0 1 0 0 10 (91%)

0 4 3 0 1 2 0 1 0 1 5

1 1 1 2 0 1 1 0 0 0 2

0 3 0 0 1 1 0 0 1 0 2

1 1 1 0 0 1 0 1 0 0 2

4.92 1.75-7.21

2.92 0.51-10.85

5.63 1.42-10.23

4.64 1.87-15

5.98 4.43-10.83

1.19 0.88-1.5

6.78 5.46-8.1

6.95 5.77-8.12

AT: ataxia telangiectasia; NBS: Nijmegen breakage syndrome; CMMRD: constitutional mismatch repair disease; XLP: X-linked lymphoproliferative disease; WAS: Wiskott-Aldrich syndrome; DS: Down syndrome; NF1: neurofibromatosis type 1; CBS n.f.sp.: chromosomal breakage syndrome not further specified; N.: number; dg.: diagnosis; B-NHL: B-cell nonHodgkin lymphoma; T-LBL: T-cell lymphoblastic lymphoma; BCP-LBL: B-cell precursor lymphoblastic lymphoma; ALCL: anaplastic large cell lymphoma; PTCL: peripheral T-cell lymphoma; DLBCL: diffuse large B-cell lymphoma; PMLBCL: primary mediastinal large B-cell lymphoma; n. f. sp.: not further specified; LDH: lactate dehydrogenase; SML: secondary malignancy; CR: complete remission. *Two of 3 patients with AT and one patient with NBS died from complications of the underlying DNA repair defect.

haematologica | 2016; 101(12)

1585

A. Attarbaschi et al.

patients died: 2 (14%) from the underlying NHL, 11 (79%) from therapy-related toxicity (including 2 SCT-related deaths), and 1 (7%) patient from complications of the underlying CVID. Eleven of 27 patients (41%) underwent allo-SCT. Ten were transplanted in first CR with only 2 deaths (25%) from toxicity. For 1 patient who died from relapsed disease, the time point of SCT was not available.

After a median follow up of 3.28 years (range 0.85-14.56 years), 5-year EFS and OS rates were 37%±11% and 43%±11%, respectively (Figure 2A and B). Five-year CIR and CID were 18%±9% and 41%±11%, respectively (Figure 3A and B).

Genetic diseases not known to be associated with an increased cancer risk Among the 40 patients within this group, 14 (35%) diagnoses were reported in more than 1 patient: glucose-6phosphate dehydrogenase deficiency (G6PD, n=4), hemophilia A (n=2), cystic fibrosis (n=2), α-1 antitrypsin deficiency (n=2), Prader-Labhart-Willi syndrome (n=2), and Williams-Beuren syndrome (n=2). Characteristics and outcomes are shown in Online Supplementary Table S3 (diagnosed in >2 patients, n=4) and Online Supplementary Table S4 (diagnosed in <3 patients, n=36), respectively.

Male-to-female ratio was 1.8:1 and median age was 10.02 years (range 1.05-18.70 years). The distribution of histological subtypes was: mature B-cell NHL (n=26, 65%), lymphoblastic lymphoma (T-LBL, n=3, 7.5% and BCP-LBL, n=2, 5%), PTCL (n=3, 7.5%), ALCL (n=5, 12.5%), and other NHL subtypes (n=1, 2.5%).

Twelve of 40 patients (30%) experienced progression/relapse and 10 (25%) died, of whom 5 (50%) from the underlying NHL and 5 (50%) from therapy-related toxicity (no SCT-related death). Two of 40 patients (5%) underwent allo-SCT. The patient with CTP synthase 1 deficiency transplanted in first remission is alive. The remaining patient with Kartagener syndrome transplanted for relapsed T-LBL died from progression. After a median follow up of 4.96 years (range 0.39-18.80 years), 5-year EFS and OS rates were 61%±8% and 72%±8%, respectively (Figure 2A and B). Five-year CIR and CID were 35%±10% and 9%±7%, respectively (Figure 3A and B).

Non-classifiable pre-existing conditions Out of the 22 patients with non-classifiable pre-existing conditions, 7 had pre-existing conditions of idiopathic or multifactorial etiology (32%), 5 pre-existing organ malformations (23%), and 10 patients were reported with symptoms and/or a developmental retardation that could not be assigned to a specific syndrome (45%). Characteristics and outcomes are shown in Online Supplementary Table S3 (developmental delay, n=8) and Online Supplementary Table S5 [all other pre-existing conditions (n=14)], respectively.

For the 22 patients with non-classifiable pre-existing conditions, both 5-year EFS and OS rates were 57%±11% (Figure 2A and B) after a median follow up of 4.95 years (range 1.39-14.68 years). The 5-year CIR and CID were 33%±9% and 3%±3%, respectively (Figure 3A and B).

Secondary malignancies after NHL Among the cohort of 151 patients with cancer predisposition syndromes and PIDs not further specified, 21 (14%) developed another malignancy with a 10-year cumulative incidence of SMLs of 24%±5% (Online Supplementary Table S6 and Online Supplementary Figure S2A). There were 1586

no SMLs among the 62 patients with genetic diseases not known to be associated with an increased cancer risk (n=40) and non-classifiable pre-existing conditions (n=32).

Eleven of 21 (52%) patients suffered from a CMMRD, 4 (19%) from NBS, 3 (14%) from AT, and 1 patient each from WAS (5%), NF1 (5%), and XLP (5%). The 15-year cumulative incidence of SMLs according to the pre-existing condition is shown in Online Supplementary Figure S2B; however, curves for WAS, NF1 and XLP have to be interpreted very cautiously due to the very small number of patients and events.

Male-to-female ratio was 15:6. The distribution of histological subtypes at primary diagnosis was: mature B-cell NHL (n=8, 38%), lymphoblastic lymphoma (T-LBL, n=11, 52%, and BCP-LBL, n=1, 5%), and 1 T-cell NHL not furTable 3. Clinical and laboratory characteristics and outcome of nonHodgkin lymphoma patients in 2 categories of immunodeficiencies not further specified.

N. patients Sex Male Female Age (years) at dg.

Median Range Histology Mature B-NHL Burkitt lymphoma DLBCL MZL B-cell NHL n.f.sp.

T-LBL BCP-LBL ALCL PTCL Stage of disease I II III IV Not available LDH ≥500 U/l <500 U/l Not available Relapse/progression Second malignancy Death Disease-related Therapy-related SML-related Underlying PID In CR Follow up (years) Median Range

PID n.f.sp.

CVID

20

7

10 (50%) 10 (50%)

4 3

4.75 0.7-15.5

9.67 1.25-17.84

16 (76%) 0 11 1 4 1 (5%) 0 1 (5%) 2 (10%)

5 1 3 1 0 0 0 1 1

1 (5%) 3 (15%) 8 (40%) 4 (20%) 4 (20%)

1 1 2 2 1

6 (30%) 8 (40%) 6 (30%) 3 (14%) 0 9 (45%) 1 8 0 0 11 (55%)

1 4 2 1 0 5 1 3 0 1 2 (29%)

3.28 0.85-14.56

3.90 0.92-6.88

PID n.f.sp.: primary immunodeficiency not further specified; CVID: common variable immunodeficiency; N.: number; dg.: diagnosis; B-NHL: B-cell non-Hodgkin lymphoma; T-LBL: T-cell lymphoblastic lymphoma; BCP-LBL: B-cell precursor lymphoblastic lymphoma; ALCL: anaplastic large cell lymphoma; PTCL: peripheral T-cell lymphoma; DLBCL: diffuse large B-cell lymphoma; MZL: marginal zone lymphoma; n.f.sp.: not further specified; LDH: lactate dehydrogenase; SML: secondary malignancy; CR: complete remission.

haematologica | 2016; 101(12)

Non-Hodgkin lymphoma and pre-existing conditions

ther specified (5%). The subtypes of SML were: an NHL other than seen at primary diagnosis (n=8, 38%), brain tumor (n=5, 24%), colorectal carcinoma (n=4, 19%), myelodysplastic syndrome (n=1, 5%), hepatoblastoma (n=1, 5%), ALL (n=1, 5%), and acute myeloid leukemia (n=1, 5%).

Eighteen patients (86%) died, 14 (78%) from the SML itself, 1 (6%) from relapse of primary NHL, and 1 (6%) from therapy-related toxicity; the cause of death was not available for 2 (11%) patients. The remaining 3 patients (14%) are surviving in CR after a median follow up of 9.19 years (range 4.10-9.26 years). The 7-year OS rate for the 21 patients was 41%Âą11% (Online Supplementary Figure S3). Analysis of 5-year OS rates according to the pre-existing condition was not possible due to the very small number of patients and events.

Outcome in patients with NHL and pre-existing conditions according to histological subtypes We compared outcome of the different histological sub-

types (mature B-cell NHL vs. T- and BCP-LBL vs. other NHLs) among the 5 largest distinct entities (diagnosis >20 patients) included in the survey: AT (n=32), NBS (n=26), CMMRD (n=21), PIDs not further specified (n=27), and genetic diseases not known to be associated with an increased cancer risk (n=40). However, due to the very small number of patients and events, along with the distribution of histological subtypes, we could not calculate EFS, OS, CIR or CID. Descriptive results are shown in Online Supplementary Table S7. We also wanted to compare outcome of the different histological subtypes (mature Bcell NHL, T- and BCP-LBL and other NHLs) across the 5 largest distinct entities (diagnosis >20 patients) included in the survey. However, again, due to the very small number of patients and events along with the distribution of histological subtypes, we could not calculate EFS, OS, CIR or CID. Descriptive results are shown in Online Supplementary Table S8.

Comparisons of the initial characteristics (sex, age, histological sub-entities and stage of disease) of the 3 major

A

B

Figure 2. Five-year event-free (A) and overall survival (B) rates of the 124 patients with a cancer predisposition syndrome, 27 patients with a primary immunodeficiency (PID) not further specified, 40 patients with genetic diseases and 22 patients with non-classifiable pre-existing conditions.

haematologica | 2016; 101(12)

1587

A. Attarbaschi et al.

histological subtypes (mature B-cell NHL, LBL and ALCL), including the 151 patients with cancer predisposition syndromes and PIDs not further specified only, with a representative cohort from the literature (NHL-BFM 95 trial for B-NHL and LBL, and EICNHL ALCL99 trial) are shown in Online Supplementary Table S9.31,34,43

Discussion The current analysis represents by far the largest cohort of children and adolescents up to 19 years of age with NHL and a pre-existing condition (n=213) reported in the literature. This retrospective study was only made possible by the collaborative effort of the two largest consortia in pediatric and adolescent NHL: the EICNHL and the iBFM SG. The results presented here not only show the wide spectrum of possible pre-existing conditions, that for further analysis had been subdivided into 4 subcategories,

but also show conditions which have not yet been observed with the development of NHL, including genetic diseases such as Îą-1 antitrypsin deficiency and CTP synthase 1 deficiency, as well as chromosomal conditions such as Smith-Magenis syndrome, Silver-Russel syndrome, Cri-du-chat syndrome, Turner syndrome or Triple X syndrome.44 Since it seems unlikely that particularly the third subcategory (i.e. genetic diseases without a known cancer risk) should include so many newly discovered lymphoma-prone disorders, the results might simply represent coincidental findings. Nevertheless, a prospective evaluation of the associated NHL subtypes, their therapy and its tolerance, as well as outcome, could be of vital importance to treating physicians, especially in view of the retrospectively high relapse rate observed.

Not surprisingly, the analysis of those 151 patients with cancer predisposition syndromes and PIDs not further specified demonstrated that patients with NHL and a pre-existing condition have an inferior EFS (5-year EFS:

A

B

Figure 3. Five-year cumulative incidence rates of relapse (A) and treatment-related death (B) of the 124 patients with a cancer predisposition syndrome, 27 patients with a primary immunodeficiency (PID) not further specified, 40 patients with genetic diseases and 22 patients with non-classifiable pre-existing conditions.

1588

haematologica | 2016; 101(12)

Non-Hodgkin lymphoma and pre-existing conditions

40%±4%) with a large proportion of deaths being therapy-related (5-year CID: 24%±4%), as compared to what is known from patients with NHL without a pre-existing condition.27,28,31,34,35,37,38 A comparative analysis showed long-term results from the NHL-BFM SG with 3-year EFS rates of 89%±1% for mature B-cell NHL in trial NHLBFM 95 [n=505 (toxic death n=10)] and 5-year EFS rates of 82%±3% and 88%±3% for stage III/IV T-LBL in trials NHL-BFM 95 [n=156 (toxic death n=2)] and 90+86 [n=163 (toxic death n=3)].31,34 Moreover, 2-year EFS rates were 74.1% (95%CI: 69.2%-78.4%) for ALCL in the EICNHL trial ALCL99 [n=352 (toxic death n=4)] and 45%±5% for an international cohort of 143 patients with PTCL.43,45 Notably, 36 of 143 (25%) had a pre-existing condition and fared very poorly, with 5-year EFS rates of 11%±7%.45 In summary, half of our patients suffered from an event, with half of these patients experiencing a relapse/progression; the other half died from therapy-related toxicity.

Moreover, our study also discovered a male predominance in children and adolescents with pre-existing conditions and NHL (approx. 65% were male) for which there is still no explanation, and further studies are warranted to reveal the biological background of why the male sex is more often affected.17 The largest subgroup of patients comprised well-known cancer predispositions (n=124) with DNA repair defects such as AT (n=32), NBS (n=26), and CMMRD (n=21), accounting for 65% of all patients in this subcategory.

Within these 3 predominant groups, we identified variations in the NHL histological subtypes. Whilst mature Bcell NHL was seen in approximately 85% of patients with AT (with DLBCL 2.8-times more frequent than Burkitt lymphoma), patients with NBS had a high incidence of PTCL (approx. 25%) and patients with CMMRD had a very high incidence of T-LBL (approx. 80%).

The association of DNA repair defects with lymphoid malignancies is well known and has been repeatedly reported with acceptable outcomes which can, however, be improved.5-8,10,16,22,46 Our analysis showed that while approximately 50% of AT patients died from therapy-related toxicity, death from NHL and therapy-related complications was evenly distributed among NBS patients. In comparison, the NHLBFM SG reported on 37 patients with chromosomal instability syndromes (AT and NBS) with 5-year EFS rates of 48%±12% for 16 patients with ALL/T-LBL (relapse/progression n=2; SML n=2; treatment-related death n=2; alive n=10) and 51%±16% for 21 patients with mature B-cell NHL (relapse/progression n=2; SML n=2; treatment-related death n=1; dead from underlying disorder n=4; alive n=12).6 Of our CMMRD patients, nearly 75% died, either from NHL or another SML. This extremely high mortality rate also confirms that CMMRD has to be diagnosed in a timely fashion to allow tumor surveillance, and that adapted chemotherapies are necessary to treat the different highly resistant childhood cancers occurring in this group of patients.9,47 In contrast, patients with XLP (n=11) and WAS (n=7), mainly suffering from mature B-cell NHL (5 of 11 and 2 of 5 with DLBCL), had a comparatively good outcome with 70%-90% of the patients in first CR. Among the other diagnoses of cancer predisposition syndromes, death from therapy-related complications accounted for most events.

Thus, although not all diagnoses can be viewed uniformly, haematologica | 2016; 101(12)

our analysis showed an inferior survival (5-year OS: 53%±5%) with a large proportion of deaths being therapy-related (5-year CID: 21%±4%) for children and adolescents with NHL and a cancer predisposition syndrome, as compared to what is known from NHL patients without cancer predispositions.27,28,31,34,35,37,38 Interestingly, despite the low number of 21 patients (13%) with a cancer predisposition who underwent allo-SCT in first CR, their outcome was rather good, with 17 (81%) of them surviving event free, suggesting that cure from both the underlying disease and NHL is possible.

The second subcategory included patients with a PID not further specified (n=27), including 7 children who were reported with a CVID. Three-quarters had mature B-cell NHLs (14 of them with DLBCL). Assuming that this group of patients could have been genetically characterized, thus belonging to defined cancer predisposition syndromes, death from therapy-related toxicity in 41%±11% of the patients highlights the same dilemma of increased therapy-related toxicity, as seen in patients with defined cancer predispositions. However, again, allo-SCT in first CR in 10 patients with 8 surviving event free suggests that long-term survival is possible.

The third subcategory was made up of patients with genetic disorders not known to be associated with an increased cancer risk (n=40) with only G6PD deficiency (n=4) seen in more than 2 patients. Notably, hemophilia A, cystic fibrosis, α-1 antitrypsin deficiency, PraderLabhart-Willi syndrome and Williams-Beuren syndrome were reported each in 2 patients, and approximately 65% of all patients had a mature B-cell NHL. Outcome analysis showed that, in this cohort of patients, therapy-related death (5-year CID: 9%±7%) was not in the same range as for the first 2 subcategories, but that the 5-year CIR of 35%±10% was rather high as compared to what is seen in the pediatric NHL population without genetic diseases.27,28,31,34,35,37,38 In comparison, long-term results from the NHL-BFM SG showed only 39 (8%) tumor failures for mature B-cell NHL in trial NHL-BFM 95 (n=505), and 18 (12%) and 14 (9%) tumor failures for stage III/IV T-LBL in trials NHL-BFM 95 (n=156) and 90+86 (n=163).31,34 In addition, 84 patients (24%) with ALCL had a tumor failure in the EICNHL trial ALCL99 (n=352).43 However, whether the observed increase in relapse rates in NHL patients with genetic disorders without a known cancer risk was due to the inability to administer sufficient anti-neoplastic therapy could not be assessed as details on therapy reduction and modification were lacking.

Although we had a unique opportunity to analyze a cohort of more than 200 patients with NHL and an underlying condition, the present study has several limitations.

1) Undoubtedly, a certain number of patients with preexisting conditions that were not documented in the participating countries upon diagnosis of NHL is missed given the inherent features of this type of project. In addition, the retrospective nature may have led to both an under- and an over-representation of certain diagnoses, thereby influencing results; this drawback was managed by subcategorizing the study cohort and limiting most outcome analyses to the 3 largest subcategories only. 2) Since we did not ask for all patients with NHL with or without a pre-existing condition within the respective countries and centers to be included in the study, we were not able to assess the relative frequency of constitutional 1589

A. Attarbaschi et al.

disorders within the general NHL population. Therefore, we were not able to compare characteristics and outcome between them. 3) The retrospective nature of the study, and lack of detailed data on chemotherapy and dose modifications, therapy-related toxicities and conditioning regimens used for SCT, did not allow us to assess the impact of these parameters on outcome. 4) As we had not asked for all patients with pre-existing conditions registered within the respective countries to be included in the study, we could also not provide the relative risk of patients with a respective pre-existing condition to develop NHL, as compared to the general population. 5) Unfortunately, we also had no information on whether the diagnosis of, in particular, some DNA repair defects and other PIDs was made before the diagnosis of NHL or whether unexpectedly severe toxicity during therapy led the treating physicians to search for underlying diseases and to subsequent diagnosis.

However, our large collaborative effort could be the first step towards establishing a multinational registry for children and adolescents with pre-existing conditions and NHL, thus not only identifying which disorders are particularly prone to develop NHL, but also facilitating therapy trials that consider the specific needs of these patients.

At least those subcategories of cancer predisposition syndromes and PIDs not further specified represent distinct subsets of patients who are especially prone to treatmentrelated toxicity and may need special vigilance when receiving standard or modified/reduced-intensity chemotherapy or allo-SCT.

References 8.

1. Bartram T, Burkhardt B, Wossmann W, et al. Childhood acute lymphoblastic leukemia-associated risk-loci IKZF1, ARID5B and CEBPE and risk of pediatric non-Hodgkin lymphoma: a report from the Berlin-Frankfurt-Munster Study Group.

Leuk Lymphoma. 2015;56(3):814-816.

2. Maris JM. Defining Why Cancer Develops in Children. N Engl J Med. 2015; 373(24):2373-2375.

3. Mody RJ, Wu YM, Lonigro RJ, et al.

Integrative Clinical Sequencing in the Management of Refractory or Relapsed Cancer in Youth. JAMA. 2015;314(9):913925.

4. 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.

5. Arico M, Mussolin L, Carraro E, et al. NonHodgkin lymphoma in children with an associated inherited condition: A retrospective analysis of the Associazione Italiana Ematologia Oncologia Pediatrica (AIEOP).

Pediatr Blood Cancer. 2015;62(10):17821789.

6. Bienemann K, Burkhardt B, Modlich S, et al. Promising therapy results for lymphoid malignancies in children with chromosomal breakage syndromes (Ataxia teleangiectasia or Nijmegen-breakage syndrome): a retrospective survey. Br J Haematol.

2011;155(4):468-476.

7. Dembowska-Baginska B, Perek D, Brozyna A, et al. Non-Hodgkin lymphoma (NHL) in children with Nijmegen Breakage syn-

1590

9.

10.

11.

12.

13.

14.

15.

16.

Acknowledgments We thank all participating institutions and physicians for their support of the study. This EICNHL and i-BFM paper was written on behalf of the Berlin-Frankfurt-Münster (BFM) Study Group (Austria, Germany, Switzerland, Czech Republic), Associazione Italiana Ematologia Oncologia Pediatrica (AIEOP), Société Française de Lutte contre les Cancers et Leucémies de l'Enfant (SFCE), United Kingdom Children’s Cancer and Leukemia Study Group (CCLG), Belgian Society of Pediatric Hematology and Oncology, Dutch Childhood Oncology Group (DCOG), Hungarian Pediatric Oncology Network, Israel's Society of Pediatric Hematology and Oncology, Slovenian Society of Hematology and Oncology, Slovakian Pediatric Association (Section of Pediatric Hemato-Oncology), Serbian Society of Hematology and Oncology, Polish Society of Pediatric Oncology and Hematology, Japanese Pediatric Leukemia/Lymphoma Study Group (JPLSG), Hong Kong Pediatric Hematology and Oncology Study Group (HKPHOSG) and three single institutions from Canada (Toronto), Belarus (Minsk) and Russia (Moscow).

Funding This work was supported by Cancer Research United Kingdom, the Forschungshilfe Station Peiper (BFM Germany), the St. Anna Kinderkrebsforschung (BFM Austria), the Czech Ministry of Health supported projects for conceptual development of research organization 00064203 and 65269705 (BFM Czech Republic), Project N. LQ1605 of the National Program of Sustainability II (Czech Republic) and the Ministry of Health, Labour and Welfare of Japan (JPLSG).

drome (NBS). Pediatr Blood Cancer.

2009;52(2):186-190.

Gladkowska-Dura M, DzierzanowskaFangrat K, Dura WT, et al. Unique morphological spectrum of lymphomas in Nijmegen breakage syndrome (NBS) patients with high frequency of consecutive lymphoma formation. J Pathol.

2008;216(3):337-344.

Lavoine N, Colas C, Muleris M, et al.

Constitutional mismatch repair deficiency syndrome: clinical description in a French cohort. J Med Genet. 2015;52(11):770-778.

Sandoval C, Swift M. Treatment of lymphoid malignancies in patients with ataxiatelangiectasia. Med Pediatr Oncol.

1998;31(6):491-497.

Schutte P, Moricke A, Zimmermann M, et al. Preexisting conditions in pediatric ALL patients: Spectrum, frequency and clinical impact. Eur J Med Genet. 2015;59(3):143151.

Seif AE. Pediatric leukemia predisposition syndromes: clues to understanding leukemogenesis. Cancer Genet. 2011;204(5):227244.

Stiller CA, Chessells JM, Fitchett M. Neurofibromatosis and childhood leukaemia/lymphoma: a population-based UKCCSG study. Br J Cancer. 1994; 70(5):969-972.

Taylor AM, Metcalfe JA, Thick J, Mak YF. Leukemia and lymphoma in ataxia telangiectasia. Blood. 1996;87(2):423-438.

Ziino O, Rondelli R, Micalizzi C, et al.

Acute lymphoblastic leukemia in children with associated genetic conditions other than Down's syndrome. The AIEOP experience. Haematologica. 2006;91(1):139-140.

Seidemann K, Tiemann M, Henze G, et al.

17.

18.

19.

20.

21.

22.

23.

24.

Therapy for non-Hodgkin lymphoma in children with primary immunodeficiency: analysis of 19 patients from the BFM trials.

Med Pediatr Oncol. 1999;33(6):536-544.

Jonkman-Berk BM, van den Berg JM, Ten Berge IJ, et al. Primary immunodeficiencies in the Netherlands: national patient data demonstrate the increased risk of malignancy. Clin Immunol. 2015;156(2):154-162.

Tran H, Nourse J, Hall S, et al.

Immunodeficiency-associated lymphomas.

Blood Rev. 2008;22(5):261-281.

Jongmans MC, Loeffen JL, Waanders E, et al. Recognition of genetic predisposition in pediatric cancer patients: An easy-to-use selection tool. Eur J Med Genet. 2016; 59(3):116-125.

Cerhan JR, Slager SL. Familial predisposition and genetic risk factors for lymphoma.

Blood. 2015;126(20):2265-2273.

Bousfiha A, Jeddane L, Al-Herz W, et al.

The 2015 IUIS Phenotypic Classification for Primary Immunodeficiencies. J Clin Immunol. 2015;35(8):727-738.

Sandlund JT, Hudson MM, Kennedy W, et al. Pilot study of modified LMB-based therapy for children with ataxia-telangiectasia and advanced stage high grade mature Bcell malignancies. Pediatr Blood Cancer.

2014;61(2):360-362.

Harris NL, Jaffe ES, Diebold J, et al. The World Health Organization classification of neoplasms of the hematopoietic and lymphoid tissues: report of the Clinical Advisory Committee meeting--Airlie House, Virginia, November, 1997. Hematol J. 2000;1(1):53-66.

Harris NL, Jaffe ES, Stein H, et al. A revised European-American classification of lymphoid neoplasms: a proposal from the

haematologica | 2016; 101(12)

Non-Hodgkin lymphoma and pre-existing conditions

25.

26.

27.

28.

29.

30.

31.

32.

International Lymphoma Study Group.

Blood. 1994;84(5):1361-1392.

Stansfeld AG, Diebold J, Noel H, et al.

Updated Kiel classification for lymphomas.

Lancet. 1988;1(8580):292-293.

Murphy SB. Classification, staging and end results of treatment of childhood nonHodgkin's lymphomas: dissimilarities from lymphomas in adults. Semin Oncol.

1980;7(3):332-339.

Pillon M, Arico M, Mussolin L, et al. Longterm results of the AIEOP LNH-97 protocol for childhood lymphoblastic lymphoma.

Pediatr Blood Cancer. 2015;62(8):1388-1394.

Pillon M, Piglione M, Garaventa A, et al.

Long-term results of AIEOP LNH-92 protocol for the treatment of pediatric lymphoblastic lymphoma: a report of the Italian Association of Pediatric Hematology and Oncology. Pediatr Blood Cancer.

2009;53(6):953-959.

Atra A, Gerrard M, Hobson R, et al.

Outcome of relapsed or refractory childhood B-cell acute lymphoblastic leukaemia and B-cell non-Hodgkin's lymphoma treated with the UKCCSG 9003/9002 protocols.

Br J Haematol. 2001;112(4):965-968.

Atra A, Imeson JD, Hobson R, et al.

Improved outcome in children with advanced stage B-cell non-Hodgkin's lymphoma (B-NHL): results of the United Kingdom Children Cancer Study Group (UKCCSG) 9002 protocol. Br J Cancer.

2000;82(8):1396-1402.

Burkhardt B, Woessmann W, Zimmermann M, et al. Impact of cranial radiotherapy on central nervous system prophylaxis in children and adolescents with central nervous system-negative stage III or IV lymphoblastic lymphoma. J Clin Oncol. 2006; 24(3):491-499.

Reiter A, Schrappe M, Ludwig WD, et al.

Intensive ALL-type therapy without local radiotherapy provides a 90% event-free survival for children with T-cell lymphoblastic lymphoma: a BFM group report.

haematologica | 2016; 101(12)

Blood. 2000;95(2):416-421.

33. Reiter A, Schrappe M, Tiemann M, et al.

Improved treatment results in childhood Bcell neoplasms with tailored intensification of therapy: A report of the Berlin-FrankfurtMunster Group Trial NHL-BFM 90. Blood.

1999;94(10):3294-3306.

34. Woessmann W, Seidemann K, Mann G, et al. The impact of the methotrexate administration schedule and dose in the treatment of children and adolescents with B-cell neoplasms: a report of the BFM Group Study NHL-BFM95. Blood. 2005; 105(3): 948-958.

35. Le Deley MC, Rosolen A, Williams DM, et al. Vinblastine in children and adolescents with high-risk anaplastic large-cell lymphoma: results of the randomized ALCL99vinblastine trial.

J Clin Oncol.

2010;28(25):3987-3993.

36. Tsurusawa M, Mori T, Kikuchi A, et al.

Improved treatment results of children with B-cell non-Hodgkin lymphoma: a report from the Japanese Pediatric Leukemia/Lymphoma Study Group BNHL03 study. Pediatr Blood Cancer.

2014;61(7):1215-1221.

37. Gerrard M, Cairo MS, Weston C, et al.

Excellent survival following two courses of COPAD chemotherapy in children and adolescents with resected localized B-cell non-Hodgkin's lymphoma: results of the FAB/LMB 96 international study. Br J Haematol. 2008;141(6):840-847.

38. Patte C, Auperin A, Gerrard M, et al.

Results of the randomized international FAB/LMB96 trial for intermediate risk B-cell non-Hodgkin lymphoma in children and adolescents: it is possible to reduce treatment for the early responding patients.

Blood. 2007;109(7):2773-2780.

39. Amos Burke GA, Imeson J, Hobson R, Gerrard M. Localized non-Hodgkin's lymphoma with B-cell histology: cure without cyclophosphamide? A report of the United Kingdom Children's Cancer Study Group on studies NHL 8501 and NHL 9001 (1985-

1996). Br J Haematol. 2003;121(4):586-591.

40. Bergeron C, Coze C, Segura C, et al.

Treatment of Childhood T-Cell Lymphoblastic Lymphoma-Long-Term Results of the SFOP LMT96 Trial. Pediatr Blood Cancer. 2015;62(12):2150-2156.

41. Goldman S, Smith L, Anderson JR, et al.

Rituximab and FAB/LMB 96 chemotherapy in children with Stage III/IV B-cell nonHodgkin lymphoma: a Children's Oncology Group report. Leukemia.

2013;27(5):1174-1177.

42. Seidemann K, Tiemann M, Schrappe M, et al. Short-pulse B-non-Hodgkin lymphomatype chemotherapy is efficacious treatment for pediatric anaplastic large cell lymphoma: a report of the Berlin-FrankfurtMunster Group Trial NHL-BFM 90. Blood.

2001;97(12):3699-3706.

43. Brugieres L, Le Deley MC, Rosolen A, et al.

Impact of the methotrexate administration dose on the need for intrathecal treatment in children and adolescents with anaplastic large-cell lymphoma: results of a randomized trial of the EICNHL Group. J Clin Oncol. 2009;27(6):897-903.

44. Martin E, Palmic N, Sanquer S, et al. CTP synthase 1 deficiency in humans reveals its central role in lymphocyte proliferation.

Nature. 2014;510(7504):288-292.

45. Mellgren K, Attarbaschi A, Abla O, et al.

Non-anaplastic peripheral T cell lymphoma in children and adolescents-an international review of 143 cases. Ann Hematol.

2016;95(8):1295-1305.

46. Suarez F, Mahlaoui N, Canioni D, et al.

Incidence, presentation, and prognosis of malignancies in ataxia-telangiectasia: a report from the French national registry of primary immune deficiencies. J Clin Oncol.

2015;33(2):202-208.

47. Ripperger T, Schlegelberger B. Acute lymphoblastic leukemia and lymphoma in the context of constitutional mismatch repair deficiency syndrome. Eur J Med Genet.

2015;53(3):133-142.

1591

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Stem Cell Trasplantation

Ferrata Storti Foundation

Haematologica 2016 Volume 101(11):1592

Comparison of graft-versus-host disease-free, relapse-free survival according to a variety of graft sources: antithymocyte globulin and single cord blood provide favorable outcomes in some subgroups Yoshihiro Inamoto,1 Fumihiko Kimura,2 Junya Kanda,3 Junichi Sugita,4 Kazuhiro Ikegame,5 Hideki Nakasone,3 Yasuhito Nannya,6 Naoyuki Uchida,7 Takahiro Fukuda,1 Kosuke Yoshioka,8 Yukiyasu Ozawa,9 Ichiro Kawano,10 Yoshiko Atsuta,11,12 Koji Kato,13 Tatsuo Ichinohe,14 Masami Inoue15 and Takanori Teshima;4 on behalf of JSHCT GVHD Working Group

Department of Hematopoietic Stem Cell Transplantation, National Cancer Center Hospital, Tokyo; 2Division of Hematology, National Defense Medical College, Tokorozawa; 3Division of Hematology, Saitama Medical Center, Jichi Medical University, Saitama; 4Department of Hematology, Hokkaido University Graduate School of Medicine, Sapporo; 5Division of Hematology, Department of Internal Medicine, Hyogo College of Medicine, Nishinomiya; 6 Department of Hematology, Gifu University Hospital; 7Department of Hematology, Toranomon Hospital, Tokyo; 8Hematology Division, Tokyo Metropolitan Cancer and Infectious Disease Center Komagome Hospital; 9Department of Hematology, Japanese Red Cross Nagoya First Hospital; 10Department of Hematology, Hamanomachi Hospital, Fukuoka; 11Department of Healthcare Administration, Nagoya University Graduate School of Medicine; 12Japanese Data Center for Hematopoietic Cell Transplantation, Nagoya; 13 Department of Hematology and Oncology, Children's Medical Center, Japanese Red Cross Nagoya First Hospital; 14Department of Hematology and Oncology, Research Institute for Radiation Biology and Medicine, Hiroshima University, and 15Department of Hematology/Oncology, Osaka Medical Center and Research Institute for Maternal and Child Health, Japan 1

ABSTRACT

Correspondence: yinamoto@ncc.go.jp

Received: May 15, 2016.

Accepted: July 27, 2016.

Pre-published: August 4, 2016. doi:10.3324/haematol.2016.149427

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/12/1592

Š2016 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.

1592

G

raft-versus-host disease-free relapse-free survival, which is defined as the absence of grade III-IV acute graft-versus-host disease, systemically treated chronic graft-versus-host disease, relapse, and death, is a novel, meaningful composite end point for clinical trials. To characterize risk factors and differences in graft-versus-host disease-free relapse-free survival according to a variety of graft sources, we analyzed 23,302 patients with hematologic malignancy that had a first allogeneic transplantation from 2000 through 2013 using the Japanese national transplant registry database. The 1-year graft-versus-host disease-free relapsefree survival rate was 41% in all patients. The rate was higher after bone marrow transplantation than after peripheral blood stem cell transplantation due to the lower risks of III-IV acute and chronic graft-versus-host disease. The rate was highest after HLA-matched sibling bone marrow transplantation. The rate after single cord blood transplantation was comparable to that after HLA-matched unrelated bone marrow transplantation among patients aged 20 years or under, and was comparable or better than other alternative graft sources among patients aged 21 years or over, due to the low risk of chronic graft-versus-host disease. Other factors associated with better graft-versus-host disease-free relapse-free survival include female patients, antithymocyte globulin prophylaxis (for standard-risk disease), recent years of transplantation, sex combinations other than from a female donor to a male patient, the absence of prior autologous transplantation, myeloablative conditioning, negative cytomegalovirus serostatus, and tacrolimus-based prophylaxis. These results provide important information to guide the choice of graft sources and are benchmarks for future graft-versus-host disease prophylaxis studies. haematologica | 2016; 101(12)

GRFS according to graft sources

Introduction Graft-versus-host disease-free relapse-free survival (GRFS), defined as the absence of grade III-IV acute graftversus-host disease (GvHD), systemically treated chronic GvHD, relapse, and death, is a novel, clinically meaningful composite end point for clinical trials evaluating GvHD prophylaxis after allogeneic hematopoietic cell transplantation (HCT).1 The GRFS end point was devised by the Blood and Marrow Transplant Clinical Trials Network to address the fact that both survival and other critical events are important in clinical trials testing new GvHD prophylaxis.2,3 Moreover, GRFS is a patient-centered measure of success, since it represents not only disease-free survival but also ideal recovery without significant morbidity related to GvHD. Recently, Holtan et al. characterized the GRFS end point in a large cohort of patients at a single center.1 The study cohort included 322 HLA-matched sibling HCT, 73 HLAmatched unrelated HCT, 135 single cord blood transplantation (CBT), and 377 double CBT between 2000 and 2012. The crude GRFS rate was 31% at 12 months after HCT. Age, disease risk, graft sources, conditioning intensity, and year of HCT were associated with GRFS events.

Notably, HLA-matched bone marrow transplantation (BMT) provided the best GRFS, while peripheral blood stem cell transplantation (PBSCT) was associated with inferior GRFS compared with BMT. Since GRFS events will vary with graft sources, further studies using cohorts with increased representation of different donor types and graft sources were warranted. In addition, studies of different ethnicities and practices, such as Japanese patients who have a lower risk of significant GvHD4 and who undergo mostly single CBT,5 are also necessary to determine a more personalized GRFS end point.

To better understand differences in GRFS according to a variety of graft sources in the Japanese population, we retrospectively analyzed national registry data collected by the Transplant Registry Unified Management Program (TRUMP) sponsored by the Japanese Society of Hematopoietic Cell Transplantation (JSHCT) and the Japanese Data Center for Hematopoietic Cell Transplantation (JDCHCT).6,7 The specific aims of this study were: 1) to determine benchmark rates for future GvHD prophylaxis studies; 2) to determine the difference in GRFS between BMT and PBSCT; 3) to characterize GRFS after single CBT and after HLA-mismatched transplantation; and 4) to characterize risk factors associated with GRFS. The results of this study will provide important information to guide the choice of graft sources.

haploidentical transplantation, or unrelated PBSCT were excluded because of their relative infrequency during the study period.

Patients gave written consent to the use of medical records for research, in accordance with the Declaration of Helsinki. This study was approved by the institutional review board of the National Cancer Center Hospital.

Study end points and definitions The primary end point was GRFS as defined by the absence of grade III-IV acute GvHD, systemically treated chronic GvHD, recurrent malignancy, and death.1 Disease risk was defined according to the 2006 American Society for Blood and Marrow Transplantation (ASBMT) schema.1 Histocompatibility data for serological and genetic typing were obtained from the transplant registry database. To reflect current practices in Japan, HLA matching for sibling and cord blood transplantation was assessed by serological data for the HLA-A, -B, and -DR loci. HLA matching for unrelated BMT was assessed by using allele data for the HLA-A, -B, -C, and -DRB1 loci.8 HLA mismatch was defined in the GvHD vector when recipient antigens were not shared by the donor. Diagnosis and clinical grading of acute and chronic GvHD were performed according to the established criteria.9,10 The intensity of conditioning regimens was defined as described elsewhere.11

Statistical analysis Probabilities of GRFS were estimated by the Kaplan-Meier method until 24 months after transplantation. Cumulative incidence estimates of individual failure events (III-IV acute GvHD, chronic GvHD, relapse, and death) were derived, treating each event as a competing risk for the other three. Weighted GRFS rates were calculated by reducing adjusted failure rates due to IIIIV acute GvHD and chronic GvHD to half. Cox models were used to examine risk factors associated with failure defined by GRFS. A backward stepwise procedure was used to develop a final model, based on a P-value threshold of 0.05. Covariates include patient age (≤20 years, ≥21 years), patient sex, patientdonor sex combination, disease risk, diagnosis, prior autologous transplantation, ABO matching, donor-patient cytomegalovirus (CMV) serostatus, conditioning intensity, GvHD prophylaxis, use of antithymocyte globulin (ATG) as GvHD prophylaxis, and year of transplantation. Proportional hazards assumption was tested for all variables considered in multivariate analysis, and no violations occurred. Competing risk regression models were used for analysis of individual failure events.12 The overall interaction of patient age, disease risk, and year of transplantation with the main effect categories of the eight graft sources was tested by allowing additional terms for each of the eight graft sources in the model, depending on the presence or absence of the factor being tested. Models with and without the interaction terms were compared using a likelihood ratio test; P=0.05 was considered significant.

Methods

Results

Patients

Patients’ characteristics

This retrospective study cohort included all patients who received a first allogeneic HCT between 2000 and 2013. Graft sources included 5-6/6 serologically HLA-matched siblings (with matching considered at HLA-A, -B, -DR), 6-8/8 allele HLAmatched unrelated bone marrow donors (with matching considered at HLA-A, -B, -C, -DRB1), and 4-6/6 serologically HLAmatched cord blood donors (with matching considered at HLAA, -B, -DR). Patients who had double cord blood transplantation,

A total of 23,302 patients were included in this study. Of these, 12,338 (53%) had standard-risk disease, 10,964 (47%) had high-risk disease, 4053 (17%) were pediatric (≤20 years old), and 19,249 (83%) were adult (≥21 years old) patients. Median patient age was 44 years (range 0-85 years). Median follow up among survivors was 48 months (range 1-176 months). Patients’ characteristics according to eight graft sources are shown in Table 1.

haematologica | 2016; 101(12)

1593

Y. Inamoto et al.

Interactions of covariates with the main effect in the analysis of GRFS

Cumulative incidence of individual failure events and GRFS rates

We first examined the overall interaction of patient age, disease risk, and year of transplantation with the main effect categories of the graft sources in the analysis of GRFS. There was a statistical interaction between patient age (≤20 years vs. ≥21 years) and the main effect (overall P<0.0001), and a statistical interaction between disease risk and the main effect (overall P=0.03). There was no statistical interaction between transplant year and the main effect (overall P=0.08). Based on these results, all analyses were stratified according to patient age and disease risk.

Cumulative incidences of individual failure events (defined as the first event) are shown in Figure 1. The GRFS rates at 12 months were 58% [95% confidence interval (CI): 56%-59%] in pediatric patients with standard-risk disease, 32% (95%CI: 30%-35%) in pediatric patients with high-risk disease, 49% (95%CI: 48%-50%) in adult patients with standard-risk disease, and 30% (95%CI: 29%-31%) in adult patients with high-risk disease. In comparing individual failure events at 12 months between graft sources (Figure 2), 6/6 HLA-matched sibling

Table 1. Patients’ characteristics.

Characteristic, n. (%) Total number Median age, years (range) Patient age ≥21 years old Patient sex Male Female Sex combination Female donor to male patient Others Unknown Disease risk* Standard High Diagnosis AML ALL ATL CML MDS MPN Lymphoma Other malignancy† Prior autologous transplantation ABO matching Match Major mismatch Minor mismatch Unknown Donor-patient CMV serostatus Either positive Both negative Unknown Conditioning Myeloablative Reduced intensity Unknown intensity GvHD prophylaxis Cyclosporine-based Tacrolimus-based Other Use of antithymocyte globulin Year of transplantation 2000-2004 2005-2009 2010-2013

6/6 SIB-BM 6/6 SIB-PB 5/6 SIB-BM 5/6 SIB-PB 8/8 UR-BM 7/8 UR-BM 6/8 UR-BM

Single CB

3153 38 (0-73) 2334 (74)

3948 47 (0-74) 3622 (92)

559 25 (0-74) 310 (55)

647 47 (0-75) 560 (87)

4960 45 (0-75) 4247 (86)

2990 43 (0-74) 2508 (84)

1075 41 (0-73) 864 (80)

5970 47 (0-85) 4804 (80)

1816 (58) 1337 (42)

2323 (59) 1625 (41)

317 (57) 242 (43)

362 (56) 285 (44)

2954 (60) 2006 (40)

1774 (59) 1216 (41)

678 (63) 397 (37)

3394 (57) 2576 (43)

807 (26) 2217 (70) 129 (4)

1083 (27) 2759 (70) 106 (3)

154 (28) 391 (70) 14 (3)

169 (26) 450 (70) 28 (4)

840 (17) 4113 (83) 7 (<1)

583 (20) 2399 (80) 8 (<1)

206 (19) 866 (81) 3 (<1)

1419 (24) 3115 (52) 1436 (24)

2019 (64) 1134 (36)

1964 (50) 1984 (50)

283 (51) 276 (49)

255 (39) 392 (61)

2845 (57) 2115 (43)

1641 (55) 1349 (45)

549 (51) 526 (49)

2782 (47) 3188 (53)

1156 (37) 874 (28) 116 (4) 233 (7) 362 (11) 76 (2) 263 (8) 73 (2) 80 (3)

1507 (38) 665 (17) 221 (6) 205 (5) 386 (10) 95 (2) 698 (18) 171 (4) 280 (7)

199 (36) 182 (33) 19 (3) 30 (5) 56 (10) 13 (2) 55 (10) 5 (<1) 18 (3)

261 (40) 116 (18) 32 (5) 22 (3) 52 (8) 14 (2) 132 (20) 18 (3) 52 (8)

1920 (39) 1100 (22) 245 (5) 296 (6) 604 (12) 107 (2) 560 (11) 128 (3) 245 (5)

1135 (38) 679 (23) 141 (5) 211 (7) 362 (12) 63 (2) 323 (11) 76 (3) 152 (5)

425 (40) 252 (23) 55 (5) 68 (6) 117 (11) 25 (2) 111 (10) 22 (2) 49 (5)

2648 (44) 1348 (23) 244 (4) 182 (3) 579 (10) 85 (1) 754 (13) 130 (2) 335 (6)

1611 (51) 458 (15) 625 (20) 459 (15)

1950 (49) 559 (14) 748 (19) 691 (18)

287 (51) 96 (17) 124 (22) 52 (9)

291 (45) 97 (15) 149 (23) 110 (17)

2855 (58) 947 (19) 1150 (23) 8 (<1)

1334 (45) 706 (24) 941 (31) 9 (<1)

492 (46) 233 (22) 340 (32) 10 (<1)

2095 (35) 1536 (26) 2327 (39) 12 (<1)

2088 (66) 233 (7) 832 (26)

2657 (67) 202 (5) 1089 (28)

388 (69) 31 (6) 140 (25)

423 (65) 30 (5) 194 (30)

4028 (81) 307 (6) 625 (13)

2466 (82) 173 (6) 351 (12)

873 (81) 55 (5) 147 (14)

4145 (69) 370 (6) 1455 (24)

2228 (71) 632 (20) 293 (9)

2198 (56) 1381 (35) 369 (9)

408 (73) 119 (21) 32 (6)

320 (49) 277 (43) 50 (8)

3501 (71) 1348 (27) 111 (2)

2118 (71) 800 (27) 72 (2)

756 (70) 289 (27) 30 (3)

3678 (62) 2276 (38) 16 (<1)

2609 (83) 415 (13) 129 (4) 38 (1)

3331 (84) 535 (14) 82 (2) 117 (3)

136 (24) 411 (74) 12 (2) 46 (8)

237 (37) 397 (61) 13 (2) 151 (23)

1249 (25) 3641 (73) 70 (1) 140 (3)

585 (20) 2354 (79) 51 (2) 168 (6)

169 (16) 882 (82) 24 (2) 85 (8)

2539 (43) 3370 (56) 61 (1) 140 (2)

991 (31) 1314 (42) 848 (27)

1324 (34) 1360 (34) 1264 (32)

172 (31) 238 (43) 149 (27)

236 (36) 215 (33) 196 (30)

1436 (29) 1397 (28) 2127 (43)

892 (30) 859 (29) 1239 (41)

397 (37) 349 (32) 329 (31)

1229 (21) 2081 (35) 2660 (45)

GvHD: graft-versus-host disease; SIB: sibling; BM: bone marrow; PB: peripheral blood stem cell; UR: unrelated; CB: cord blood; AML: acute myeloid leukemia; ALL: acute lymphoblastic leukemia; ATL: adult T-cell leukemia/lymphoma; CML: chronic myeloid leukemia; MDS: myelodysplastic syndrome; MPN: myeloproliferative neoplasm; CMV: cytomegalovirus. *According to American Society for Blood and Marrow Transplantation 2006 schema: acute leukemia in first or second complete remission, chronic myeloid leukemia in first chronic phase, Hodgkin or non-Hodgkin lymphoma in complete or partial chemotherapy sensitive remission, chronic lymphocytic leukemia in first remission, myelodysplastic syndrome, and myeloproliferative disorder without excess blasts were all considered standard risk. All others were defined as high-risk diseases. †Plasma cell neoplasms and unclassified leukemia.

1594

haematologica | 2016; 101(12)

GRFS according to graft sources

Table 2. Multivariate analysis on risk of failure defined by graft-versus-host disease (GvHD)-free relapse-free survival.

Disease risk

Pediatric (age â&#x2030;¤ 20) HR (95% CI) P

N

Adult (age â&#x2030;Ľ 21) HR (95% CI)

P

618 1.00 (reference) 165 1.74 (1.36-2.21) <0.001 135 1.90 (1.46-2.46) <0.001 40 2.72 (1.81-4.09) <0.001 476 1.33 (1.10-1.60) 0.003 325 1.91 (1.57-2.33) <0.001 129 2.22 (1.70-2.88) <0.001 754 1.35 (1.14-1.60) <0.001

1403 1802 148 215 2370 1316 420 2028

1.00 (reference) 1.27 (1.15-1.39) 1.27 (1.01-1.59) 1.93 (1.62-2.31) 1.06 (0.97-1.17) 1.18 (1.06-1.31) 1.58 (1.37-1.82) 1.20 (1.09-1.33)

<0.001 0.04 <0.001 0.20 0.002 <0.001 <0.001

1566 1.00 (reference) 1075 0.83 (0.74-0.93) 0.002 107 0.62 (0.44-0.86) 0.005

5450 4247 335

1.00 (reference) 0.89 (0.83-0.94) 0.81 (0.69-0.95)

<0.001 0.008

1011 1.00 (reference) 833 0.87 (0.76-0.99) 0.03 797 0.79 (0.69-0.91) 0.001

2394 3217 4086 2012 413

1.00 (reference) 0.92 (0.85-0.99) 0.85 (0.80-0.92) 1.10 (1.02-1.18) 1.19 (1.05-1.36)

0.02 <0.001 0.01 0.006

6434 2851 412

1.00 (reference) 1.08 (1.01-1.15) 0.86 (0.74-1.00)

0.02 0.05

202 1.00 (reference) 162 1.27 (0.98-1.64) 0.07 114 1.78 (1.36-2.34) <0.001 47 1.69 (1.18-2.42) 0.004 238 1.46 (1.16-1.85) 0.001 157 1.44 (1.11-1.85) 0.005 82 1.51 (1.11-2.06) 0.009 412 1.27 (1.02-1.57) 0.03

933 1826 162 345 1880 1193 444 2776

1.00 (reference) 1.29 (1.18-1.42) 1.33 (1.10-1.62) 1.58 (1.37-1.82) 1.12 (1.02-1.24) 1.40 (1.26-1.56) 1.57 (1.37-1.80) 1.35 (1.23-1.48)

<0.001 0.004 <0.001 0.02 <0.001 <0.001 <0.001

405 1.00 (reference) 525 0.95 (0.83-1.11) 0.53 1 NA NA 52 0.54 (0.36-0.80) 0.002 140 0.55 (0.42-0.71) <0.001 114 0.72 (0.55-0.93) 0.01 111 0.91 (0.71-1.17) 0.46 64 0.63 (0.45-0.89) 0.009

3551 710 1072 588 1433 364 1275 559

1.00 (reference) 1.21 (1.11-1.32) 0.99 (0.91-1.07) 0.70 (0.62-0.77) 0.76 (0.70-0.82) 0.78 (0.69-0.89) 0.96 (0.88-1.03) 0.77 (0.68-0.86)

<0.001 0.73 <0.001 <0.001 <0.001 0.26 <0.001

583 1.00 (reference) 453 0.76 (0.65-0.89) 0.001 376 0.67 (0.56-0.79) <0.001

2683 3310 3553

1.00 (reference) 0.84 (0.79-0.89) 0.80 (0.76-0.85)

<0.001 <0.001

5765 3787 729

1.00 (reference) 0.83 (0.79-0.87) 1.12 (1.01-1.23)

<0.001 0.03

4484 4926 142

1.00 (reference) 0.90 (0.85-0.95) 1.13 (0.93-1.37)

<0.001 0.21

Characteristic N

Standard

High

Graft source 6/6 SIB / BM 6/6 SIB / PB 5/6 SIB / BM 5/6 SIB / PB 8/8 UR / BM 7/8 UR / BM 6/8 UR / BM Single CB Patient sex Male Female Use of antithymocyte globulin Year of transplantation 2000-2004 2005-2009 2010-2013 Female donor to male patient Prior autologous transplantation Conditioning Myeloablative Reduced intensity Unknown intensity Graft source 6/6 SIB / BM 6/6 SIB / PB 5/6 SIB / BM 5/6 SIB / PB 8/8 UR / BM 7/8 UR / BM 6/8 UR / BM Single CB Diagnosis AML ALL ATL CML MDS MPN Lymphoma Other malignancy* Year of transplantation 2000-2004 2005-2009 2010-2013 Donor-patient CMV serostatus Both negative Either positive Unknown Patient sex Male Female Prior autologous transplantation GvHD prophylaxis Cyclosporine-based Tacrolimus-based Other

122 1.00 (reference) 867 1.33 (1.04-1.72) 423 1.21 (0.92-1.60)

0.03 0.18

N: number; GvHD: graft-versus-host disease; GRFS: graft-versus-host disease (GvHD)-free relapse-free survival; HR: hazard ratio; CI: confidence interval; SIB: sibling; UR: unrelated; BM: bone marrow; PB: peripheral blood stem cell; CB: cord blood; AML: acute myeloid leukemia; ALL: acute lymphoblastic leukemia; ATL: adult T-cell leukemia/lymphoma; CML: chronic myeloid leukemia; MDS: myelodysplastic syndrome; MPN: myeloproliferative neoplasm; CMV: cytomegalovirus; NA: not applicable due to insufficient numbers of events for analysis. *Plasma cell neoplasms and unclassified leukemia

haematologica | 2016; 101(12)

1595

Y. Inamoto et al.

BMT was notable for the low proportion of III-IV acute GvHD, 6/6 HLA-matched sibling PBSCT was notable for the high proportion of chronic GvHD, HLA-mismatched HCT was notable for the high proportion of III-IV acute GvHD and the low proportion of relapse, and CBT was notable for the low proportion of chronic GvHD and the high proportion of death without relapse or significant GvHD.

Multivariate analyses for GRFS events Multivariate Cox models showed that 6/6 HLAmatched sibling BMT compared with most of other graft sources, and recent years of HCT were factors associated with better GRFS in all stratified cohorts (Table 2). The use of ATG as GvHD prophylaxis was associated with better GRFS among patients with standard-risk disease. Prior autologous transplantation was associated with worse GRFS among adult patients. Certain diagnoses in the highrisk group were associated with better or worse GRFS. Other factors associated with better GRFS include female patients, sex combinations other than from a female donor to a male patient, myeloablative conditioning, negative CMV serostatus, and tacrolimus-based GvHD prophylaxis.

Adjusted GRFS rates Adjusted GRFS rates according to graft sources are shown in Figure 3. The 6/6 HLA-matched sibling BMT showed the highest GRFS rate in all stratified cohorts.

Among adult patients with standard-risk disease, the GRFS rate after 8/8 HLA-matched unrelated BMT was

comparable to that after 6/6 HLA-matched sibling BMT (HR 1.06, 95%CI: 0.97-1.17; P=0.20).

We next compared GRFS rates after CBT with other graft sources. Among pediatric patients with standard-risk disease, the GRFS rate after CBT was similar compared with 8/8 HLA-matched unrelated BMT (HR 1.02, 95%CI: 0.86-1.21; P=0.84) and higher than other graft sources.

Among pediatric patients with high-risk disease, the GRFS rate after CBT was comparable to that after 6-8/8 HLAmatched unrelated BMT, and was better than that after 5/6 HLA-matched sibling BMT (HR 0.71, 95%CI: 0.560.90; P=0.004) and possibly after PBSCT (HR 0.75, 95%CI: 0.54-1.04; P=0.09). Among adult patient, the GRFS rate after CBT was comparable to that after 5/6 HLA-matched sibling BMT and 7/8 HLA-matched unrelated BMT, and was better than that after 5/6 HLA-matched sibling PBSCT and 6/8 HLA-matched unrelated BMT (data not shown).

Comparison of PBSCT with BMT Associations of PBSCT with the risks of individual GRFS events, compared with BMT, are shown in Table 3.

Among children with standard-risk disease who had HCT from a 6/6 HLA-matched sibling donor, PBSCT was associated with a higher risk of GRFS events (HR 1.81, 95%CI: 1.42-2.31; P<0.001) and failure due to chronic GvHD (HR 2.98, 95%CI: 1.93-4.58; P<0.001). Among adult patients with both standard and high-risk disease who had HCT from a 6/6 HLA-matched sibling donor, PBSCT was associated with a higher risk of GRFS events and failure due to III-IV acute GvHD and chronic GvHD, although PBSCT

High-risk disease

Standard-risk disease

A

B

Chronic GvHD

Chronic GvHD III-IV Acute GvHD

C

III-IV Acute GvHD

D

Chronic GvHD III-IV Acute GvHD Months after transplantation

Chronic GvHD III-IV Acute GvHD Months after transplantation

Figure 1. Cumulative incidence of individual failure events (defined as the first event). Each area represents recurrent malignancy, death without other failure events, onset of systemically treated chronic graft-versus-host disease (GvHD), and onset of grade III-IV acute GvHD. The white area represents GvHD-free relapsefree survival (GRFS). (A) Patients aged 20 years or under with standard-risk disease, (B) patients aged 20 years or under with high-risk disease, (C) patients aged 21 years or over with standard-risk disease, and (D) patients aged 21 years or over with high-risk disease.

1596

haematologica | 2016; 101(12)

GRFS according to graft sources

was associated with a lower risk of failure due to relapse among the same group of patients. Among adult patients with standard-risk disease who had HCT from a 5/6 HLAmatched sibling donor, PBSCT was associated with a higher risk of GRFS events (HR 1.59, 95%CI: 1.21-2.09;

P<0.001) possibly due to higher risks of III-IV acute GvHD (HR 1.51, 95%CI: 0.92-2.48; P=0.10) and chronic GvHD (HR 1.56, 95%CI: 0.97-2.52; P=0.07). Other subgroups did not show statistically significant differences in GRFS events.

Table 3. Comparison of sibling peripheral blood stem cell transplantation with sibling bone marrow transplantation.

Age ≤ 20

HLA 6/6

Disease risk Standard* High†

5/6

Standard* High†

≥ 21

6/6

Standard‡ High§

5/6

Standard‡ High§

N. of Any GRFS event PB/BM HR (95% CI) P 165/617

1.81 (1.42-2.31) 161 / 202 1.28 (0.98-1.68) 40 / 135 1.49 (0.94-2.35) 47 / 114 1.03 (0.70-1.53) 1799 / 1402 1.28 (1.17-1.42) 1823 / 932 1.30 (1.18-1.43) 215 / 148 1.59 (1.21-2.09) 345 / 162 1.23 (0.99-1.53)

Death HR (95% CI)

<0.001 0.07 0.09

0.46 (0.10-2.09) 1.39 (0.73-2.65) NA

0.87

0.58 (0.21-1.59) <0.001 1.09 (0.85-1.41) <0.001 1.06 (0.84-1.32) 0.001 1.17 (0.66-2.08) 0.07 1.34 (0.85-2.10)

Relapse HR (95% CI) P

P 0.32 0.32 NA 0.29 0.50 0.64 0.59 0.21

1.30 (0.91-1.85) 0.73 (0.50-1.07) 1.04 (0.32-3.41) 1.25 (0.63-2.49) 0.77 (0.64-0.91) 0.85 (0.72-1.01) 1.13 (0.54-2.34) 0.80 (0.53-1.21)

0.15 0.10 0.95 0.52 0.003 0.06 0.75 0.29

Type of failure III-IV acute GvHD HR (95% CI) P

Chronic GvHD HR (95% CI) P

1.34 0.37 2.98 <0.001 (0.71-2.55) (1.93-4.58) 1.74 0.16 2.05 0.02 (0.80-3.77) (1.11-3.76) 1.50 0.26 1.76 0.14 (0.74-3.07) (0.84-3.69) 1.04 0.90 1.78 0.18 (0.55-1.98) (0.76-4.17) 1.60 <0.001 1.49 <0.001 (1.27-2.03) (1.27-1.75) 1.75 <0.001 1.21 0.04 (1.42-2.17) (1.01-1.44) 1.51 0.10 1.56 0.07 (0.92-2.48) (0.97-2.52) 1.27 0.29 1.13 0.59 (0.81-1.98) (0.72-1.77)

HLA: human leukocyte antigen; N: number; PB: peripheral blood stem cell; BM: bone marrow; GRFS: graft-versus-host disease-free relapse-free survival; GvHD: graft-versus-host disease; HR: Hazard Ratio; CI: confidence interval; NA: not applicable due to insufficient numbers of events for analysis. *Adjusted for patient sex, use of antithymocyte globulin, and year of transplantation. †Adjusted for diagnosis, year of transplantation, donor-patient cytomegalovirus (CMV) serostatus. ‡Adjusted for patient sex, use of antithymocyte globulin, year of transplantation, a female donor to a male patient, prior autologous transplantation, and conditioning intensity. §Adjusted for diagnosis, year of transplantation, patient sex, prior autologous transplantation, and GvHD prophylaxis.

A

B

C

D

Death

Relapse

III-IV acute GvHD

Chronic GvHD

Figure 2. Proportions of failure events at 12 months. (A) Patients aged 20 years or under with standard-risk disease, (B) patients aged 20 years or under with highrisk disease, (C) patients aged 21 years or over with standard-risk disease, and (D) patients aged 21 years or over with high-risk disease. SIB: sibling; BM: bone marrow; PB: peripheral blood stem cell; UR: unrelated; CB: cord blood; GvHD: graft-versus-host disease.

haematologica | 2016; 101(12)

1597

Y. Inamoto et al.

Association of antithymocyte globulin prophylaxis with risks of individual failure events The use of ATG prophylaxis is a modifiable factor. Since ATG prophylaxis was associated with a lower risk of GRFS events among patients with standard-risk disease (Table 2), we examined its association with risks of individual failure events (Table 4). In children, ATG prophylaxis was not statistically associated with risks of any individual failure events. In adult patients, ATG prophylaxis was associated with lower risks of failure due to III-IV acute GvHD (HR 0.36, 95%CI: 0.23-0.56; P<0.001) and chronic GvHD (HR 0.59, 95%CI: 0.42-0.83; P=0.002), while it was associated with higher risks of failure due to death (HR 1.63, 95%CI: 1.27-2.09; P<0.001) and relapse (HR 1.35, 95%CI: 1.00-1.82; P=0.05). Causes of death were similar between patients with and without ATG prophylaxis (data not shown). Further subgroup analyses according to graft sources are shown in Table 4. Among adult patients, ATG prophylaxis was associated with a lower risk of GRFS events after sibling PBSCT and after 6/8 HLA-matched unrelated BMT. These associations appeared to be derived from lower risks of failure due to III-IV acute GvHD and chronic GvHD; however, ATG was associated with a higher risk of failure due to relapse after 5/6 HLA-matched sibling PBSCT. Interestingly, the benefit of ATG was not evident after 8/8 HLA-matched unrelated BMT due to a higher risk of failure due to death. ATG prophylaxis was associated with a higher risk of GRFS events

A

Standard-risk disease

C

after CBT, which was derived from the higher risk of failure due to death and possibly also due to relapse.

Subgroup analysis in children was inconclusive due to the limited number of patients who had ATG prophylaxis.

Weighted GRFS and long-term survival according to graft sources Since the onset of grade III-IV acute GvHD and systemically treated chronic GvHD may not necessarily hamper the long-term success of HCT, we went on to perform a weighted comparison of GRFS according to graft sources.

The subsequent 4-year survival rates among patients who had failure due to III-IV acute GvHD and chronic GvHD before 12 months were 68% and 69%, respectively.

Considering these results and impaired utility values in patients who developed significant GvHD,13 we reduced failure rates due to III-IV acute GvHD and chronic GvHD by half in the weighted analyses (Table 5). In addition, adjusted 10-year overall survival rates according to graft sources are shown in Table 5. The relative relationship among graft sources remained almost similar in these analyses. We further compared risk of secondary solid cancer according to graft sources. Among adult patients with high-risk disease, risk of secondary solid cancer was higher after 6/6 HLA-matched sibling PBSCT (HR 2.23, 95%CI: 1.20-4.13; P=0.01) and after 5/6 HLA-matched sibling PBSCT (HR 3.32, 95%CI: 1.45-7.57; P=0.004), and after 6/8 HLA-matched unrelated BMT (HR 2.46, 95%CI:

B

High-risk disease

D

Months after transplantation

Months after transplantation

Figure 3. Adjusted graft-versus-host disease (GvHD)-free relapse-free survival (GRFS) according to graft sources. (A) Patients aged 20 years or under with standardrisk disease. Results are adjusted for patient sex, use of antithymocyte globulin prophylaxis, and year of transplantation. (B) Patients aged 20 years or under with high-risk disease. Results are adjusted for diagnosis, cytomegalovirus (CMV) serostatus and year of transplantation. (C) Patients aged 21 years or over with standard-risk disease. Results are adjusted for diagnosis, CMV serostatus and year of transplantation. (D) Patients aged 21 years or over with high-risk disease. Results are adjusted for diagnosis, CMV serostatus and year of transplantation, patient sex, prior autologous transplantation, and GvHD prophylaxis. SIB: sibling; PB: peripheral blood stem cell; BM: bone marrow; UR: unrelated; CB: cord blood.

1598

haematologica | 2016; 101(12)

GRFS according to graft sources

1.05-5.76; P=0.04), compared with 6/6 HLA-matched sibling BMT. There was no statistical difference in the risk of secondary cancer among graft sources in other subgroups.

Discussion We analyzed a composite end point, GRFS, in the Japanese population using the national registry database, which includes different donor types and graft sources.

The 1-year GRFS rates were 58% in pediatric patients with standard-risk disease, 49% in adult patients with standard-risk disease, and approximately 30% in both pediatric and adult patients with high-risk disease. These rates were higher than both the rate reported in the Holtan study, which included mostly Caucasians at a single center, and the 23% GRFS in 628 adult patients registered to

the Center for International Blood and Marrow Transplant Research (CIBMTR).1,3 The GRFS rate was similar to that reported in the study of adult acute myeloid leukemia patients in remission registered to the European Society for Blood and Marrow Transplantation (EBMT).14 These differences may reflect the lower incidence of severe GvHD in the Japanese population derived from genetic homogeneity than in the Caucasian population,15 suggesting the importance of calculating benchmark rates for GRFS in patients of different ethnicities.

Consistent with the results of the Holtan,1 BMT provided remarkably higher GRFS rates than PBSCT in most subgroups. We extended analysis to differences in details of failure type and to HLA-mismatched subgroups. The higher GRFS rates associated with BMT were accounted for by the lower risks of failure due to III-IV acute GvHD and chronic GvHD. Although PBSCT was associated with

Table 4. Association of antithymocyte globulin prophylaxis with risks of individual failure events among patients with standard-risk disease.

Graft source

N. of ATG/ Any GRFS event wo ATG HR (95% CI) P

Age ≤ 20 All graft sources* 107 / 2534 6/6 SIB / BM†

18 / 599

6/6 SIB / PB†

8 / 157

5/6 SIB / BM†

13 / 122

5/6 SIB / PB†

8 / 32

8/8 UR / BM†

18 / 458

7/8 UR / BM†

17 / 308

6/8 UR / BM†

8 / 121

Single CB†

17 / 737

Age ≥ 21 All graft sources‡ 335 / 9362 6/6 SIB/BM§ 6/6 SIB/PB§

7 / 1395 41 / 1758

5/6 SIB/BM§

11 / 137

5/6 SIB/PB§

46 / 169

8/8 UR/BM§

62 / 2307

7/8 UR/BM§

84 / 1232

6/8 UR/BM§

36 / 384

Single CB§

48 / 1980

0.62 (0.44-0.86) 0.61 (0.23-1.64) 0.82 (0.29-2.28) 0.49 (0.17-1.39) 1.00 (0.29-3.42) 0.56 (0.23-1.37) 0.55 (0.24-1.26) 0.67 (0.24-1.85) 0.63 (0.26-1.52)

0.005

0.81 (0.69-0.95) NA 0.54 (0.34-0.86) 0.42 (0.13-1.38) 0.59 (0.37-0.93) 1.12 (0.80-1.57) 0.72 (0.51-1.01) 0.59 (0.35-1.00) 1.65 (1.18-2.30)

0.008

Type of failure Death Relapse HR (95% CI) P HR (95% CI)

0.70

1.39 (0.61-3.20) 2.97 (0.36-24.5) NA

0.18

NA

1.00

NA

0.21

3.03 (0.73-12.6) NA

1.19 (0.19-7.43) 0.46 0.96 (0.29-3.15)

0.85 0.95

NA

1.63 <0.001 1.35 (1.27-2.09) (1.00-1.82) NA NA NA NA 0.01 1.01 0.98 0.90 (0.36-2.84) (0.38-2.15) 0.15 NA NA 1.93 (0.20-18.2) 0.02 2.24 0.08 2.29 (0.90-5.58) (1.01-5.19) 0.51 2.23 0.003 1.61 (1.32-3.74) (0.87-2.95) 0.06 0.89 0.72 1.40 (0.47-1.69) (0.72-2.72) 0.05 2.23 0.06 0.36 (0.98-5.08) (0.05-2.78) 0.003 2.53 <0.001 1.74 (1.64-3.90) (0.94-3.35)

0.05

0.36 (0.23-0.56) NA 0.18 (0.03-1.33) 0.68 (0.14-3.35) 0.23 (0.08-0.68) 0.29 (0.07-1.16) 0.40 (0.16-0.98) 0.42 (0.15-1.17) 0.48 (0.16-1.44)

0.15 0.44 0.30

2.18 (0.32-14.7) 1.72 (0.42-7.08)

0.59 (0.32-1.11) 0.31 NA

0.10

1.91 (0.58-6.33) 0.51 (0.06-4.05) 2.68 (0.18-39.1) 0.13 0.38 (0.05-2.68) NA

0.29

III-IV acute GvHD HR (95% CI) P 0.59 (0.33-1.08) 1.08 (0.14-8.04) 1.34 (0.17-10.6) 0.77 (0.18-3.33) 0.97 (0.19-4.97) 0.38 (0.05-2.58) 1.03 (0.39-2.76) NA

0.33

0.43

P

0.42

0.52 0.47 0.33

NA 0.81 0.57 0.05 0.13 0.32 0.33 0.08

0.09 0.94 0.78 0.72 0.97 0.32 0.95

<0.001 NA 0.09 0.63 0.007 0.08 0.04 0.10 0.19

Chronic GvHD HR (95% CI) P 0.66 (0.33-1.33) 1.45 (0.36-5.81) NA

0.25

0.49 (0.07-3.59) 0.63 (0.04-9.53) 0.46 (0.06-3.34) 0.61 (0.14-2.62) 3.41 (0.82-14.1) NA

0.49

0.60

0.74 0.44 0.51 0.09

0.59 (0.42-0.83) NA 0.63 (0.31-1.28) NA

0.002

0.36 (0.12-1.09) 0.62 (0.29-1.33) 0.71 (0.37-1.36) 0.38 (0.12-1.21) 0.26 (0.04-1.92)

0.07

NA 0.20 NA

0.22 0.31 0.10 0.19

ATG: antithymocyte globulin; GvHD: graft-versus-host disease; wo: without; GRFS: graft-versus-host disease-free relapse-free survival; HR: hazard ratio; CI: confidence interval; SIB: sibling; UR: unrelated; BM: bone marrow; PB: peripheral blood stem cell; CB: cord blood; NA: not applicable due to insufficient numbers of events for analysis. *Adjusted for graft source, patient sex and year of transplantation. †Adjusted for patient sex and year of transplantation. ‡Adjusted for graft source, patient sex, year of transplantation, a female donor to a male patient, prior autologous transplantation, and conditioning intensity. §Adjusted for patient sex, year of transplantation, a female donor to a male patient, prior autologous transplantation, and conditioning intensity.

haematologica | 2016; 101(12)

1599

Y. Inamoto et al.

a lower risk of failure due to relapse only in adult patients with standard-risk disease who underwent 6/6 HLAmatched sibling HCT, its benefits were offset by the higher risk of significant GvHD. These results are consistent with the results of randomized studies and registry studies comparing PBSCT with BMT.16-18 We also found that PBSCT was associated with a higher risk of secondary solid cancer compared with BMT among adult patients with high-risk disease. Previous studies found that chronic GvHD was a major factor associated with risk of secondary solid cancer.19-22 Although the absence of significant GvHD may not be a long-term goal, particularly for patients with high-risk disease, the relationship among graft sources remained similar even in the weighted analyses. These results favor the use of bone marrow graft for sibling HCT to promote ideal recovery of patients without significant morbidity in the Japanese population.

The results of this study highlighted relative merits of single CBT as an alternative donor source from the perspective of the GRFS end point, although long-term overall survival did not show large differences among alternative donor sources. The merits of CBT are likely related to the low incidence of significant GvHD despite an increase in early mortality due to delayed hematopoietic and immunological recovery and graft failure after single CBT.23-25 In the Holtan study, relative risks of GRFS events after CBT using mostly double units were approximately 2.0 compared with 6/6 HLA-matched sibling BMT, while hazard ratios after CBT compared with 6/6 HLA-matched sibling BMT in our study were lower at ranges between 1.20 and 1.35 regardless of patient age and disease risk.

The difference between the studies could be accounted for by the lower risk of severe GvHD after single CBT com-

pared with double CBT,23,24 and by the lower risk of severe GvHD in the Japanese population compared with the Caucasian population.4 Consistent with the Holtan study,1 our study found that 6/6 HLA-matched sibling BMT was associated with higher GRFS compared with other graft sources. We also confirmed that myeloablative conditioning and more recent HCT were both associated with higher GRFS. The higher GRFS in recent years is likely related to the decreased incidence of non-relapse mortality and severe GvHD.26,27 With the larger analytical power permitted by the registry database, we found that better HLA matching, female patients, ATG prophylaxis, sex combinations other than a female donor to a male patient, no prior autologous HCT, certain diagnoses in the high-risk group, CMV-negative donor and recipient, and tacrolimus-based GvHD prophylaxis were associated with higher GRFS. These factors have been associated with the risks of GvHD and overall mortality in previous studies.28-36 Antithymocyte globulin prophylaxis is a modifiable factor and our results suggest the potential merits of ATG prophylaxis for patients with standard-risk diseases, although the risk of failure due to relapse might be increased in some patients. Subgroup analysis according to graft sources was inconclusive for pediatric patients, but identified several groups of adult patients who may benefit or suffer from ATG prophylaxis. ATG prophylaxis is likely to improve GRFS among adult patients with standard-risk disease who undergo 5-6/6 HLA–matched sibling PBSCT and 6/8 HLA–matched unrelated BMT, although an increased risk of failure due to relapse was observed after 5/6 HLA–matched sibling PBSCT. These results were consistent with the results of a recent ran-

Table 5. Adjusted graft-versus-host disease-free relapse-free (GRFS) rates and weighted GRFS rates at 12 months, and adjusted overall survival rates at ten years according to graft sources.

Graft source Age ≤20 6/6 SIB / BM 6/6 SIB / PB 5/6 SIB / BM 5/6 SIB / PB 8/8 UR / BM 7/8 UR / BM 6/8 UR / BM Single CB Age ≥21 6/6 SIB / BM 6/6 SIB / PB 5/6 SIB / BM 5/6 SIB / PB 8/8 UR / BM 7/8 UR / BM 6/8 UR / BM Single CB

Adjusted GRFS

Standard-risk disease* Weighted Adjusted GRFS‡ 10y OS

Adjusted GRFS

High-risk disease† Weighted Adjusted GRFS‡ 10y OS

0.68 0.50 0.49 0.33 0.59 0.47 0.44 0.61

0.77 0.64 0.64 0.52 0.71 0.62 0.60 0.72

0.71 0.61 0.63 0.61 0.72 0.68 0.61 0.65

0.42 0.34 0.23 0.19 0.31 0.31 0.31 0.36

0.57 0.51 0.43 0.40 0.49 0.49 0.49 0.52

0.48 0.44 0.43 0.44 0.39 0.36 0.36 0.39

0.54 0.45 0.45 0.32 0.53 0.50 0.40 0.49

0.61 0.53 0.53 0.42 0.60 0.57 0.49 0.56

0.58 0.52 0.54 0.40 0.55 0.51 0.46 0.47

0.36 0.28 0.27 0.23 0.34 0.27 0.24 0.29

0.50 0.44 0.43 0.40 0.48 0.43 0.41 0.45

0.38 0.30 0.26 0.22 0.32 0.24 0.25 0.24

GRFS: graft-versus-host disease-free relapse-free; 10y: 10-year; OS: overall survival; SIB: sibling; UR: unrelated; BM: bone marrow; PB: peripheral blood stem cell; CB: cord blood. *Adjusted for patient sex, use of antithymocyte globulin prophylaxis, and year of transplantation for patients aged 20 years or under. Adjusted for patient sex, use of antithymocyte globulin prophylaxis, and year of transplantation, donor-recipient sex combination, prior autologous transplantation, and conditioning intensity for patients aged 21 years or over. † Adjusted for diagnosis, cytomegalovirus (CMV) serostatus and year of transplantation for patients aged 20 years or under. Adjusted for diagnosis, CMV serostatus and year of transplantation, patient sex, prior autologous transplantation, and graft-versus-host disease (GvHD) prophylaxis for patients aged 21 years or over. ‡Failure rates due to III-IV acute GvHD and chronic GvHD were reduced to half in the weighted model.

1600

haematologica | 2016; 101(12)

GRFS according to graft sources

domized study.37 ATG prophylaxis is likely to have detrimental effects after single CBT due to increased risks of death and relapse, a result that agrees with a recent study using the European transplant registry database.38 This study has several limitations. First, poor GRFS may not justify avoidance of a particular graft source, since the absence of significant GvHD may not be a long-term goal for some patients. Thus, we performed weighted analysis and found that the relative relationship among graft sources remained similar even if the failure rates due to significant GvHD were reduced by half. Second, the results of ATG analysis would require careful interpretation, since the proportion of patients who had had ATG prophylaxis was relatively small in this study, and the doses, schedule, and types of ATG were not collected in the registry database. Prospective studies of ATG prophylaxis with pre-specified doses and schedules using the GRFS end point are warranted to clarify the merits of ATG prophylaxis for specific conditions. Third, some subgroup analyses are inconclusive, particularly for pediatric patients. Lastly, we did not include unrelated PBSCT, hap-

References 1. Holtan SG, DeFor TE, Lazaryan A, et al.

Composite end point of graft-versus-host disease-free, relapse-free survival after allogeneic hematopoietic cell transplantation.

Blood. 2015;125(8):1333-1338.

2. Sankoh AJ, Li H, D'Agostino RB Sr. Use of composite endpoints in clinical trials. Stat Med. 2014;33(27):4709-4714.

3. Schacke H, Docke WD, Asadullah K. Mechanisms involved in the side effects of glucocorticoids. Pharmacol Ther. 2002; 96(1):23-43.

4. Oh H, Loberiza FR Jr, Zhang MJ, et al.

Comparison of graft-versus-host-disease and survival after HLA-identical sibling bone marrow transplantation in ethnic populations. Blood. 2005;105(4):1408-1416.

5. Takahashi S, Ooi J, Tomonari A, et al.

Comparative single-institute analysis of cord blood transplantation from unrelated donors with bone marrow or peripheral blood stem-cell transplants from related donors in adult patients with hematologic malignancies after myeloablative conditioning regimen. Blood. 2007;109(3):1322-1330.

6. Atsuta Y. Introduction of Transplant Registry Unified Management Program 2 (TRUMP2): scripts for TRUMP data analyses, part I (variables other than HLA-related data). Int J Hematol. 2016;103(1):3-10.

7. Kanda J. Scripts for TRUMP data analyses.

Part II (HLA-related data): statistical analyses specific for hematopoietic stem cell transplantation. Int J Hematol. 2016; 103(1):1119.

8. Morishima Y, Sasazuki T, Inoko H, et al.

The clinical significance of human leukocyte antigen (HLA) allele compatibility in patients receiving a marrow transplant from serologically HLA-A, HLA-B, and HLA-DR matched unrelated donors. Blood. 2002;99 (11):4200-4206.

9. Przepiorka D, Ippoliti C, Koberda J, et al.

Interleukin-2 for prevention of graft-versushost disease after haploidentical marrow transplantation. Transplantation. 1994; 58(7):858-860.

haematologica | 2016; 101(12)

loidentical HCT, or double CBT because these graft sources were recently introduced in Japan and we have not yet had sufficient numbers of patients for analysis.

Further data collection is required to address these graft sources. The results of this study were derived from the national registry database collected from multiple centers, and thus will benchmark future GvHD prophylaxis trials in the Japanese population. The use of a large database allowed us to examine a variety of donor types and graft sources, and to identify robust risk factors associated with GRFS events. Our results will also inform physicians of the merits and demerits of a particular graft source from the perspective of the GRFS end point that measures ideal recovery without ongoing morbidity.

Funding This work was supported by the grant 15K19563 from the Japan Society for the Promotion of Science (JSPS), Friends of Leukemia Research Fund, and the grant 15Aek0510012h0001 from the Japan Agency for Medical Research and Development (AMED).

10. Shulman HM, Sullivan KM, Weiden PL, et al. Chronic graft-versus-host syndrome in man. A long-term clinicopathologic study of 20 Seattle patients. Am J Med. 1980;69 (2):204-217.

11. Giralt S, Ballen K, Rizzo D, et al. Reducedintensity conditioning regimen workshop: defining the dose spectrum. Report of a workshop convened by the center for international blood and marrow transplant research. Biol Blood Marrow Transplant.

2009;15(3):367-369.

12. Fine J, Gray R. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94(446):496-509.

13. Lee SJ, Klar N, Weeks JC, Antin JH. Predicting costs of stem-cell transplantation.

J Clin Oncol. 2000;18(1):64-71.

14. Ruggeri A, Labopin M, Ciceri F, Mohty M, Nagler A. Definition of GvHD-free, relapsefree survival for registry-based studies: an ALWP-EBMT analysis on patients with AML in remission. Bone Marrow Transplant. 2016;51(4):610-611.

15. Kanda J, Brazauskas R, Hu ZH, et al. Graftversus-Host Disease after HLA-Matched Sibling Bone Marrow or Peripheral Blood Stem Cell Transplantation: Comparison of North American Caucasian and Japanese Populations. Biol Blood Marrow Transplant.

2016;22(4):744-751.

16. Anasetti C, Logan BR, Lee SJ, et al.

Peripheral-blood stem cells versus bone marrow from unrelated donors. N Engl J Med. 2012;367(16):1487-1496.

17. Eapen M, Logan BR, Confer DL, et al.

Peripheral blood grafts from unrelated donors are associated with increased acute and chronic graft-versus-host disease without improved survival. Biol Blood Marrow Transplant. 2007;13(12):1461-1468.

18. Nagafuji K, Matsuo K, Teshima T, et al.

Peripheral blood stem cell versus bone marrow transplantation from HLA-identical sibling donors in patients with leukemia: a propensity score-based comparison from the Japan Society for Hematopoietic Stem Cell Transplantation registry. Int J Hematol.

2010;91(5):855-864.

19. Savani BN, Stratton P, Shenoy A, Kozanas E, Goodman S, Barrett AJ. Increased risk of cervical dysplasia in long-term survivors of allogeneic stem cell transplantation--implications for screening and HPV vaccination.

Biol Blood marrow Transplant.

2008;14(9):1072-1075.

20. Rizzo JD, Curtis RE, Socie G, et al. Solid cancers after allogeneic hematopoietic cell transplantation. Blood. 2009;113(5):11751183.

21. Atsuta Y, Suzuki R, Yamashita T, et al.

Continuing increased risk of oral/esophageal cancer after allogeneic hematopoietic stem cell transplantation in adults in association with chronic graft-versus-host disease. Ann Oncol. 2014; 25(2):435-441.

22. Inamoto Y, Shah NN, Savani BN, et al.

Secondary solid cancer screening following hematopoietic cell transplantation. Bone Marrow Transplant. 2015;50(8):1013-1023.

23. Wagner JE Jr, Eapen M, Carter S, et al. Oneunit versus two-unit cord-blood transplantation for hematologic cancers. N Engl J Med. 2014;371(18):1685-1694.

24. Ruggeri A, Sanz G, Bittencourt H, et al.

Comparison of outcomes after single or double cord blood transplantation in adults with acute leukemia using different types of myeloablative conditioning regimen, a retrospective study on behalf of Eurocord and the Acute Leukemia Working Party of EBMT. Leukemia. 2014;28(4):779-786.

25. Laughlin MJ, Barker J, Bambach B, et al.

Hematopoietic engraftment and survival in adult recipients of umbilical-cord blood from unrelated donors. N Engl J Med. 2001;344(24):1815-1822.

26. Gooley TA, Chien JW, Pergam SA, et al.

Reduced mortality after allogeneic hematopoietic-cell transplantation. N Engl J Med. 2010;363(22):2091-2101.

27. Kurosawa S, Yakushijin K, Yamaguchi T, et al. Changes in incidence and causes of nonrelapse mortality after allogeneic hematopoietic cell transplantation in patients with acute leukemia/myelodysplastic syndrome: an analysis of the Japan Transplant Outcome Registry. Bone

1601

Y. Inamoto et al.

Marrow Transplant. 2013;48(4):529-536.

28. Atkinson K, Horowitz MM, Gale RP, et al.

Risk factors for chronic graft-versus-host disease after HLA-identical sibling bone marrow transplantation.

Blood.

1990; 75(12):2459-2464.

29. Nash RA, Pepe MS, Storb R, et al. Acute graft-versus-host disease: analysis of risk factors after allogeneic marrow transplantation and prophylaxis with cyclosporine and methotrexate. Blood. 1992;80(7):1838-1845.

30. Hahn T, McCarthy PL Jr, Zhang MJ, et al.

Risk factors for acute graft-versus-host disease after human leukocyte antigen-identical sibling transplants for adults with leukemia. J Clin Oncol. 2008;26(35):57285734.

31. Flowers ME, Inamoto Y, Carpenter PA, et al.

Comparative analysis of risk factors for

1602

acute graft-versus-host disease and for chronic graft-versus-host disease according to National Institutes of Health consensus criteria. Blood. 2011;117(11):3214-3219.

32. Lee SJ, Klein J, Haagenson M, et al. High-resolution donor-recipient HLA matching contributes to the success of unrelated donor marrow transplantation.

Blood.

2007;110(13):4576-4583.

33. Morishima Y, Kashiwase K, Matsuo K, et al.

Biological significance of HLA locus matching in unrelated donor bone marrow transplantation. Blood. 2015;125(7):1189-1197.

34. 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.

35. Jagasia M, Arora M, Flowers ME, et al. Risk factors for acute GVHD and survival after hematopoietic cell transplantation. Blood.

2012;119(1):296-307.

36. Inamoto Y, Flowers ME, Wang T, et al.

Tacrolimus versus Cyclosporine after Hematopoietic Cell Transplantation for Acquired Aplastic Anemia. Biol Blood Marrow Transplant. 2015;21(10):1776-1782.

37. Kroger N, Solano C, Wolschke C, et al.

Antilymphocyte Globulin for Prevention of Chronic Graft-versus-Host Disease. N Engl J Med. 2016;374(1):43-53.

38. Pascal L, Tucunduva L, Ruggeri A, et al.

Impact of ATG-containing reduced-intensity conditioning after single- or double-unit allogeneic cord blood transplantation. Blood.

2015;126(8):1027-1032.

haematologica | 2016; 101(12)

ARTICLE

Complications in Hematology

Metabolic syndrome in long-term survivors of childhood acute leukemia treated without hematopoietic stem cell transplantation: an L.E.A. study Paul Saultier,1 Pascal Auquier,2 Yves Bertrand,3 Camille Vercasson,2 Claire Oudin,1,2 Audrey Contet,4 Dominique Plantaz,5 Marilyne Poirée,6 Stéphane Ducassou,7 Justyna Kanold,8 Marie-Dominique Tabone,9 Jean-Hugues Dalle,10 Patrick Lutz,11 Virginie Gandemer,12 Nicolas Sirvent,13 Sandrine Thouvenin,14 Julie Berbis,2 Hervé Chambost,1 André Baruchel,10 Guy Leverger9 and Gérard Michel1,2

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2016 Volume 101(12):1603

Department of Pediatric Hematology and Oncology, Timone Enfants Hospital, APHM and Aix-Marseille University, Marseille; 2Research Unit EA 3279 and Department of Public Health, Aix-Marseille University and Timone Hospital, Marseille; 3Department of Pediatric Hematology and Oncology, University Hospital of Lyon; 4Department of Pediatric Onco-Hematology, Hôpital d’Enfants de Brabois, Vandoeuvre-les-Nancy; 5 Department of Pediatric Hematology-Oncology, University Hospital of Grenoble; 6 Pediatric Hematology and Oncology department, University Hospital L’Archet, Nice; 7 Department of Pediatric Hematology and Oncology, University Hospital of Bordeaux; 8 Department of Pediatric Hematology and Oncology, CIC Inserm 501, University Hospital of Clermont-Ferrand; 9Pediatric Hematology Department, Trousseau Hospital, Paris; 10 Pediatric Hematology Department, Robert Debré Hospital, Paris; 11Department of Pediatric Hematology-Oncology, Hospital University, Strasbourg; 12Department of Pediatric Hematology and Oncology, University Hospital of Rennes; 13Pediatric Hematology and Oncology department, University Hospital, Montpellier and 14Pediatric Hematology, University Hospital, Saint Etienne, France

1

ABSTRACT

C

ardiovascular conditions are serious long-term complications of childhood acute leukemia. However, few studies have investigated the risk of metabolic syndrome, a known predictor of cardiovascular disease, in patients treated without hematopoietic stem cell transplantation. We describe the overall and age-specific prevalence, and the risk factors for metabolic syndrome and its components in the L.E.A. (Leucémie de l'Enfant et de l’Adolescent) French cohort of childhood acute leukemia survivors treated without hematopoietic stem cell transplantation. The study included 650 adult patients (mean age at evaluation: 24.2 years; mean follow-up after leukemia diagnosis: 16.0 years).

The prevalence of metabolic syndrome was 6.9% (95% CI 5.1-9.2). The age-specific cumulative prevalence at 20, 25, 30 and 35 years of age was 1.3%, 6.1%, 10.8% and 22.4%, respectively. The prevalence of decreased high-density lipoprotein cholesterol, increased triglycerides, increased fasting glucose, increased blood pressure and increased abdominal circumference was 26.8%, 11.7%, 5.8%, 36.7% and 16.7%, respectively. Risk factors significantly associated with metabolic syndrome in the multivariate analysis were male sex (OR 2.64; 95% CI 1.32-5.29), age at last evaluation (OR 1.10; 95% CI 1.04-1.17) and body mass index at diagnosis (OR 1.15; 95% CI 1.01-1.32). The cumulative steroid dose was not a significant risk factor. Irradiated and non-irradiated patients exhibited different patterns of metabolic abnormalities, with more frequent abdominal obesity in irradiated patients and more frequent hypertension in non-irradiated patients. Survivors of childhood acute leukemia are at risk of metabolic syndrome, even when treated without hematopoietic stem cell transplantation or central nervous system irradiation. A preventive approach with regular screening for cardiovascular risk factors is recommended. clinicaltrials.gov identifier:01756599. haematologica | 2016; 101(12)

Correspondence: paul.saultier@gmail.com

Received: May 4, 2016.

Accepted: August 5, 2016.

Pre-published: August 11, 2016. doi:10.3324/haematol.2016.148908

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/100/12/1603

©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.

1603

P. Saultier et al.

Introduction Acute leukemia (AL) accounts for one third of childhood cancers.1 A significant improvement in patient survival has led to a heightened focus on the long-term complications associated with this disease and corresponding treatments. Notably, childhood AL survivors exhibit a more than 4-fold increase in cardiovascular-related mortality rates, including congestive heart failure, coronary artery disease including myocardial infarction, cardiac arrest and stroke, compared with siblings or the general population.2,3 Anthracyclines, which have been linked to cardiac toxicity, are only partly responsible for this increased cardiovascular mortality.4 Furthermore, childhood AL survivors have been shown to have early signs of atherosclerotic lesions.5,6 The natural evolution of atheromatous disease begins years before the onset of a lesion with clinical impacts. Therefore, clinicians and researchers have aimed to identify early markers associated with an increased risk of developing cardiovascular disease. One of the most studied markers in the general population is metabolic syndrome (MetS).7 It is defined as a combination of cardiovascular risk factors, including abdominal obesity, dyslipidemia, glucose intolerance and hypertension. This composite marker predicts coronary artery disease and stroke risk better than each of its components.8 Most of the published data concerning MetS prevalence among childhood AL survivors have been focused on patients treated with hematopoietic stem cell transplantation (HSCT). In this population, the prevalence of MetS is particularly high.9–11 Previous studies have investigated MetS prevalence among childhood AL patients treated without HSCT,12–18 although most of these studies had distinct limitations: (I) child or very young adult populations, (II) self-administered questionnaires, (III) single-center studies, (IV) small cohort size, or (V) retrospective design.

Thus, the reported MetS prevalence among patients treated without HSCT is quite variable, with values ranging from 8.3% to 31.7%.13,15,17,18 Consequently, the risk of MetS remains controversial, especially for patients treated without central nervous system (CNS) irradiation. This matter is all the more important as modern protocols tend to strongly limit the treatment modality of CNS irradiation.19 Additionally, the pathophysiology and risk factors of increased metabolic and cardiovascular risk in this population remain poorly understood.

A previous study on MetS from the French Leucémie de l'Enfant et de l’Adolescent L.E.A. program20 has described a cohort of patients treated either with or without HSCT. However, the number of patients was relatively small and precluded any specific analysis of each therapeutic subgroup. The further expansion of the L.E.A. cohort allowed us to perform the study herein, which focused on 650 patients treated without HSCT. Its aim was to prospectively describe the overall and age-specific prevalence and risk factors for MetS and its components among adult survivors of childhood AL who did not receive HSCT and were included in the L.E.A. multicenter French cohort.

up clinic at predefined dates, starting one year after completion of chemotherapy. These visits are repeated every two years until the age of 20 and at least 10 years of complete remission, and every four years thereafter. The program began in 2004 and rests on the constitution of a multicenter historical and prospective cohort, which includes both incident cases (diagnosed after the start date of the participation of the center in the L.E.A program) and prevalent cases (diagnosed between 01/01/1980 and the start date of the participation of the center in the L.E.A program). Since 2007, adult patients have systematically undergone a complete evaluation for MetS. Patients were eligible for the present study if they had been included in the L.E.A. program between 2007 and 2013, were more than 18 years of age and were treated without HSCT. The eligible patients with complete evaluation for MetS were included in the study. All patients provided written informed consent. The study was approved by the French National Program for Clinical Research and the National Cancer Institute.

MetS was defined according to the National Cholesterol Education Program - Adult Treatment Panel III (NCEP-ATPIII) criteria revised in 200522 as the combination of at least three of the following criteria: (I) increased waist circumference (≥102 cm in men and ≥88 cm in women), (II) increased blood pressure (systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥85 mmHg) or treatment for hypertension, (III) decreased HDL cholesterol (<1.03 mmol/l in men and <1.3 mmol/l) in women), (IV) increased fasting glucose (≥5.5 mmol/l) or treatment for hyperglycemia, and (V) increased triglycerides (≥1.7 mmol/l) or treatment for hypertriglyceridemia. The body mass index (BMI) at diagnosis was expressed as standard deviation from the mean value of children of the same age in French references (z-score).

Adult BMI was expressed in kg.m-2. Overweight and obese patients were defined as BMI =25-30 kg.m-2 and BMI ≥30 kg.m-2, respectively. The cumulative steroid dose was calculated in equivalent of prednisone for each patient using the following formula: cumulative steroid dose = cumulative prednisone dose + (cumulative dexamethasone dose x 6.67) in mg.m-2 as previously described.23 Statistical analysis was performed using SPSS 20.0 software (SPSS Inc., Chicago, IL, USA) and Intercooled Stata 9.0 (StataCorp., College Station, TX, USA). Quantitative variables were expressed as mean ± standard deviation. Categorical variables were compared using the χ2 test or the Fisher’s exact test.

Quantitative variables were compared using the Student’s t-test or the Mann-Whitney test. Prevalence rates are displayed with 95% confidence interval (CI). The associations between MetS or its components and potential risk factors were initially analyzed via univariate logistic regression. The variables associated with MetS in the univariate analysis with a P-value <0.05 were then included in a multivariate analysis. Odds ratios are displayed with 95% CI. The age-specific cumulative prevalence of MetS was estimated using the Kaplan-Meier method.

Results Comparison between included patients and eligible but not included patients

Methods L.E.A. is a long-term follow-up program involving all childhood AL survivors treated in the French participating centers since 1980.

As detailed elsewhere,21 participants are summoned to the follow1604

The study flow chart is presented in Figure 1. Eight hundred and seventy patients from the L.E.A. program were eligible for the study. Among them, 650 underwent a complete evaluation for MetS and were ultimately included in the study cohort. The comparison between the characterhaematologica | 2016; 101(12)

Metabolic syndrome in childhood leukemia survivors

istics of patients included and those who were eligible but not included is presented in Table 1. Sex, age at diagnosis, age at last evaluation, time from diagnosis to last evaluation, BMI z-score at diagnosis, relapse rate, CNS irradiation rate and type of irradiation were similar in the two groups. In the studied cohort, the percentage of myeloblastic leukemia cases was significantly higher (9.5% versus 5.0%, P=0.033) and the mean cumulative steroid dose was significantly lower (4494±2578 mg/m2 versus 5050±2216 mg/m2, P=0.016) than among the eligible but not included patients.

Description of the study cohort The characteristics of the 650 included patients are shown in Table 1. The study population comprised 52.2% women and 47.8% men. The mean age at diagnosis was 8.2±4.8 years. The mean age at last evaluation was 24.2±5.2 years, and the time from diagnosis to last evaluation was 16.0±6.8 years. The mean BMI z-score at diagnosis was -0.15±1.72. CNS irradiation was performed in 18.0% of patients (n=117). When administered, the irradiation dose was 18 Gy (14.5%, n=94/650) or 24 Gy (3.2%; n=21/650). The mean BMI at last evaluation was 23.3±4.2

2623 patients included in the L.E.A. program (2007-2013) 550 patients treated with HSCT 2073 patients treated without HSCT

Figure 1. Flow chart of the cohort. HSCT: hematopoietic stem cell transplantation.

Table 1. Description of the patient cohort and comparison with eligible but not included patients.

Patient characteristics

Sex Female Male Leukemia type ALL AML Biphenotypic Age at diagnosis (years) mean ± SD Age at last evaluation (years) mean ± SD Time from diagnosis to last evaluation (years) mean ± SD BMI z-score at diagnosis mean ± SD Relapse No Yes CNS irradiation No 18 Gy 24 Gy Unknown‡ Type of irradiation† Cranial Craniospinal Unknown Cumulative prednisone-equivalent dose (mg/m2)mean ± SD

Total eligible patients n=870 (100%) n (%) or mean ± SD

Eligible but not included patients n=220 (25.3%) n (%) or mean ± SD

Patient cohort (included patients) n=650 (74.7%) n (%) or mean ± SD

P

456 (52.4%) 414 (47.6%)

117 (53.2%) 103 (46.8%)

339 (52.2%) 311 (47.8%)

0.792

791 (90.9%) 73 (8.4%) 6 (0.7%)

209 (95.0%) 11 (5.0%) 0 (0.0%)

582 (89.5%) 62 (9.5%) 6 (0.9%)

0.033

8.33±4.80

8.66±4.80

8.22±4.80

0.245

24.22±5.23

24.22±5.39

24.23±5.18

0.983

15.89±6.89

15.56±7.20

16.00±6.79

0.409

-0.16±1.66

-0.19±1.43

-0.15±1.72

0.807

828 (95.2%) 42 (4.8%)

210 (95.5%) 10 (4.5%)

618 (95.1%) 32 (4.9%)

0.821

700 (80.5%) 132 (15.2%) 32 (3.7%) 6 (0.7%)

170 (77.3%) 38 (17.3%) 11 (5.0%) 1 (0.5%)

530 (81.5%) 94 (14.5%) 21 (3.2%) 5 (0.8%)

0.270

123 (73.7%) 42 (25.1%) 2 (1.2%)

36 (72.0%) 13 (26.0%) 1 (2%)

87 (74.4%) 29 (24.8%) 1 (0.9%)

0.837

4623±2508

5050±2216

4494±2578

0.016

SD: standard deviation; ALL: acute lymphoblastic leukemia; AML: acute myeloblastic leukemia; BMI: body mass index; CNS: central nervous system; ‡unknown irradiation status or dose of irradiation; †expressed as % of irradiated patients (n=117); significant values (P<0.05).

haematologica | 2016; 101(12)

1605

P. Saultier et al.

kg.m-2 (Online Supplementary Table S1). The number of complete MetS evaluations was one for 512 patients and two or more for 138 patients. The prevalence of overt diabetes was 0.8% (n=5/650).

Overall and age-specific cumulative prevalence of MetS and its components Overall, the prevalence of MetS was 6.9% (95% CI 5.19.2%; n=45/650). The Kaplan-Meier estimation describing the age-specific cumulative prevalence of MetS is shown in Figure 2. The age-specific cumulative prevalence at 20, 25, 30 and 35 years of age was 1.3% (95% CI 0.6-2.7), 6.1% (95% CI 4.0-9.1), 10.8% (95% CI 7.2-15.9) and 22.4% (95% CI 15.1-32.6), respectively. The prevalence of each component of MetS is shown in Table 2.

Furthermore, 385 patients (59.2%) had at least one abnormal MetS component. A description of the BMI values as well as the overweight and obesity prevalence with respect to MetS status is reported in the Online Supplementary Table S1. The mean BMI at last evaluation was 22.9±3.7 kg/m2 in patients without MetS and 29.5±5.8 kg/m2 in patients with MetS (P<0.001). The rate of obese patients was 3.7% (n=22) in patients without MetS versus 45.2% (n=19) in patients with MetS (P<0.001).

Univariate and multivariate analysis of risk factors for MetS and its components The results of the univariate analysis of potential MetS risk factors are presented in the Online Supplementary Table S2. Leukemia type, age at diagnosis, relapse rate, type of CNS irradiation and cumulative steroid dose were not significantly associated with MetS in the univariate analysis.

Consequently, these factors, along with the factor “time from diagnosis to last evaluation”, which was highly correlated to the factor “age at last evaluation”, were not included in the multivariate analysis. The results of the multivariate analysis are provided in Table 3. Three variables were found to be significantly associated with MetS in the multivariate analysis: male sex (OR 2.64; 95% CI 1.32-5.29; P=0.006), age at last evaluation (OR 1.10 per each additional year of follow-up; 95% CI 1.04-1.17; P=0.001) and BMI z-score at diagnosis (OR 1.15 per each additional z-score unit; 95% CI 1.01-1.32; P=0.037).

When each component of MetS was separately examined via multivariate analysis, several risk factors were

highlighted. Undergoing 24 Gy CNS irradiation was a risk factor for a decreased HDL cholesterol (OR 2.76; 95% CI 1.03-7.40; P=0.044). Male sex was a risk factor for increased triglycerides and increased fasting glucose: OR 1.69 (95% CI 1.01-2.84; P=0.045) and 2.71 (95% CI 1.255.84; P=0.011), respectively. Risk factors for increased blood pressure were male sex (OR 3.47; 95% CI 2.41-4.99; P<0.001) and age at last evaluation (OR 1.08 for each additional year of follow-up; 95% CI 1.04-1.12; P<0.001).

Undergoing 18 Gy and 24 Gy CNS irradiation were negatively associated with increased blood pressure, with OR 0.48 (95% CI 0.28-0.82; P=0.007) and 0.21 (95% CI 0.070.64; P=0.006), respectively. Age at last evaluation, BMI z-score at diagnosis and 24 Gy CNS irradiation were risk factors for increased waist circumference: OR 1.07 (95% CI 1.02-1.12; P=0.008), 1.44 (95% CI 1.23-1.70; P<0.001) and 4.84 (95% CI 1.63-14.44; P=0.005), respectively.

Males were negatively associated with increased waist circumference (OR 0.38; 95% CI 0.22-0.64; P<0.001).

Metabolic profile with regard to irradiation The prevalence of each MetS component with respect to whether or not the patient had received CNS irradiation and radiation dose was analyzed for all patients with at least one abnormal component (n=385; 59.2%). The results are presented in Figure 3. Increased blood pressure was significantly more frequent in the non-irradiated group than in the 18 Gy and 24 Gy irradiated groups (63.4%, 50.0% and 35.7%, respectively; P=0.028). In contrast, increased waist circumference was significantly more frequent in the 24 Gy irradiated group than in the 18 Gy irradiated group and the non-irradiated group (64.3%, 37.3% and 22.4%, respectively; P<0.001). The other components of MetS were not significantly different with regard to the history of irradiation and corresponding dosage.

Discussion Our study aimed to precisely describe the overall and age-specific cumulative prevalence, and risk factors for MetS in a large cohort of childhood AL survivors treated without HSCT. Notably, the proportion of children who received CNS radiation was relatively low in our cohort.

Therefore, the subgroup treated with chemotherapy alone

Figure 2. Age-specific cumulative prevalence of the metabolic syndrome.

1606

haematologica | 2016; 101(12)

Metabolic syndrome in childhood leukemia survivors

(i.e., without CNS irradiation) is one of the largest ever published. As current protocols include very limited CNS irradiation indications, it is critical to evaluate MetS in non-irradiated patients.

The proportion of AML (acute myeloblastic leukemia) patients was higher among the included patients than among the eligible but not included patients, thereby explaining the lower mean cumulative steroid dose. This difference could be explained by the high severity of AML, which could have improved compliance for followup programs.

We report an overall MetS prevalence of 6.9% (95% CI 5.1-9.2) in our cohort, which involved relatively young patients (mean age of 24.2 years). This result is consistent with the results of a previously published L.E.A. study.20 Prevalence was approximately two-fold higher than that observed in the adult French general population under 40 years of age, which ranges from 2.2% to 5.0%.24,25 In the literature, the reported prevalence of MetS among adult survivors of childhood AL without HSCT is widely varied, ranging from between 8.3% and 31.7%.13,15,17,18,20 Apart from the intrinsic limits of previous studies, direct comparison between the reported prevalence rates of MetS is difficult due to the heterogeneity of the studied populations with respect to leukemia type, age, follow-up duration and treatment modalities. Concerning the treatment modalities, some authors have reported data of patients primarily treated using protocols including CNS irradiation, which increases the risk of MetS.13,18 Furthermore, the mean age of patients in previous studies ranged from 19 to 31.7 years, thereby rendering comparative analyses difficult as MetS prevalence tends to increase with age.13,15,17,18,20 Overall however, the prevalence reported in our study appears lower than those reported in the literature, even after making an approximate correction for differences of age. We suggest that this difference could be partly explained by the observation that the prevalence of MetS observed in the general population in France is lower than that in other industrialized countries.25 We report a high rate of decreased HDL cholesterol in our cohort, which is consistent with the literature.18,26 Little is known regarding the pathophysiology of HDL cholesterol diminution among childhood leukemia survivors. The commonly reported risk factors in the general population are genetic predisposition, smoking habits, low physical activity and obesity, and some of these factors are frequent among childhood acute leukemia survivors. Furthermore, interestingly, Canadian investigators have reported that several lipid abnormalities, including decreased HDL cholesterol, are already displayed by childhood leukemia patients at diagnosis.27 To our knowledge, no previous study has reported an age-specific analysis of MetS prevalence in this type of population. The age-specific cumulative prevalence of MetS was found to increase markedly, reaching 22.4% at 35 years of age, thereby highlighting the importance of a prolonged follow-up duration. The risk factors for MetS in adult survivors of childhood AL treated without HSCT have not been clearly established. Our multivariate analysis identified three significant risk factors of MetS: male sex, age at last evaluation and BMI at diagnosis.

Interestingly, cumulative steroid dose was not significantly associated with MetS in this L.E.A. cohort. Moreover, MetS prevalence among acute myeloblastic leukemia survivors was not lower than that observed among acute haematologica | 2016; 101(12)

lymphoblastic leukemia survivors, while AML protocols did not usually include any steroid therapy. Ongoing corticosteroid therapy induces many metabolic abnormalities, although limited data is available concerning the role that steroids play in the occurrence of MetS several years after treatment termination. In accordance with our results, the St Jude lifetime cohort study18 has also suggested that there was no significant association between cumulative steroid dose and MetS occurrence.

In the present study, male sex was significantly associated with MetS (OR 2.64; 95% CI 1.32-5.29; P=0.006). This increased risk among men was also observed in the French general population. Indeed, the reported prevalence of MetS in French adult females and males under 40 years of age ranged from 2.2% to 3.1% and from 4.3% to 5.0%, respectively.24,25 In the American population however, no difference between females and males has been reported,28 and interestingly, sex was not a significant risk factor in the St Jude lifetime cohort study.18 Notably, recent data suggest that the impact of cardiovascular risk factors may be more important for women.29 Consequently, it remains necessary to monitor cardiovascular and metabolic risk in both male and female patients. In our study, male sex was also a risk factor for three components of MetS: the elevation of fasting glucose, elevated triglycerides and elevated blood pressure. Investigators from the St Jude lifetime cohort study also found male sex was a significant risk factor for elevated fasting glucose.18 Inversely, investigators from Thailand have only reported “age at evaluation” as a significant risk factor for impaired glucose tolerance.30 Several teams have shown that male survivors exhibit an increased risk of elevated triglyceride levels and other lipid abnormalities.18,31 In our study, elevated blood pressure was significantly associated with male sex, which is also a known risk factor in the general population, especially among young and middle-aged adults.32 This risk factor has already been described in adult survivors of childhood AL.18,33 We found no clear evidence to explain a particularly increased risk of elevated blood pressure among male sur-

Table 2. Prevalence of the metabolic syndrome and its components.

Number of cases (%) MetS prevalence 45 (6.9%) No. of abnormal MetS components ≥1 385 (59.2%) ≥2 149 (22.9%) MetS components prevalence (among the entire cohort) Decreased HDL cholesterol 165 (26.8%) Increased triglycerides 75 (11.7%) Increased fasting glucose 36 (5.8%) Increased blood pressure 228 (36.7%) Increased waist circumference 93 (16.7%) Decreased HDL cholesterol 32 (5.2%) and increased triglycerides MetS components prevalence (among MetS patients) Decreased HDL cholesterol 38 (88.4%) Increased triglycerides 30 (68.2%) Increased fasting glucose 10 (23.3%) Increased blood pressure 37 (88.1%) Increased waist circumference 33 (76.7%) Decreased HDL cholesterol 24 (55.8%) and increased triglycerides

95% CI 5.1 - 9.2 55.3 - 63.0 19.8 - 26.4 23.4 - 30.5 9.3 - 14.5 4.2 - 8.1 32.9 - 40.7 13.7 - 20.1 3.6 - 7.3

74.1 - 95.6 52.3 - 80.9 12.3 - 39.0 73.6 - 95.5 61.0 - 87.7 40.0 - 70.6

CI: confidence interval; MetS: metabolic syndrome; HDL: high-density lipoprotein.

1607

P. Saultier et al.

Figure 3. Metabolic profile with regard to central nervous system irradiation and corresponding dosage in patients with at least one abnormal component of the metabolic syndrome. HDL: high-density lipoprotein; CNS: central nervous system; *P<0.05; ***P<0.001

vivors of childhood AL. The St Jude lifetime cohort team showed that this increased risk was probably not related to chronic kidney disease, for which the prevalence was 28% in their cohort.18 Interestingly, investigators in the USA have demonstrated that flow-mediated vasodilatation was better in female survivors of childhood AL compared with their male counterparts.34 Finally, in our study, male sex was a protective factor of increased abdominal circumference (OR 0.38; 95% CI 0.22-0.64; P<0.001). It was striking to find this predisposition of abdominal obesity among the female subjects because such abdominal fat distribution is rather a male characteristic in the general population, especially among young and middle-aged adults.35 It has already been reported that female adult survivors of childhood AL display an increased risk of obesity.36 Female patients may have higher leptin levels.37 Furthermore, it has been shown that female and male patients display significantly different BMI kinetics during treatment.38 Age at last evaluation was found to be a risk factor for MetS, increased blood pressure and increased abdominal circumference. This observation is consistent with published data concerning adult survivors of childhood AL18 and other childhood cancers. Furthermore, as indicated above, MetS prevalence is also known to increase with age in the general population.

Increased BMI at diagnosis was also a risk factor for MetS in our study. We have shown that it was also a significant risk factor for increased abdominal circumference.

Obesity, particularly abdominal obesity, is a major determinant of MetS in the general population.22,39 It has been suggested that abdominal obesity could be a marker of impaired fat storage capacity in the subcutaneous adipose tissue, thereby resulting in the accumulation of ectopic visceral fat leading to metabolic disturbances.40 This heightened risk of increased abdominal circumference and MetS could be explained by several observations. Children with an elevated BMI at diagnosis may have a genetic predisposition to obesity or certain metabolic disturbances.

Likewise, such patients may have a familial and social environment that renders them more vulnerable to obesity or metabolic complications. A study has shown that being overweight or obese at the time of diagnosis of childhood AL was associated with a higher risk of obesity during long-term follow-up.41 However, to our knowledge, 1608

BMI at diagnosis was never reported in the literature as a risk factor for MetS among adult survivors of childhood AL. These results indicate the importance of both monitoring the metabolic risk of children with an elevated BMI at the time of leukemia diagnosis and improving prevention and early treatment of associated complications.

In the univariate analysis, CNS irradiation was a significant risk factor for MetS only when administered at the 24 Gy dose, which could suggest a dose-effect relationship.

However, in the multivariate analysis, CNS irradiation was not found to be a risk factor for MetS, neither at the 18 Gy nor at the 24 Gy dose. We hypothesize that the loss of effect in the univariate analysis can be explained by the observation that â&#x20AC;&#x153;age at last evaluationâ&#x20AC;? was a confounding factor that has been corrected in the multivariate analysis. Indeed, CNS irradiation was more widely applied in older treatment protocols. In the literature however, brain irradiation has been frequently reported as a risk factor for MetS.18,42,43 This can probably be explained in part by the observation that our irradiated patients displayed a lower risk of elevated blood pressure. We conducted an analysis of the prevalence of each MetS component among patients who had at least one abnormal component (Figure 3). We noted a greater prevalence of increased abdominal circumference among irradiated patients and, inversely, a greater prevalence of elevated blood pressure among non-irradiated patients. The irradiated patients may therefore have a different metabolic risk profile compared with the non-irradiated patients, thereby suggesting varying mechanisms of pathogenesis. The St Jude lifetime cohort study showed that CNS irradiation is a risk factor for MetS.18 They did not find that irradiation had a significant effect on blood pressure; however, in agreement with our study, they found an increased risk of reduced HDL cholesterol and increased abdominal circumference among irradiated patients. Investigators in the USA have recently shown that CNS irradiation affects lipid profiles and in particular HDL cholesterol levels.31 However, the long-term effect of CNS irradiation on blood pressure among adult survivors of childhood AL remains poorly understood. In several studies concerning acute lymphoblastic leukemia survivors, no significant difference in blood pressure between irradiated and non-irradiated patients has been described.44,45 Another team has haematologica | 2016; 101(12)

Metabolic syndrome in childhood leukemia survivors

Table 3. Multivariate analysis of potential risk factors for the metabolic syndrome and its components.

Patients’ characteristics

Sex Female Male Age at last evaluation (years) BMI z-score at diagnosis CNS irradiation No 18 Gy 24 Gy

Metabolic syndrome OR P (95% CI)

Decreased HDL cholesterol OR P (95% CI)

1 2.64 (1.32-5.29) 1.10† (1.04-1.17)

1 0.74 0.006 (0.51-1.07) 0.99† 0.001 (0.96-1.03)

1.15‡ (1.01-1.32)

1.10‡ 0.037 (0.99 - 1.21)

1 0.92 (0.37-2.29) 1.87 (0.56-6.27)

1 1.11 0.866 (0.65-1.90) 2.76 0.309 (1.03-7.40)

Increased triglycerides OR P (95% CI)

Increased fasting glucose OR P (95% CI)

Increased blood pressure OR P (95% CI)

0.799

1 1 1.69 2.71 (1.01-2.84) 0.045 (1.25-5.84) 0.011 1.04† 1.05† (0.99-1.09) 0.097 (0.98-1.12) 0.132

1 3.47 (2.41-4.99) 1.08† (1.04-1.12)

0.076

1.02‡ 0.92‡ (0.90 - 1.17) 0.741 (0.73 - 1.16) 0.482

1.00‡ (0.90 - 1.11)

1 1 1.27 1.33 (0.64-2.50) 0.495 (0.54-3.32) 0.534 1.37 1.31 (0.41-4.60) 0.61 (0.26-6.64) 0.745

1 0.48 (0.28-0.82) 0.21 (0.07-0.64)

0.112

0.712 0.044

Increased waist circumference OR P (95% CI)

<0.001

1 0.38 (0.22-0.64) <1.07† (1.02-1.12)

0.984

1.44‡ (1.23-1.70)

<0.001

0.007 0.006

1 1.67 (0.87-3.24) 4.84 (1.63-14.44)

<0.001 0.008

<0.001

0.122 0.005

HDL: high-density lipoprotein; OR: odds ratio; CI: confidence interval; BMI: body mass index; CNS: central nervous system; †OR per each additional year of follow-up; ‡OR per each additional z-score unit; significant values (P<0.05).

reported higher systolic blood pressure values among nonirradiated patients, albeit without performing a statistical analysis of the difference.42 Finally, one study has shown that blood pressure was significantly higher in non-irradiated patients, which is consistent with our results.45 Overall, the consequences of CNS irradiation on cardiovascular risk factors seem complex. We suggest that both irradiated and non-irradiated patients are at elevated risk for metabolic and cardiovascular disturbances, although their metabolic risk profile may vary.

Our study has several limitations. Other potential risk factors for MetS, such as genetic factors and behavioral factors (dietary factors, physical activity and smoking status), were not accounted for in the analysis herein.

Furthermore, despite the comparison of the reported data to the latest available French data in the general population, this study lacks an appropriate comparison group.

The results of our study confirm the need for early, close and prolonged follow-up of adult survivors of childhood AL, even when treated without HSCT and without CNS irradiation. This prerequisite could enable both the early detection of metabolic abnormalities and the implementation of all appropriate therapeutic procedures to reduce the morbidity and mortality associated with cardiovascular complications in such patients. Upon initiation of the follow-up, the patient and his family should be advised to adopt general public health recommendations to reduce cardiovascular risk (e.g., increasing the consumption of fruit and vegetables, lowering the consumption of fat and

References 1. Kaatsch P. Epidemiology of childhood cancer.

Cancer Treat Rev. 2010;36(4):277-285.

2. Mertens AC, Liu Q, Neglia JP, et al. Causespecific late mortality among 5-year survivors

haematologica | 2016; 101(12)

carbohydrates, regular physical activity and quitting smoking). Indeed, adherence to these public health policies are known to be inversely associated with MetS among survivors of childhood cancer.46 Several other therapeutic interventions can be considered, of which some have already been reported in the literature. A meta-analysis study has examined the effect of physical exercise programs for children with cancer.47 Other investigators have examined the interest of growth hormone supplementation in a select population of patients.48,49 A group of American investigators has shown that preventive therapeutic interventions limiting the increase in BMI and insulin resistance may be particularly important during maintenance therapy.50 These therapeutic interventions will however require more expensive prospective studies for accurate evaluation.

Funding The study was funded in part by the French National Clinical Research Program, the French National Cancer Institute (INCa), the French National Research Agency (ANR), the Cancéropôle PACA, the Regional Council PACA, the Hérault departmental committee of the Ligue contre le Cancer and the French Institute for Public Health Research (IReSP).

Acknowledgments The authors would like to thank the L.E.A. study group (Online Supplemetary Data) for data collection and the patients and family members for their kind cooperation in this study.

of childhood cancer: the Childhood Cancer Survivor Study. J Natl Cancer Inst.

2008;100(19):1368-1379.

3. Mody R, Li S, Dover DC, et al. Twenty-five– year follow-up among survivors of childhood acute lymphoblastic leukemia: a report from the Childhood Cancer Survivor Study. Blood.

2008;111(12):5515-5523.

4. Chen MH, Colan SD, Diller L. Cardiovascular disease: cause of morbidity and mortality in adult survivors of childhood cancers. Circ Res. 2011; 108(5):619-628.

5. Gurney JG, Ojha RP, Ness KK, et al.

Abdominal aortic calcification in young adult

1609

P. Saultier et al. survivors of childhood acute lymphoblastic leukemia: results from the St. Jude Lifetime Cohort study. Pediatr Blood Cancer.

2012;59(7):1307-1309.

6. Dengel DR, Kelly AS, Zhang L, Hodges JS, Baker KS, Steinberger J. Signs of early subclinical atherosclerosis in childhood cancer survivors. Pediatr Blood Cancer. 2014; 61(3):532-537.

7. Gami AS, Witt BJ, Howard DE, et al.

Metabolic syndrome and risk of incident cardiovascular events and death: a systematic review and meta-analysis of longitudinal studies. J Am Coll Cardiol. 2007; 49(4):403414.

8. Isomaa B, Almgren P, Tuomi T, et al.

Cardiovascular morbidity and mortality associated with the metabolic syndrome.

Diabetes Care. 2001;24(4):683-689.

9. Paris C, Yates L, Lama P, Zepeda AJ, Gutiérrez D, Palma J. Evaluation of metabolic syndrome after hematopoietic stem cell transplantation in children and adolescents. Pediatr Blood Cancer. 2012;59(2): 306-310.

10. Bajwa R, Skeens M, Garee A, et al. Metabolic syndrome and endocrine dysfunctions after HSCT in children. Pediatr Transplant.

2012;16(8):872-878.

11. Oudin C, Auquier P, Bertrand Y, et al.

Metabolic syndrome in adults who received hematopoietic stem cell transplantation for acute childhood leukemia: an L.E.A. study.

Bone Marrow Transplant. 2015;50(11):14381444.

12. Kourti M, Tragiannidis A, Makedou A, Papageorgiou T, Rousso I, Athanassiadou F. Metabolic syndrome in children and adolescents with acute lymphoblastic leukemia after the completion of chemotherapy. J Pediatr Hematol Oncol. 2005; 27(9):499-501.

13. Gurney JG, Ness KK, Sibley SD, et al.

Metabolic syndrome and growth hormone deficiency in adult survivors of childhood acute lymphoblastic leukemia. Cancer.

2006;107(6):1303-1312.

14. Trimis G, Moschovi M, Papassotiriou I, Chrousos G, Tzortzatou-Stathopoulou F. Early indicators of dysmetabolic syndrome in young survivors of acute lymphoblastic leukemia in childhood as a target for preventing disease. J Pediatr Hematol Oncol.

2007;29(5):309-314.

15. van Waas M, Neggers SJCMM, Pieters R, van den Heuvel-Eibrink MM. Components of the metabolic syndrome in 500 adult long-term survivors of childhood cancer. Ann Oncol.

2010;21(5):1121-1126.

16. Aldhafiri F, Al-Nasser A, Al-Sugair A, AlMutairi H, Young D, Reilly JJ. Obesity and metabolic syndrome in adolescent survivors of standard risk childhood acute lymphoblastic leukemia in Saudi Arabia. Pediatr Blood Cancer. 2012;59(1):133-137.

17. Blijdorp K, van Waas M, van der Lely A-J, Pieters R, van den Heuvel-Eibrink M, Neggers S. Endocrine sequelae and metabolic syndrome in adult long-term survivors of childhood acute myeloid leukemia. Leuk Res. 2013;37(4):367-371.

18. Nottage KA, Ness KK, Li C, Srivastava D, Robison LL, Hudson MM. Metabolic syndrome and cardiovascular risk among longterm survivors of acute lymphoblastic leukaemia - From the St. Jude Lifetime Cohort. Br J Haematol. 2014; 165(3):364374.

19. Pui C-H, Campana D, Pei D, et al. Treating childhood acute lymphoblastic leukemia without cranial irradiation. N Engl J Med. 2009;360(26):2730-2741.

20. Oudin C, Simeoni M-C, Sirvent N, et al.

1610

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

Prevalence and risk factors of the metabolic syndrome in adult survivors of childhood leukemia. Blood. 2011;117(17):4442-4448.

Berbis J, Michel G, Baruchel A, et al. Cohort Profile: the French childhood cancer survivor study for leukaemia (L.E.A. Cohort). Int J Epidemiol. 2015;44(1):49-57.

Grundy SM, Cleeman JI, Daniels SR, et al.

Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation.

2005;112(17):2735-2752.

Strauss AJ, Su JT, Dalton VM, Gelber RD, Sallan SE, Silverman LB. Bony morbidity in children treated for acute lymphoblastic leukemia. J Clin Oncol Off J Am Soc Clin Oncol. 2001;19(12):3066-3072.

Pannier B, Thomas F, Eschwège E, et al.

Cardiovascular risk markers associated with the metabolic syndrome in a large French population: the ‘SYMFONIE’ study. Diabetes Metab. 2006;32(5 Pt 1):467-474.

Vernay M, Salanave B, de Peretti C, et al.

Metabolic syndrome and socioeconomic status in France: The French Nutrition and Health Survey (ENNS, 2006–2007). Int J Public Health. 2013;58(6):855-864.

Oeffinger KC, Adams-Huet B, Victor RG, et al. Insulin Resistance and Risk Factors for Cardiovascular Disease in Young Adult Survivors of Childhood Acute Lymphoblastic Leukemia. J Clin Oncol. 2009;27(22):36983704.

Halton JM, Nazir DJ, McQueen MJ, Barr RD. Blood lipid profiles in children with acute lymphoblastic leukemia.

Cancer.

1998;83(2):379-384.

Park Y-W, Zhu S, Palaniappan L, Heshka S, Carnethon MR, Heymsfield SB. The Metabolic Syndrome. Arch Intern Med. 2003;163(4):427-436.

Appelman Y, van Rijn BB, Ten Haaf ME, Boersma E, Peters SAE. Sex differences in cardiovascular risk factors and disease prevention. Atherosclerosis. 2015; 241(1):211-218.

Surapolchai P, Hongeng S, Mahachoklertwattana P, et al. Impaired glucose tolerance and insulin resistance in survivors of childhood acute lymphoblastic leukemia: prevalence and risk factors. J Pediatr Hematol Oncol. 2010; 32(5):383-389.

Malhotra J, Tonorezos ES, Rozenberg M, et al. Atherogenic low density lipoprotein phenotype in long-term survivors of childhood acute lymphoblastic leukemia. J Lipid Res. 2012;53(12):2747-2754.

Whelton PK. Epidemiology of hypertension.

Lancet. 1994;344(8915):101-106.

Janiszewski PM, Oeffinger KC, Church TS, et al. Abdominal obesity, liver fat, and muscle composition in survivors of childhood acute lymphoblastic leukemia. J Clin Endocrinol Metab. 2007;92(10):3816-3821.

Dengel DR, Ness KK, Glasser SP, Williamson EB, Baker KS, Gurney JG. Endothelial function in young adult survivors of childhood acute lymphoblastic leukemia. J Pediatr Hematol Oncol. 2008; 30(1):20-25.

Dixon AK. Abdominal fat assessed by computed tomography: sex difference in distribution. Clin Radiol. 1983;34(2):189-191.

Harper RLC, Breene RAL, Gattens M, Williams RM, Murray MJ. Non-irradiated female survivors of childhood acute lymphoblastic leukaemia are at risk of long-term increases in weight and body mass index. Br J Haematol. 2013;163(4):510-513.

Kohler JA, Moon RJ, Wright S, Willows E, Davies JH. Increased adiposity and altered adipocyte function in female survivors of

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

childhood acute lymphoblastic leukaemia treated without cranial radiation. Horm Res Pædiatrics. 2011;75(6):433-440.

Didi M, Didcock E, Davies HA, Ogilvy-Stuart AL, Wales JK, Shalet SM. High incidence of obesity in young adults after treatment of acute lymphoblastic leukemia in childhood. J Pediatr. 1995;127(1):63-67.

Carr DB, Utzschneider KM, Hull RL, et al.

Intra-abdominal fat is a major determinant of the National Cholesterol Education Program Adult Treatment Panel III criteria for the metabolic syndrome.

Diabetes.

2004;53(8):2087-2094.

Virtue S, Vidal-Puig A. Adipose tissue expandability, lipotoxicity and the Metabolic Syndrome--an allostatic perspective. Biochim Biophys Acta. 2010; 1801(3):338-349.

Asner S, Ammann RA, Ozsahin H, BeckPopovic M, von der Weid NX. Obesity in long-term survivors of childhood acute lymphoblastic leukemia. Pediatr Blood Cancer.

2008;51(1):118-122. van Waas M, Neggers SJCMM, Uitterlinden AG, et al. Treatment factors rather than genetic variation determine metabolic syndrome in childhood cancer survivors. Eur J Cancer Oxf Engl 1990. 2013;49(3):668-675.

Chow EJ, Simmons JH, Roth CL, et al.

Increased Cardiometabolic Traits in Pediatric Survivors of Acute Lymphoblastic Leukemia Treated with Total Body Irradiation. Biol Blood Marrow Transplant.

2010;16(12):1674-1681.

Veringa SJE, van Dulmen-den Broeder E, Kaspers GJL, Veening MA. Blood pressure and body composition in long-term survivors of childhood acute lymphoblastic leukemia. Pediatr Blood Cancer. 2012; 58(2):278-282.

Geenen MM, Bakker PJM, Kremer LCM, Kastelein JJP, van Leeuwen FE. Increased prevalence of risk factors for cardiovascular disease in long-term survivors of acute lymphoblastic leukemia and Wilms tumor treated with radiotherapy. Pediatr Blood Cancer.

2010;55(4):690-697.

Smith WA, Li C, Nottage KA, et al. Lifestyle and metabolic syndrome in adult survivors of childhood cancer: A report from the St. Jude Lifetime Cohort Study: Metabolic Syndrome in Childhood Cancer. Cancer. 2014;120(17): 2742-2750.

Braam KI, van der Torre P, Takken T, Veening MA, van Dulmen-den Broeder E, Kaspers GJL. Physical exercise training interventions for children and young adults during and after treatment for childhood cancer. Cochrane Database Syst Rev. 2013;4:CD008796.

Bulow B, Link K, Ahren B, Nilsson A-S, Erfurth EM. Survivors of childhood acute lymphoblastic leukaemia, with radiationinduced GH deficiency, exhibit hyperleptinaemia and impaired insulin sensitivity, unaffected by 12 months of GH treatment.

Clin Endocrinol (Oxf). 2004;61(6):683-691.

Follin C, Thilén U, Ahrén B, Erfurth EM. Improvement in cardiac systolic function and reduced prevalence of metabolic syndrome after two years of growth hormone (GH) treatment in GH-deficient adult survivors of childhood-onset acute lymphoblastic leukemia. J Clin Endocrinol Metab.

2006;91(5):1872-1875.

Esbenshade AJ, Simmons JH, Koyama T, Lindell RB, Friedman DL. Obesity and insulin resistance in pediatric acute lymphoblastic leukemia worsens during maintenance therapy: Obesity/Insulin Resistance Prevalent in ALL. Pediatr Blood Cancer. 2013;60(8):12871291.

haematologica | 2016; 101(12)